S 157 PL-1 EXPLORING THE FUNCTIONAL ROLE OF SYNAPTIC PROTEINS WITH RAPID TECHNIQUES E. Neher, J.B. Sørensen, G. Nagy It is widely accepted that SNARE proteins (such as Syntaxin, synaptobrevin, and SNAP-25) form a tight complex, which plays an essential role in membrane-to-membrane fusion during intracellular trafficking, neurotransmitter release, and hormone secretion. In regulated secretion the final step of vesicle fusion and release of contents is very rapid, when triggered by an abrupt increase in intracellular free calcium concentration ([Ca ++] i ). We made use of carbon fiber amperometry and membrane capacitance measurement, as rapid assays of secretion, and combined these techniques with flash-photolysis of caged-Ca++ to rapidly trigger exocytosis in bovine adrenal chromaffin cells. Several distinct kinetic phases of exocytosis can be clearly separated. The first phase is a rapid ‘exocytotic burst‘ immediately following the [Ca ++] increase, which releases about 200 granules. This burst is interpreted to represent the exocytosis of all those granules, which had been in a primed and release-ready state at the time of the flash. Since two components of this burst can be distinguished, we postulate that there are indeed two different states from which release can occur. Molecularly, we tentatively associated these two states with a ‘loose‘ and ‘tight‘ form of the SNARE complex (Xu et al., 1999). Following the exocytotic burst, there is a slow, sustained increase in capacitance, corresponding to a steady release-rate of about 50 fF/sec or 20 vesicles/sec. We interpret this as the arrival (or more exactly the priming) of new vesicles, which are being released instantly by the persistently high [Ca++] in the very moment when they become releasecompetent. Specifically, we think that SNARE complex formation is the rate-limiting step for this process. We employed these techniques to study cells from a variety of transgenic animals, which lacked one or the other of the synaptic proteins (knockouts). In addition we mutated some of the SNARE proteins and overexpressed them in chromaffin cell cultures. When we manipulated secretory proteins in this way, we saw changes in the kinetic components of the secretory response, which loosely can be grouped into three categories: A ‘Loss of Function‘ phenotype, i.e. an unspecific reduction in all three kinetic components. The manipulations, which lead to this change include some mutations of residues at the outer surface of the SNARE complex, knockout of Munc18 (Voets et al. 2001a) and ATP depletion. Particularly we would like to point out that mutating polarizable groups in the so-called zerolayer of the SNARE complex slow down the sustained component and reduce the size of the burst. In the framework of our model this indicates a role of the zero layer in the formation or stability of trans-SNARE complexes, or in the disassembly of cisSNARE complexes (Wei et al., 2000). A ‘Gain of Function’ phenotype, i.e. an increase in all three kinetic components. The most prominent effect of this kind is obtained by overexpressing Munc 13-1 (Ashery et al., 2000). This indicates that this protein has a priming function or else it leads to formation or stabilization of more SNARE complexes. Other manipulations, which increase the size of the exocytotic burst, include overexpression of Munc18, increased basal level of [Ca ++] and stimulation by phorbol esters. A specific loss of the rapid burst component. Surprisingly this phenotype is generated by a number of manipulations, which affect the outer surface of the SNARE-complex or potential binding partners. We first observed this phenotype when adding an antibody against SNAP25 and also after action of Botulinum neurotoxin A (Xu et al., 1999; Xu et al., 1998). Meanwhile, we see this phenotype also for a number of quite diverse manipulations such as modifications of charged residues on the outside of the SNARE-complex and when examining cells from Synaptotagmin knockout mice (Voets et al., 2001b). For interpretation of these results we think in terms of two forms of complexes, one being matured and involving several interaction partners (such as synaptotagmin). Only this complex can mediate the very rapid and highly Ca ++-sensitive burst component. Many manipulations might prevent the formation of this delicate entity and leave the vesicle with a more rudimentary form of the SNARE complex, which represents the slow burst component. References: ASHERY, U., VAROQUEAUX, F., VOETS, T., BETZ, A., THAKUR, P., KOCH, H., NEHER, E., BROSE, N. AND RETTIG, J. (2000). Munc13-1 acts as a priming factor for large densecore

vesicles in bovine chromaffin cells. EMBO J. 19, 3586-3596. XU, T., BINZ, T., NIEMANN, H. and E. NEHER (1998). Multiple kinetic components of exocytosis distinguished by neurotoxin sensitivity. Nature neuroscience 1, 192-200. XU, T., RAMMNER, B., MARGITTAI, M., ARTALEJO, A.R., NEHER, E. AND R. JAHN (1999). Inhibition of SNARE complex assembly differentially affects kinetic components of exocytosis. Cell 99, 713-722. VOETS, T., TOONEN, R.F., BRIAN, E.C., De WIT, H., MOSER, T., RETTIG, J., SUEDHOF, T.C., NEHER, E.and VERHAGE, M. (2001a). Munc18-1 promotes larger dense-core vesicle docking. Neuron 31, 581-591. VOETS, T., MOSER, T., LUND, P.-E., CHOW, R.H., GEPPERT, M., SUEDHOF, T.C. and NEHER, E. (2001b). Intracellular calcium dependence of large dense-core vesicle exocytosis in the absence of synaptotagmin I. PNAS 98, 11680-11680. WEI, S.-H., XU, T., ASHERY, U., KOLLEWE, A., MATTI, U., ANTONIN, W., RETTIG, J. AND NEHER, E. (2000). Exocytotic mechanism studied by truncated and zero layer mutants of the C-terminus of SNAP-25. EMBO J. 19, 12791289. Max-Planck-Institut für biophysikalische Chemie, D-37077 Goettingen, Germany PL-2 ON THE WAY TO UNDERSTAND THE CAPRICIOUS ATPSENSITIVITY OF ATP-SENSITIVE K+ CHANNELS T. Baukrowitz ATP-sensitive potassium channels (KATP channels) are metabolic sensors that couple electrical activity to cellular metabolism. They play an important role in processes such as insulin secretion, cardioprotection, neuroprotection, or vasodilatation. KATP channels are heteromultimers composed of inwardly rectifying K + channel subunits (Kir6.2) and sulphonylurea receptor subunits (SUR)1. Central to their function as metabolic sensors is their sensitivity to inhibition by intracellular ATP which is antagonised by MgADP. Although this ATP/ADP dependency provided a straightforward explanation for much of the observed metabolic regulation, it has been puzzling for long why the channels’ sensitivity to ATP largely varies among various types of cells and even within the same cell. Work of the last few years identified phosphatidyinositol phosphates (PIPs), components of the membrane bilayer, as regulators of many ion transporters and channels2. It is now clear that most, if not all, members of the superfamily of inwardly rectifying K+ channels require the presence of PIPs in the membrane to be functional. In KATP channels, PIPs proofed to be potent modulators of ATP inhibition that are able to shift the ATP sensitivity by more than three orders of magnitude 3. The membrane concentrations of PIPs are regulated by a set of enzymes comprising phospholipases, phospholipid-phosphatases and phospholipid-kinases. These enzymes might represent a link of K ATP channel activity to many signal transduction pathways4. In line with this concept, KATP channels were shown to be regulated by phospholipase C and phospholipid-kinases. Recently, long-chain acyl-coenzyme A esters such as oleyl-CoA were identified as another class of modulators of ATP sensitivity tying KATP channel activity to fatty acid metabolism5. The main structural feature of those lipids that affect ATP-inhibition appears to be a highly negatively charged head group as found in PIPs and oleyl-CoA. Biochemical and electrophysiological experiments indicated that PIPs interact electrostatically with a number of basic residues in the Cterminus and N-terminus of the Kir6.2 subunits. These results show that lipids of the cell membrane do not only serve as scaffolds, but instead represent a potent machinery to regulate the function of ion channels. References: 1. Babenko, A. P., Aguilar-Bryan, L., and Bryan, J. (1998) A view of SUR/KIR6.X, KATP channels. Annu Rev Physiol 60, 667-87 2. Hilgemann, D. W., Feng, S., and Nasuhoglu, C. (2001) The complex and intriguing lives of PIP2 with ion channels and transporters. Sci STKE. 2001 Dec 4;2001(111):RE19 3. Baukrowitz, T., Schulte, U., Oliver, D., Herlitze, S., Krauter, T., Tucker, S. J., Ruppersberg, J. P., and Fakler, B. (1998) PIP 2 and PIP as determinants for ATP inhibition of K ATP channels. Science 282, 1141-4 4. Baukrowitz, T., and Fakler, B. (2000) K(ATP) channels: linker

S 158 between phospholipid metabolism and excitability. Biochem Pharmacol 60(6), 735-40 5. Liu, G. X., Hanley, P. J., Ray, J., and Daut, J. (2001) Long-chain acyl-coenzyme A esters and fatty acids directly link metabolism to K(ATP) channels in the heart. Circ Res 88(9), 918-24. Physiologisches Institut der Universität Jena, Germany PL-3 EPITHELIAL AMINO ACID TRANSPORTERS: MEMBRANE PROTEINS WITH MULTIPLE FUNCTIONS C.A.R. Boyd Net transport of aminonitrogen across epithelia is central to the physiological basis of nutrition and hence to intestinal, renal, placental and mammary function. Amino acid transport systems were originally recognised by their functional properties, specifically by their substrate specificities, energy-coupling and pathophysiology when recognised as ”inborn errors of metabolism”. Many (but not yet all) of the genes encoding the integral membrane proteins that form the basis of such systems and that act to catalyse amino acid transport across the plasma membrane of epithelial cells have been discovered in the last ten years through expression cloning. Their functional properties have then been studied in heterologous expression systems. These transporters form gene families based on sequence homology and hence on presumptive structure (at present there are no structures of such molecules available). Asymmetry of the distribution of transporters is essential to epithelial function and thus a major goal is to discover how these recently identified proteins are distributed between the apical and basal surfaces of epithelial cells. Since asymmetric distribution of such molecules coupled to their differing functional properties is a prerequisite for generating net transport, at present a major question in epithelial cell biology arises from attempts to understand the rules governing the origin of polarity of transporter distribution; thus in addition to work on transporters studied individually in non-polarised expression system there are real reasons for wishing to study the properties of such molecules when expressed in model epithelial cells (see Verrey 2002). The heteromeric family of amino acid transporters form one particular fascinating group that is the focus of very substantial current research (well reviewed by Verrey et al 2000, Wagner et al 2001). At least four different well-described amino acid transport systems are members of this family. These have been cloned and through this, novel isoforms recognised some with obviously distinct functional properties. The discovery that such transporters are distributed to the apical or basal surface of intestinal and renal epithelia depending on the nature of the associated heavy chain has provided a major insight. This has also contributed greatly to understanding the molecular basis of certain rare genetic disorders of amino acid transport (the aminoacidurias cystinuria and lysine protein tolerance, LPI). Here the contribution of physiology, molecular genetics and disease phenotype to understanding the mechanisms of pathophysiology have been remarkable (Chillaron et al 2001). Physiologically an important question in the field has been to establish whether transport is electroneutral or electrogenic, the extent to which it is coupled to ion gradients (and if so for which ion), and the underlying transport mechanisms. Mechanistic questions have then focused on whether the transporter is capable of generating net transport or whether the protein acts as an obligatory exchanger. This has been an area that was ignored for many years but which is now seen to be central to the unravelling the physiological design of the process of transepithelial amino acid transport. In this context, recent theoretical insight (Daniel and Boyd 2002, unpublished) into transporter mechanism will be discussed. Additionally, for heteromeric amino acid transporters the multiple roles of the heavy chains is another area full of experimental challenge. Recent work will be reviewed (Deves and Boyd 2000) focusing on the possible roles of CD98hc (known also as membrane antigen 4F2). It seems possible that this protein is involved, in addition to or as part of its function in amino acid transport, in transmembrane signalling. It is intriguing that this protein is also known to be a receptor for extracellular glycoprotein ligands, to be involved in regulating cell adhesion via interaction with specific integrins and to generate tyrosine kinase based signalling within the cell. This raises the issue of whether the family of CD98 related amino acid transporters may be involved not only in transepithelial amino acid transport but also in processes such as epithelial cell migration, cell adhesion to the basal lamina, stem cell

proliferation and nutrient signalling. The various amino acid transport systems responsible for cationic amino acid transport will be used as an exemplar showing functional properties of such systems in the intact epithelium. Subtle interactions between the transport of amino acids at different faces of the epithelial cell have been known for more than 40 years. The basis of these interactions is now becoming clear as the properties of systems hitherto studied in isolation in expression systems are put together, providing a paradigm of the opportunities and challenges of modern post-genomic physiology. References: Chillaron J, Roca R, Valencia A, Zorzano A, Palacin M. (2001) Heteromeric amino acid transporters: biochemistry, genetics, and physiology. Am J Physiol Renal Physiol. 281 F995-F1018. Deves R, Boyd CAR. (2000) Surface antigen CD98(4F2): not a single membrane protein, but a family of proteins with multiple functions. J Membr Biol.; 173:165-177. Verrey F, Meier C, Rossier G, Kuhn LC. (2000) Glycoprotein-associated amino acid exchangers: broadening the range of transport specificity. Pflugers Arch.; 440:503-512. Verrey F. (2002), Epithelial amino acid transport by heterodimeric amino acid exchangers. Journal of Physiology (in press). Wagner CA, Lang F, Broer S. (2001) Function and structure of heterodimeric amino acid transporters. Am J Physiol Cell Physiol. ; 281:C1077-C1093. Department of Human Anatomy and Genetics, University of Oxford, South Parks Road, Oxford, OX1 3QX, UK PL-4 CYTOCHROME P450 (CYP) EPOXYGENASES: DEFINING THEIR ROLE IN VASCULAR SIGNALLING AND HOMEOSTASIS R. Busse Most cytochrome P450 enzymes (CYPs) are expressed in the liver, however some CYPs, such as members of the CYP 2J, CYP 2C and CYP 4A subfamilies, can be detected in extrahepatic tissues particularly in the cardiovascular system. Recent data suggest that specific CYPs localised in the vascular smooth muscle and endothelium, contribute to the acute and long-term regulation of vascular tone and homeostasis.1,2 A link between CYP activity and the generation of an endotheliumderived hyperpolarizing factor (EDHF) that mediates nitric oxide (NO)and prostacyclin (PGI2)-independent relaxation, has been intensively investigated over the last decade. Indeed, the hyperpolarizing factor produced by coronary and renal arteries from several species, including humans and pigs displays characteristics similar to those of a CYPderived metabolite of arachidonic acid. The arachidonic acid metabolites in question are the epoxyeicosatrienoic acids (EETs); 5,6-EET, 8,9-EET, 11,12-EET and 14,15-EET which are generated by endothelial CYP epoxygenases. EETs induce relaxation either by activating Ca2+-dependent K+ (K +Ca) channels on smooth muscle cells, or in the case of many EDHF-mediated responses, by activating K+Ca channels on endothelial cells and inducing endothelial cell hyperpolarization which is a crucial step initiating the EDHF response. The correlation between CYP activity and NO/PGI2-independent relaxation has been addressed by enhancing the expression of CYP enzymes, as well as by attenuating CYP expression. Utilizing the former approach, CYP inducers, such as β-naphthoflavone or nifedipine, were found to increase the synthesis of EETs by cultured and native endothelial cells, and to enhance the agonist-induced, EDHF-mediated hyperpolarization and relaxation of intact coronary artery segments. 3,4 However, the most convincing evidence obtained in support of the hypothesis that a CYP 2C epoxygenase is intimately involved in the EDHF response in porcine coronary arteries, was provided by the use of antisense oligonucleotides directed against the coding region of CYP 2C8/9. Incubation of porcine coronary arteries with antisense, but not sense or scrambled, oligonucleotides markedly reduced CYP 2C mRNA and protein and attenuated bradykinin-induced, EDHF-mediated hyperpolarization and relaxation without compromising responsiveness to endogenously produced NO or an NO donor.4 In addition to affecting local vascular tone, EETs facilitate communication between vascular cells by modulating gap junctional communication. In CYP 2Cexpressing porcine coronary endothelial cells, bradykinin, which enhances EET formation, elicits a biphasic effect on the electrical cou-

S 159 pling and transfer of Lucifer Yellow between endothelial cells, consisting of a transient increase in coupling followed by a sustained uncoupling. The initial phase is sensitive to the CYP 2C9 inhibitor sulfaphenazole, and the PKA inhibitor Rp-cAMPS, and can be mimicked by caged cyclic AMP and the PKA activator Sp-cAMPS. Gap junction uncoupling in bradykinin-stimulated porcine coronary endothelial cells can be attributed to the activation of extracellular regulated protein kinase, Erk1/2, and the subsequent phosphorylation of connexin 43 (the most abundant connexin expressed in the cells under investigation). Bradykinin-induced cell uncoupling could be prevented by inhibiting the activation of Erk1/2. In human endothelial cells, which express little CYP 2C, bradykinin elicited only an Erk1/2-mediated inhibition of inter-cellular communication. However in these cells, the CYP 2C9 product 11,12-EET, also exerted a dual effect on both the electrical and dye coupling which was sensitive to Rp-cAMPS. Thus, CYP -dependent pathways and 11,12-EET can positively regulate inter-endothelial gap junctional communication, via the activation of PKA, an effect which is curtailed by the subsequent activation of Erk1/2. Although humoral stimuli, such as bradykinin, can elicit the generation of EETs, hemodynamic stimuli such as fluid shear stress and pulsatile stretch are the physiologically most important determinants of the continuous production of these autacoids in vivo. 5 Indeed, pulsatile changes in transmural pressure, enhance the production of EETs by native porcine coronary endothelial cells. Prolonged exposure to either stretch or fluid shear stress also increases the expression of CYP 2C mRNA and protein in cultured and native endothelial cells, identifying the CYP 2C as a mechano-sensitive endothelial cell gene product.6 Not all of the cellular effects elicited by the EETs can be linked to the activation of K +Ca channels. Therefore, it follows that these CYPderived metabolites of arachidonic acid may be more than modulators of vascular tone. Indeed, EETs, in particular 11,12- and 14-15-EET, modulate the activity of protein tyrosine phosphatases and activate several intracellular protein kinases including tyrosine kinases, the p38 MAP kinase, Erk1/2, as well as Akt/PKB. 7 11,12-EET is generally assumed to be an anti-inflammatory substance as its infusion into the carotid artery of mice attenuated the tumour necrosis factor (TNF)induced increase in NF-κB activity and prevented the TNF-induced expression of the pro-arterosclerotic adhesion molecule vascular cell adhesion molecule-1 (VCAM-1).8 Overexpression of CYP 2C9 in cultured human endothelial cells markedly increases proliferation, an effect paralleled by an upregulation of the G1 phase regulatory protein, cyclin D1. The specific CYP 2C9 inhibitor, sulfaphenazole, prevents both the enhanced cell proliferation and upregulation of cyclin D1. Cyclin D1 expression appears to be linked to changes in the activity of the c-Jun N-terminal kinase (JNK) as CYP 2C9-overexpression is associated with a decrease in JNK activity. Indeed, co-expression of wild type JNK with CYP 2C9 attenuates the CYP 2C9-induced increase in cyclin D1 expression as well as CYPinduced endothelial cell expression. In contrast, cotransfecting dominant negative JNK with CYP 2C9, restores the CYP 2C9-mediated upregulation of cyclin D1. The CYP 2C9-induced inactivation of JNK is the result of its dephosphorylation by dual specificity MAP kinase phosphatases (MKPs) and either the transfection of endothelial cells with CYP 2C9 or incubation of cells with 11,12-EET markedly enhances the expression of MKP-1. While such data demonstrate a link between CYP 2C9 activation and endothelial cell proliferation mainly by altering JNK activity, EETs exert additional cellular effects which also promote endothelial cell proliferation. For example, there is also evidence of an intracellular cross-talk between EETs and the EGF receptor. This crosstalk seems to play an important role in regulating CYP 2C9-induced angiogenesis (formation of capillary tubes). Microsomal CYP epoxygenases continuously produce reactive oxygen species, implying that similar enzymes expressed in endothelial cells may contribute significantly to the vascular production of oxygenderived free radicals. Indeed, the CYP 2C which generates EETs and mediates NO/PGI2-independent relaxation in porcine coronary arteries also produces reactive oxygen species.9 The consequences of this CYPderived free radical generation range from the impairment of NOmediated relaxation to a chronic elevation in the activity of the redoxsensitive transcription factor, NF-κB and the expression of VCAM-1. As endothelial CYP activity and expression can be stimulated by hemodynamic forces, such as cyclic stretch, the activation of CYP 2C in endothelial cells may contribute to the stretch-induced generation of O2-, which has until now been attributed to the activation of the NADPH oxidase. It follows therefore that the effects observed following the exogenous application of a CYP product to vascular cells may be different to those elicited as a consequence of endogenous CYP

activation. For example, although the direct application of EETs to endothelial cells attenuates NF-κB activity, overexpressing CYP 2C9 protein in these cells to increase endogenous CYP activity has the opposite effect. Thus, despite the potential anti-inflammatory properties of the EETs, enhanced activation of vascular CYPs may eventually be detrimental to vascular homeostasis, as a consequence of the simultaneous generation of EETs and reactive oxygen species. References: 1. Fleming I. Cytochrome P450 enzymes in vascular homeostasis. Circ Res. 2001;89:753-762. 2. Roman RJ. P-450 metabolites of arachidonic acid in the control of cardiovascular function. Physiol Rev. 2002;82:131-185. 3. Fisslthaler B, Hinsch N, Chataigneau T, Popp R, Kiss L, Busse R, Fleming I. Nifedipine increases cytochrome P4502C expression and EDHF-mediated responses in coronary arteries. Hypertension. 2000;36:270-275. 4. Fisslthaler B, Popp R, Kiss L, Potente M, Harder DR, Fleming I, Busse R. Cytochrome P450 2C is an EDHF synthase in coronary arteries. Nature. 1999;401:493-497. 5. Popp R, Fleming I, Busse R. Pulsatile stretch elicits the release of the endothelium-derived hyperpolarizing factor from isolated coronary arteries: a modulator of arterial compliance. Circ Res. 1998;82:696-703. 6. Fisslthaler B, Popp R, Michaelis UR, Kiss L, Fleming I, Busse R. Cyclic Stretch Enhances the Expression and Activity of the Coronary EDHF Synthase. Hypertension. 2001;38:1427-1432. 7. Fleming I, Fisslthaler B, Michaelis UR, Kiss L, Popp R, Busse R. The coronary EDHF stimulates multiple signalling pathways and proliferation in vascular cells. Pfluger’s Arch Eur J Physiol. 2001;442:511-518. 8. Node K, Huo Y, Ruan X, Yang B, Spiecker M, Ley K, Zeldin DC, Liao JK. Anti-inflammatory properties of cytochrome P450 epoxygenase-derived eicosanoids. Science. 1999;285:1276-1279. 9. Fleming I, Michaelis UR, Bredenkotter D, Fisslthaler B, Dehghani F, Brandes RP, Busse R. Endothelium-derived hyperpolarizing factor synthase (Cytochrome P450 2C9) is a functionally significant source of reactive oxygen species in coronary arteries. Circ Res. 2001;88:44-51. Institut für Kardiovaskuläre Physiologie, Klinikum der J.W.G.-Universität, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany PL-5 THE CYTOKINE RESPONSE TO EXERCISE B.K. Pedersen In relation to physical exercise, interleukin (IL)-6 is produced in larger amounts than any other cytokine and the production of IL-6 during exercise is related to the intensity and duration (6-11). Recent studies suggest that IL-6 is produced locally in working skeletal muscle (1,3,4,13,15) and can account for the exercise-induced increase in plasma IL-6 during exercise (17). Furthermore, circulating monocytes do not contribute to the IL-6 production during exercise (12,14). Muscle-derived IL-6 is released from the working muscle into the circulation in high amounts, the IL-6 release being 17 fold higher than the amount accumulated in plasma during exercise (17). Adrenaline does not play an important role in stimulating the release of IL-6 (16), whereas the glycogen content appears to be a determining factor for the production of IL-6 (3,15). Thus, in a recent study one leg was glycogen-depleted by exercise prior to two-legged kneeextensor exercise. In the latter design both legs were exposed to the same concentration of blood-glucose and hormones. The IL-6 release, mRNA-IL-6 and mRNA (15) and heat shock protein (HSP)72 (2) in muscle were faster and greater in the leg with low glycogen compared to the control leg (2, 15). In another study, it was demonstrated that the transcription rate for IL-6 in muscle nuclei isolated from muscle biopsies obtained before, during and after exercise, was very fast and further enhanced when muscle glycogen content was low (3). We have also recently shown that cultured human primary muscle cells are capable of increasing IL-6mRNA when incubated with the calcium ionophore ionomycin (4). Therefore, it is likely that myocytes produce Il-6 in response to muscle contraction and production of Il-6 by such tissue accounts for the exercise-induced increase in plasma levels of this cytokines. Recently, we demonstrated that 3 h of non-damaging exercise was sufficient to induce both transcription of the IL-6 and HSP72 genes. Anti-oxidant treatment blunted the exercise-induced

S 160 HSP72, but not the IL-6 response, suggesting that in contracting muscle, IL-6 is not mediated via an HSP72 dependant pathway (unpublished data). The biological roles of muscle-derived IL-6 have been investigated in studies, where human recombinant IL-6 was infused into healthy volunteers to closely mimic the IL-6 concentrations observed during prolonged exercise. We have demonstrated that physiological concentrations of IL-6 clearly induce lipolysis. Although, we are yet to determine the precise biological action of muscle-derived IL-6, our date support the hypothesis that the role of IL-6 release from contracting muscle during exercise is to act in a hormone like manner to mobilize extracellular substrates and/or augment substrate delivery during exercise. Another possible biological role of IL-6 is to inhibit the production of TNF and thereby to inhibit TNF-induced insulin resistance. Of note, Wallenius et al (18) have recently demonstrated that IL-6 deficient mice develop obesity and glucose intolerance, while chronic treatment of these animals with IL-6 partially attenuates these metabolic perturbations. Recently, we demonstrated a small insignificant increase in TNFmRNA in working skeletal muscle and no release of a TNF protein, whereas IL-6mRNA increased 100 fold with the release of a protein to the circulation (unpublished data). These data support the idea that muscle-derived IL-6 impairs TNF expression in skeletal muscle. Previous studies have shown that IL-6 inhibits the production of TNF from LPS-treated monocytes and that TNF levels are elevated in anti-IL-6 treated mice and in IL-6 deficient knock-out mice. To test the hypothesis that muscle-derived IL-6 would inhibit the production of TNF, we recently exposed normal healthy volunteers to a small bolus of E.Coli endotoxin (LPS) following either rest or three hours of exercise. The results showed a 3 fold increase in plasma TNF in response to LPS administration in vivo in the resting group, whereas physical exercise totally inhibited the TNF production (unpublished data). These data further support the idea that muscle-derived IL-6 may inhibit the production of TNF and thereby inhibit insulin resistance. Muscle-derived IL-6 may have several biological functions (10): 1) During muscle work, classical target genes for IL-6, C/EBPbeta and C/EBPdelta are up-regulated and the STAT3 system is activated indicating the muscle-derived IL-6 play a local role. 2) Muscle-derived IL-6 is released into the circulation in high amount and is likely to work in a hormone-like fashion, exerting an effect on the liver and adipose tissue, thereby contributing to maintain glucose homeostasis during exercise and mediating exercise-induced lipolysis. 3) Musclederived IL-6 may also work to inhibit the effects of pro-inflammatory cytokines such as TNF-alpha. The latter cytokine is produced by adipose tissue and inflammatory cells and plays a pathogenetic role in insulin resistance and atherogenesis.

interleukin-6 mRNA production in rat skeletal muscles. Journal of Physiology (London), 528, 157-163 2. Febbraio MA, Steensberg A, Walsh R, Koukoulas I, van Hall G, Saltin B, Pedersen BK (2001). Effect of glycogen availability on HSP72 expression in contracting human skeletal muscle. Journal of Physiology (London), in press 3. Keller C, Steensberg A, Pilegaard H, Osada T, Saltin B, Pedersen BK, Neufer PD (2001). Transcriptional activation of the IL-6 gene in human contracting skeletal muscle – influence of muscle glycogen content. The FASEB Journal 15: 2748-2750 4. Keller C, Hellsten Y, Pilegaard H, Febbraio M, Pedersen BK. Human muscel cells express IL-6 via a Ca dependent pathway. J.Physiology (in press) (abstr) 2002 5. Ostrowski, K., Rohde, T., Zacho, M., Asp, S., & Pedersen, B. K. (1998a). Evidence that IL-6 is produced in skeletal muscle during intense long-term muscle activity. Journal of Physiology (London) 508, 949-953. 6. Ostrowski K, Hermann C, Bangash A, Schjerling P, Nielsen JC, Pedersen BK (1998b). A trauma-like production of cytokines in humans in response to treadmill running. Journal of Physiology (London) 508: 889-894 7. Ostrowski K, Rohde T, Asp S, Schjerling P, Pedersen BK (1999) Pro- and anti-inflammatory cytokine balance in strenuous exercise in humans. Journal of Physiology (London), 515.1, 287-291 8. Pedersen BK & Hoffman- Goetz, L. (2000). Exercise and The Immune System: Regulation, Integration and Adaption. Physiological Reviews 80, 1055-1081. 9. Pedersen B.K., Steensberg A., & Schjerling P. (2001). Exercise and interleukin-6. Current Oppinion in Hematology 8, 137-141. 10. Pedersen BK, Steensberg A, Schjerling P (2001). Muscle-derived interleukin – possible metabolic effects. Topical Review. Journal of Physiology (London), 536.2. 329-337. 11. Pedersen BK, Woods JA, Nieman DC (2001). Exercise-induced immune changes - an influence on metabolism? Trends in Immunology formerly Immunology Today, 22: 473-375 12. Starkie, R.L., Angus, D.J., Rolland, J., Hargreaves, M., & Febbraio, M. (2000). Effect of prolonged submaximal exercise and carbohydrate ingestion on monocyte intracellular cytokine production in humans. Journal of Physiology (London) 528, 647-655. 13. Starkie R.L., Arkinstall M.J., Koukoulas I., Hawley J.A., & Febbraio M.A. (2001) Carbohydrate ingestion attenuates the increase in plasma interleukin 6, but not skeletal muscle interleukin 6 mRNA, during exercise in humans. Journal of Physiology (London), 533, 585-91 14. Starkie, R.L., Rolland, J., Angus, D.J., Anderson, M.J., Febbraio, M.A. (2001) Circulating monocyes are not the source of elevations in plasma IL-6 and TNF-alpha levels after prolonged running. American Journal of Physiology 280, C769-74 15. Steensberg A, Febbraio MA, Osada T, Schjerling P, van Hall G, Saltin B, Pedersen BK. Low glycogen content increases interleukin-6 production in contracting human skeletal muscle. Journal of Physiology (London), 537: 633-639, 2001 16. Steensberg A, Toft AD, Schjerling P, Halkjær-Kristensen, Pedersen BK (2001). Plasma interleukin-6 during strenuous exercise – role of adrenaline. American Journal of Physiology, 281(3):C1001-4, 17. Steensberg A., van Hall, G., Osada T, Sacchetti, M., Saltin, B., & Pedersen BK (2000). Production of IL-6 in contracting human skeletal muscles can account for the exercise-induced increase in plasma IL-6. Journal of Physiology (London) 529, 237-242. 18. Wallenius V, Wallenius K, Ahren B, Rudling M, Calsten H, Dickson H, Dickson SL, Ohlsson C, Jansson JO (2002). Interleukin-6deficient mice develop mature-onset obesity. Nature Medicine 8: 75-79 The Department of Infectious Diseases and the Copenhagen Muscle Research Centre, Rigshospitalet, University of Copenhagen, Denmark

PL-6 Referrences: 1. Jonsdottir, I., Schjerling, P., Ostrowski, P., Asp, S, Richter, E.A. & Pedersen, B.K. (2000). Muscle contractions induces

ATP-SENSITIVE K-CHANNELS: REGULATION AND ROLES IN HEALTH AND DISEASE F.M. Ashcroft

S 161 ATP sensitive K-channels (KATP channels) play important roles in a diverse range of tissues (including pancreatic β−cells, neurones, and cardiac, skeletal and smooth muscles) by coupling the metabolic state of the cell to its electrical activity (Ashcroft and Gribble, 1999). In pancreatic β-cells, K ATP closure in response to glucose metabolism produces membrane depolarization, leading to Ca2+ influx and insulin secretion. KATP channels are also involved in neuronal glucose sensing, glucose uptake in skeletal muscle, seizure protection, regulation of vascular tone and the response to cardiac and cerebral ischemia. Metabolic regulation is mediated by changes in intracellular ATP (which blocks the channel) and MgADP (which activates the channel). KATP channels are inhibited by sulphonylurea drugs, which stimulate insulin secretion and are used to treat type 2 diabetes, and activated by KATP-channel openers, a structurally diverse group of drugs with a wide range of potential therapeutic applications. The β-cell KATP channel is 4:4 heteromeric complex of pore-forming Kir6.2 and regulatory SUR1 subunits. SUR1 endows the channel with sensitivity to the stimulatory effects of MgADP and K ATP -channel openers and the inhibitory action of sulphonylureas. ATP closes the channel by binding to Kir6.2. Although mutation of both N and C terminal residues of Kir6.2 reduces ATP inhibition (Tucker et al., 1988), it is not clear whether the nucleotide interacts directly with these residues or if the reduction in ATP sensitivity is mediated allosterically. To determine which residues contribute directly to the ATP-binding site, we used a cysteine substitution approach, in combination with thiol modification. Thiol reagents bind irreversibly to cysteine residues and provide a means to introduce either a positive (MTSEA) or negative (MTSES) charge into the protein. If the ATPbinding site lies sufficiently close to the modified cysteine, the introduced charge will exert an electrostatic effect on ATP binding and thus on the ability of ATP to close the channel. Comparison of the ATP sensitivity of cysteine-mutant channels (expressed in Xenopus oocytes) before and after thiol modification revealed that the side-chains of several residues (eg. R50, K185) lie within a few angstroms of the phosphate tail of the bound ATP molecule. Thus, the ATP-binding site contains contributions from both the amino (R50) and carboxy (K185) termini of Kir6.2. Binding of ATP to Kir6.2 subunit is allosterically linked to closure of the channel pore. To elucidate the location of the conformational changes associated with channel closing, we examined the block by Ba2+ ions. Barium produces a voltage-dependent block of Kir6.2/SUR1 suggesting it acts by plugging the pore. Mutation of V129 to threonine (V129T) markedly reduced Ba2+ block suggesting Ba2+ binds at or near this residue, which is predicted to lie just below the selectivity filter. In both wild-type and V129T mutant channels, the on-rate of Ba2+ block was not slowed by the presence of ATP, as would be expected if access of Ba2+ to its blocking site is restricted when the channel is shut. This argues that the gate closed by ATP lies above the Ba2+ binding site, close to the selectivity filter. This is in marked contrast to voltage-gated K+ channels, where the helix bundle crossing at the inner mouth of the KATP channel serves as a gate for K+ ions. Interestingly, KATP channels also exhibit an intrinsic fast gating that governs the short closed times within a burst of openings: this gate can be modulated independently of the liganded gate yet also appears to lie within the selectivity filter (Proks et al., 2001). Mutations in SUR1 or Kir6.2 that result in channel closure produce congenital hypoglycaemia of infancy in man, a disease of excessive insulin secretion (Aguilar-Bryan and Bryan, 1999). Conversely, impaired cell metabolism, or mutations in Kir6.2, which lead to enhanced channel activity, result in diabetes. Thus the KATP channel plays a key role in the regulation of insulin secretion in both health and disease. References Aguilar-Bryan and Bryan (1999) Molecular biology of adenosine triphosphate-sensitive potassium channels. Endocrine Reviews 20, 101-135. Ashcroft FM, Gribble FM (1999) ATP-sensitive K+ channels in health and disease. Diabetologia 42, 903-919. Proks P, Capener C, Jones P, Ashcroft FM (2001). Mutations within the P-loop of Kir6.2 modulate the intraburst kinetics of the ATP-sensitive potassium channel. Journal of General Physiology 118, 341-353. Tucker SJ, Gribble FM, Proks P, Trapp S, Ryder TJ, Haug T, Reimann F, Ashcroft FM (1998) Molecular determinants of KATP channel inhibition by ATP. EMBO Journal 17, 3290-3296. University Laboratory of Physiology, Parks Road, Oxford, OX1 3PT, UK

PL-7 P. AERUGINOSA INFECTS MAMMALIAN CELLS VIA MEMBRANE RAFTS E. Gulbins, H. Grassmé Bacteria employ many mechanisms to trigger apoptosis of infected host cells. Those include the secretion and transfer of factors into the host cells that prevent protein synthesis, form pores or activate the endogenous death machinery in the infected cell. Induction of apoptosis in host cells might be a part of the bacterial attack against the host, however, it might be also part of the host defense. For instance apoptosis of macrophages upon infection may protect the bacterium against phagocytosis and digestion and, thus, helps the bacteria to infect the host. If bacteria kill macrophages, which guard protect the mucosa, such as M-cells on Peyer’s Patches, this mechanisms may permit the bacteria to pass the epithelial cell barrier. On the other hand, induction of apoptosis via activation of endogenous death receptors might result in the release of some cytokines, that protect the host. In addition, packaging of internalized bacteria into apoptotic bodies might facilitate phagocytosis by other cells and, thus, also contribute to the host defense. Here, we present data on the infection of mammalian cells with P. aeruginosa. P. aeruginosa infections belong to the most important clinical bacteriological problems, since P. aeruginosa are resistant to many antibiotics and are crucial in life threatening infections of immuno-compromised patients. Most important, almost all patients with cystic fibrosis suffer from pulmonary P. aeruginosa infections, which significantly contribute to the reduced life expectance of these patients. Our studies demonstrate that infection of epithelial cells with P. aeruginosa up-regulates CD95 and CD95 ligand on the cell surface (1). Although many details of this process are unknown, the upregulation of CD95 and CD95 ligand on P. aeruginosa infected cells has been shown to require a type III secretion system and bacteria lacking a functional type III secretion system almost fail to trigger apoptosis in epithelial cells (2). P. aeruginosa triggered binding of CD95 ligand to its receptor finally results in an activation of caspases 8 and 3, release of mitochondrial cytochrome C release, stimulation of JNK (1, 2) and release of apoptosis inducing factor (V. Jendrossek and F. Lang, personal communication). Genetic studies on cells or mice genetically deficient for functional CD95 or CD95 ligand reveal the significance of the CD95/CD95 ligand system for P. aeruginosa triggered cell death apoptosis: Deficiency of CD95 or CD95 ligand in epithelial cells in vivo or ex vivo fibroblasts resulted in a resistance of these cells to P. aeruginosa induced apoptosis (1). Further in vivo studies revealed that apoptosis is a central part of the host defense pulmonary infections of CD95 or CD95 deficient mice resulted in a sepsis and death of the animals, while normal mice completely cleared the infection within a few days. A protective effect of bacterial triggered apoptosis is also demonstrated by the recent finding that infection of the worm Caenorhabditis elegans with S. typhimurium triggers a ced 3 and ced 4 regulated death of worm gonad cells (3). Inhibition of apoptosis in those cells by mutations of ced 3 and ced 4, respectively, hyper-sensitizes C. elegans to the S. typhimurium infection. Additional data show that sphingolipid rich rafts that have been shown to be able to form larger signaling platforms and to cluster receptor molecules, are central in the response to P. aeruginosa and in the regulation of cellular apoptosis as well as bacterial internalization upon infection (Heike Grassmé and Erich Gulbins, unpublished data). Infection of mammalian cells with different P. aeruginosa strains results in the activation of the acid sphingomyelinase and a concomitant release of ceramide. Activated ASM is localized on the outer leaflet of the cell membrane upon fusion of intracellular vesicles with the cell membrane after infection with P. aeruginosa. Extracellularly oriented ceramide re-organizes pre-existing small sphingolipid-enriched rafts to larger platforms, which are required for internalization of P. aeruginosa and for induction of apoptosis in infected cells. Those events are likely mediated by the CD95 receptor and the cystic fibrosis conductance regulator (CFTR) molecule, which cluster in P. aeruginosa triggered ceramide enriched membrane platforms. Deficiency of ASM or ceramide prevents membrane platform formation, and mice lacking the asm gene are highly sensitive to P. aeruginosa. These data indicate the significance of ceramide enriched membrane platforms for P. aeruginosa infections and might be a paradigm for many other host pathogen interactions. The work was supported by DFG grant Gu 335/10-2, the National Institutes of Health Grant CA 21765 and the American Lebanese Syrian Associated Charities (ALSAC) to E.G.

S 162 References: Grassmé H, Kirschnek S, Riethmueller J, Riehle A, von Kürthy G, Lang F, Weller M, Gulbins E. Host defense to Pseudomonas aeruginosa requires CD95/CD95 ligand interaction on epithelial cells. Science 290: 527-530, 2000. Jendrossek V, Grassmé, Mueller I, Lang F, Gulbins E. P. aeruginosa induced apoptosis involves mitochondria and stress activated protein kinases. Infect Immun 69: 2675-2683, 2001. Aballay A, Ausubel FM. Programmed cell death mediated by ced-3 and ced-4 protects Caenorhabditis elegans from Salmonella typhimurium-mediated killing. Proc. Natl. Acad. Sci. USA 98: 2735-2739, 2001. Dept. of Molecular Biology, University of Essen, Hufelandstrasse 55, 45122 Essen PL-8 THE COCHLEA: WHERE MECHANICS AND MOLECULES MEET. J.F. Ashmore It has long been appreciated that the mammalian cochlea is designed to select frequencies from a complex sound with a high degree of fidelity. It is also known the design works with stability and reliability at frequencies that may extend well above 10 kHz. The underlying explanation is that the cochlea functions as a mechanical spectrum analyser that exploits both the macroscopic structure of the organ as well as the design features of the cells from which it is constructed. This requirement places demands on neuronal signalling mechanisms and biological structure within the cochlea. The basic architecture of the cochlea can be described as a fluid filled tube divided along its length by a partition, the basilar membrane. As explored by Bekesy, the stiffness gradient along the partition sets up a frequency map so that differing frequencies are associated with different excitation sites. The cochlear wave motion induced by sound can now be modelled with reasonable accuracy on a small computer [1]. The dynamical response of such models matches the more recent interferometry measurements made in vivo and helpful insights can be obtained from this computational approach. Since a travelling wave would normally be highly damped by the viscosity of the cochlear fluid and of the organ of Corti, it has been appreciated since the early 1980s both theoretically and by measurement that there are mechanisms in the normal cochlea that compensate for dissipative viscous forces [2]. The more recent models have served to emphasise that the macroscopic mechanics is enhanced by a population of sensory hair cells, the outer hair cells within the cochlear partition by a process of regulated positive feedback. At auditory threshold, the power in a sound stimulus is on the order of attowatts. Despite the low power levels, the threshold amplitude of the basilar membrane is about 0.3 nm, about 100 x greater than would be the case if the cochlea were a passive device. At each site along the cochlea, the vibration is signalled by the primary sensory cells of the cochlea, inner hair cells. Despite considerable effort, the molecular identity of the transducer in mammalian cochlear hair cells is unknown. It is associated with apical stereocilia, (for a recent review, see [3]) but it seems increasingly clear that the transducer is likely to be a complex, containing linkage and anchoring proteins. The functional assay for potential candidate molecules is complicated by this complexity. There may even be differences between auditory and vestibular transduction mechanisms with sensory hair cells within the cochlea containing channels modified for ultrafast response times and little adaptation. Equivalent populations of hair cells, supporting cells and their organisation can be relaiably identified between species, providing a basis for extrapolating between mouse models and human hearing. This interplay has been fuelled by current genomic programmes. There are differences of detail, however, which reflect the slightly different hearing ranges found between species. The region of the cochlea that seems to be most susceptible to insult is the basal end where the cells have proved the most difficult to study. Data from several laboratories have emphasised that there are also gradients of morphology and molecular type along the length of the cochlea that may correlate with the selective hearing losses of both genetic and environmental origin. As and example, there are higher levels of potassium channels expressed in basal end outer hair cells [4], a channel that contains a novel KCNQ4 sub-unit as part of its structure. It is also likely that basal end outer hair cells express higher levels of the ATP receptor P2X2 in their apical membrane [5], a receptor whose function in the cochlea is unclear but

which may be correlated with the larger transducer currents found in cells encoding higher frequencies. There have been significant recent developments in understanding how sound is amplified in the cochlea. For each input sound frequency, a cluster of about 300 outer hair cells amplify the basilar membrane vibration, a process named ‘cochlear amplification’ The favoured hypothesis is that outer hair cells are force generating elements that are stimulated by opening of the mechano-electric transducer channels in their apical pole. The cells are also positioned to feed back energy into the basilar membrane vibration. The molecular basis for this force is a dense array of motor molecules in the basolateral membrane of the outer hair cells which are able to extract energy from the electric field across the membrane. This still remains a contentious issue, since the process has to operate at acoustic frequencies in excess of 10 kHz. Although under experimental conditions outer hair cells can be driven at frequencies in excess of 70 kHz, the question is whether the membrane electric field would be large enough at these frequencies since it will be attenuated by the cells’ own low pass electrical characteristics. For this reason it has been suggested that cochlear amplification may arise through the reaction of the hair cell stereocilia during deflection at the apical pole of the cell. Although described in acoustic organs in amphibia [6] and in turtles [7] designed to operate at frequencies below 1 kHz the evidence for this mechanism in mammalian hearing is not direct. As a substrate for outer hair cell force generation, the cell’s basolateral membrane does exhibit properties that might be termed ‘piezoelectric’. Each cell contains in excess of 107 motor molecules that can be detected through their electrical properties, through their freeze fracture profiles and through their ability to confer ‘electromotility’ on patches of membrane. A recent significant advance has been the identification of an underlying protein from a subtractive outer hair cell cDNA library [8]. The protein, named ‘prestin’ exhibits the same biophysical properties as the outer hair cell motor when expressed in heterologous cell systems. The protein contains 12 alpha-helices that form a hydrophobic core embedded in the membrane. It is a member of an anion transporter superfamily SLC26 that includes a number of bicarbonate transporters that are expressed in other epithelial systems. On the basis of this structure, ‘prestin’, thus seems able to undergo changes of area in the plane of the membrane. Its ability to detect changes in the electrical field seems to have an economical explanation as arising from the induced movement of intracellular anions into a deep pore of the protein structure on the cytoplasmic surface of the molecule [9]. It may thus be described as a transporter that exhibits an incomplete transport cycle. Prestin may also share some properties of a sugar transporter. The high density of this transporter in hair cells makes this an attractive system in which to study the interplay between molecular structure, cell function and system properties. As well as its growing interest to sensory physiologists, the cochlea thus has the potential to be a laboratory for testing proposals about how transport molecules work References: [1] Nobili R, Mammano F, & Ashmore JF (1998) T.I.N.S. 21:159-67 [2] Ruggero, M & Robles L (2001) Physiol Rev. 81:1305-52 [3] Gillespie, PG & Walker, RG (2001) Nature 413:194-202 [4] Mammano F, & Ashmore, JF (1996) J Physiol. 496:639-46. [5] Raybould N, & Housley G J (1997) Physiol. 498:717-27 [6] Martin P, Mehta, AD & Hudspeth, AJ (2000) P.N.A.S 97:1202631. [7] Fettiplace R, & Crawford AC (1985) J Physiol. 364:359-79. [8] Zheng J et al. (2000) Nature 405:149-155 [9] Oliver D. et al, (2001) Science 292:2340-3. Department of Physiology, University College London, Gower Street, London WC1E 6BT, UK O 01-1 ICln ION CHANNEL SPLICE VARIANTS IN C. ELEGANS: VOLTAGE DEPENDENCE AND INTERACTION WITH AN OPERON PARTNER PROTEIN M. Ritter, J.G. Danzl, J. Rudzki, E. Scandella, J. Fürst, S. Chwatal, C. Bertocci, J. Provenzano, M. Jakab, M. Paulmichl The ICln protein has been shown to act as ion channel and to be essential for cell volume regulation. The transcript of the ICln gene from C. elegans is alternatively spliced to yield two protein variants, termed IClnN1 and IClnN2. IClnN1 is highly homologues to all ICln

S 163 proteins identified in other species, whereas IClnN2 bears additional 20 AAs (encoded by exon 2a) located on the cytosolic side in close vicinity to the inner mouth of the putative channel pore. This string of AAs carries five consecutive positively charged AAs. Both variants were reconstituted in planar lipid bilayers and currents were measured under voltage clamp conditions using the tip-dip technique. Both ICln variants exhibit identical single channel conductance of 3 pS. The open probabilities (Po) of both variants are identical at negative (pipette vs. bath) holding potentials (0.3-0.4). However, at positive potentials the Po of IClnN2 rapidly drops to zero, whereas the Po of IClnN1 remains unaffected, indicating voltage dependent inactivation of IClnN2. Reconstitution of IClnN1 in the presence of synthetic peptides comprising the entire sequence from exon 2a or of fragments thereof bearing the positively charged AAs confers voltage dependent inactivation to the IClnN1 protein. These observations suggest the existence of a „ball and chain“ like inactivation structure. Reconstitution of IClnN2 and protein Nx, which is transcribed along with ICln from the nematode operon, reveals suppression of voltage dependent inactivation of IClnN2, pointing to a close interaction of the ICln protein and its genomic partner Nx. Institute of Physiology, Fritz Pregl Str. 3, University of A-6020 Innsbruck, Austria

tion induced by ENaC is dependent on Na+. When oocytes coinjected with both channels are incubated in amiloride (10 µM) during expression, promotion of CFTR stimulation by ENaC is drastically reduced. However, short-time application of amiloride during the measurements alone has no effects. On the other hand, CFTR shows no influence on ENaC: Also, varying the intracellular Cl - concentration by clamping to different holding potentials has no effect on ENaC. Amiloride-sensitive ENaC currents are not influenced neither by silent CFTR nor by CFTR activated by cAMP-cocktail. CFTR currents stimulated by cAMP-cocktail are inhibited by DPC (250 µM, Vm = -100 mV) to the same amount in CFTR expressing oocytes and in ENaC/CFTR coexpressing oocytes, demonstrating that ENaC indeed promotes the expression of CFTR and not activation of an endogenous current. Nevertheless, the drastic increase in CFTR current and conductance induced by coexpression of ENaC could be oocyte-specific since in CF airway epithelia the enhanced ENaC activity does obviously not promote the expression of CFTR. Our data clearly show that CFTR does not inhibit ENaC and that ENaC-CFTR interactions are independent of [Cl-]i. However, from the present data we cannot conclude whether the observed ENaC-CFTR interactions are unique for the oocyte expression system or whether those effects could be reproduced using another system for heterologous protein expression. Laboratory of Physiology, KU Leuven, Campus Gasthuisberg, B-3000 Leuven, Belgium

O 01-2 TIME DEPENDENCE OF CFTR EXPRESSION IN XENOPUS LAEVIS OOCYTES A. Segal, N.D. Ullrich, J. Eggermont, W. Van Driessche, W.-M. Weber We recently showed that the cystic fibrosis transmembrane conductance regulator (CFTR) expressed in oocytes of Xenopus laevis is activated by cAMP via different pathways: a direct activation of CFTR channels already present in the plasma membrane and exocytotic delivery of preformed CFTR molecules from intracellular pools to the oocyte membrane, mediated by special isoforms of protein kinase A (PKA). Here we report that these different pathways of CFTR activation are strongly dependent on expression time. We measured changes in membrane current (Im), conductance (Gm) and capacitance (Cm), the latter being a direct measurement of the membrane surface area, following exposure to a ‘cocktail’ consisting of chlorophenylthio-cAMP (100 µM) and IBMX (1 mM). However, this cAMP-cocktail has nearly no effect on day 1 after mRNA injection, only tiny increases in Im and Gm can sometimes be observed while Cm always remains unchanged. On day 2 after injection, Im and Gm are always significantly increasing, yet Cm still remains uninfluenced, demonstrating that only CFTR channels in the plasma membrane are activated without detectable exocytotic insertion of preformed CFTR. However, after 3 days both activation pathways are observed and Im, Gm and Cm are increased simultaneously by cAMP-cocktail, indicating that exocytosis of CFTR to the plasma membrane occurs only after 3 days. From these data we conclude that after mRNA translation, CFTR molecules are first directly inserted into the oocyte plasma membrane until a certain threshold is reached. These CFTR channels are directly activated by cAMP-dependent phosphorylation, thereby increasing Im and Gm without effect on Cm. Then intracellular pools are filled with CFTR that, upon stimulation by cAMPdependent PKA, can be delivered via exocytosis to the oocyte membrane, leading to increased plasma membrane surface area mirrored in increases in Cm. Our data do not allow to conclude whether these expression patterns are ubiquitous or specific for the Xenopus oocyte expression system. Laboratory of Physiology, KU Leuven, Campus Gasthuisberg, B-3000 Leuven, Belgium

O 01-4 FUNCTIONAL ALTERATIONS OF HUMAN CLC-2 CL CHANNELS IN IDIOPATHIC EPILEPSY M. Warnstedt1, A. Alekov 2, K. Haug3, Ch. Fahlke 1, H. Lerche2, A. Heils3 Idiopathic generalized epilepsy (IGE) is a common form of human epilepsy characterized by generalized seizures without any detectable brain lesion. Recently, in two IGE families co-segregation of the disease status with two distinct mutations in CLCN2 was demonstrated: a missense mutation predicting the exchange of G to E at codon 715, and a single nucleotide insertion in bp-position 597 (597insG) causing a premature stop codon (M200fsX231) (Heils et al., Swiss/Austrian/ German ILAE meeting 2001). To get further insights into the pathophysiology of IGE and to identify the role of ClC-2 in the regulation of excitability of the human brain, we functionally characterized WT and mutant human ClC-2 in tsA201 cells using the patch-clamp technique. WT hClC-2 is closed at positive potentials and activates upon membrane hyperpolarization. The voltage dependence of activation changes with intracellular [Cl-] in a way that ClC-2 only opens negative to the Cl - reversal potential (V ½ = –80 ± 8 mV (n=4) for [Cl-]int=124 mM, -131 ± 11 mV for [Cl-]int=34 mM, and -154 ± 12 mV for [Cl -] int=4 mM, n=4). This gating behavior makes ClC-2 a sole efflux pathway for anions and allows ClC-2 to contribute to the establishment and maintenance of the low internal [Cl-] that is essential for the inhibitory action of GABA-ergic synapses. M200fsX231 abolished the formation of functional channels with a dominant negative effect on the WT and G715E altered the effect of [Cl -]int on ClC-2 gating. At [Cl-]int of 34 mM, G715E channels activated at less negative potentials (V ½ = -99 ± 5 mV, n=5) permitting mutant channels to conduct Cl- into the cell during action potentials causing a depolarization of the anion reversal potential. Hence, expression of the two disease-causing mutations may depolarize the anion reversal potential by distinct mechanisms thereby affecting inhibitory inputs in certain neurons and thus cause epileptic seizures. 1 Institut für Physiologie, RWTH Aachen; 2Abteilung Angewandte Physiologie, Universität Ulm; 3Institut für Humangenetik, Universität Bonn

O 01-3 O 01-5 ENAC PROMOTES EXPRESSION OF CFTR WHILE CFTR DOES NOT INHIBIT ENAC A. Segal, N.D. Ullrich, J. Eggermont, W. Van Driessche, W.-M. Weber We coexpressed the cystic fibrosis transmembrane conductance regulator (CFTR) and the amiloride-sensitive epithelial Na+ channel (ENaC) in oocytes of Xenopus laevis. In oocytes expressing ENaC and CFTR activation of CFTR with a cAMP-cocktail of chlorophenylthio-cAMP (100 µM) and IBMX (1 mM) is more than 10 times larger than in oocytes solely injected with CFTR as measured by increases in whole cell current (Im) and conductance (Gm). The increased CFTR stimula-

THE MALARIA PARASITE PLASMODIUM FALCIPARUM UPREGULATES A STRETCH-ACTIVATED CFTR-LIKE CHANNEL S.L.Y. Thomas, S. Egée, F. Lapaix, G. Decherf, H.M. Staines, J.C. Ellory, C. Doerig The patch-clamp technique was used to investigate the membrane conductance of human red blood cells (RBCs) infected by Plasmodium falciparum. The impermeable monovalent cation N-methyl-Dglucamine (NMDG) was used in all pipette and bathing solutions to prevent cationic currents. In control uninfected RBCs, the mean whole-

S 164 cell (± SEM) conductance (Gm) was 44 ± 1 pS (n = 32). With PKA (100 nM) and ATP (10 mM) in the pipette, Gm was 1700 ± 40 pS (n = 58) between –100 mV and Er. This increase did not occur in the presence of 5-Nitro-2-(3-phenylpropylamino)–benzoate (NPPB, 100 µM), glibenclamide (100 µM) and diphenylamine-2-carboxylic acid (DPC, 1 mM) in the bath solution. The permeability sequence was I- > Br- ≥ Cl> F-. In infected cells, Gm was 5300 ± 140 pS (n = 24) between –100 mV and Er. The permeability sequence was I- > NO3- > Br≥ Cl- > F- > gluconate. NPPB, niflumic acid and glibenclamide reduced Gm with IC50 values of 0.8 ± 0.3 µM (n = 4), 2.5 ± 0.3 µM (n = 4), 35.0 ± 8.7 µM (n = 3), respectively. DPC induced 56 ± 11 % (n = 3) inhibition at 1 mM. DIDS (4,4'-diisothiocyanostilbene-2,2'-disulphonic acid, 100 µM) produced no significant inhibition. Furosemide (200 µM) reduced Gm by 45 ± 1 % (n = 3). In the cell-attached configuration, no channel activity was detected in uninfected cells (n > 50), but spontaneous channel activity, showing slight inward rectification, was measured in all (n > 25) infected RBCs. The average conductance was 18.0 ± 1.1 pS (n =25 ) between –100 mV and –20 mV. Er was around +12 mV. This channel exhibited voltage-dependent gating, with increasing open probability (Po) between +50 mV and -100 mV. Together with the voltage-dependent opentime duration and number of active channels, Po-dependency on Vm provides the explanation for the inward rectification of the whole-cell I-V curves observed in malaria-infected RBCs. After excision, a cystic fibrosis transmembrane regulator (CFTR)-like channel was observed in membrane patches from both uninfected (n > 100) and infected (n > 50) RBCs. This channel can be activated in normal RBCs by intracellular application of PKA (12.3 ± 0.5 pS, n = 8 ) in the presence of ATP, or by membrane deformation (13.4 ± 0.5 pS, n = 10 ). In infected cells, the channel was always spontaneously active after excision (15.7 ± 0.8 pS, n = 7) and exhibited voltage-dependent gating consistent with the observations in the cell-attached configuration. These findings are of obvious interest to the field of cystic fibrosis biology and give a new insight into the way in which the parasite modifies the host membrane to survive intracellularly.

O 02-1 THE ANTI-ATHEROSCLEROTIC CYTOKINE IL-10 INDUCES ENDOTHELIAL NITRIC OXIDE SYNTHASE EXPRESSION IN HUMAN ENDOTHELIAL CELLS M. Cattaruzza, W. Stodowski, M. Stojakovic, R. Krzesz , M. Hecker Background: lnterleukin-10 (IL-10) is a potent anti-inflammatory cytokine in T helper 1 (T H1) cell-dependent chronic inflammatory responses thought to act mainly by opposing the effects of the prototypic TH1 cytokine interleukin-12 (IL-12). In addition, IL-10 has been shown to limit the progression of atherosclerosis, presumably by influencing endothelial cell function. Methods: To analyse by which mechanism IL-10 causes this increase in endothelial NO release, we cultured human umbilical cord endothelial cells (HUVEC) and exposed them to IL-10 (1-30 ng/ml). Transcriptional regulation, expression and function of the IL-10 receptor and of the constitutive NO synthase (eNOS) were analysed by standard techniques. Results: Here we demonstrate that various pro-inflammatory cytokines induce expression of the IL-10 receptor in human cultured endothelial cells. Subsequent exposure to IL-10 resulted in a transient but prominent (5fold) increase in eNOS expression. Both gel shift analyses and decoy oligonucleotide experiments suggested that this effect of IL-10 was mediated through activation of the transcription factor STAT-3. One functional consequence of IL-10 up-regulation of eNOS abundance was the attenuation of CD154-induced IL-12 p40 expression in the cultured endothelial cells. Moreover, CD-154-induced IL-12 p40 expression in these cells was enhanced following blockade of eNOS activity but attenuated in the presence of exogenous nitric oxide. Conclusions: Thus, enhanced eNOS expression under pro-inflammatory conditions might be one mechanism by which IL-10 exerts its antiinflammatory and anti-atherosclerotic effects. Department of Cardiovascular Physiology, Heart Center, University of Göttingen, Humboldtallee 23, 37073 Göttingen, Germany

CNRS, UPR 9042, Station Biologique, B.P.74, 29682 ROSCOFF cedex, France O 01-6 O 02-2

ANION AND ORGANIC OSMOLYTE CHANNELS OF HUMAN ERYTHROCYTES INFECTED WITH PLASMODIUM FALCIPARUM. S.M. Huber, C. Duranton, A.-C. Uhlemann, P. Kremsner, F. Lang Intraerythrocytic survival of the malaria parasite Plasmodium falciparum requires that host cells supply nutrients and dispose of waste products. This solute transport is accomplished by an infectioninduced new permeability pathway (NPP) in the erythrocyte membrane. The NPP has been demonstrated to be anion-selective and permeable for organic solutes (e.g., sorbitol). Consequently, infected erythrocytes but not control cells hemolyze in isosmotic sorbitol solution. In patch clamp experiments, infected cells have high activities of an inwardly and an outwardly rectifying anion conductance which can be induced in non-infected cells by membrane oxidation. To explore which of the two anion conductances is generating the organic osmolyte permeability, the sensitivities of both conductances to NPPB, DIDS, furosemide and glybenclamide were determined by whole-cell recording and compared to those of hemolysis in isosmotic sorbitol. These blockers inhibited the inwardly rectifying current with IC50s of >1 µM (NPPB), >10µM (furosemide), >100 µM (DIDS), and >1 mM (glybenclamide), respectively. The outwardly rectifying current, in contrast, was blocked with IC 50s of 100 nM (NPPB) and 1-10 µM (DIDS, glybenclamide and furosemide), respectively, which was similar to the blocking effect on hemolysis in isosmotic sorbitol (IC50s of 1 µM for NPPB and 10-30 µM for DIDS, glybenclamide and furosemide), suggesting that the outwardly rectifying conductance but not the inwardly rectifyer participated in the generation of the sorbitol pathway. Moreover, partial replacement of NaCl (50 mM) in the bath solution by sorbitol, mannitol or sucose (100 mM) differentially changed the whole-cell outwardly rectifying conductance which might hint to a competition of carbohydrates and anions within the channel pore of the outwardly rectifying channel. In conclusion, the infection-induced outwardly-rectifying anion conductance of human erythrocytes most probably is permeable for organic osmolytes.

NOREPINEPHRINE-INDUCED EXPRESSION OF PROINFLAMMATORY CYTOKINES IN ISOLATED BIVENTRICULAR WORKING RAT HEARTS W. Briest, C. Elsner, J. Hemker, G. Müller-Strahl, H.-G. Zimmer The norepinephrine (NE)-induced hypertrophy of the left ventricle (LV) in the rat is associated with increased interleukin (IL)-6 and ILlβ expression. In the present study, the isolated biventricular working rat heart (IBWH) was used to examine whether NE may directly induce cytokine mRNA expression in a preparation devoid of circulating hormonal and humoral factors. Representative hemodynamic parameters and the expression of various cytokines of the IBWH were compared with the respective in vivo results. Systolic pressure (SP) of the right ventricle (RVSP) was higher in the IBWH than in the intact anesthetized rat (42.9 ± 1.89 vs. 33.3 ± 0.86). However, heart rate (HR), LVSP and the increase in the rate of pressure development of both ventricles (LV- and RV dP/dtmax) were lower. After NE infusion (30 nM), SP and dP/dtmax were increased in the LV and RV, by 30% and 90%, respectively. In vivo, the ventricles showed a different response to NE (0.1 mg/kg·h): LVSP increased by 19%, RVSP was doubled, LV dP/dtmax was tripled, and RV dP/dtmax was elevated by 85%. The analysis of cytokine mRNA expression with the RNase protection assay revealed that in vivo EL-6 and IL-lβ were increased 80- and 12-fold in the LV and RV between 4 and 12 h, respectively, while there was weak expression under control conditions. In the IBWH IL-lα, IL-lβ, IL-6 and TNFα were increased already during control perfusion. The increase of these stress-activated cytokines indicates that the isolation and perfusion procedure may exert a stress on the heart. NE induced an additional time-dependent increase of IL-lβ and IL-6 mRNA in the LV and of IL-lα, IL-lβ and IL-6 in the RV after 1 h of infusion. Thus, NE has a direct effect on the cardiac IL-6 expression, which occurred earlier in the in vitro preparation than in the rat heart in vivo. (Supported by DFG Grant Zi 199/10-3).

Department of Physiology I, University of Tübingen, Gmelinstr. 5, 72072 Tübingen.

Carl-Ludwig-Institut für Physiologie, Universität Leipzig, Liebigstr. 27, 04103 Leipzig

S 165 O 02-3

O 02-5

ENHANCED PLATELET RECRUITMENT BY NADPH-OXIDASE DEPENDENT PLATELET SUPEROXIDE RELEASE F. Krötz, H.Y. Sohn, T. Gloe, T. Riexinger, M. Keller, S. Zahler, U. Pohl Platelets have been suggested to release superoxide (O 2- ) during activation. However, the source of platelet-derived O2- and its effects on platelet function are unclear. We studied the enzymatic source, the effect of agonists on its release and the role of platelet-derived O2- in thrombus formation. Collagen (6µg/mL), but not ADP (1-50µM) or thrombin (0,1-1 U/ml), increased O2 - formation in washed human platelets (by 59.4%, n=24, p<0.01). Inhibition of NAD(P)H-oxidase using diphenyleneiodoniumchloride (DPI, 100µM) or the specific peptide-antagonist gp91ds-tat (100µM) abolished O 2 - production, whereas a scrambled peptide did not. Direct evidence of a platelet NAD(P)H-oxidase was confirmed by immunoblotting experiments showing expression of its subunits p47phox and p67phox (n=5). Collagen, in the presence of increasing doses of NADH and, to a lesser degree of NADPH, added to platelet lysates increased O2- formation (n=3 each). Activation of PKC (PMA, 1µM) also induced O2- release (by 100.3%, n=21, p<0.01). Exogenous O2- induced aggregation in the presence of subthreshold concentrations of ADP (1µ M, n=9). Platelet-derived O2also significantly enhanced platelet recruitment which was dependent on release from a preformed thrombus, as recruitment following collagenstimulation was attenuated by superoxide dismutase (-19%, p<0.01, n=10) or DPI (-16%, p<0.01, n=10) and was completely blocked by the ectonucleotidase apyrase (lU/mL, n=9). These results indicate that NAD(P)H-oxidase dependent O2- release increases stimulatory effects of ADP on platelets thereby increasing platelet recruitment. We suggest that O2- dependent inhibition of a collagen-activated ectonucleotidase might be involved.

DELAYED NEUTROPHIL APOPTOSIS IN CD18 NULL MICE: A NOVEL MECHANISM FOR THE CONTROL OF NEUTROPHIL HOMOEOSTASIS B. Walzog, P. Weinmann The control of neutrophil turnover in the circulation is a key event in homoeostasis and inflammation. Using CD18-deficient (CD18-/-) mice which show a massive neutrophilia under non-inflammatory conditions, we provide evidence that β2 integrins (CD11/CD18) contribute to the control of apoptosis of peripheral neutrophils: When compared to neutrophils obtained from the circulation of wild-type animals (CD18+/ + ), neutrophils form CD18-/- mice revealed a profound delay of apoptosis as measured by detection of CD16 expression on the cell surface, by analysis of DNA content and nuclear morphology as well as by detection of DNA degradation. In contrast, no difference in apoptosis was observed in neutrophils derived from the bone marrow revealing that the delay of neutrophil apoptosis in CD18 -/- mice occurs in the peripheral blood. Plasma experiments with isolated neutrophils confirmed that the observed delay of apoptosis was not due to the presence of soluble mediators in the plasma. The analysis of the molecular mechanism underlying the observed effect revealed an up-regulation of the anti-apoptotic gene bcl-xl and a down-regulation of the proapoptotic gene bax-alpha in CD18-/- neutrophils compared to CD18+/ + neutrophils suggesting that the shift of balance between pro- and anti-apoptotic members of the bcl-2 family may be responsible for the delayed neutrophil apoptosis in CD18-/- mice. This was confirmed by the finding that treatment of CD18+/+ neutrophils with bax-alpha antisense oligonucleotides induced a delay of apoptosis and prolonged neutrophil survival. Altogether, these results suggest that the delayed apoptosis in CD18 -/- neutrophils causes an alteration of neutrophil homoeostasis which may be responsible for the massive increase of neutrophil counts in the circulation of these animals. Thus, the β 2 integrins seem to be critically involved in the control of neutrophil turnover in the circulation. Supported by SFB 366/C3.

Institute for Physiology, LMU, Schillerstr, 44, 80335 München

Dept. of Physiology, Freie Universität, Arnimallee 22, D-14195 Berlin, Germany O 02-4 O 02-6 WITHDRAWAL OF STATINS INCREASES REDOX-MEDIATED GENE EXPRESSION IN VASCULAR CELLS R.P. Brandes, T. Ha, S. Beer, C. Vecchione, R. Busse Unrelated to their cholesterol-lowering properties, HMG-CoA-reductase inhibitors (statins) have potent anti-arteriosclerotic properties. We determined the effects of statins and statin withdrawal on gene expression and signalling in cultured aortic vascular smooth muscle cells (SMC) and on endothelium-dependent vasodilation in mouse aortic rings. Exposure of SMC to cerivastatin (0.01 - 1 µM), lovastatin (110 µM) and atorvastatin (1-10 µM) prevented the agonist-induced translocation of Rac to the plasma membrane and induced a time- and dose-dependent attenuation of agonist (thrombin, angiotensin II, PDGF)-induced oxygen-derived free radical (ROS) generation. In addition, agonist-induced activation of Erk 1/2 and p38 MAP kinases as well as induction of tissue factor (TF) and vascular endothelial growth factor (VEGF) mRNA were attenuated by statins. Withdrawal of statins induced a rebound phenomenon as repetitive washing of the cells lead to an increase in constitutive and agonist-induced TF expression to levels higher than those observed in untreated cells. Moreover, following statin withdrawal, agonist-induced MAP kinase activation and ROS generation were enhanced. This effect could be blocked by the NADPH oxidase inhibitors diphenylene iodonium and apocynin, whereas exogenous administration of geranylgeranyl-pyrophosphate enhanced it. In vivo treatment with atorvastatin (10 mg/kg, 10 days) or cerivastatin (2 mg/kg, 10 days) also improved acetylcholine-induced, endotheliumdependent, but not nitroglycerin-induced, endothelium-independent relaxation of mouse aortic rings. Following statin withdrawal, acetylcholine-induced relaxations deteriorated. Seven days after withdrawal, relaxations were significantly attenuated in comparison to relaxations observed in aortae of placebo-treated animals. These observations suggest that enhanced ROS formation after statin withdrawal increases redox-sensitive signalling, gene expression and leads to an attenuation of endothelium-dependent relaxation. Institut für Kardiovaskuläre Physiologie, Klinikum der J.W. GoetheUniversität, Theodor-Stern-Kai 7, 60596 Frankfurt am Main, Germany

ANGIOTENSIN II REGULATES NAD(P)H OXIDASE EXPRESSION IN A DOSE-DEPENDENT MANNER IN ENDOTHELIAL CELLS: ANTIOXIDATIVE POTENTIAL OF AT 1 RECEPTOR BLOCKADE IN PATIENTS WITH CORONARY ARTERY DISEASE U. Rueckschloss, J. Holtz, H. Morawietz Objective of study Angiotensin II (Ang II) can increase proatherosclerotic oxidative stress by augmented superoxide anion formation in endothelial cells. A NAD(P)H oxidase has been identified as a major endothelial source of superoxide anions. However, the molecular mechanism underlying this increased NAD(P)H oxidase activity in response to Ang II is not well understood. Methods and Results Comparing relative expression of NAD(P)H oxidase subunits by multistandard-assisted competitive RT-PCR we identified gp91phox as limiting subunit in primary cultures of human umbilical vein endothelial cells (HUVEC). Ang II induces gp91phox mRNA expression 2.6fold in a dose of 100 nmol/L (max. at 7 h). This induction is AT 1 receptor-mediated. In contrast, a higher dose of Ang II (1 µmol/L) induces gp91phox to a significantly lower extent. This partial inhibition can be blocked with AT2 receptor antagonist PD123319. Dose-dependent regulation of gp91phox mRNA expression is paralleled by a similar biphasic Ang II-mediated increase and partial inhibition of superoxide formation in HUVEC. In further studies, we measured gp91phox expression in internal mammary artery biopsies from patients with or without AT 1 receptor blocker or ACE inhibitor therapy before coronary bypass surgery. Treatment of patients with AT 1 receptor blockers resulted in downregulation of gp91 phox expression in internal mammary arteries. In contrast, ACE inhibitor therapy in prescribed dosages had no significant effect on gp91 phox expression. Conclusions These data support a dose-dependent induction of proatherosclerotic oxidative stress in human endothelial cells in response to Ang II. AT1 receptor blockade has an antioxidative potential by reduction of oxidative stress in the vessel wall. Institute of Pathophysiology, Martin Luther University HalleWittenberg, Magdeburger Str. 18, D-06097 Halle, Germany

S 166 O 03-1 SHARP WAVE-RIPPLE COMPLEXES IN MOUSE HIPPOCAMPAL SLICES IN VITRO N. Maier, H. Siegmund, A. Draguhn The rodent hippocampus exhibits characteristic patterns of neural network oscillations at frequencies of ~5-10 Hz (Θ), ~30-70 Hz (γ) and at ~200 Hz (“ripples”). The latter oscillations occur as brief bursts of voltage deflections in the extracellular field on top of slower positive or positive-negative waves (“sharp waves”) which originate in CA3 and travelling through CA1 to the entorhinal cortex. Based on in vivo recordings, sharp wave-ripple complexes have been suggested to serve a function in memory consolidation during rest and sleep. However, the cellular and molecular mechanisms underlying this synchronised activity are poorly understood. We have performed extra- and intracellular recordings in hippocampal slices from adult mice and observed spontaneous network discharges reminiscent of sharp waveripple (SW-R) complexes in vivo. Ripples were most pronounced in the pyramidal cell layers of CA1 and CA3 and propagated from CA3 through CA1 to the subiculum. SW-R complexes were blocked by bicuculline, CNQX and by the gap junction blocking agents carbenoxolone and quinine. In the presence of GABA A receptor antagonists, epileptiform high-frequency discharges at ~200 Hz were still present in the pyramidal cell layers and could be suppressed by uncoupling agents. The intracellular correlate of field SW-R complexes in CA1 pyramidal cells revealed a strong hyperpolarisation with superimposed phasic potential deflections at ~200 Hz. Our data show that network activity reminiscent of sharp wave-ripple complexes can be recorded in rodent hippocampal slices in vitro and is dependent on intact glutamatergic excitation, GABAergic inhibition and electrical communication via gap junctions.

characterized, since most intracellular recordings have been limited to anaesthetized animals. We investigated postsynaptic responses of neurons in the rat somatosensory (barrel-) cortex while the animals performed fast rhythmic whisker movements (whisking). Whole-cell recordings were obtained in awake, head-fixed animals by standard whole-cell recording techniques. Whisker movements were sampled at a rate of 150 Hz by an infrared stroboscope, coupled to an infrared sensitive video camera and by labeling individual whiskers with reflex foil. We found that in a majority of cells the ongoing subthreshold activity was suppressed during whisking, i.e. the membrane potential traces became flat during whisking. In a smaller subset of S1 neurons (20-30%) we observed that during whisking episodes the membrane potential exhibited periodic modulations of up to 15 mV amplitude that followed the phase and frequency of the whisker movement. Relative to the whisker movement phase and amplitude of the whisking associated membrane potential modulations differed from cell to cell. Cells whose membrane potential followed the whisker movement often showed depolarizing responses before an upcoming whisker movement suggesting that these cells receive motor-related inputs. AP rates during whisking episodes were low (< 1 Hz). Cells strongly modulated by the whisker movement are likely to discharge preferentially at certain points in the whisker cycle. This could provide the animal with clues about the spatial position of obstacles. Max-Planck-Institut für medizinische Forschung, Jahnstr. 29, 69120 Heidelberg, Germany

O 03-4

Institute of Physiology, Bühlplatz 5, CH – 3012 Bern, Switzerland

FUNCTIONAL MICROCIRCUITRY OF THE DEVELOPING CEREBRAL CORTEX: EVIDENCE FOR A SYNAPTIC NETWORK IN THE SUBPLATE I.L. Hanganu, W. Kilb, H.J. Luhmann The subplate, a transient layer of neurons located directly under the cortical plate, plays an important role in the formation of thalamocortical connections during early development. In order to identify and characterize the presence and origin of synaptic activity onto subplate neurons (SPn), we performed whole-cell recordings from visually identified and biocytin labeled SPn in somatosensory cortical slices from postnatal day (P) P0-P3 rats. SPn show three distinct classes of spontaneous postsynaptic currents (sPSCs): i) Fast sPSCs mediated by AMPA receptors. ii) Slow sPSCs mediated by NMDA receptors. iii) Chloride-driven sPSCs mediated by GABAA receptors. Local electrical stimulation of the thalamocortical afferents elicited in 68% of the SPn monosynaptic, unitary and CNQX-sensitive PSCs with fast kinetic and unimodal latency distribution. These PSCs were mediated by AMPA receptors, since they were prolonged by cyclothiazide and blocked by GYKI 52466. Stimulation of the cortical plate evoked in 72% of the SPn monosynaptic AMPA receptor-mediated PSCs and in 40% of the investigated cells polysynaptic responses, which were reduced in amplitude by the GABAA antagonist gabazine and blocked by the NMDA receptor antagonist CPP. Electrical stimulation of the subplate elicited in 67% of the SPn dual-component monosynaptic PSCs mediated by AMPA and NMDA receptors. Monosynaptic GABAergic responses could be observed in 68% of the SPn following stimulation of the subplate. All stimulus-evoked PSCs recorded in SPn required action potential-dependent transmitter release. In gramicidin-perforated patch recordings, bath application of GABA caused a membrane depolarization from -70 to -40 mV and elicited action potentials. These results demonstrate that SPn receive distinct functional synaptic inputs arising from the thalamus, cortical plate and subplate, indicating that SPn are capable to integrate and process information from cortical and subcortical regions via activation of specific postsynaptic receptors. Supported by DFG grants Lu375/3, SFB 194/B4 and neuroscience graduate program.

O 03-3

Neurophysiology, Heinrich-Heine-University, D-40001 Düsseldorf, Germany

Johannes-Müller-Institut für Physiologie der Charité, Tucholskystr. 2, 10117 Berlin O 03-2 GABAA RECEPTORS AND Ih NARROW THE TIME WINDOW FOR COINCIDENCE DETECTION IN LAYER V PYRAMIDAL CELLS OF THE RAT SOMATOSENSORY CORTEX T. Berger, H.R. Lüscher The ability of a neuron to generate action potentials in response to different synaptic inputs within a narrow time window is referred to as coincidence detection. An essential prerequisite for coincidence detection are the spatial and temporal summation properties of the neuron. In order to study the impact of different conductances on the summation properties of layer V pyramidal cells of the rat somatosensory cortex, we patch-clamped their somata and dendrites and induced postsynaptic potentials using extracellular stimulation. The activation of GABA A receptors narrowed the time window for coincidence detection and prevented burst firing in response to synaptic input. In addition, uneven weights of the GABAergic components in both inputs could lead to an inhomogeneous shunting and consequently to a time shift of the peak of the coincidence detection window. Ih showed an even more pronounced impact because coincidence detection was practically abolished within a time window of 30 ms after blockade of this conductance. Both conductances curtailed the decaying phase of the postsynaptic potential, albeit by different mechanisms, and thus impaired the summation of incoming inputs. In addition, block of GABAA receptors or Ih could switch the origin of spike generation and reverse its direction of propagation. Coincidence detection in layer V pyramidal cells is therefore influenced by at least two mechanisms: i. modulation of I h as a function of the postsynaptic membrane potential, ii. activation of the presynaptic inhibitory network. Supported by SNF 3100-061335.00.

WHOLE-CELL RECORDINGS IN THE BARREL CORTEX OF AWAKE RATS DURING WHISKING M. Brecht, B. Sakmann Acquisition of sensory information is an active process often involving exploratory movements of eyes, ears, fingers or feelers. However, synaptic responses evoked by exploratory activity have not been

O 03-5 FEEDBACK CONTROL OF REPRESENTATIONS IN PRIMARY VISUAL CORTEX R.A.W. Galuske, K.E. Schmidt, R. Goebel, S.G. Lomber, B.R. Payne

S 167 In V1, neurons with similar response preferences are grouped into domains forming continuous maps of different stimulus features. Stimulus orientation and direction of movement are prominent examples of such maps. So far, the elaboration of orientation and direction preference has mainly been attributed to forward and lateral connectivity without substantial consideration of feedback from higher visual areas that also densely innervate V1. In this study, we examined the functional impact of feedback arising from the posterior parts of middle suprasylvian (pMS) sulcal cortex. To this end, reversible cooling deactivation of pMS cortex was combined with the recording of orientation and direction preference maps in V1 using optical imaging of intrinsic signals. Feedback deactivation severely degraded orientation and direction maps; both the absolute level of activity as well as the specificity of responses were markedly decreased in the absence of pMS feedback signals. Even so, the basic structure of orientation maps was still preserved although the global layout of the direction maps was lost during inactivation of pMS cortex. Electrophysiological recordings revealed that the observed changes in the maps are based on the silencing of highly direction selective neurons during the feedback removal. These data reveal that feedback projections from higher order visual areas contribute substantially to the elaboration of the basic response properties and feature maps in V1. They suggest that even the earliest steps in cortical visual processing rely strongly on interactions between components of a network of several cortical areas positioned at multiple hierachical levels. Supported by MPG, DAAD, NSF, NINDS. Max-Planck-Institute for Brain Research, Deutschordenstrasse 46, 60528 Frankfurt am Main, Germany O 03-6 SPATIO-TEMPORAL DISTRIBUTION OF NEURONAL ACTIVITY: GEOMETRIC PATTERN ANALYSIS OF OPTICAL IMAGING DATA R. Köhling1, J. Reinel2 J. Vahrenhold2, K. Hinrichs2, E.-J. Speckmann1 Optical imaging of neuronal network activity yields information of spatial dynamics which generally is analyzed visually. The transient appearance of activity patterns are difficult to gauge in a quantifiable manner, or may even altogether escape detection. Here, we employ geometric shape matching using Fréchet distances or straight skeletons to search for pre-selected patterns in optical imaging data with adjustable degrees of tolerance. Data were sampled from fluorescence changes of a voltage-sensitive dye incorporated into cellular membranes and recorded with a 464-photodiode array in a neuronal network in vitro. Neuronal activity prompting fluorescence fluctuations consisted of spontaneous epileptiform discharges in neocortical slices from patients undergoing epilepsy surgery. The experiments show that (a) spatial activity patterns can be detected in optical imaging data using both algorithms, (b) shapes such as “mini-foci” appear in close correlation to bioelectric discharges monitored by field potentials in a reproducible manner, (c) Fréchet distances yield more conservative matches regarding rectangular, and less conservative hits with respect to radially symmetric shapes than the straight skeleton approach and (d) tolerances of 0.030.1 are suited to detect faithful images of pre-selected shapes, whereas values >0.8 will report matches with any polygonal pattern. In conclusion, the methods reported here are able to detect and analyze spatial, geometric dynamics in optical imaging data. Institut für Physiologie1 and Institut für Informatik2, Westfälische Wilhelms Universität Münster, Germany O 04-1 HIF1 PROLYL HYDROXYLASES ARE OXYGEN-, IRON-, AND ALPHA-KETOGLUTARATE-DEPENDENT DIOXYGENASES E. Metzen 1, J. Gleadle 2 , A.C.R. Epstein 2, T. Hellwig-Buergel l, W. Jelkmann1, P.H. Maxwell2, C.W. Pugh2, P.J. Ratcliffe2 Hypoxia-Inducible-Factor-1 (HIF1) is a heterodimeric transcription factor which is critical for the adaptation of cells as well as of organisms to hypoxia. The activity of the HIF1 complex is regulated by proteolytic destruction of the α subunit in normoxia. The degradation signal is hydroxylation of amino acid residue Pro564. The hydroxylated protein binds the von-Hippel-Lindau tumor suppressor protein (VHL) which targets HIF1α for proteasomal degradation. In hypoxia, however, HIF1α accumulates and induces expression of a large number of hypoxia-

responsive genes. We have found that the components of this regulation process are also present in the nematode worm C. elegans. The C. elegans protein EGL-9 could be shown to modify C. elegans HIF by prolyl hydroxylation. The human homologs of EGL-9, termed Prolyl Hydroxylase Domain containing protein 1, 2, and 3 (PHD1, PHD2, and PHD3) were cloned, expressed by in-vitro-transcription-translation and used in vitro to modify Gal-HIF fusion proteins. PHD1, 2, and 3 were found to be dependent on the concentration of ambient oxygen and sensitive to desferrioxamine or cobalt application while a mutant PHD1 and a PHD1 splice variant were ineffective. It is well established that HIF is induced in vivo by iron chelation and cobalt treatment, thus these characteristics of the PHD enzymes in vitro parallel features of the oxygen sensing process in vivo. α-ketoglutarate is required as a cosubstrate for dioxygenase function. Indeed organic α-ketoglutarate analogs inhibited enzyme function in our experiments. Transient cotransfection of cells with PHD1 and a HIF responsive luciferase reporter gene resulted in a threefold reduction of luciferase activity in hypoxia. Our data suggest that C. elegans EGL-9 and its mammalian homologs PHD1, PHD2, and PHD3 regulate HIF1α by prolyl hydroxylation and serve as oxygen sensors in vivo. 1

Institute of Physiology, Medical University of Luebeck, Ratzeburger Allee 160, D-23538 Luebeck, Germany; 2The Henry Wellcome Building of Genomic Medicine, Roosevelt Drive, Oxford OX3 7BN, United Kingdom O 04-2 ROLE OF RETINOIC ACID FOR HIF-1 DEPENDENT GENE EXPRESSION M. Cremer, Y. Ma, S. Frede, J. Fandrey In low oxygen conditions hypoxia inducible factor-1 (HIF-1) enhances the expression of various genes such as vascular endothelial growth factor (VEGF) or erythropoietin (EPO). Two phenomenona of the response to retinoic acid (RA) are known: EPO gene expression is upregulated by RA under hypoxia in cell culture models and RA is able to differentiate the human leukaemia cell line HL 60 into macrophage like cells. We wanted to investigate how this treatment affects the HIF-1 response to hypoxia. Methods: 1. HL 60 cells were grown with or without the addition of all-trans retinoic acid (1µM, five days) and then incubated under normoxic or hypoxic conditions. 2. HL 60 cells were treated for 6h with RA and simultaneously incubated under normoxic or hypoxic conditions. The VEGF secretion in the culture supernatant was measured using an ELISA-system and the alpha subunit of the HIF-1 protein was detected by Western blot analysis. Results: The differentiation of the HL60 cells lead to a decreased accumulation of HIF-1 under hypoxia compared to blastic cells, although the amount of the VEGF secretion remained unchanged. Short term treatment of HL 60 with RA did not influence VEGF secretion and left HIF accumulation nearly unchanged. Conclusions: The release of VEGF from leukaemia cell lines is independent of the oxygen tension and the HIF 1 accumulation. Furthermore, RA displays two different effects, a long term and a short term effect. The differentiation of HL 60 cells lead to a decreased HIF response, whereasr the short term treatment had no influence of the HIF accumulation. Supported by Stiftung VERUM Dept. of Physiology, Universität Essen, Hufelandstr. 55, 45122 Essen, Germany O 04-3 ISOFORM-SPECIFIC EXPRESSION OF THE HYPOXIAα IN THE TAIL OF MOUSE INDUCIBLE FACTOR-1α SPERMATOZOA H. H. Marti1, D. M. Katschinski, K. F. Wagner, L. Schaffer2, B. Stier , R. H. Wenger The lumen of the seminiferous tubules of the testis is among the most hypoxic regions of the body with oxygen partial pressures as low as 2 mmHg. Thus, spermatogenesis, the differentiation of spermatogonial precursor cells to mature spermatozoa, takes place along a steep oxygen gradient. Sperm capacitation, motility changes, acrosome reaction and fertilization is exclusively dependent on anaerobic glycolysis and can occur under strictly anaerobic conditions. To satisfy the anaerobic ATP production, various testis-specific isoforms of glycolytic enzymes are expressed in the haploid stages of spermatogenesis and are still active in mature spermatozoa. The heterodimeric hypoxia-induc-

S 168 ible factor (HIF)-l is a transcriptional master regulator of several genes involved in mammalian oxygen homeostasis, including erythropoietin, vascular endothelial growth factor and factors involved in glucose transport and metabolism. The mouse Hifla gene is expressed from two distinct promoter/first exon combinations resulting in tissue-specific (mHIF-lαI.l) and ubiquitous (mHIF-lαI.2) mRNA isoforms. By in situ hybridization, we detected mHIF-lαI.l mRNA exclusively in the elongated spermatids of the testis. In vitro studies indicated that the switch from mHIF-lαI.2 to mHIF-lαI.l mRNA expression does not occur at the pre-meiotic stages of mouse spermatogenesis. Exposure of mice to hypoxic conditions induced mHIF-lαI.2 protein in spermatocytes and probably in Sertoli cells but not in spermatogonia. In contrast, expression of the putative mHIF-lαI.l protein in spermatozoa of the testis and epididymis was oxygen-independent, and located to the principal piece of the spermatozoal flagellum. Both the switch in transcript expression during spermiogenesis and the unexpected protein localization in mature sperm cells suggest a so far unrecognized function of HIF-lα. Institute of Physiology, Medical University of Lübeck, D-23538 Lübeck, Germany; 1Institute of Physiology and 2Department of Obstetrics and Gynaecology, University of Zürich, CH-8000 Zürich, Switzerland O 04-4 TISSUE SPECIFIC ERYTHROPOIETIN GENE REGULATION: COMPARISON BETWEEN HUMAN NEUROBLASTOMA AND HEPATOMA CELL LINES I. Stolze, U. Berchner-Pfannschmidt, P. Freitag, S. Frede, C. Wotzlaw, H. Acker, J. Fandrey Erythropoietin (EPO) gene expression is regulated by the transcriptional factor complex hypoxia inducible factor-1 (HIF-1) which is composed of two subunits: the O2-labile α- and the constitutive βsubunit. In addition, hepatocyte nuclear factor 4α (HNF-4α) is required for EPO expression in human hepatoma cells (HepG2). Herein, we compare EPO gene regulation in human neuroblastoma (NB) and HepG2 cells. Methods: EPO production in SH-SY5Y, Kelly and HepG2 cells was measured by real time PCR and EPO Elisa. Intracellular distribution of HIF-1α, HIF-1ß and HNF-4α was studied by Western blot and immunohistochemistry. To analyse HIF-1 DNA binding to regulatory elements in the EPO enhancer EMSAs were performed. Transcriptional activity of HIF-1 was investigated by transient transfection of a HIF-1 driven luciferase reporter construct. To determine the role of HNF-4α a dominant negative (dn) HNF-4 plasmid was transiently expressed. Results: Hypoxia induced EPO gene expression in NB cells was maximal at <0.1% O2. HIF-1α protein accumulated and translocated to the nucleus at an oxygen threshold of 7% O2 in HepG2 cells but at 3% O2 in NB. HIF-1 DNA binding and transcriptional activity of HIF1 was maximally induced at <0.1% O2 in SH-SY5Y and at 3% O2 in HepG2 cells. Differences were also seen in HIF-1α protein degradation υpon reoxygenation. Very little HNF-4α was expressed in NB, but did not colocalize with HIF-1α under hypoxic conditions. Transfection with dnHNF-4 did not affect hypoxic induction of EPO in NB cells. Conclusions: EPO production in neuroblastoma like in hepatoma cells is hypoxia-inducible. The O 2-sensitivity of EPO production in NB cells, however, was shifted to lower pO2 levels than in HepG2. Furthermore EPO gene regulation in human NB cell lines appears to be HNF4α independent.

In order to induce severe tumor hypoxia, one group of animals was housed in a hypoxic atmosphere containing 8% O2. Tumor oxygenation was assessed polarographically using O 2-sensitive needle electrodes. Afterwards, tumors were excised and disaggregated and the expression of the glucose transporter 1 (GLUT-1) was determined by flow cytometry using antibodies against the extracellular and cytoplasmic domains of GLUT-1. In a second series, the FDG-uptake of DS-sarcomas was determined either by using 14C- FDG (148 kBq) or by positron emission tomography (PET) with 18F-FDG (5 MBq). The expression of the GLUT-1 molecule did not change with increasing tumor volume. Neither the intracellular expression of GLUT-1 nor the expression of GLUT-1 in the cell membrane showed a correlation with tumor volume (which is closely related to the oxygenation status of the experimental sarcoma used). Also in tumors made severely hypoxic the expression was not significantly different from that found in DS-sarcomas of control animals. In parallel, the FDG-uptake did not vary systematically with tumor volume neither in the experiments where the uptake was determined directly (14C-FDG) nor by PET imaging (after partial volume correction) indicating that the FDG-uptake was independent of the extent of hypoxia in the tumor. These results reveal that GLUT-1 expression and glucose uptake are not affected by the extent of hypoxia. Based on these findings, it is unlikely that FDG-uptake is a suitable indicator for indirect assessment of tumor hypoxia. Institute of Physiology and Pathophysiology and Department of Nuclear Medicine, University of Mainz, 55099 Mainz O 04-6

Institut für Physiologie, Universität Essen, Hufelandstr. 55, 45147 Essen

HYPEROXIA INDUCES ENDOTHELIAL CALCIUM SIGNALING IN LUNG VENULAR CAPILLARIES W.M. Kuebler, C. Brueckl, H. Habazettl, A.R. Pries, K. Messmer Introduction: Although chronic hyperoxic ventilation is known to induce lung structural and functional damage, effects of short hyperoxic episodes <4 h are poorly characterized. Transient hyperoxia may induce oxidative signaling across the blood-gas barrier, in that reactive oxygen species (ROS) generated by NADPH oxidase evoke complex cellular responses. Here, the endothelial Ca2+ concentration ([Ca2+]i) is of particular interest, since it regulates vascular permeability and triggers inflammatory reactions. Methods: In intact capillaries of the isolated blood-perfused rat lung, we determined the endothelial [Ca2+]i by our in situ fura-2 ratio imaging technique (J Clin Invest 104:495, 1999). Images were obtained at normoxia (FiO2 0.21), then during 90 min of hyperoxia (FiO2 1.0), and for 45 min after return to normoxia. Experiments were replicated after addition of the irreversible NADPH oxidase blocker diphenyleneiodonium (DPI, 20 µM) to the blood. In a separate set of experiments, we analyzed the endothelial concentration of ROS by fluorescence imaging of the dye H2DCF. Results: During hyperoxic ventilation, [Ca2+]i increased within 45 min from 99±5 nM to 167±26 nM (p<.05). Concomitantly, endothelial H2DCF fluorescence increased from 123±15 fluorescence units (FU) to 294±23 FU (p<.05). Return to normoxic ventilation re-established baseline [Ca2+]i-profile of 107±3 nM and H2DCF fluorescence of 115±10 FU within 30 min. Over timecourse of experiments, [Ca2+]i and H2DCF fluorescence correlated significantly (rs = 0,79, p<.005). DPI did not affect baseline [Ca 2+ ] i (90±3 nM), but completely blocked the hyperoxia-induced [Ca2+]i increase (100±1 nM after 45 min, p<.05 vs control). Conclusion: Hyperoxia induces endothelial [Ca2+]i signaling in lung capillaries within < 1 h, presumably via ROS generated by NADPH oxidase. This second messenger response may contribute to hyperoxic lung injury.

O 04-5

Institute of Physiology, Free Univ. of Berlin, Arnimallee 22, 14195 Berlin, Germany

HYPOXIA DOES NOT AFFECT GLUT-1 EXPRESSION AND GLUCOSE UPTAKE IN EXPERIMENTAL TUMORS O. Thews, S. Kraus, A. Helisch, D.K. Kelleher, P. Bartenstein, P. Vaupel Hypoxia is a common phenomenon in tumors and is often combined with an increased glycolytic activity of the cells. The aim of the study was to analyze the impact of tumor hypoxia on mechanisms of cellular glucose uptake. Since in nuclear medicine, glucose uptake can be assessed using labeled fluoro-deoxyglucose (FDG) the question arises of whether FDG-uptake can be used as a surrogate parameter for noninvasive detection of tumor hypoxia. DS-sarcomas with a volume of between 0.5 and 2.5 ml were grown on the hind foot dorsum of SD-rats.

O 05-1 INTRACELLULAR CALCIUM STORES REGULATE ACTIVITYDEPENDENT NEUROPEPTIDE SECRETION FROM DENDRITES M. Ludwig, N. Sabatier, P.M. Bull, R. Landgraf, G. Dayanithi, G. Leng Classically, information in neurons flows from the synapses, through the dendrites and soma, and finally along the axon as spikes of electrical activity that cause the release of neurotransmitters from axonal terminals. This concept is now subject to many qualifications. In particular the dendrites of many neurons have an active secretory

S 169 role, transmitting information back from the post-synaptic cell to presynaptic nerve endings. By combining microdialysis with extracellular single cell recording in vivo we have shown that agents that induce liberation of calcium from intracellular stores induce oxytocin and vasopressin secretion from dendrites of magnocellular neurones of the hypothalamus without increasing the electrical activity of the cell body, and thus without inducing neuropeptide secretion from nerve terminals in the posterior pituitary. Conversely, electrical discharge activity in the cell bodies induces release from nerve terminals that is accompanied by little or no release of vasopressin and oxytocin from the dendrites. However, liberation of calcium from intracellular stores also ”primes” the releasable pool of neurosecretory material in the dendrites, making these stores available for secretion triggered by electrical discharge. Department of Biomedical and Clinical Laboratory Sciences, University of Edinburgh Medical School, George Square, Edinburgh EH8 9XD, UK O 05-2 IDENTIFICATION OF FUNCTIONALLY DISTINCT TYPES OF ASTROCYTES IN MOUSE HIPPOCAMPUS K. Matthias, F. Kirchhoff, G. Seifert, K. Hüttmann, M. Matyash H. Kettenmann and C. Steinhäuser Astrocytes in the mammalian brain express voltage- and ligand-gated ion channels. Previous analysis of glial cells in acute brain slices revealed that their functional properties significantly changed during development and in response to brain damage and disease. It is still unclear if this heterogeneity reflects variable stages of cellular maturation from precursors to more mature cells or rather indicates a heterogeneity among astrocytes. To unequivocally identify astrocytes, we took advantage of the availability of recently described transgenic mice with GFAP promoter-controlled EGFP-expression. Cells were freshly isolated from the hippocampal CA1 stratum radiatum of 6 to 20-days-old transgenic mice, and astrocytes were identified by their bright green fluorescence. The fluorescent cells displayed different current patterns, similar to those of astrocytes in wildtype mice. A fast application technique was used to identify glutamate receptors (GluRs) and glutamate transporters (n = 198). We find compelling evidence for a segregated expression of GluRs and glutamate transporters in cells with distinct current phenotypes (Fig. 1) and morphology. Not a single astrocyte co-expressed receptors and transporters, suggesting a coexistence of different types of astrocytes in the hippocampus. Thus, astrocytes can no longer be considered a homogenous cell population but, similar to neurons, have to be defined according to their specific functional properties. Supported by DFG (SFB 400, 515) and BMBF G luta m a te G lutam ate +N BQ X

0 .1 nA 1 nA

25 m s 15 m s A M PA

D-A sp artate

50 pA 10 m s

Fig.l. Expression of GluRs (top) vs. glutamate transporters (bottom) in two hippocampal astrocytes Experimental Neurobiology, Neurosurgery, Bonn University, Sigmund-Freud-Str. 25, 53105 Bonn, Germany O 05-3 EVIDENCE FOR GLUTAMATE IN AXON-GLIAL SIGNALLING IN THE DEVELOPING RAT OPTIC NERVE A.M. Butt, K.T. Tsang It has been proposed that the activity-dependent release of potassium (K+) or glutamate (Glu) is a way in which axons communicate with glia in white matter tracts of the central nervous system (CNS). Previously,

we have presented evidence that application of exogenous Glu induced raised glial calcium and K+ release in the developing rat optic nerve. The aim of the present study was to test the hypothesis that Glu is released by electrically active axons. Rat pups, aged postnatal day (P) 10 to P25 were humanely killed by anaesthetic overdose and optic nerves were isolated intact and maintained alive for 1-3 h in a brain slice chamber, continuously bathed in oxygenated artificial cerebrospinal fluid (aCSF). Optic nerves were held in two suction electrodes, one for stimulation and a second for measuring the compound action potential (CAP). During development, there was a large increase in the CAP amplitude and velocity, which developed from a simple, slow monophasic CAP to a complex, multiphasic rapid CAP. In immature nerves, the CAP could be sustained for many hours during continuous stimulation at low frequency of 2Hz, whereas stimulation at frequencies of 20Hz caused an activity-dependent decrease in the CAP. Tiring of the CAP was measured as peak amplitude after 120s expressed as a percentage of the amplitude at time 0s (2Hz) in the same nerve; all values are expressed as a MEAN ± SEM and significance tested using unpaired t-tests. The effects of tetraethyl ammonium chloride (TEA, 1 mM), glutamate (1mM), quisqualate (0.1 mM), kainate (0.1 mM), (CNQX, 10 µM), dihydrokainic acid (DHK, 50 µM) and l-trans-pyrrolidine2,4-dicarboxylic acid (PDC, 50 µM) were tested by bath incubation for 10 min. At 20 Hz, there was a decrease in CAP amplitude to 35 ± 11% of control (n=12), which was significantly less in the presence of the K+ channel blocker TEA (CAP amplitude 8 ± 4% of control; n=4) or the ionotropic Glu receptor (iGluR) antagonist CNQX (10 µM; 10 ± 7% of control; n=4). Incubation in Glu or the iGlur agonists quisqualate or kainate had no significant effect on axon tiring, whereas both PDC and DHK, which respectively block axonal (EAAC1) and glial (GLT1) Glu uptake, protected nerves against tiring (8 ± 6 % and 7 ± 3% of control, respectively; n=3). The results indicate that Glu is released during axonal activity and that it promulgates axonal tiring by activation of iGluR together with Glu uptake into axonal and glial compartments. Neural Damage and Repair Group, Centre for Neuroscience Research, Hodgkin Building, GKT Guy’s Campus, King’s College, London SE1 1UL, UK

O 05-4 ATP RELEASE UPON MECHANICAL STRAIN: AUTOCRINE AND PARACRINE PURINERGIC RECEPTOR STIMULATION AND PROPAGATION OF CALCIUM WAVES H. Sauer, J. Hescheler, M. Wartenberg Mechanical strain applied to confluent prostate cancer cells induced a [Ca 2+] i wave spreading radially with a velocity of about 15 µm s -1. [Ca2+]i waves were not dependent on extracellular Ca2+ and membrane potential since high K+ and Ca2+-free solution did not impair wave propagation. Waves did not depend on an intact cytoskeleton or functional gap junctions, since pretreatment with cytochalasin B, and nocodazole which disrupt microfilaments and microtubules, respectively, and 2.5 mM 1-heptanol, which uncouples gap junctions, were without effects. Fluorescence recovery after photobleaching (FRAP) experiments revealed absence of gap-junctional coupling. [Ca2+]i waves were inhibited by the purinergic receptor antagonists basilen blue and suramin, by pretreatment with ATP, UTP, ADP, UDP, 2-methylthioATP (2-MeS-ATP), benzoylbenzoyl-ATP (Bz-ATP), after depletion of intracellular ATP by 2-deoxyglucose and ATP scavenging by apyrase. Furthermore [Ca2+]i waves were inhibited by the anion channel inhibitors 5-nitro-2-(3-phenylpropylamino)benzoic acid (NPPB), tamoxifen, 4,4‘-diisothiocyanatostilbene-2,2‘-disulfonic acid (DIDS), niflumic acid, and gadolinium. Bioluminescence experiments revealed that ATP was released following membrane strain. This was significantly inhibited in the presence of anion channel blockers. We hypothesize that, upon mechanical strain, ATP is secreted via mechanosensitive anion channels and activates purinergic receptors on the same cell or neighboring cells in autocrine and paracrine manner, thus leading to [Ca2+]i wave propagation. Institut für Neurophysiologie, Universität zu Köln, Robert-Koch-Str. 39, D-50931 Köln, Germany

S 170 O 05-5 HIF-1α, A NEW HSP90 CLIENT PROTEIN: IMPLICATIONS FOR OXYGEN SENSING L. Le*, D. Heinrich, K.F. Wagner, S.G. Schindler, T. Hofer, R.H. Wenger, D.M. Katschinski Hypoxia-inducible factor (HIF)-1α is the oxygen-sensitive subunit of HIF-1, a transcriptional master regulator of oxygen homeostasis. Oxygen-dependent prolyl hydroxylation targets HIF-1α for ubiquitinylation and proteasomal degradation. Previously an interaction of HSP90 with HIF-1α in vitro has been described. To elucidate the physiological function of the HSP90/HIF-1α interaction in vivo, we determined the influence of endogenous (via exposure to 37°C–42°C) or transient ectopic (pcDNAHSP90) overexpression of HSP90 on HIF-1α expression and activity in HepG2 cells. Interestingly overexpression of HSP90 or heat was sufficient to stabilize mainly non-phosphorylated HIF-1α, whithout affecting HIF-1α mRNA expression. This effect could be antagonized with HSP90 inhibitors (geldanamycin and novobiocin). Heat-induced HIF-1α expression was also seen in vivo, when mice were exposed to elevated temperatures. HIF-1α, which was induced via endogenous or ectopic overexpression of HSP90, accumulated in the nucleus but neither bound to DNA nor trans-activated reporter or target gene expression (Glut-1 and aldolase), demonstrating the need of other interaction partners than HSP90 or posttranslational modification for these functions. Coimmunoprecipitaton of HSP90 and HIF1α as well as nuclear accumulation of HSP90 under hypoxic conditions demonstrates a general role of HSP90 for HIF-1α stabilization and therefore for oxygen sensing. This was further substantiated by destabilization of hypoxia-induced HIF-1α and concomittant decreased expression of HIF-1 target genes by HSP90 inhibitors. To further elucidate this hypothesis we recently adopted a genetic HSP90β deficient cell model.

R. Preisig-Müller, G. Schlichthörl, T. Görge, S. Heinen, A. Brüggemann, S. Rajan, C. Derst, R.W. Veh, J. Daut Andersen’s syndrome, an autosomal dominant disorder related to mutations of the potassium channel Kir2.1, is characterised by cardiac arrhythmias, periodic paralysis and dysmorphic bone structure. The aim of our study was to find out whether heteromerization of Kir2.1 channels with wild-type Kir2.2 and Kir2.3 channels contributes to the phenotype of Andersen’s syndrome. The following results show that Kir2.x channels can form functional heteromers: (1) HEK293 cells transfected with Kir2.2-Kir2.1 concatemers expressed inwardly rectifying K+ channels with a conductance of 30.3 ± 0.8 pS. (2) Expression of Kir2.x-Kir2.y concatemers in Xenopus oocytes produced inwardly rectifying, Ba2+ sensitive currents. (3) When Kir2.1 and Kir2.2 channels were co-expressed in Xenopus oocytes the IC50 for Ba2+ block of the inward rectifier current differed substantially from the value expected for independent expression of homomeric channels. (4) Coexpression of non-functional constructs, in which the GYG region of the pore region was replaced by AAA, with Kir2.x channels produced both homomeric and heteromeric dominant-negative effects. (5) Kir2.1 channels could be co-immunoprecipitated with Kir2.3 channels from membranes of isolated guinea-pig cardiomyocytes. (6) Yeast-two-hybrid-analysis showed interaction between the N- and C-terminal intracellular domains of different Kir2.x subunits. Co-expression of Kir2.1 mutants related to Andersen’s syndrome with wild-type Kir2.x channels showed a dominant negative effect the extent of which varied between different mutants. Our results suggest that differential tetramerization of the mutant allele of Kir2.1 with wild-type Kir2.1, Kir2.2 and Kir2.3 channels represents the molecular basis of the extraordinary pleiotropy of Andersen’s syndrome. Institute of Physiology, Marburg University, Deutschhausstraße 1-2, D-35037 Marburg, Germany

*Institute of Physiology, Medical University of Lübeck, D-23538 Lübeck, Germany O 05-6 O 06-2 INTRARENAL LOCALIZATION AND REGULATION OF THE SERUM AND GLUCOCORTICOID DEPENDENT KINASE ISOFORMS SGK1, SGK2 AND SGK3 Y. Feng, B. Friedrich, K. Klingel, M. Rizzo, G. Capasso, A. Vandewalle, S. Wärntges, T. Risler, D. Pearce, F. Lang The serum and glucocorticoid dependent kinases sgk1 was originally cloned as glucocorticoid dependent gene and later shown to be regulated by cell volume and mineralocorticoids. Homology cloning yielded the two isoforms sgk2 and sgk3. As shown separately, sgk1 upregulates several transport proteins relevant for renal electrolyte excretion, including ENaC, ROMK, Na+,K+,2Cl--cotransport and Na+/K+ ATPase, a property partially shared by sgk2 and sgk3. Thus, the three isoforms are probably involved in the regulation of renal electrolyte transport. In the present study, in situ hybridization, RT-PCR and Northern Blotting have been performed to determine intratubular localization and regulation of the three sgk isoforms. As a result sgk1 mRNA is mainly expressed in the renal collecting system. Sgk2 mRNA is found mainly in proximal tubule and thick ascending limb. Sgk3 mRNA abundance is low throughout the nephron. In a stable collecting duct cell line sgk1 mRNA, but not sgk2 or sgk3 mRNA is significantly upregulated by aldosterone (1 µM) and dexamethasone (0.1 µM). Sgk1 transcription is further increased by angiotensin II (10 µM) and TGF? (10 ng/ml). In mice, a three day low Na+-diet significantly increases the transcription of sgk1 but not of sgk2 and sgk3. Treatment with indomethacin further increases sgk1 transcription. In conclusion, while sgk1, sgk2 and sgk3 show overlapping functional activity, only sgk1 is up-regulated by Na+ depletion and mineralocorticoids. The stimulating effect of indomethacin on sgk1 transcription may relate to the beneficial effect of this drug in Bartter’s syndrome. Department of Physiology, University of Tübingen, Gmelinstr. 5, D 72076 Tübingen O 06-1 HETEROMERIZATION OF KIR2.x POTASSIUM CHANNELS CONTRIBUTES TO THE PHENOTYPE OF ANDERSEN’S SYNDROME

BLOCK AND PERMEATION OF THE CARDIAC MUSCARINIC K + CHANNEL, Kir3.1/Kir3.4, BY EXTRACELLULAR POLYAMINES S.Y. Makary, K. Dibb, T. Claydon, R. Leach, C.G. Nicholls , M.R. Boyett Voltage-dependent block by intracellular polyamines is in part responsible for the inward rectification displayed by the Kir family of K+ channels. We have previously shown that mutation of residues around the selectivity filter of Kir3.1/Kir3.4 (the cardiac muscarinic K+ channel) reduces inward rectification. It is possible, therefore, that polyamines interact with the selectivity filter. To test this, we have investigated block and permeation of Kir3.1/Kir3.4 by extracellular polyamines. Kir3.1, Kir3.4 and hD2 (human dopamine D2 receptor) were expressed in Xenopus oocytes and ionic current was measured with the two microelectrode voltage clamp technique in the presence of 10 µM dopamine and 90 mM extracellular K +. Extracellular putrescine and spermine blocked the wild-type channel with a KD at -130 mV of 9.5 ± 0.8 (mean ± SEM; n=6) and 1.2 ± 0.1 (n=6) mM; the block exhibited voltage-dependence (δ=0.16 and 0.11 for putrescine and spermine, respectively). To test whether the wild-type channel is permeable to polyamines, the 90 mM K+ in the bathing solution was substituted by 90 mM putrescine or spermine. The wild-type channel was slightly permeable to both polyamines: with putrescine and spermine, current at -100 mV was 0.20 ± 0.06 (n=6) and 0.15 ± 0.02 (n=9), respectively, times that carried by K+ (reversal potential, Erev, ~0 mV under all conditions). The mutations that reduced inward rectification (Kir3.1/Kir3.4-E145Q; Kir3.1/Kir3.4-R2155E; Kir3.1-R149E/ Kir3.4-E145R) reduced or abolished polyamine block and increased polyamine permeability. For example, putrescine and spermine failed to block Kir3.1-R149E/Kir3.4-E145R and this mutant channel was markedly permeable to both polyamines: with putrescine and spermine, current at -100 mV was 2.2 ± 0.2 (n=7) and 18 ± 5 (n=8), respectively, times that carried by K + (E rev, ~0 mV). These results support the hypothesis that polyamines can access the narrow selectivity filter of the channel. School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK

S 171 O 06-3 EFFECTS OF OVEREXPRESSION OF GRK2 AND ß-ARRESTIN2 ON DESENSITIZATION OF THE MUSCARINIC K+ CURRENT AND INTERNALISATION OF THE M2 MUSCARINIC RECEPTOR IN NEONATAL RAT ATRIAL CELLS Z. Shui, T.T. Yamanushi , M.R. Boyett In the heart, acetylcholine (ACh) binds to the m2 muscarinic receptor, causing activation of Gi protein and, hence, the muscarinic K+ channel. During a continuous application of ACh, muscarinic K+ current declines (desensitization) perhaps as a result of internalisation of the receptor. To investigate the involvement of G protein coupled receptor kinase (GRK) and P-arrestin in the desensitization process, we used the patch clamp technique to record whole-cell muscarinic K+ current in neonatal rat atrial cells transfected with GRK2 or β-arrestin2. In cells transfected with GRK2 or β-arrestin2, during a 3 min application of 10 µM ACh, the muscarinic K+ current declined by 51.9 ± 4.3 % (n=10) and 58.5 ± 4.9 % (n=12), respectively, which is significantly higher than that in untransfected cells (40.4 + 3.6 %, n=13). Immunocytochemistry was used to study the localization of the m2 receptor in neonatal rat atrial cells with and without pre-treatment by 10 µM CCh for 3 min. The m2 muscarinic receptor label in the untreated cells was on the cell membrane, whereas in the pretreated cells a substantial portion of the m2 receptor label was internalised within the cytoplasm. After CCh pretreatment, the mean intensity of m2 receptor labelling on the cell membrane in the cells transfected with GRK2 was significantly lower (34.4 ± 3.6 %, n=87) than in untransfected cells (55.2 ± 3.7 %, n=91). We conclude that both GRK2 and β-arrestin2 may play a major role in internalisation of the m2 receptor and thus desensitization of the cardiac muscarinic K+ channel.

TASK2 is a member of the two-P-domain K+ channel family and is predominantly expressed in epithelial tissues. According to its properties in heterologous expression systems, it is suggested that TASK2 serves as background K+ channel sensitive to extracellular pH. Here we examined the renal phenotype of TASK2 knockout mice. Knockout mice have a reduced bodyweight and litters are smaller. They present no other obvious phenotype. In the kidney TASK2 is strongly expressed. X-gal coloration of the kidney using the TASK2-promotordependent expression of beta-galactosidase in heterozygous and knockout mice revealed an expression of TASK2 in proximal tubular cells and kidney papilla. The staining in renal papilla is located in collecting duct cells, however, a staining of capillary endothelium cannot be excluded. In inulin clearance experiments fractional Na+ excretion of TASK2 knockout mice was slightly enhanced. In order to unmask the role of alkalosis-activated TASK2 channel for salt reabsorption in the kidney NaHCO3 was administered to increase blood pH and Na+ load. After one hour of NaHCO3 infusion, blood pH was similarly increased in wildtype and knockout mice. However, knockout mice showed a markedly increased urinary flow and salt loss. These data suggest a role of TASK2 for salt and water reabsorption of proximal tubular cells. One could speculate that during NaHCO3 reabsorption an increase of interstitial pH stimulates TASK2 K+ channels leading to a hyperpolarisation. Since Na+/3HCO3- cotransport is electrogenic this hyperpolarization would support further Na+ and HCO3- reabsorption in proximal tubules of wildtype mice. The role of TASK2 in papillary collecting ducts awaits further investigation. Physiologisches Institut, Winterthurerstr. 190, CH-8057 Zürich and IPMC, CNRS, 660 Route des Lucioles, F-06560 Valbonne Sophia Antipolis

School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK O 06-6 O 06-4 EXPRESSION AND FUNCTIONAL ROLE OF TWO PORE DOMAIN K+ CHANNELS IN THE DORSAL THALAMUS T. Budde, S. Meuth, T. Kanyshkova, T. Munsch, H.-C. Pape The process of waking up is associated with the change from rhythmic burst activity to tonic single spike firing in many neurons of the thalamocortical system. This change is mediated by transmitters of the ascending brainstem system by reduction of leak K+ channels, resulting in steady membrane depolarization. Until now the molecular basis of these channels has not been analyzed in the thalamus. Therefore the expression of TWIK-related acid-sensitive K+ (TASK) channels, which contribute to setting the membrane potential in many mammalian neurons, was probed in the dorsal lateral geniculate nucleus (dLGN) of the rat by combining whole cell patch-clamp recordings in brain slices, molecularbiological, and immunohistochemical techniques. Setting the membrane potential of thalamocortical relay neurons to a value of -20 mV induced a persistent outward current of 200-400 pA, which was significantly reduced by lowering the external pH from 7.2 to 6.4 and application of ACh, muscarin, and the local anaesthetic bupivercaine. In addition, the steady state outward current was increased by bath application of the general anaesthetic halothane, thereby completing the typical pharmacological and regulatory profile of TASK channels. RT-PCR analysis, antibody staining, and in situ hybridization demonstrated the expression of TASK1 and TASK3 in dLGN, thereby further confirming the electrophysiological findings. During current clamp recordings, application of halothane hyperpolarized the membrane potential from about -60 to -70 mV, an effect which was accompanied by a change in firing pattern from tonic single spike to burst activity. Taken together these data indicate that several TASK channel subtypes are involved in setting the membrane potential of dLGN relay neurons and contribute to the production of anesthesia in vivo. (Supported by DFG BU-1019/4-2; Leibniz-Programm; Graduiertenkolleg “Biol. Grundl. Erkr. NS”)

ALTERATION OF K+ SWELLING-ACTIVATED CHANNELS BY CFTR KNOCK-OUT IN DCT AND CCT CELLS IN PRIMARY CULTURES OF RENAL TUBULES R. Belfodil, H. Barrière, M. Tauc, C. Poujeol, P. Poujeol Objective Cultured proximal, distal and collecting duct cells from the kidney of CFTR-/- mice exhibited an impaired cell volume regulation due to an absence of regulatory volume decrease process (RVD). As we have previously demonstrated, this defect could be explain by the loss of swelling-activated Cl- currents. However, since RVD is mediated by Cl- and K+ conductances it was interesting to know whether the CFTR knock-out induced also alterations of the K+ swelling-activated channels in renal cells. Method: Segments of renal tubules were microdissected to obtain primary cultures of proximal convoluted (PCT), distal convoluted (DCT) and cortical collecting tubules (CCT) from wild type (CFTR+/+) and CFTR knock-out (CFTR-/-) mice. K+ conductances were characterised using the whole cell clamp technique in 7 day-old primary cultures. Results: In cultured DCT and CCT from CFTR+/+ mice, a hypo-osmotic shock (30%) induced outwardly rectifying K+ currents. These currents were blocked by TEA (1mM), charybdotoxin (10nM) and quinidine (500µM). In PCT the K+ swelling-activated channels were insensitive to TEA and charybdotoxin and they were activated at extracellular pH 8.0 and inhibited at pH 6.0. PCT cells from CFTR-/- exhibit K+ swelling-activated channels having identical characteristics. By contrast in DCT and CCT cells from CFTR-/- mice, the hypotonic shock was completely inefficient in increasing K + currents. In conclusion: The inactivation of the CFTR gene induces a strong decrease in swelling-activated K + conductances in DCT and CCT but not in PCT cells. In PCT cells, the pharmacology and the regulation of K+ swelling-activated channels are different compared to DCT and CCT cells suggesting the presence of two types of swellingsensitive K+ channels. In PCT cells, the K+ swelling-activated channels could belong to the TASK channels family. UMR-CNRS 6548, Université de Nice Sophia Antipolis, Parc Valrose, 06108 Nice cedex, France

Otto-von-Guericke-Universität Magdeburg, Institut fur Physiologie, Leipziger Str. 44, D-39120 Magdeburg O 07-1 O 06-5 +

TASK2 K CHANNEL KNOCKOUT MICE SUFFER FROM RENAL SALT AND WATER LOSS R. Warth, M. Tauc, P. Poujeol, N. Guy, J. Barhanin

PROTEIN KINASE G-STIMULATED DEPHOSPHORYLATION OF MYOSIN LIGHT CHAINS IS DUE TO AN ACTIVATION OF PROTEIN PHOSPHATASES IN ENDOTHELIAL CELLS S. Nau, D. Gündüz, K. Bindewald, D. Klingenberg, H. M. Piper, T. Noll Endothelial cells possess a contractile apparatus similar to smooth

S 172 muscle cells. The state of the contractile activation depends on the phosphorylation of the regulatory myosin light chains (MLC~P) which is under control of myosin light chain kinase and myosin light chain phosphatase. Here the influence of the cGMP-Protein kinase G (PKG) signal pathway on the MLC~P (Western blot) was studied in cultured aortic endothelial cells (pig). In non-stimulated endothelial cells MLC was phosphorylated by 78±5 % (mean±SD, n=5, P<0.05; maximum of MLC~P= 200%, since MLC are phosphorylated at two sites). Activation of the soluble guanylate cyclase by the ΝΟ−Donor SNAP (10 µM) or direct activation of PKG by 8-Br-cGMP (50 µM) reduced MLC-P to 27±6 % and 24±5 % (for both P<0.05) after 20 min, respectively. Inhibition of guanylate cyclase by ODQ (5 µM) or PKG by KT 5823 (50 µM) prevented the dephosphorylation of MLC. Inhibition of myosin light chain kinase by ML-7 (50 µM) decreased MLC~P to 34±3% (P<0.05) after 20 min. The simultaneous addition of 50 µM ML-7 plus 10 µM SNAP caused a further reduction of MLC~P to 9±4% (P<0.05), indicating that inhibition of myosin light chain kinase does not influence the cGMP-evoked reduction of MLC~P. Inhibition of the protein phosphatase 1 and 2A by calyculin A (10 nM) increased MLC~P to 117±9% significantly (P<0.05), indicating that MLCK is still active under basal conditions in endothelial cells. In contrast, simultaneous addition of 10 nM calyculin A plus 10 µM SNAP abolished the SNAP-induced reduction of MLC~P, indicating that cGMP may influence the state of MLC~P via modulation of protein phosphatase activity. Conclusion: In endothelial cells, phosphorylation of MLC is reduced via cGMP-PKG signal pathway. The data of the present study show that the cGMP effect is due to an activation of protein phosphatases and does not rely on an inhibition of the myosin light chain kinase. Physiologisches Institut, Justus-Liebig-Universität Giessen, D-35392 Giessen

tension and organ damage in mice. Methods: We studied mice with 1, 2 or 3 copies of the Agt gene originally described by Kim, Smithies and coworkers. Mice with renin-dependent two-kidney, one-clip renovascular hypertension (2K1C, 8 weeks) or renin-independent deoxycorticosterone-acetat salt hypertension (DOCA, 6 weeks) were compared with controls (sham-operation, or uninephrectomy and saltloading) of the respective genotype Systolic blood pressure was measured weekly (tail cuff). Intraarterial blood pressure (carotid artery) and relative left ventricular weight were measured at the end of the experiment. Kidney renin was measured by counting the percentage of glomeruli with a juxtaglomerular apparatus staining positive for renin. Agt-0 mice were not included because of congenital kidney injury whereas the experiments in Agt-4 mice are not yet complete. Results: The 2K1C procedure increased blood pressure significantly by 21 mmHg in Agt-2 (n=15) and by 16 mmHg in Agt-3 (n=12), compared to sham-operated animals (n=7-12 per genotype). In contrast, Agt-1 with 2K1C had only marginally higher blood pressure than sham-operated Agt-1 (116 versus 109 mmHg, n=7 each), and did not develop left ventricular hypertrophy (3.4 versus 3.3 mg/g in shamoperated Agt-1) which was present in 2K1C Agt-2 (3.8 versus 3.2 mg/ g in sham-operated Agt-2) and Agt-3. Renin was induced in the clipped and suppressed in the contralateral kidney (60 versus 11 % of glomeruli) in 2K1C Agt-1, indicating that the 2K1C procedure itself had been succesfull. DOCA treatment increased left ventricular weight and blood pressure in all genotypes (by 33, 19, and 35 mmHg in Agt-1, -2, and -3, respectively), compared to controls of the same genotype. Conclusion: Mice with only one copy of the angiotensinogen gene are protected against renin-dependent hypertension. Thus, the angiotensinogen genotype modulates the development of high-renin but not low-renin hypertension. IV Medizinische Klinik, Universität Erlangen-Nürnberg, Universitätsstraße 17, 91054 Erlangen

O 07-2 REDISTRIBUTION OF BLOOD FLOW DUE TO AN ENHANCED SCHRETZENMAYR EFFECT IN THE FOREARM OF PATIENTS WITH INCREASED CA-REPEATS IN INTRON 13 OF THE ENDOTHELIAL NITRIC OXIDE SYNTHASE GENE H. Jepsen, M. Laule, K. Pagonidis, V. Stangl, I. Cascorbi*, I. Roots*, G. Baumann, K. Stangl Stangl et al. (Pharmocogenetics, 2000) have shown an association between high CA repeat copy number and an excess risk of coronary artery disease. Since CA-repeats are known as mRNA-splicing enhancers NO-metabolism might be affected. In order to analyse endothelium dependent vasodilatation in the forearm, blood flow measurements were performed according to Doppler‘s principle (HDI 3000, ATL, USA, 10 MHz probe, HDD-mode) and by Laser Doppler (LD) Fluxmetry (Perimed, Sweden) in 18 patients with CA-repeats < 32 (low #) and in 30 patients with at least 1 allele containing > 34 CA-repeats (high CA #). Measuring sites: brachial artery proximal to an infusion catheter and the microvasculature of the superficial and deep compartment of the finger skin, the distal forearm skin and the M. ext. digit. comm. Changes of blood flow were observed following infusion of 5 dosages of SNP and ACH. Sensitivity of flow-dependent diameter regulation (% change in vessel diameter divided by imposed % change of shear rate) was enhanced in the brachial artery of high CA # -patients (4.3 + 0.4 % (high) vs 2.9 + 0.4 % (low), p < 0.05). In addition, flux-flow ratio was increased within the proximal extensor muscle, whereas it was reduced at all distal skin sites at all dosages in high CA # -patients. Conclusion: In high CA #-patients blood flow is redistributed to the proximal vasculature due to enhanced sensitivity to shear stress in upstream vessels. Medizinische Klinik (Kardiologie, Angiologie, Pneumologie), *Institut für Klinische Pharmakologie, Charité, Humboldt Universität zu Berlin O 07-3 EFFECT OF THE ANGIOTENSINOGEN GENOTYPE ON EXPERIMENTAL HYPERTENSION IN MICE C. Handtrack, B. Klanke, N. Cordasic, R. Veelken, K.F. Hilgers Background: Genetic variants of the angiotensinogen gene (Agt) may be associated with hypertension in men. We tested the hypothesis that the number of Agt gene copies affect the development of hyper-

O 07-4 CYTOCHROME P450 2C9 INDUCES ENDOTHELIAL CELL PROLIFERATION AND ANGIOGENESIS VIA CROSS-TALK WITH THE EPIDERMAL GROWTH FACTOR RECEPTOR (EGFR) U.R. Michaelis, B. Fisslthaler, R. Busse, I. Fleming Cytochrome P450 (CYP) epoxygenase products, such as 11,12epoxyeicosatrienoic acid (EET), exert a proliferative effect on endothelial cells. The aim of the present study was to identify the signal transduction cascade linking EET generation to enhanced proliferation and to determine whether or not EETs are angiogenic. CYP 2C9 over-expressing human umbilical vein endothelial cells (HUVEC) generate 11,12- and 14,15-EET and proliferate faster than endothelial cells transfected with a control vector. Cell number as well as BrdU incorporation into DNA were increased in these cells and both effects were sensitive to the selective CYP 2C9 inhibitor, sulfaphenazole. CYP 2C9 over-expression was associated with an increase in the phosphorylation of Akt, an effect which was abolished by AG1478, a specific inhibitor for the EGFR tyrosine kinase. AG1478 also prevented the CYP 2C9-induced increase in cell proliferation and BrdU incorporation. Since activation of the EGFR tyrosine kinase with EGF induces angiogenesis, we determined whether or not the over-expression of CYP 2C9 was associated with angiogenesis in an in vitro assay. HUVEC were infected with adenoviral constructs encoding CYP 2C9 or an antisense CYP 2C9 and seeded onto a fibrin gel. After approximately 14 days, the formation of endothelial cell tubes was assessed. The CYP 2C9 antisense virus did not affect endothelial tube formation whereas CYP 2C9 over-expression marked increased tube formation (by 6 fold). A similar phenomenon was observed in endothelial cells treated with exogenous 11,12-EET and was comparable with the tube formation stimulated by EGF. These results indicate that CYP 2C9 plays an important role in cell proliferation and in the angiogenic response via the generation of EETs. EET signalling involves crosstalk with the EGFR and the activation of Akt. Institut für Kardiovaskuläre Physiologie, Klinikum der J.W.G.-Universität, D-60590 Frankfurt am Main, Germany

S 173 O 07-5

O 08-1

DECREASED EXPRESSION OF SOLUBLE GUANYLYL CYCLASE IN GENETIC HYPERTENSION IS MEDIATED BY THE mRNA STABILIZING PROTEIN HUR S. Klöss, R. Bordel, K. Suna, A. Mülsch The expression of soluble guanylylcyclase (sGC) subunits α 1β 1 is critically controlled by degradation of the respective mRNAs. In rat aortic smooth muscle cells we previously found that the mRNA stabilizing protein HuR binds to AU-rich motifs in the 3’-UTR of the sGCα1 mRNA and protects this subunit from degradation by an AUspecific endonuclease. The objective of the present study was to clarify whether reduced expression of HuR could account for decreased expression of sGC in aortic tissue of hypertensive rats (SHR) as compered to normotensive, aged-matched control rats (WKY). In aortic protein extracts from 16 month old SHR the HuR binding activity, as detected by electrophoretic mobility shift analysis (RNAEMSA) using an AU-rich, biotin-labelled oligonucleotide, was significantly reduced as compared to pre-hypertensive young SHR (2 month) and age-matched normotensive WKY. sGCα1 mRNA enriched from lung tissue of normotensive rats was degraded by aortic protein extract from aged SHR with an increased rate as compared to WKY and young SHR. The expression of HuR in aortae from aged SHR was decreased by about 50 % at the mRNA (RT-PCR) and protein level (W-Blot) as compared to age-matched WKY. The expression (WBlot) and activity (DNA-EMSA) of AP-1, a transcription factor involved in downregulation of HuR, was increased in aged SHR compared to WKY. All parameters tested were not different between young pre-hypertensive SHR and young WKY. These findings suggest that in this model of genetic hypertenision reduced expression of sGC is mediated by downregulation of the mRNA stabilizing protein HuR.

IMAGING CALCIUM TRANSIENTS OF OLFACTORY NEURONS EXPRESSING DISTINCT ODORANT RECEPTORS U. Pehl, J. Kaluza, H. Breer, J. Strotmann The mammalian olfactory system is able to recognize and distinguish thousands of odorants that vary widely in structure. This discriminatory power is supposed to be due to the differential responsiveness of defined olfactory sensory neurons (OSNs) based on their distinct odorant receptor type. To characterize the response spectrum of odorant receptor types combined functional and molecular analyses of individual isolated OSNs are most promising; transgenic mouse lines have now provided additional options. In these mice, OSNs expressing a distinct receptor type co-express GFP and thus can be identified by their green fluorescence. A suitable imaging set-up allowed to monitor the GFP-fluorescence of these cells and in addition register their calcium concentration using appropriate indicators. This approach was employed trying to determine ligands for the OR37-receptor subtypes which are unique not only based on their clustered expression pattern on an exposed topographic area of the nasal epithelium but also due to characteristic structural features. Employing the imaging approach the viability and general responsiveness of isolated OR37-cells was confirmed by calcium signals induced by depolarization with KCl. The reactivity of the second messenger cascade was confirmed by superfusing the cells with membrane-permeable 8Br-cAMP or 8Br-cGMP which elicited calcium signals in a dose-dependent manner. Subsequently, the responsive cells were exposed to a large variety of representative odorants; none of these odorants elicited responses in OR37-neurons suggesting that according to the unique features of the OR37-receptors, unconventional compounds may be the appropriate ligands. Out of a variety of „unusual“ odorous compounds, certain long chain aliphatic acids induced significant calcium responses in OR37 cells. The specificity of these reactions is underlined by the fact that non-OR37cells did not respond. The data provide first evidence that long chain aliphatic acids may be appropriate ligands for the OR37 receptors.

Institute of Cardiovascular Physiology, J.W.G.-Universität, D-60590 Frankfurt

Department of Physiology, University of Hohenheim, Garbenstr. 30, D-70599 Stuttgart, Germany

O 07-6 O 08-2 CYCLIC STRAIN-DEPENDENT CD40 RECEPTOR EXPRESSION IN HUMAN ENDOTHELIAL CELLS: ROLE OF REACTIVE OXYGEN SPECIES A.H. Wagner, S. Baumgarten, K. Schreiber, M. Hecker In endothelial cells the balance between shear stress-dependent nitric oxide (NO) and cyclic stretch-dependent superoxide (O2-) formation seems to be important for the functional integrity of these cells. Thus, in the early phase of arteriosclerosis expression of chemokines and adhesion molecules and, as a consequence, the recruitment and activation of circulating leukocytes at the luminal surface of the endothelium appears to be due to an imbalance between these radicals. The interaction of activated T cells expressing the CD40 ligand (CD154) with CD40 constitutively expressed on the surface of endothelial cells is thought to play a crucial role in the recruitment of leukocytes, namely monocytes in chronic inflammatory diseases including arteriosclerosis. This study was designed to assess the effects of oxidative stress and cyclic stretch (0.5 Hz, 20% elongation) on the expression of CD40 in human umbilical vein cultured endothelial cells (HUVEC) by using semiquantitative RT-PCR and Western blot analysis. Cyclic strain triggered a transient increase in the formation of reactive oxygen species (O2-, H2O2) both in cultured (HUVEC, 2',7'-dichlorofluorescein fluorescence) and native endothelial cells (rabbit carotid artery, dihydroethidine oxidation). Exposure of HUVEC to both endogenous (NO synthase blockade) and exogenous oxidative stress (generated by xanthine oxidase/ xanthine system) up-regulated the expression of CD40 2-3fold after 6 h. Moreover, exposure of the cells to cyclic stretch for 3-6 h also upregulated expression of CD40 2fold. These findings suggest that in addition to the known induction of CD40 expression by cytokines, cyclic stretch may trigger the formation of CD40 in a redox-sensitive manner. As a consequence, the increased CD40/CD154-mediated interaction between endothelial cells and circulating leukocytes may contribute to initiation and/or progression of arteriosclerosis.

INTRACELLULAR ANIONS AS THE VOLTAGE-SENSOR OF PRESTIN, THE OUTER HAIR CELL MOTOR PROTEIN D. Oliver, D.Z.Z. He, N. Klöcker, P. Dallos, B. Fakler Outer hair cells (OHCs) of the mammalian cochlea actively change their cell length in response to changes in membrane potential. This electromotility, thought to be the basis of cochlear amplification, is mediated by a voltage-sensitive motor molecule recently identified as the membrane protein prestin. Structural rearrangements of the protein are driven by the movement of a charged voltage sensor through the electric field, giving rise to the characteristic voltage dependent charge movement or non-linear capacitance (NLC) of the OHC. We attempted to identify this voltage sensor, a key element in electromechanical transduction by prestin. Mutational neutralization of candidate charged amino acid residues of prestin failed to eliminate NLC and thus to identify a protein domain as the voltage sensor. In contrast, we show that voltage-sensitivity is conferred to prestin by the intracellular anions Cl- and HCO3-. Removal of these anions abolished NLC as well as fast voltage-dependent motility. Upon removal of intracellular Cl, prestin adopted its contracted conformation, independent of the membrane voltage. When Cl- was replaced by other monovalent anions, the characteristics of charge movement were dependent on the anion species present at the cytoplasmic side of the membrane. In inside-out patches from OHCs, the amount of translocated charge systematically varied with different monovalent anions (I- > Br- > NO3- > Cl- > HCO3> F-) and the steepness of voltage dependence decreased with increasing size of the anion (formate > acetate > propionate > butyrate). No charge movement occurred with the divalent sulfate. The results support a model in which anions act as extrinsic voltage-sensors, that bind to an intracellular site of the prestin molecule and thus trigger the conformational changes required for motility of OHCs. Supported by DFG (SFB 430/A1) and NIH (DC00089).

Department of Cardiovascular Physiology, University of Goettingen, Humboldtallee 23, D-37073 Goettingen, Germany

Physiologisches Institut II, Hermann-Herder-Str. 7, 79104 Freiburg, Germany

S 174 O 08-3 UPTAKE OF SUGAR BY CELLS EXPRESSING THE OUTER HAIR CELL MOTOR PROTEIN PRESTIN J.-M. Chambard, N. O’Shea, J.F. Ashmore In the mammalian cochlea, auditory sensitivity is enhanced by a population of outer hair cells (OHC) that can generate forces at acoustic rates. The mechanism depends on an assembly of “motor” molecules in the cell’s basolateral membrane allowing OHCs to change length at acoustic frequencies upon membrane polarization. The motor protein has been identified as Prestin [1], a member of a transporter superfamily SLC26A that operates in a cycle where transmembrane chloride transport is arrested [2]. OHCs also take up sugars and label with antibodies raised against the fructose transporter GLUT5 [3]. A Cterminal GFP-tagged prestin construct (a gift from B. Fakler) was expressed in HEK-293 cells. To investigate sugar uptake, cells were exposed to isotonic replacement of the sugar in the bathing medium. The fluorescent profile of the cells was imaged and stored for subsequent analysis. Isotonically replacing external glucose with 30 mM fructose led to a reversible increase in the diameter of spherical cells by approximately 2.9 ± 1.7%.(mean ± sd, n=18). The cell area increased by 5.9 ± 3.6%. The data are consistent with an influx of water following fructose exposure. The recovery occurred with a time constant of approximately 140 seconds. There was no area increase in cells expressing the GFP alone nor was any increase seen when extracellular solution alone was applied. The data suggest that HEK-293 cells can take up fructose when transfected with Prestin. Although other subunits may co-assemble with Prestin, the most economical hypothesis is that the alpha helical structure of this protein allows neutral substrate transport. [1] Zheng et al. (2000) Nature 405, 149-155; [2] Oliver et al. (2001) Science 292, 2340-2343; [3] Geleoc et al.(1999) Nature Neurosci. 2, 713-719. Supported by the Wellcome Trust. Department of Physiology, University College London, Gower Street, WC1E 6BT London, UK O 08-4 IDENTIFICATION OF A NOVEL PROTEIN EXPRESSED IN SENSORY HAIR CELLS AND CNS NEURONS E. Reisinger, U. Zimmermann, J. Ludwig, M. Knipper, B. Fakler, D. Oliver The exquisite sensitivity of hearing in mammals depends on highly specialised sensory and accessory cell types within the cochlea. We are interested in proteins that define these functional and structural specialisations to understand inner ear physiology and pathophysiology on a molecular level. In rodents, differentiation of many cochlear cell types occurs during postnatal development prior to or concomitantly with the onset of hearing, allowing for a developmental approach to the identification of such proteins. Sequencing of cDNA libraries prepared from organs of Corti from 1 and 9 days old rats (PD1 or 9, respectively) revealed an EST with an open reading frame (ORF) that lacks homology to characterised genes. This EST was consistently found at PD9 but absent at PD1. Using race methods, we obtained the full-length cDNA with an ORF of 2.8 kb. The deduced amino acid sequence indicates a mostly hydrophilic protein of 106kDa. Alignments with NCBI nucleotide database showed similarity to a hypothetical human protein of unknown function. In situ hybridisation (ISH) in the rat cochlea revealed the exclusive expression of the mRNA in sensory hair cells with high abundance in outer hair cells (OHCs) and lower expression in inner hair cells. Expression showed a developmental pattern with the earliest onset in OHCs of the basal cochlear turn at PD4. Moreover, ISH showed abundant expression of this new protein in CNS neurons, most prominent in cortex, hippocampus and cerebellum. Neuronal expression was detectable as early as postnatal day 1. This novel protein might play a specific role in the function of sensory receptors and neurons.

hair cells because they help shape their receptor potentials. In particular, the functional maturation of inner hair cells in mice around postnatal day 12 (P12; onset of hearing) coincides with the expression of a BK conductance. In hearing organs of lower vertebrates frequency tuning depends on BK channels with different kinetics and Ca2+ sensitivity. These features are varied by alternative splicing of the pore forming α subunit (encoded by the slo α or KCNMA gene) as well as combination with modulating β subunits (KCNMB1). It was unclear if similar mechanisms “fine tune” mammalian hair cells. We used RT-PCR and “single cell” RT-PCR to screen rat cochleae for KCNMA splice variants. Mainly, we isolated mRNA coding for a minimal variant of the KCNMA subunit. Occasionally we found additional splice variants in three of the seven known internal splice sites. We conclude that there is no extensive usage of multiple KCNMA subunits with differing kinetics and Ca2+ sensitivities in the rat cochlea. However, we isolated mRNA coding for six KCNMA splice variants varying in their C-terminal amino acid sequences. We speculate that these direct the channel protein to different intracellular domains. In RT-PCR experiments we also demonstrated the expression of KCNMB1, 2 and 4 specific mRNA. The cellular expression patterns of several KCNM specific mRNAs was further investigated by in situ hybridisation. The experiments revealed a similar staining pattern for KCNMA and KCNMB1 specific probes. In inner hair cells strong staining was observed emerging shortly before the onset of hearing. Staining of outer hair cells set in some days later and appeared generally weaker. Several types of supporting cells, basal parts of the stria vascularis and spiral ganglion cells proved also positive for KCNMA and KCNMB1 specific mRNA. Our molecular data show a strict coexpression of KCNMA and KCNMB1 mRNA and confirm earlier electrophysiological results in the cochlea. The expression pattern of the KCNMB4 specific mRNA differed strongly from those of the other KCNM subunits. It seemed to be expressed only in immature organs of Corti before the onset of expression of KCNMA and KCNMB1 specific mRNAs but persisted in apical turns of the spiral ganglia. Its physiological role is obscure. In summary, variation of BK channel kinetics and Ca2+ sensitivity do not seem to be used for tuning of mammalian hair cells as done in lower vertebrates. Patricia Langer, Physiologisches Institut Abt. II und Forschungslabor der HNO Tübingen, Röntgenweg 11, D-72076 Tübingen, Germany O 08-6 THE BK CURRENT IS NECESSARY FOR A GRADED RECEPTOR POTENTIAL IN MOUSE INNER HAIR CELLS: A SIMULATION STUDY S. Münkner1, J. Engel1, C. J. Kros2 Inner hair cells (IHCs) in the organ of Corti transform acoustical vibrations into transmitter release. The current flowing through the transduction channels depolarizes the IHCs, which activates Ca2+ channels. The resulting Ca2+ influx triggers neurotransmitter release that is phase coupled to the acoustical stimuli up to several kHz. Before the onset of hearing IHCs in mice are unable to show this phase coupling, because they react to current injection with slow action potentials rather than a graded receptor potential (Kros et al., 1998, Nature, 394:281-284). We demonstrate in a computer model that a fast acting BK-type K+ current that develops during the onset of hearing is crucial for the IHCs to transform the transduction current into a graded receptor potential and thus a phase coupled transmitter release. With a neonatal composition of ionic currents (a delayed rectifier K+ current, an L-type Ca2+ current and a leak conductance) without the BKtype K+ current, the model shows the slow action potentials found in the experimental data. Supported by the Deutsche Forschungsgemeinschaft (DFG En 294/2-1, to J.Engel) and a Royal Society joint project grant to C.J.Kros and J.Engel. 1

Inst. of Physiology II, Gmelinstr 5, 72076 Tubingen, Germany; School of Biol. Sciences, Sussex Univ., Falmer, Brighton BN1 9QG, UK

2

Physiol. Institut II, Hermann-Herder-Str. 7, 79104 Freiburg, Germany O 09-1 O 08-5 IDENTIFICATION AND LOCALISATION OF BK-SPECIFIC mRNA IN THE RAT COCHLEA P. Langer, S. Gründer, A. Rüsch Ca2+ activated K+ channels of the BK type are important for cochlear

GENDER DIFFERENCES IN THE CHEMICAL CONTROL OF BREATHING IN MAN T. Schäfer, W.T. Ulmer, M.E. Schläfke We have compared hypercapnic ventilatory responses (HCVR) in 15 men and 15 women, machted for age (24.0 ± 4.8 yrs, range 19 – 37

S 175 yrs) using a steady-state method with the subjects sitting on a chair and breathing through a mouthpiece with a nose clip in place. Ventilation has been quantified by a heated pneumotachograph (Jaeger, Würzburg). In addition to minute volume the early mouth occlusion pressure 100 ms after onset of inspiration (P0.1) has continuously been monitored as a measure of central inspiratory drive. PCO2 has been determined by a rapid infrared analyser in side-stream technique. Three steps of increased FiCO2 have been applied with maximum PETCO2 values of 9 Torr above resting P ETCO 2 . Finally a complete spirometric and bodyplethysmographic lung function testing has been performed. All subjects were healthy and had normal lung functions (forced vital capacity (FVC): 106.0 ± 12.1 % predicted, FEV 1: 105.7 ± 10.6 % predicted, total airway resistance Rt: 0.17 ± 0.04 kPa*s/l). Men had a steeper HCVR than women (2.85 ± 1.65 l/min/Torr versus 1.23 ± 0.45 l/min/ Torr, p < 0.002, t-test). This difference remained even after correction for body weight, body height, body surface area, total lung capacity, and FRC. In contrast, the slope of the hypercapnic P0.1 response was similar in men and women (35.7 ± 18.4 versus 25.9 ± 10.4 Pa/Torr, P > 0.08), indicating that the difference in the ventilatory response may be due to mechanical factors rather than to different central CO 2 dependent respiratory drives. This is supported by the fact that the ratio of minute ventilation per P 0.1 was significantly higher in men than in women (e.g. 94.7 ± 23.2 versus 61.4 ± 15.0 l/min/kPa, even after correction for FRC. We conclude that the hypercapnic ventilatory response in healthy subjects is 2.3 fold higher in men than in women and that mechanical factors responsible for the conversion of inspiratory negative pressure into respired volume rather than differences in central chemical control are responsible for this difference. Dept. of Applied Physiology, Ruhr-Universität Bochum, D-44780 Bochum O 09-2 ISOLATED VENTILATED HEMOPERFUSED PORCINE LUNGS AS MODEL FOR RESPIRATORY RESEARCH D. A. Groneberg, C. Grosse-Siestrup, S. Dittrich, C. Peiser, A. Fischer Background: The rapidly evolving field of transplantation research with a focus on reperfusion injuries and graft rejection has gained importance since the methodology of organ preservation significantly limits graft survival. Numerous models of isolated perfused lungs have been established in the past years but limitations such as organ size, perfusate and ethical standards have restricted a wide-spread research in this area. Methods: A model of isolated ventilated autologouslyhemoperfused porcine lungs was established which encompasses the advantages of autologous blood as optimal perfusate and the pig as donor species. Results: The size and geometry of the porcine lung is more comparable to human conditions and various respiratory functions and blood parameters can easily be accessed in the present model. Stable organ function can be maintained over 3 hours with an amount of 500 to 1000 ml of autologous blood which is metabolically controlled via a dialysis system. Conclusion: The present model describes a new and economic approach for targeting respiratory functions in transplantation research by combining autologous blood as optimal perfusate with a well defined organ geometry.

lung size completely compensates for the observed gender differences. There was great diversity between strains for all of the lung function parameters studied, for example, the total lung capacity as well as the pulmonary diffusing capacity for carbon monoxide varied by 50 % and the static lung compliance by almost a factor of two among the strains. Little, but statistical significant variability was detectable for the dead space volume and the respiratory system resistance. There was no clear-cut evidence for any strain exhibiting either the smallest or the largest values for all parameters studied suggesting that there were no simple allometric relationships of lung size between the strains. Well established genealogical relationships among strains were obviously not reflected in phenotype similarities of pulmonary function. Therefore, these data strongly support a complex heritable genetic trait for pulmonary function. Generous funds from AstraZeneca were used to defray the cost of mice through the Mouse Phenome Project (The Jackson Laboratory, Bar Harbor, Maine, USA). GSF-Institute for Inhalation Biology, Institute for Experimental Genetics, 85758 München/Neuherberg, Germany O 09-4 EFFECTS OF ALPHA- AND BETA-ADRENERGIC STIMULATION ON THE DEVELOPMENT OF PULMONARY EDEMA AND PLEURAL EFFUSION IN RATS. B. Raßler, A. Tannapfel *, H.-G. Zimmer Continuous intravenous infusion of norepinephrine (NE; 0.1 mg/kg/ h) causes transient pleural effusion in rats. The first occurrence of the pleural effusion after 6h of stimulation was associated with an increase in total peripheral resistance (TPR). When NE was administered in combination with the alpha-adrenergic antagonist prazosin (PZ; 0.1 mg/kg/h), neither increase in TPR nor development of pleural effusion was observed (Raßler et al., Basic Res. Cardiol. 96: 265-273, 2001). In the present study, we investigated the effects of alpha- and betaadrenergic stimulation over time-intervals from 90min to 72h. Alpha-adrenergic stimulation (A) was performed using phenylephrine (PE; 0.5 mg/kg/h; n = 23) or NE plus propranolol (PR; 1.0 mg/kg/h; n = 30). For beta-adrenergic stimulation (B), we used isoproterenol (ISO; 0.024 mg/kg/h; n = 32) or NE plus PZ (n = 35). Saline-infused rats served as controls. Hemodynamic parameters such as right ventricular systolic pressure (RVSP) and TPR were obtained by heart catheterization in vivo. Additionally, we collected pleural fluid (PF) and excised lung tissue for histological analysis. Treatment A increased RVSP from 34 (controls) to 42mmHg and TPR from 0.43 to 0.64 mmHg·min·kg/l. After 6 to 24h of stimulation, moderate amounts of PF (0.85±0.21ml) were found, occurring preferably after 6 and 8h. Treatment B increased RVSP to 52mmHg. TPR decreased to 0.36 mmHg·min·kg/l. PF occurred after 16h of treatment and reached a peak of 4.28±1.13ml after 24h of stimulation. Lung histology revealed alveolar edema in animals treated with A over 8h, but not in B-treated animals. After 24h, pulmonary edema was also found in B-treated rats. It was less pronounced than in A, but was accompanied by severe inflammation. The results demonstrate that both alpha- and beta-adrenergic mechanisms contribute in a different way to the pulmonary fluid shifts induced by NE-infusion.

Dept. of Pediatric Pneumology and Immunology, Charite Campus Virchow, BMFZ Forum 4, Augustenburger Platz 1, D-13353 Berlin, Germany

Carl-Ludwig-Institute of Physiology and * Institute of Pathology, University of Leipzig, D-04103 Leipzig

O 09-3

O 09-5

VARIABILITY IN RESPIRATORY FUNCTION OF INBRED MOUSE STRAINS C. Reinhard, G. Eder, A. Ziesenis, H. Fuchs, J. Heyder, H. Schulz The purpose of the present study was to determine the strain specific phenotype variance of lung function parameters among common inbred laboratory mouse strains. In accordance with the ”Mouse Phenome Project” run by the Jackson Laboratory, lung volumes, lung mechanics and diffusing capacity of 16 males and 16 females of each strain (C3H/ HeJ, BALB/cByJ, C57Bl/6J, A/J, FVB/J, 129SV/ImJ and SWR/J) were determined in a standardized manner. The defined respiratory maneuvers for lung function testing were performed with a custom-made computer-controlled servo-ventilator in anesthetized animals. Sex differences within the strains were found in most (83 %) of the absolute lung function parameters. Usually, normalization to bodyweight or

PULMONARY EDEMA AND PLEURAL EFFUSION IN NOREPINEPHRINE-STIMULATED RATS – HEMODYNAMIC OR INFLAMMATORY EFFECT? B. Raßler, C. Reißig, M. Leicht, W. Briest, H.-G. Zimmer Stimulation with norepinephrine (NE) leads to pulmonary edema and pleural effusion in rats (Raßler et al., Basic Res. Cardiol. 96: 265-273, 2001). These pulmonary fluid shifts may result from pulmonary congestion due to the hemodynamic effects of NE and/or inflammation with an increase in vascular permeability. The contribution of these two factors were investigated in the present study. Female SpragueDawley rats received continuous intravenous infusion of either NE (0.1mg/kg/h) or saline (controls) over time-intervals between 20min and 72h. We measured left (LVSP) and right ventricular systolic pressure (RVSP) and total peripheral resistance (TPR) by heart catheter-

S 176 ization in vivo. Pulmonary edema was shown histologically. Inflammation was assessed by measuring the expression of interleukin (IL)-6 as one of the most potent acute-phase protein mediators in the serum and the pleural fluid (PF) using ELISA and in the lung tissue using Western Blotting. NE increased RVSP to 200% of control throughout all stimulation intervals, but had no effect on LVSP. After 6h of NE treatment, TPR increased significantly, and pleural effusion occurred for the first time. PF reached maximum amounts between 8 and 16h (2.2±0.3ml, controls <0.5ml) and disappeared within 48h of stimulation. The amount of PF correlated with TPR (r = 0.63, p<0.001). IL6 concentration increased from 80±30pg/ml (serum of controls) to 300 ±110 (serum) and 5500 ±1890 (PF) after 8h of NE stimulation. In the lung tissue, IL-6 concentration started to increase after 12h (190%) and reached 330% of the control value after 48h of NE treatment. We conclude that hemodynamic factors play an important role in the pulmonary reaction to NE treatment. An inflammatory reaction is involved, but seems to be a consequence rather than a cause of the pulmon-ary fluid shift. Carl-Ludwig-Institute of Physiology, University of Leipzig, Liebigstr. 27, D-04103 Leipzig O 09-6 LONG-TERM FOLLOW-UP OF HYPERCAPNIC VENTILATORY RESPONSES IN CONGENITAL CENTRAL HYPOVENTILATION SYNDROME T. Schäfer, C. Schäfer, M.E. Schläfke Respiratory CO2-sensitivity guarantees the acid base homeostasis of brain extracellular fluid during different states of behaviour and drives ventilation from birth on. In children with congenital central hypoventilation syndrome (CCHS), however, the HCVR during sleep and wakefulness is strongly reduced or missing. Studies on the plasticity of central chemosensitivity in animals and in children with CCHS have shown that stimulation on the basis of classical conditioning could increase the slope of the HCVR at least on a short-term scale. We wanted to assess whether the HCVR can recover spontaneously on a long-term scale. Nine children with congenital central hypoventilation syndrome have been followed up during 1 to 15 years (mean: 7.5 yrs) and compared to 9 healthy children measured once using the same technique. The hypercapnic ventilatory response has been measured using the steady-state technique during wakefulness with the children sitting comfortably in a semi-recumbent position, the infants lying supine. During the tests end-tidal PCO2 and PO 2 have continuously been measured. Ventilation has been quantified by means of calibrated respiratory inductive plethysmography, with the children breathing through the nose avoiding nose clip and mouthpiece. Mean slopes (in percent of mean resting ventilation) in CCHS were –0.08 ± 3.01 %/ Torr (range –6.5 to 6.0 %/Torr) versus 41.7 ± 11.4 %/Torr (range 21.6 to 59.1 %/Torr) in controls (P < 0.001, t-test). There were no significant age trends in all but one child, who showed an increase of 0.6 %/Torr per year (p < 0.05), but did not exceed 4.6 %/Torr or 6.7 ml/min/Torr/ kg. Mean resting PCO2 in CCHS was 44.2 ± 8.3 Torr versus 38.0 ± 6.3 Torr in control children (n.s.). There were no CO2-related changes in breathing frequency (23.3 ± 3.6 breaths per minute (bpm) during resting ventilation versus 23.0 ± 3.4 bpm at highest PCO2, p = 0.63, t-test for paired samples). We conclude that there is no spontaneous recovery of the HCVR in these children with CCHS and that they may be dependent on artificial ventilation or other ventilatory support in the long term. Dept. of Applied Physiology, Ruhr-Universität Bochum, D-44780 Bochum

about the relation of molecular mechanisms and the change in migration, it is necessary to translate the assumptions in physical models and mathematical equations. We use a first order differential equation where the cell, which is linked to the substrate via a ‘friction’ parameter, is moving under the influence of protrusive and retractive forces. Four different models of these forces are derived from spatially resolved shape changes at the front and at the rear part of the cell (leading edge [I], mean deformation [II], protrusions [III], protrusions and retractions [IV]). A second parameter additionally controls the effective influence of the forces on the cell centre. Furthermore, the parameters of the model dependent on the two typical states of a migrating cell which are ‘moving’ and ‘reorientation’. Bayesian probability theory and the maximum entropy principle are applied in order to find the best of the four models describing cell migration in view of the experimental data. In addition, Bayesian analysis assesses the quality of experimental information against the complexity of the model (e.g. the number of parameters) preferring simpler models, if the experimental data may not allow to draw complicated conclusions. This principle is called Occam’s Razor. Thus, we can now characterise cell migration with as much parameters as necessary for the best model. The technique applied to migrating MDCK-F cells confirms model [IV] with the inclusion of protrusive and retractive forces. In addition, all cells of this type show the same parameter of cell-substrate interaction and the parameter of force-coupling is related to a fractal length function indicating the efficiency of cell migration. In conclusion, the current approach is able to summarise data from migration experiments within a mathematical model, to judge about hypothesis of the migration mechanism in a quantitative manner and to detect quantitatively effects arising from particular molecular manipulations. 1

Institut für Physiologie, TU-Dresden, Fetscherstraße 74, D-01307 Dresden, 2Institut für Physiologie der Universität Würzburg, 3MaxPlanck-Institut für Plasmaphysik, Garching, Germany O 10-2 OXIDATIVE STRESS ACTIVATES HETEROLOGOUSLY EXPRESSED rClC-2 IN XENOPUS OOCYTES G. Henke, C. Böhmer, S.M. Huber, F. Lang Oxidative stress leads to marked shrinkage of hepatocytes, an effect due to electrolyte release through K +-and/or Cl- channels. Oxidative stress of erythrocytes activates inwardly rectifying Cl- channels. The present study has been performed to test, whether the inwardly rectifying Cl- channel ClC-2 is sensitive to oxidation. To this end, rat ClC-2 was heterologously expressed in Xenopus oocytes and the ClC2 induced currents determined by dual electrode voltage clamp. In confirmation of earlier studies, the ClC-2 induced current displayed inward rectification with larger currents at hyperpolarized (-789 ± 166 nA at -120 mV, n = 24) than at depolarized cell membrane (+135 ± 19 nA at +40 mV, n = 24). Furthermore, similar to earlier reports, a reduction of extracellular osmolarity by removal of 100 mM sucrose (which before replaced 50 mM NaCl) stimulated the current in ClC-2 expressing oocytes. Oxidation by addition of 1mM t-butyrylhydroxyperoxide (tBHP) to the isotonic bath solution resulted within 15 minutes in a more than twofold activation of ClC-2 currents (to – 1624 ± 211 nA, n = 24 at –120 mV and to 467 ± 60 nA, n = 24 at +40 mV). The tBHP-induced rClC-2 activation was fully reversed upon administration of 1mM Dithiothreitol (DTT), a reducing substance. H2O-injected oocytes treated with tBHP displayed a comparably small outwardly rectifying current (-56 ± 14 nA, n=4 at -120 mV and 143 ± 25 nA at +40 mV). In conclusion, ClC-2 is not only activated by osmotic cell swelling but as well by oxidation. The redox sensitivity of ClC-2 may participate in the altered regulation of cell volume following oxidative stress.

O 10-1 MATHEMATICAL MODELLING OF CELL MIGRATION WITH BAYESIAN ANALYSIS P. Dieterich1, A. Schwab2, H.-J. Schnittler1, R. Preuss3 Cell migration plays an important role in (patho-)physiological processes such as immune defence, wound healing or formation of tumour metastases. On a molecular level, cell migration requires a coordinated dynamic reorganisation of the cytoskeleton, cell-substrate interactions and the activity of ion channels and transporters. Typical migration experiments encompass specific modifications of the cellular migration machinery. The resulting effects are observed using timelapse video microscopy. In order to validate quantitatively hypotheses

Physiologisches Institut der Universität Tübingen, Gmelinstr. 5, 72076 Tübingen, Germany O 10-3 OXIDATION INDUCES A CL--DEPENDENT CATION CONDUCTANCE IN HUMAN RED BLOOD CELLS C. Duranton, V. Tanneur, S.M. Huber, F. Lang Oxidative stress induces complex alterations of membrane proteins in red blood cells (RBCs) eventually leading to hemolysis. To study changes of membrane ion permeability induced by oxidative stress, whole-cell

S 177 patch-clamp recordings and hemolysis experiments were performed in control and oxidised human RBCs. Control RBCs exhibited a small cation-selective whole-cell conductance (236±38 pS; n=8) which was highly sensitive to the external Cl - concentration: replacement of NaCl in the bath by Na-gluconate, induced an increase of this cation conductance by about 85%. Exposing RBCs to t-butylhydroxyperoxide (1mM for 10min) induced a two-fold increase in this cation conductance which was further stimulated after replacement of extracellular Cl- by gluconate, Br-, I- or SCN-. In addition, lowering the ionic strength of the bath solution by isosmotic substitution of NaCl by sorbitol activated the cation conductance. The Cl --sensitive and oxidationinduced cation conductance was Ca 2+-permeable, exhibited a permselectivity of Cs+ > K+ > Na+ = Li+ >> NMDG+, and was inhibited partially by amiloride (1mM) and almost completely by GdCl3 (150µM), but was insensitive to TEA, BaCl2, NPPB, flufenamic acid or quinidine. DIDS (100µM) reversibly blunted the activation of the cation conductance by removal of external Cl-. Oxidation induced hemolysis in NaClbathed human RBCs. This hemolysis was attenuated by amiloride (1mM) and inhibited by replacement of bath Na+ by the impermeant cation NMDG+. The Na+- and Ca2+-permeable conductance might be involved in hemolytic diseases induced by elevated oxidative stress such as glucose-6-phosphate dehydrogenase deficiency.

37°C as determined by FACS-analysis of FITC-Annexin stained cells. Infection with P. aeruginosa 762 caused depolarisation of the mitochondrial membrane potential (∆ψm) (42.6% after 30 min). In addition, upregulation of CD95-receptor on the surface of the infected cells was observed (23% after 30 min). In contrast, when the infection was performed at 22°C, P. aeruginosa failed to induce significant apoptosis (7.8% Annexin-positive cells after 30 min of infection) and mitochondrial alterations (14.9% cells with low ∆ψm after 30 min of infection). Longer infection times at 22°C led only to a small increase in the amount of apoptotic cells (14.4% Annexinpositive cells after 2h) whereas infection at 37°C induced complete cell death within 1h. The loss of the apoptotic response at 22°C when determined after 30 min was not due to reduced bacterial growth or to diminished secretion of bacterial exotoxins. Most importantly, only a small upregulation of CD95 on the surface of the epithelial cells was observed when the infection was performed at 22°C (7.1% after 30 min). From our data we conclude that P. aeruginosa 762 induced apoptosis is dependent on the temperature during infection and that cellular pathways leading to CD95-upregulation are inhibited at low temperature Dept. of Physiology I, University of Tübingen, Gmelinstrasse 5, D72076 Tübingen, Germany.

Department of Physiology I, University of Tübingen, Gmelinstr. 5, 72072 Tübingen, Germany O 10-6 O 10-4 PSEUDOMONAS AERUGINOSA INDUCED SWELLING OF HOST CELLS C. Böhmer, S. Fillon, K.S. Lang, E. Gulbins, V. Jendrossek, F. Lang Pseudomonas aeruginosa induce CD95 mediated epithelial cell death, an effect critical for the course of the infection [Grassmé et al. Science 290: 527, 2000]. As one of the hallmarks of apoptotic cell death is cell shrinkage, and as cell shrinkage is one of the known triggers of apoptotic cell death, the influence of P. aeruginosa on cell volume has been tested. The experiments have been performed in the Chang human conjunctiva cell line. Cell volume was estimated from the forward scatter in FACS analysis. Surprisingly, the P. aeruginosa strain PAOI leads to cell swelling by 32 ± 8 %, an effect paralleling annexin binding. The P. aeruginosa induced cell swelling was not sensitive to neither Na + /H + exchange inhibitor HOE 694 (10 µM) nor Na +,K +,2Cl cotransport inhibitor furosemide (100 µM) but was significantly inhibited by amiloride (100 µmol/l), which at those concentrations inhibits both, Na+/H+ exchange and Na+-channels. None of those drugs modified cell volume significantly in the absence of P. aeruginosa. The Cl- channel inhibitor NPPB (100 µM) significantly increased cell volume but did not prevent further swelling in the presence of P. aeruginosa. Interestingly, in the presence of P. aeruginosa, Chang cell volume was virtually identical in the presence and absence of NPPB pointing to inhibition of NPPB-sensitive Cl- channels and/or dissipation of the anion gradient. In conclusion, the exposure to P. aeruginosa PAOI leads to amiloride sensitive swelling of Chang epithelial cells, an effect presumably participating in the pathophysiology of P. aeruginosa infection. Department for Physiology, University of Tübingen, Gmelinstr. 5, D 72076 Tübingen O 10-5 CELL DEATH OF HUMAN EPITHELIAL CELLS AFTER INFECTION WITH PSEUDOMONAS AERUGINOSA IS DEPENDENT ON THE TEMPERATURE DURING INFECTION S. Fillon, B. Puttkammer, I. Müller, F. Lang, V. Jendrossek P. aeruginosa, a gram-negative facultative pathogen, causes severe infections particulary in cystic fibrosis or immunocompromised patients. We showed earlier that P. aeruginosa induces apoptosis of human epithelial cells. Efficient induction of apoptosis was dependent on functional CD95-receptor and CD95-ligand and was essential for survival of mice infected intranasally with P. aeruginosa. Since P. aeruginosa belongs to bacteria which are still growing at low temperature, the aim of the present study was to define cytotoxic effects of the clinical isolate P. aeruginosa 762 and to study the effect of temperature on signal transduction during apoptosis. Infection of human conjunctiva epithelial Chang cells with P. aeruginosa 762 resulted in a rapid induction of apoptosis. 64.8% of the cells underwent apoptosis within 30 min at

CALCIUM-INDUCED PRODUCTION OF REACTIVE OXYGEN SPECIES MODULATES MITOCHONDRIAL RESPIRATORY CHAIN REDOX EQUILIBRIUM IN LIVING NEURONS T. Gillessen, C. Grasshoff, D. Gutzeit, L. Szinicz The formation of reactive oxygen species (ROS) has been demonstrated convincingly under conditions associated with neuronal excitotoxicity. This mechanism may be involved in the collapse of the mitochondrial membrane potential (∆Ψm) and subsequent bioenergetic failure induced by sustained stimulation of glutamate receptors in excitotoxicity. Most ROS exhibit a high reactivity and are assumed to react close to the site of their production. Since the mitochondrial respiratory chain is an important source of ROS formation, redox reactions with inner mitochondrial membrane constituents could disturb the function of proteins within the inner mitochondrial membrane. We studied in cultured rat cortical neurons whether the redox equilibrium within the respiratory chain is affected by glutamate receptor-mediated calcium loading and subsequent calciumstimulated ROS formation. Intracellular calcium [Ca2+]i, the production of superoxide and ∆Ψm were assessed using calcium green-5N, hydroethidin, and tetramethylrhodamineethylester, respectively. The respiratory chain redox equilibrium was monitored noninvasively by measuring mitochondrial flavoprotein autofluorescence. Cytoplasmic calcium uptake resulted in increased superoxide formation and a collapse of ∆Ψ m that partially persisted even under blockade of the mitochondrial permeability transition pore. This ∆Ψm decrease was sensitive to treatment with ROS scavengers, resulting in complete prevention of mitochondrial depolarization. Moreover, calcium loading induced a shift of the respiratory chain redox state to a more reduced equilibrium, that could be inhibited by blockade of ROS formation or ROS scavenging, indicating that electron flux within the respiratory chain can be inhibited by ROS. Pharmacological experiments revealed that both Ca2+ and ROS are implicated in the modulation of the respiratory chain electron transfer. Institut fuer Pharmakologie und Toxikologie, BSW, Neuherbergstrasse 11, 80937 München, Germany O 11-1 FIBROBLAST GROWTH FACTOR-2 STIMULATES CELL LOCOMOTION AND PROMOTES THE RELOCATION OF CALCIUM SENSITIVE K+ CHANNELS IN MIGRATING CELLS W. Keßler, B. Hinkes, M. Römer, C. Schulz, A. Wulf, M. Gekle, A. Schwab Cell migration plays an important role in immune defense, epithelial wound healing and formation of tumor metastases. Several ion channels and transporters whose activity is required for cell locomotion, are distributed in a characteristic way in migrating cells. The aim of the present study was to define a tool that allows the study of the dynamic redistribution of Ca2+ sensitive K+ channels (hIK1) channels in migration and to visualize hIK1 channel proteins in living cells. We

S 178 measured migration of single renal epithelial (MDCK-F) cells in the absence and presence of fibroblast growth factor 2 (FGF-2). The subcellular distribution of hIK1 was shown by indirect immunofluorescence (fixed cells) or by fusing hIK1 to a green fluorescent protein (GFP). ERK 1/2 phosphorylation was shown by Western blot analysis with phosphospecific antibodies. The rate of migration rises by up to 71% in the presence FGF-2. The increase in velocity is maximal after 6 min. At that time ERK 1/2 phosphorylation has doubled. FGF-2 does not stimulate migration when ERK 1/2 phosphorylation is inhibited. Blockers of hIK1 channels attenuate the stimulatory effect of FGF-2. However, migration is not slowed down to the same extent as by hIK1 blockade alone. FGF-2 treatment also increases the fraction of cells in which hIK1 channel proteins are concentrated at the cell front. In living migrating cells GFP-labeled hIK1 channel proteins are transported to the cell front by vesicle transport. Our results indicate that FGF-2 is a potent chemokine which modulates migration and subcellular distribution of hIK1 channels. Therefore, it is a valuable tool to visualize the intracellular transport of GFP-labeled hIK1 channel proteins to the plasma membrane of migrating cells.

Cre/loxP-technique. To ensure that channel activity is completely abolished, the pore coding exon was targeted. Germ line transmission of the mutation produced viable heterozygous pubs. Unexpectedly, mating of heterozygous pairs with constitutive knock out resulted in viable mutants that were homozygous for the BK channel deletion. Since the gene deletion of the BK channel smooth muscle βl subunit led to severe hypertension (1, 2), it might be speculated that the general knock out of the BK channel α subunit produces a much more serious phenotype, even when the deletion does not compromise viability. (1) Plueger S et al. (2000) Circ Res 87, E53-E60. (2) Brenner R et al. (2000) Nature 407, 870-876.

Physiologisches Institut, Röntgenring 9, 97070 Würzburg, Germany

bFGF AND VEGF INDUCE UP-REGULATION OF THE INTERMEDIATE CA2+-ACTIVATED K+ CHANNEL (IK1) IN HUMAN ENDOTHELIUM I. Grgic, S. Brakemeier, I. Eichler, A. Kersten, J. Hoyer, R. Köhler Objective: Ca 2+-activated K+ channels (KCa) are important regulators of endothelial function and have also been proposed to promote cell proliferation by enhancing membrane potential driven Ca2+-influx. Here we tested whether the IK1 is involved in endothelial cell proliferation in response to the angiogenic factors bFGF and VEGF. Methods and Results: IK1 function and expression were determined by use of the patch-clamp technique and quantitative real time RT-PCR in human umbilical vein endothelial cells. A 48h exposure to bFGF (50 ng/ml) or VEGF (50 ng/ml) resulted in a significant increase in CTX and clotrimazole-sensitive IK1-currents (bFGF: 1.92 ± 0.43 pA/pF, p<0.01; VEGF: 1.79 ± 0.67 pA/pF, p<0.05 vs control: 0.36 ±0.09 pA/ pF), which was paralleled by a sixfold and fourfold increase in IK1 expression, respectively. Inhibition of PKC by staurosporine (10 nM) and of MAPKK by PD98059 (25 µM) prevented bFGF and VEGFinduced up-regulation of IK1. Moreover, bFGF and VEGF-induced cell proliferation was almost completely abolished in the presence of CTX (100 nM) or clotrimazole (1 µM). Conclusions: The angiogenic factors bFGF and VEGF stimulate expression and function of IK1 in endothelial cells. This up-regulation of IK1 seems to promote mitogen-induced cell proliferation and might therefore represent a new therapeutic target to prevent angiogenesis.

O 11-2 BIOPHYSICAL SIGNATURE OF Ca2+ ACTIVATED Cl- CHANNEL IN RENAL INNER MEDULLARY COLLECTING DUCT (mIMCDK2) CELLS S.H. Boese, M.A. Gray, N.L. Simmons Mouse renal inner medullary collecting duct (mIMCD-K2) cells express a Ca2+-activated Cl - conductance (CaCC) involved in transepithelial (Na+) Cl- secretion (Boese et al., J. Physiol. 523, 325-338, 2000). We used whole-cell as well as cell-attached and inside-out patch-clamp current recordings to characterise the biophysical fingerprint of CaCC in mIMCD-K2 cells. In whole cell recordings current activation started at [Ca2+]i levels as low as 50-80nM and maximal activation was achieved at levels of 800-1000nM. However, the kinetics of the stimulated current was dependent on [Ca2+]i. At intermediate levels (100-600nM) of [Ca 2+]i, currents displayed marked outward rectification with timedependent activation of outward current and inactivation of inward current. At maximal current activation, minimal time-dependent behaviour was present and the I/V plot was almost linear. In cellattached recordings application of the Ca2+-ionophore ionomycin (0.11µM) activated channel activity. Characterisation of single channel activity in inside-out patches, bathed in symmetrical NMDG-Cl solutions (2mM cytosolic Ca2+) gave a single channel conductance of 30.8±2.7pS at positive membrane potentials (mean ± SEM, n=9) and 19.2±2.6pS at negative membrane potentials (n=9). Channel open probability (Po) was 0.69±0.10 during membrane depolarisation and 0.38±0.08 during membrane hyperpolarisation (n=4). In conclusion, the single channel data alone cannot explain the properties of the fully active macroscopic CaCC in mIMCD-K2 cells. Even at high [Ca2+]i no loss of strong outward rectification could be observed at the single channel level. These data suggest that excised patch recordings may lead to changes in the biophysical properties of CaCC , perhaps due to the loss of a regulatory protein. Alternatively, another CaCC channel could be present in whole cell recordings, that is absent in excised recordings. Department of Physiological Sciences, University Medical School, Newcastle upon Tyne, NE2 4HH, UK O 11-3 DELETION OF THE Ca2+-ACTIVATED K+ CHANNEL OF THE BKTYPE IN MICE C. Arntz1, S. Kühbandner, S. Kamm, R. Feil, M. Sausbier, P. Ruth Ca2+- and voltage dependent K+ channels of the BK type are assumed to play important roles as negative feedback regulators that reset membrane potential in response to local increases of Ca2+. Native BK channels are tetrameric complexes of the pore forming α subunit that are mostly assembled with modulatory β subunits. BK channels are expressed in neurones, endocrine, endothelial and smooth muscle cells where they serve as major targets for the cAMP and the NO/cGMP pathway. In order to decipher the biological role of the BK channel in different cell types and to evaluate its contribution to cAMP and cGMP mediated cellular effects we deleted the gene of the BK channel α subunit in mice both constitutively and conditionally by using the

1

lnstitut für Pharmakologie und Toxikologie, Technische Universitaet, 80802 München, Germany, new address: Pharmazeutisches Institut der Universitaet, Lehrstuhl Pharmakologie, 72076 Tuebingen, Germany O 11-4

R. Köhler, Universitätsklinikum Benjamin Franklin, Hindenburgdamm 30, 12200 Berlin, Germany O 11-5 SHEAR STRESS-INDUCED UP-REGULATION OF CA2+-ACTIVATED K+-CHANNELS IN HUMAN ENDOTHELIUM S. Brakemeier, A. Kersten, I. Grgic, I. Eichler, A. Zakrzewicz, H. Hopp, R. Köhler, J. Hoyer Alterations of vascular Kca have been proposed to be associated with hypertension. In the endothelium, hyperpolarizing KCa-currents provide the electrochemical driving force for Ca2+-influx, which stimulates the synthesis of vasodilatory factors. We tested the hypothesis, whether expression and function of endothelial K Ca are modified by alterations in laminar shear stress (LSS). Human umbilical vein endothelial cells (HUVEC) were exposed to LSS of 5 or 15 dyn/cm2 for 4, 24, or 48 h. Expression of the KCa genes hIK1 and hSK3 were determined by quantitative real-time RT-PCR. Signal transduction mechanisms were investigated using MAPK and PKC inhibitors. KCa function was determined in whole-cell patch-clamp experiments and by measurements of agonist-induced hyperpolarization.A 3-fold (p<0.05) and 16-fold (P<0.01) increase in hIK1-expression was observed after 24 h and 48 h exposure to LSS of 5 dyn/cm2, respectively. Similar to hIK1, a 2-fold and 8-fold increase in hSK3-expression was detected after 24 h and 48 h of LSS of 5 dyn/cm2 (p<0.05). KCa whole-cell currents as well as hyperpolarization responses to ATP and 1-EBIO were significantly increased. Inhibition of the p38 and ERK1/2 MAPK pathways did not prevent KCa gene upregulation after LSS exposure. In contrast, inhibition of PKC completely abolished up-regulation of KCa, suggesting that LSS induces KCa expression in a AP-1 dependent manner via stimulation of PKC and downstream JNK-pathways. Exposure to LSS

S 179 upregulates expression and function of endothelial KCa . This might represent a novel important mechanism in endothelial adaptation to altered hemodynamics and in hypertension. S. Brakemeier, Dept. of Nephrology, UKBF, Hindenburgdamm 30, 12200 Berlin, Germany

rounded by an area of relatively decreased lCBF indicating local cerebral edema and/or steal. In summary, the adenoviral mediated expression of VEGF in a defined brain region induces local angiogenesis which translates into increased ICBF. However, this has as side effect a reduction of lCBF in adjacent brain tissue. Department of Physiology and Pathophysiology, University of Heidelberg, Im Neuenheimer Feld 326, D-69120 Heidelberg, Germany

O 11-6 CALCIUM MODULATED POTASSIUM CHANNELS IN ENTERIC SENSORY NEURONS W Kunze, M Mueller The excitability of intrinsic enteric sensory neurons (ESNs) is strongly modified by calcium activated potassium currents. To gauge the roles of individual KCA channels, we have recorded from single channels from guinea-pig sensory neurons in intact myenteric ganglia using pipettes filled with potassium-rich saline. K channels had conductances of 200, 159-229 (median, 95% Cl) (maxi-K Ca) and 59, 31-92 (IKCa) pS; the maxi-KCa but not the IKCa channel was voltage dependent. Maxi-K Ca PO, measured in cell attached patches increased with patch depolarization, voltage of half activation: -43 (95% Cl = -33 - -26) mV and a voltage sensitivity of 11.2, 10.2-12.8 mV). Both channel types were sensitive to calcium applied to inside out patches. [Ca2+] of half activation was 4 µM (adjusted for patch potential = -65 mV) and 0.9 µM for maxi−Κ Ca and IKCa channels respectively. The Hill coefficients were 1.9 and 1.7. Also, the maxi-KCa but not the IKCa opened with positive pressure (Pr of half activation = 23 mBar, sensitivity =12 mBar). 50 nM charybdotoxin applied to outside-out patches silenced both channels types, 50 nM iberiotoxin suppressed only maxi-KCa channels. MaxiKCa dwells were represented by 3 closed and 2 open time exponentials, IKCa required 3 closed and 1 open time exponentials to fit the dwell time distribution. Rate constants of optimal Markov chain kinetic models were estimated using the maximum likelihood method. Charybdotoxin, iberiotoxin, pressure, voltage, and calcium sensitive transitions were located in the models. Unitary currents for both channels were recorded at RMP with P0 ≤ 0.2, suggestive of a contribution to the resting current. Ensemble averages of channel currents after single action potentials demonstrated that the IKCa significantly contributes to the slow afterhyperpolarization (onset 80,65-112 ms, duration 2, 0.3-3.2 s). The maxi-KCa channel contributes to the fast AHP and to the late component of the compound slow afterhyperpolarization evoked by action potential bursts. Low frequency presynaptic stimulation induced long term excitation (LTE), which when repeated could produce depolarizing inactivation. Induction of LTE was accompanied by complete suppression of IKCa and Maxi-KCa opening. Our results suggest that the ESN KCa channels would be outstanding targets for modulating intestinal sensory and reflex excitability. Zentrum Medizinische Forschung (ZMF), Allgemeine Chirurgie, Waldhörnlestr. 22, 72072 Tubingen, Germany O 12-1 EFFECTS OF ADENOVIRAL MEDIATED TRANSFECTION OF HUMAN VEGF-165 CDNA INTO RAT BRAINS ON LOCAL CEREBRAL BLOOD FLOW (lCBF) J. Vogel, C. Hörner, C. Haller, W. Kuschinsky Following cerebral ischemia survival of neurons could be improved by increasing blood flow in areas surrounding the ischemic core. Therefore we tested the hypothesis that lCBF can be increased through local angiogenesis mediated by VEGF. A recombinant adenovirus carrying the VEGF-165 cDNA was stereotactically injected into the caudate nucleus of 48 halothane-anesthetized rats (1.035 * 10 7 to 1.035 * 109 pfus (4 rats/dose)). NaCl and adenovirus carrying the cDNA for Bgalactosidase were used as controls. The animals were re-anesthetized 11 days later. Blood brain barrier permeability was assessed by i.v. injection of Evans blue. LCBF was measured using the quantitative iodo-[14C]antipyrine technique. Rats that had received the highest dose of VEGF (1.035 * 109 pfu) died from cerebral edema whereas all other animals showed no apparent effects. After local transfection with adenovirus carrying the VEGF-cDNA a dose-dependent increase in vessel density was found at the injection site. Local Evans blue extravasation showed an increased permeability of the blood brain barrier which was not apparent with NaCl or virus carrying the cDNA for β-galactosidase. The 14C-iodantipyrine autoradiography revealed a dose-dependent, up to 6-fold increase in lCBF. The hypervascular area was sur-

O 12-2 ANGIOPOIETIN-2 AND ITS RECEPTOR TIE-2 ARE CONVERSELY REGULATED BY SHEAR STRESS: IMPLICATIONS FOR ANGIO-GENESIS A. Zakrzewicz, M. Bongrazio, O. Baum, C. Baumann, AR Pries. Introduction: The architecture of microvascular networks is continuously adapted to changing blood-flow conditions. Since the angiopoietin/tie-2 system plays a key role in vessel remodelling, this study was done to investigate a possible correlation between the angiopoietin/tie-2 system in endothelial cells and wall shear stress. Methods: First passage human umbilical vein endothelial cells (HUVEC) were exposed to shear stress (0.6-6.0 dyn/cm2) for up to 48 h using a cone-and-plate apparatus. Shear stress-dependent expression of angiopoietin-1, -2, and tie-2 was investigated by semi-quantitative RT-PCR, northern blot, and Western blot. Results: Under shear stress (6 dyn/cm2) angiopoietin-2-mRNA expression is progressively downregulated for at least 48 h. Above a threshold level of about 1 dyn/cm2, this effect correlates with the wall shear stress applied. Following flow-arrest, expression of angiopoietin-2 is back to control levels after two hours. In contrast, expression of tie-2-mRNA is significantly enhanced by shear stress. Both, shear stress-induced downregulation of angiopoietin-2 as well as up-regulation of tie-2, were confirmed on the protein level. Expression of angiopoietin-1 showed no modulation by shear stress. Conclusions: Angiopoietin-2, as an inhibitor of Tie-2, will block Tie-2-effects under no-flow conditions, while expression of Tie-2 is itself turned to lower levels. Activation of Tie-2 is known to stabilize vessel walls by recruitment of perivascular cells and its inhibition induces loosening of the vessel wall. Therefore, non-perfused vessels during sprouting angiogenesis would be more succeptible to angiogenic factors, whereas well perfused vessels during collateralization, could be stabilized. In summary, modulation of the angiopoietin/tie-2 system by wall shear stress is a suitable mechanism to mediate vessel adaptation to flow. Supported by DFG (Za 184/1-2) Dept. of Physiology, Freie Universität Berlin, Arnimallee 22, 14195 Berlin. O 12-3 THE CORONARY ENDOTHELIUM-DERIVED HYPERPOLARIZING FACTOR (EDHF) MODULATES INTERENDOTHELIAL GAP FUNCTIONAL COMMUNICATION R.P. Brandes, R. Popp, R. Busse, I. Fleming A role for gap junctional communication between vascular cells has been implicated in the endothelium-derived hyperpolarizing factor (EDHF)-mediated relaxation and conducted vasodilatation in many vascular beds. Here, we assessed the mechanisms by which the EDHF, generated by a cytochrome P450 2C (CYP 2C) epoxygenase in porcine coronary endothelial cells, controls inter-endothelial coupling. Coupling was determined by measuring the electrical capacitance as well as Lucifer Yellow dye coupling between cultured endothelial cells. In porcine coronary endothelial cells, bradykinin, which elicits the generation of EDHF, evoked a biphasic effect consisting of a transient increase in coupling followed by a sustained uncoupling. The initial phase was sensitive to the CYP 2C9 inhibitor sulfaphenazole, which prevents the generation of EDHF. The subsequent gap junction uncoupling was preceeded by activation of Erk 1/2, and an Erk 1/2-mediated phosphorylation of connexin 43 and could be prevented by inhibiting Erk 1/2 activation. In human endothelial cells, which do not produce significant amounts of the EDHF, bradykinin elicited only an Erkl/2mediated attenuation of inter-cellular communication. In these cells however, the CYP 2C9 product and putative EDHF, 11,12epoxyeicosatrienoic acid (11,12-EET), exerted a dual effect on coupling similar to that of bradykinin in porcine coronary endothelial cells. In addition, exogenous application of 11,12-EET to endothelial cells enhanced intracellular cyclic AMP levels. Liberation of cyclic

S 180 AMP from a caged analogue markedly increased the intra-endothelial dye transfer. In coronary endothelial cells Rp-8-cpt-cAMPS, a protein kinase A inhibitor, prevented the initial bradykinin-induced increase in gap junctional communication between coronary endothelial cells. These results indicate that an agonist-activated, CYP 2C9-/EDHF synthase-dependent pathway increases inter-endothelial gap junctional communication via the production of EET and cyclic AMP. The subsequent uncoupling of endothelial cells can however be attributed to the activation of Erk1/2. Thus, EDHF activates both stimulatory and inhibitory protein kinase pathways which control gap junctional communication. Institut für Kardiovaskuläre Physiologie, Klinikum der J.W. GoetheUniversität, Theodor-Stern-Kai 7, 60596 Frankfurt am Main, Germany O 12-4 DEFICIENCY OF CX40 IS ASSOCIATED WITH ARTERIAL HYPERTENSION INDEPENDENTLY OF NO, EDHF AND ANGIOTENSIN-II C. de Wit, F. Roos, U. Pohl The important role of intercellular communication via gap junction forming connexins (Cx) in the vascular system is highlighted by our previous findings that the loss of one member of the connexin family, namely Cx40, is associated with hypertension in anesthetized animals. To further corroborate these findings we extended our measurements to conscious animals of both genders and investigated possible mechanisms that may lead to hypertension. To this end, arterial pressure was measured in conscious animals via an arterial line and the effect of vasoactive compounds was tested. To characterize endothelial function in these mice, acetylcholine (ACh) was applied systemically before and after inhibition of endothelial NO-synthase. Cx40-deficient mice exhibited a significantly elevated pressure in the conscious state in both genders (male: 117±10 vs. 92±4 mm Hg, female: 115±7 vs. 97±7 mm Hg, n=6 each group, p < 0.05 in both genders). In contrast, heart rates were similar in these 2 genotypes. Despite blockade of NOsynthase (L-NA 100 mg/kg i.v.) or increasing dosages of an AT 1receptor antagonist (candesartan, up to 10 mg/kg i.v.), the pressure differences between genotypes were still observed. Systemically applied boli of ACh and SNP were equally potent and effective to induce pressure decreases in Cx40–/– and wildtype animals before as well as after L-NA. We conclude that the deletion of Cx40 not only alters conduction along arterioles, as shown previously, but is also associated with arterial hypertension. This hypertension is unlikely to be caused by the action of angiotensinII and not related to an altered efficacy of NO or other endothelial factors. We propose that Cx40 has a significant role in the control of peripheral vascular resistance and that defects in cell-coupling within the vascular tree should be regarded as a possible cause of hypertension in humans. Physiologisches Institut, LMU, Schillerstraße 44, 80336 München O 12-5 THE INHIBITORY EFFECT OF NO ON GAP JUNCTION COUPLING IS SPECIFICALLY RELATED TO CONNEXIN 37 P. Kameritsch, U. Pohl Endothelial and smooth muscle cells are rich of gap junctions which are formed by connexins. Previous studies have shown that NO reduces cell-to-cell coupling in endothelial cells, and we now investigated whether one of the connexins is specifically targeted. Therefore, HeLa cells that do not have endogenous Cx were transfected with either Cx37 or Cx43. Gap junctional communication (GJC) was analyzed by injecting a fluorescent dye (Alexa Fluor 488) into single cells (n=123) of subconfluent monolayers and determining the number of stained adjacent cells. This was normalized to the total number of cells in the microscopic field (% coupled cells). 10 min application of the NO donor SNAP (2 µM) did not affect GJC of HeLa cells expressing Cx43 (70% coupling). HeLa cells expressing only Cx37 showed such a low basic coupling that inhibitory effects of NO were hardly detected. In contrast, in mixed cultures of Cx37 and Cx43 cells (identified by different membrane staining), 77% of the cells were coupled under control conditions (after injection of Cx37). This coupling was reduced to 10% by SNAP, whereas the coupling of the Cx43 cells (70% after injection in Cx43 cells) in the same dishes remained unaffected. To

analyze, whether NO acts via an increase of cGMP or alternative pathways, cells were pretreated with the inhibitor of the soluable guanylyl cyclase, ODQ (10µM). In the presence of ODQ, the SNAPeffect was reduced by about 50%. Pretreatment with the cGMP elevating compounds CNP and ANP (1µM) did affect GJC of Cx37/Cx43 cells. These results suggest, that NO inhibits GJC by a specific influence on Cx37. This NO effect is only partially mediated by cGMP. To reveal the importance of other signaling pathways – like direct phosphorylation of Cx – further experiments have to be done. Physiologisches Institut der Universität München, Schillerstr. 44, 80336 München O 12-6 INVESTIGATION OF GROWTH FACTOR FUNCTION DURING VASCULARIZATION IN EMBRYONIC STEM CELLS D. Wenzel, *S. Kazemi, P. Sasse, E. Kolossov, *A. Reible, *K. Addicks, J. Hescheler, *W. Bloch, B.K. Fleischmann Embryonic vascularization comprises different processes such as proliferation, migration, differentiation and tube-formation of endothelial cells. To date little is known about morphogenetic changes of endothelial cells and the molecular mechanisms involved occurring during early stages of embryonic development We have therefore established stably transfected mouse embryonic stem (ES) cell lines, where the endothelial-specific PECAM promoter drives the expression of the live reporter enhanced green fluorescent protein (EGFP). This approach enables investigation of morphogenetic changes and related signalling cascades in endothelial cells during early embryonic development. ES cells are differentiated into embryoid bodies (EBs) applying the hanging drop technique. Morphogenetic changes of endothelial cells in the presence of key regulatory molecules, i.e. fibroblast growth factor (FGF) and vascular endothelial growth factor (VEGF) are monitored employing time lapse microscopy. Three days post plating EGFP positive cells are detected in the EBs. Their endothelial nature is confirmed with PECAM-1 and Flk-1 immunostaining. At first, clusters of angioblasts are predominant which later develop into elongated network-like structures. Time lapse microscopy reveals a surprisingly high rate of vascular remodelling in clusters of angioblasts. VEGF induces proliferation and pronounced sprouting of angioblasts whereas FGF leads to stabilization of preformed endothelial structures. We conclude that the ES cell system in combination with endothelialspecific EGFP expression is a valid tool to investigate early events of vascular differentiation in vitro. We conclude that VEGF plays an important role for early vessel formation whereas FGF is crucial for endothelial survival. Departments of Neurophysiology and *Anatomy I, University of Cologne, Robert-Kochstr. 39, D-50931 Koeln O 13-1 THE EXPRESSION OF NEUROKININ 1 (NK1) RECEPTORS ON CULTURED DRG NEURONES OF THE ADULT RAT IS SIGNIFICANTLY UPREGULATED BY PROSTAGLANDIN E 2 (PGE2) INDUCED ACTIVATION OF PKA A. Scholze, G. Segond von Banchet, H.-G. Schaible In a previous study we have shown a pronounced upregulation of NK1 and sst4 receptors in the lumbar DRG neurones in a model of antigeninduced arthritis (AIA) in one knee joint of the rat. The expression of NK1 receptors was markedly upregulated only in the acute phase of the AIA ipsi- and contralateral to the inflammation. The upregulation of sst4 receptors was also bilateral but persistent in the chronic phase of AIA. We have provided evidence that PGE2 could be one of the mediators that may be responsible for the upregulation of NK1 receptors in DRG neurones. The aim of this study was to identify intracellular pathways that may be involved in the PGE2 induced upregulation of NK1 receptors. DRG neurones from all spinal levels were removed from normal Wistar rats and cultured overnight in Ham‘s F12 medium. The following two days every two hours PGE 2, somatostatin (SOM), H-89 (PKA inhibitor) or PGE2 together with H-89 was added to the cultures. An anti-NK1 antibody was used to label the NK1 receptors on the neurones. As already shown, PGE 2 enhanced the proportion of neurones expressing the NK1 receptor from 10 to 40 %. When H-89 was added together with PGE 2 to the neurones the upregulation of NK1 receptors was not present. Preliminary data even

S 181 suggest that H-89 or SOM (latter leads to an inhibition of adenylate cyclase) induced a downregulation of NK1 receptors in DRG neurones. In conclusion the expression of the NK1 receptor can be influenced by PGE2 via the activation of the AC-PKA cascade. This raises the possibility that the downregulation of the NK1 receptors in the chronic phase of AIA may be caused at least in part by the upregulation of the SOM system. Supported by the DFG, Scha 404/9-2. Institut für Physiologie Ι, Universität Jena, Teichgraben 8, D-07743 Jena, Germany

C-fibre evoked field potentials can be dose-dependently inhibited by alpha2A-adrenoceptor agonists and µ-opiate receptor agonists. The proteinkinase C-inhibitor H7 (50µM) added 60 minutes prior to either DAMGO (1µM, n=6, 84±6% of control) or guanfacine (50µM, n=6; 78.5±9.8% of control) showed in both cases a significant inhibiton of the depressive effects of DAMGO or guanfacine on C-fibre evoked field potentials. These findings suggest that proteinkinase C is one of the intracellular effective targets for the analgetic effects of alpha2Aadrenoceptor and µ-opiate receptor agonists. It could also be demonstrated that C-fibre evoked field potentials in-vitro can serve to test pharmacological properties at different sites of nociception within the spinal dorsal horn.

O 13-2 THE ANESTHETIZED MOUSE: A MODEL FOR IN-VIVO ELECTROPHYSIOLOGICAL STUDIES OF TRIGEMINAL NOCICEPTION J. Ellrich, M. Teigeler Like many other complex biological phenomena, pain is starting to be studied at the level of the gene. The technique of manipulating the genome of the mouse is well established. Thus, most laboratories have switched from the classical animal model rat to the smaller mouse. Whereas this change is only a small step for in-vitro or behavioral studies, only a very few groups took the plunge to introduce the mouse into in-vivo electrophysiological studies. Here we describe the attempt to apply the techniques of in-vivo electrophysiology from the rat to the mouse. Male balb/c mice between 60 and 80 days old (weight: 20 to 25 g) were chosen. The mouse was anesthetized with an initial intraperitoneal dose of 50 mg/kg pentobarbital sodium salt (5 mg/ml). After tracheotomy a fine bore polythene tubing (0.76 mm inner diameter, 1.22 mm outer diameter) was introduced into the trachea. The left jugular vein was exposed and catheterized by a portex intravenous cannula (0.7 mm outer diameter). Teflon-coated stainless steel wires (0.125 mm diameter) were inserted into the right anterior digastric muscle to record the jaw opening reflex (JOR). The JOR was evoked by bipolar electric stimulation (500 µs) via two steel needles inserted into the tip of the tongue. The mouse was placed in a stereotaxic frame and artificially respired with oxygen-enriched room air (respiration rate 180/min, tidal volume 0.5 ml). Heart rate and arterial blood oxygen saturation were noninvasively and continuously monitored by an infrared sensor clip. General anesthesia was maintained by continuous intravenous administration of methohexital sodium salt (10 mg/ml). After exposure of the caudal brainstem and the cerebellar cortex 1 mm behind lambda recordings were made from single units using tungsten microelectrodes.Applying this technique it is possible to record from single neurons within the rostral ventromedial medulla and the spinal trigeminal nucleus. Furthermore, reflex studies can be done in the trigeminal system.

Institute of Physiology and Pathophysiology, University of Heidelberg, D-69120 Heidelberg, Germany O 13-4 INFLUENCE OF TWO SEROTONIN RECEPTOR ANTAGONISTS IN NEUROPATHIC RATS H. Wei, A. Pertovaara Introduction: There is some previous evidence indicating that serotonin receptors may play an important role in induction and maintainance of neuropathic pain. In the present study we investigated the influence of two serotonin receptor antagonists methysergide and tropisetron in rats with nerve ligation. Methods: Neuropathy was induced by unilateral ligation of two spinal nerves L5 and L6 in rats. The compounds tested were methysergide, a non-selective serotonin receptor antagonist, and tropisetron, a 5-HT3 receptor antagonist. Physiological saline was used as control. Drugs were administered systemically (s.c.) or intrathecally (i.t.) via a chronic cannula. Tactile allodynia was assessed by monofilaments, mechanical hyperalgesia was meassured by paw pressure test, and thermonociception was determined by heatinduced tail flick reflex. Results: Following i.t. administration methysergide produced a significant dose-dependent antiallodynic, antihyperalgesic and antinociceptive effect. Following s.c. administration methysergide induced only an antihyperalgesic effect. Tropisetron did not modulate nociceptive behaviors of neuropathic rats independent of the route of drug administration. Conclusions: The results indicated that spinal serotonin receptors are involved in the induction of hypersensitivity by nerve injury. This pronociceptive effect at the spinal cord level was not due to action on 5-HT3 receptors. Biomedicum Helsinki, Institute of Biomedicine/Physiology, P.O. Box 63 (Haartmaninkatu 8), FIN-00014 Helsinki University, Finland O 13-5

Department of Experimental and Clinical Pharmacology and Toxicology, University of Erlangen-Nuremberg, Emil-Fischer-Center, Fahrstr. 17, D-91054 Erlangen; [email protected] O 13-3 ALPHA2A-RECEPTOR AND µ-OPIATE RECEPTOR AGONISTS SHOWING PROTEINKINASE C-DEPENDENT EFFECTS ON CFIBRE EVOKED FIELD POTENTIALS IN LAMINA I/II OF THE SPINAL DORSAL HORN OF THE RAT C. Schott*, H. Seller The intracellular mechanisms for antinociceptive and analgetic effects of µ-opiate-receptor and alpha2A-receptor agonists are not yet well understood. There are several intracellular pathways which may be responsible for the above effects. As a first step, the interactions between proteinkinase-C and the specific alpha2A-adrenoceptor agonist guanfacine or the µ-opiate specific agonist DAMGO on C-fibre evoked field potentials within lamina I/II of the spinal dorsal horn of the rat were investigated. The lumbosacral segment (L1±S1) of the spinal cord of the rat (p28-32) with dorsal roots attached was removed under deep anesthesia. It was then cut into transverse slices of 1000µm with a dorsal root attached. Each slice was placed in a recording chamber and superfused with Krebs solution equilibrated with 95% O2 and 5% CO2 at 24 °C. C-fibre evoked potentials were recorded in the superficial laminae I/II of the lumbar spinal cord (glass electrodes, 3-5 MΩ) in response to electrical stimulation of the attached ipsilateral dorsal root (2 times of threshold, pulsewidth 0.5 ms, 30 sec intervals). As described previously (Schott C., Seller H., Eur J Physiol 2001; 441 Suppl, R157)

THE PROINFLAMMATORY CYTOKINE IL-1β FACILITATES IHEAT IN RAT SENSORY NEURONS O. Obreja, P. Rathee, K. Lipps, C. Distler, M. Kress Interleukin-1β (IL-1β) contributes to the development of inflammatory hyperalgesia and facilitates heat-induced release of calcitonin gene-related peptide from rat cutaneous nociceptors, in vitro. In the present study, we investigated the effect of IL-1β on heat-activated inward currents (Iheat) in isolated rat sensory neurons, and the pathways possibly mediating the IL-1β effect. After dissociation, neurons from rat dorsal root ganglia were cultured in the presence of nerve growth factor. In whole-cell voltage-clamp recordings (V h = -80 mV), membrane currents were recorded from single neurons loaded with the fluorescent dye Fura-2 via the patch pipette, for simultaneous measurements of [Ca2+]i. A fast drug application system was used to apply different test solutions and for feedback-controlled heat stimulation. Neurons were repetitively exposed to identical ramp-shaped heat stimuli at 60 s intervals. IL-1β (20 ng/ml) was used as conditioning intermittent stimulus and kinase inhibitors or their negative controls were continuously applied starting 2-3 min before IL-1β application. No significant changes of [Ca2+]i were elicited during IL-1β application. However, IL-1β induced a significant increase in Iheat amplitude (466 ± 198 pA to 914 ± 218 pA, n = 7) and a significant drop of threshold temperatures, of about 2°C. These changes were followed by partial recovery. The IL-1β-induced sensitisation of Iheat was unaffected by suramin. The non-specific protein kinase inhibitor staurosporine, the specific protein kinase C (PKC) inhibitor BIM1 as well as the tyrosine kinase (TyrK) inhibitor genistein significantly reduced the sensitising effect of IL-1β on Iheat, while negative controls

S 182 were ineffective. RT-PCR and in situ hybridisation revealed IL-1RI (but not IL-1RII) expression in sensory neurones of rat DRG. Together, our results suggest that IL-1β can directly sensitise rat sensory neurons to noxious heat, via an IL-1RI/TyrK/PKC-dependent mechanism, which does not involve G-protein coupled receptors.

mechanism supports a component of LTP that increases agedependently. 1

Institut für Pharmakologie und Toxikologie der Technischen, Universität, Biedersteiner Str. 29, 80802 München, Germany; 2 MaxPlanck-Institut für Experimentelle Medizin, Herman-Rein- Strasse

Inst. f. Physiologie u. Exp. Pathophysiologie, Universitaetsstr. 17, 91054 Erlangen, Germany O 14-2 O 13-6 HERITABLE DIFFERENCES IN NOXIOUS HEAT BUT NOT MECHANICAL SENSITIVITY ARE REFLECTED IN NEUROSECRETORY, TRANSCRIPTIONAL AND SENSORY CAPACITIES OF PRIMARY AFFERENT NEURONS IN INBRED MOUSE STRAINS F. Miermeister *, F. Seifert, N. Bernardini, M. Petersen, H. Hofmann, A. Siegling, M. Michaelis, M. Devor, P.W. Reeh Differences in sensitivity to painful stimuli have a heritable component as was found in inbred mouse strains that display major and consistent differences in behavioral tests (Mogil et al., 1999). However, the cellular basis of the genetic traits is yet unknown. We tested neurosecretory, transcriptional and sensory capacities of the primary nociceptive neuron in two inbred mouse strains, AKR exhibiting low heat and high mechanical sensitivity versus C57BL/6 behaving vice versa. Heat-induced iCGRP release from the isolated mouse skin well reflected the behavioral strain differences in heat responsiveness and even a higher degree of inflammatory hyperalgesia in AKR mice could be reproduced in vitro by pre-administering the direct PKC activator PMA. Different responses to KCl (C57>AKR) indicated an involvement of neurosecretory in addition to sensory differences. Quantitative gene expression for CGRP, measured by the Taqman technique, matched the total content of iCGRP in the skin and was about 1.7-fold higher in the C57BL/6 mice, whereas the VR1 gene was 4-times less expressed in AKR. Immunocytochemical staining of lumbar DRGs was similar in both strains pointing to a difference in neuronal CGRP content rather than number of CGRP containing neurons. The results from singlefiber recordings in the isolated mouse skin-nerve preparation showed a much lower heat responsiveness in AKR mice, in accordance with the behavioral studies, but did not reflect higher mechanical sensitivity. Heat-induced cobalt staining of cultured lumbar DRGs supported these findings and indicate a possible role of NGF in up-regulating heat transduction. Our results show that phenotypical differences in noxious heat but not mechanical sensitivity are well reflected in the first sensory neuron which may foster the search for responsible quantitative trait loci in the mouse genome. Department of Physiology and Experimental Pathophysiology, University of Erlangen/Nürnberg, Universitätsstr. 17, D-91054 Erlangen, Germany O 14-1 AGE-DEPENDENT REDUCTION OF LONG-TERM POTENTIATION IN THE HIPPOCAMPAL CA1 REGION OF MICE WITH A BRAIN-SPECIFIC DELETION OF THE CGMPDEPENDENT PROTEIN KINASE I GENE T. Kleppisch 1, W. Wolfsgruber 1, R. Allmann 1, S. Kühbandner 1, S. Goebbels2, K.A. Nave2, R. Feil1, F. Hofmann1 Nitric oxide is thought to support long-term potentiation (LTP) in the CA1 region of the hippocampus via cGMP-dependent protein kinase I (cGKI). However, LTP is normal in cGKI-deficient mice generated with the classical gene targeting technique. This finding might result from limitations of the conventional knockout mice, e.g. the reduced life span restricting the analysis to young animals. Here, we used a conditional brain-specific gene deletion strategy sparing vital functions of cGKI in other tissues and reevaluated the role of cGKI. Immunoblots demonstrate a marked and selective reduction of the cGKI protein in the hippocampus of both juvenile (3- to 4-weeks-old) and adult (at least 12-weeks-old) conditional cGKI knockout mice proving effective gene inactivation early during development. Remarkably, LTP in response to a repetitive weak theta burst was significantly decreased solely in adult brain-specific cGKI mutants compared to control mice. In contrast, LTP induced by the same stimulation protocol was normal in juvenile conventional cGKI knockout mice as well as in juvenile conditional cGKI mutants. We conclude that a cGKI-dependent

BIDIRECTIONAL CHANGES OF VESICLE RELEASE PROBABILITY IN SYNAPTIC PLASTICITY IN VITRO P.K. Stanton, U. Heinemann, W. Müller* Long-term, activity-dependent synaptic plasticity is an attractive candidate mechanism for learning and memory. Whether the sites of expression are presynaptic, postsynaptic or both is of fundamental neurological and computational importance. Recently, we demonstrated cGMP-dependent long-term depression of vesicular release from Schaffer collaterals in field CA1 of the hippocampus (J. Neurosci. 21:RC167,1-6). Here we use two-photon laser scan microscopy of FM1-43-labeled presynaptic vesicles to study changes in their release probability during synaptic plasticity in acute hippocampal slices. After induction of long-term depression (LTD) and staining of the total vesicle pool by 40 mM K + -perfusion, the rate of destaining of presynaptic terminals was reduced by about 50 %. FM1-43 labeling was reduced by two thirds and release kinetics was enhanced when the readily releasable pool (RRP) was selectively loaded by perfusion of hyperosmotic sucrose in ACSF (800 mosm/l). Indeed, the effect of LTD on the total vesicle pool was completely explained by a reduction in release probabality from the RRP. Conversely, strong long-term potentiation (LTP, 4x1s/100Hz, doubled intensity) caused enhanced release of vesicles after induction of LTP, while weak LTP could be expressed in the absence of any increase of vesicle release. During LTD, release from the reserve pool was primarily unaffected, indicating that synaptic plasticity specifically modulates release from the RRP. Our evidence supports bidirectional control of vesicle release probability that may play an important role in both dynamic information processing and storage in the hippocampal neuronal network. AG Molekulare Zellphysiologie des ZNS, NWFZ der Charité, Schumannstrasse 20/21, 10117 Berlin, Germany O 14-3 GAMMA RANGE MODULATION OF HUMAN OCCIPITAL EEG: EVIDENCE FOR SACCADE RELATED SHORT TERM PLASTICITY? I. Bodis-Wollner, H. von Gizycki, M. Avitable, V. Amassian, Z. Hussain, A. Javeid, A. Habib, A. Raza, M. Sabet Functional MRI studies suggest the striate cortex as a site for dynamic interaction of visual processing and saccades ; however, little is known of their timing interrelationships. We studied gamma activity which is believed to reflect local circuit binding of task related neurons. Our hypothesis is that perisaccadic gamma EEG has a necessary relation to saccade direction and timing. Ten normals (mean age 38 yrs.) were studied. Self paced saccades and eye position were recorded by the EOG and infrared scan. 14 channels of EEG were recorded over posterior occipital and parietal channels. The continuous EEG and eye movement recordings were visually inspected on the monitor and cursors placed bracketing acceptable saccades that subtended 60 degrees and had a velocity exceeding 200 degrees/s. The relevant EEG segments were analyzed with Continuous Wavelet Transform –wEEG- using Matlab tools.The analysed periods were 150 ms presaccadic, two 75 ms intrasaccadic and 150 ms post-saccadic. The wavelet coefficient of the 37 Hz band for each period was entered into further analysis using standard statistics. The statistical analysis compared gamma range coefficient variation as a function of time (pre-saccade, intrasaccade and new fixation) and hemisphere. Side to side wEEG correlation (left vs. right) of the gamma band was low preceding the saccade, high during the saccade (preceding the new fixation) and then high once new fixation was achieved. All subjects showed time dependent modulation of the perisaccadic gamma occurring 30-40 ms following the initiation of the saccade and prior to new fixation. Synchrony ceased roughly 60 ms after the eye came to rest. Occipital changes preceding an eye movement may be consistent with the notion of corollary discharge, while occipital activity following the eye

S 183 movement could signify short term plasticity changes of the visual cortex to a new,shifted visual field. Our results suggest direction dependent preparatory neuronal binding prior to new fixation. Depts of Neurology and Scientific Computing, State U. of NY, Downstate Med. Center, 450 Clarkson Avenue, Brooklyn NY 11203, USA

O 14-4 A SPATIAL ORIENTATION TASK INDUCES CHANGES IN PROTEIN EXPRESSION PATTERNS IN MOUSE HIPPOCAMPAL AND NEOCORTICAL BRAIN REGIONS H. Adelsberger, M. Güntner, S. Kathan, S. Seeber* In this study changes in the protein expression patterns induced by a spatial orientation task were investigated in different brain regions with two-dimensional gel electrophoresis. This method allows detection of both quantitative changes of protein amounts and posttranslational modifications of existing proteins.The learning task consisted of a labyrinth made from translucent acrylic glass tubes. In this labyrinth the animals had to make nine decisions to go left or right to find the exit where a small dark box, usually placed in their home cage, was positioned. The mice were released in the labyrinth six times with 30 min break between the trials. After each trial the tubes were changed randomly to avoid the influence of olfactory cues. The first trial was an inspection trial lasting usually 5-15 min. Both the number of ways and the time the mice needed to find the end of the labyrinth decreased rapidly from the second trial on and after six trials most of the animals performed the task without making more than 2 or 3 mistakes. From controls not released into the labyrinth and at different times after the beginning of the learning task protein extracts were made from the hippocampus and the neocortex and analysed by two-dimensional gel electrophoresis at different pH-ranges. In the hippocampus 8 increases of protein amounts, 2 decreases and 5 shifts of protein spots on the isoelectric point axes, reflecting posttranslational modifications were found. These changes could be seen in all six independent series of mice tested. In the neocortex 3 increases and 12 shifts of protein spots were found. In contrast to the hippocampal extracts the probably due to the more complex involvement of the neocortex in such a learning and memory task. Identification of 6 changed protein spots by MALDITOF spectroscopy started. Zoologisches Institut der LMU München, Luisenstr. 14, 80333 München. *Institut für Biochemie, Universität Erlangen, Fahrstr. 17, 91054 Erlangen

pattern of protection of synapses, with morphologically intact nerve terminals and EPPs and/or MEPPS in 13-50% of muscle fibres (N=11 muscles). Western analysis of Wld gene expression showed Wld protein levels unaltered with age. Thus, the declining protection of neuromuscular synapses from the effects of axotomy with age in WldS mice is not due to changes in the pattern of Wld gene expression, or to age of the animals per se. Rather, the persistence and/or progressive withdrawal of regenerated synapses following axotomy in old mice appears to be due to the immaturity of the synaptic terminals. The function and mechanism of this form of synapse withdrawal could be the same as that activated physiologically, during competitive synapse elimination following reinnervation, as well as during normal postnatal development. Support: The Wellcome Trust. Department of Neuroscience, 1 George Square, Edinburgh University, Edinburgh EH8 9JZ, UK.

O 14-6 MYOKYMIA AND NEONATAL EPILEPSY CAUSED BY A MUTATION IN THE VOLTAGE SENSOR OF THE KCNQ2 K+-CHANNEL K. Dedek, B. Kunath, C. Kananura, U. Reuner, T.J. Jentsch, O.K. Steinlein KCNQ2 and KCNQ3 are two homologous K+ channel subunits that can combine to form heterotetrameric channels with properties of neuronal M channels. Loss-of- function mutations in either subunit can lead to benign familial neonatal convulsions (BFNC), a generalized, idiopathic epilepsy of the newborn. We now describe a new syndrome in which BFNC is followed later in life by myokymia, involuntary contractions of skeletal muscles. All affected members of the myokymia/BFNC family carried a mutation (R207W) that neutralized a charged amino-acid in the S4 voltage-sensor segment of KCNQ2. This led to a shift of voltage-dependent activation of KCNQ2, and to a dramatic slowing of activation upon depolarization. Myokymia is thought to result from hyperexcitability of the lower motoneuron, and indeed both KCNQ2 and KCNQ3 mRNAs were detected in the anterior horn of the spinal cord where the cells of the lower motoneurons arise. We propose that a difference in firing patterns between motoneurons and central neurons, combined with the drastically slowed voltage-activation of the R207W mutant, explains why this particular KCNQ2 mutant causes myokymia in addition to BFNC. Zentrum für Molekulare Neurobiologie (ZMNH), Universität Hamburg, Falkenried 94, D-20246 Hamburg, Germany

O 14-5 RECAPITULATION OF AXOTOMY-INDUCED SYNAPSE WITHDRAWAL AT REGENERATED NEUROMUSCULAR JUNCTIONS IN MATURE WLDS MICE R.R. Ribchester, M.P. Coleman , D. Thomson, T.H. Gillingwater, T.G.A. Mack, D. Wagner The Wld S mutant mouse expresses a chimeric protein not found in wild-type mice, comprising a hybrid of the 70 N-terminal amino acids of the ubiquitination factor Ube4b and the complete sequence of NMNAT, the enzyme that synthesises NAD (Ube4b/Nmnat; Wld protein). The Wld protein protects distal axons from degeneration after injury, and this phenotype is reproduced in transgenic mice expressing the Wld chimeric gene (Mack et al., 2001; Nature Neuroscience 4, 1199-1206). In WldS mice younger than 2 months, neuromuscular synapses undergo progressive withdrawal from endplates in response to axotomy. However, in older animals (> 7 months old) axotomised synapses degenerate at almost the same rate in as in wildtype mice. Here we asked whether the loss of synapse protection by Wld was due to age-dependence of the pattern of Wld gene expression, or to the local state of differentiation of the synapses. To distinguish these possibilities we examined the response to axotomy of regenerated neuromuscular junctions (“young” synapses) in old Wld S mice. The sciatic nerve in adult WldS mice aged 7-12 months was crushed, under halothane/N20 anaesthesia, and allowed to regenerate for 8-10 weeks. The mice were then re-anaesthetised and the tibial nerve was cut. Tibial nerve section alone (without a previous sciatic lesion) in mice of this age caused all neuromuscular transmission and synaptic morphology to be lost within 2 days. Muscles in which the tibial nerve was lesioned following prior sciatic nerve injury and regeneration showed a consistent

O 15-1 DIFFERENTIAL CONTROL OF VESICLE PRIMING AND SHORT-TERM PLASTICITY BY MUNC13 ISOFORMS C. Rosenmund 1*, A. Sigler1, I. Augustin2, K. Reim2, N. Brose 2, J.S. Rhee 1 Presynaptic short-term plasticity is an important adaptive mechanism regulating synaptic transmitter release at varying action potential frequencies. However, the underlying molecular mechanisms are unknown. We examined genetically defined (using knockouts) and functionally unique axonal subpopulations of synapses in excitatory hippocampal neurons that utilize either Munc13-l or Munc13-2 as synaptic vesicle priming factor. Primary autaptic cultures from single and double Munc13-1/2 deficient mice were prepared and synaptic properties were measured using standard patch clamp techniques. In 90% of excitatory wildtype synapses, Munc13-l is employed as a priming factor, while 10% of synapses utilize Munc13-2 for priming. Munc13 synapses showed wildtype like depression of EPSC amplitudes during high frequency stimulation. In contrast, Munc13-2 driven synapses show pronounced and transient augmentation of synaptic amplitudes during and following high frequency stimulation. This augmentation is caused by a Ca2+-dependent increase in release probability and an increase in the size of the releasable vesicle pool. Augmentation was not affected by modulators of various kinases, phosphatases or cytoskeletal dynamics, but required phospholipase C activity, indicating an involvement of the diacylglycerol binding C1 domain of Munc13-2 in the expression of augmentation. Thus, differential expression of Munc13 isoforms at individual synapses represents a gen-

S 184 eral mechanism that controls short-term plasticity and controlled distribution of Munc13 isoforms may contribute to the heterogeneity of synaptic information coding.

Institute of Life Science, The Hebrew University of Jerusalem, Jerusalem, 91904 Israel

1 MPI für biophysikalische Chemie, Abt. Membranbiophysik; 2MPI für experimentelle Medizin, Abt. Molekulare Neurobiologie, 37070 Göttingen

O 15-4

O 15-2 α-NEUREXINS AS KEY MODULATORS OF CALCIUM CHANNEL FUNCTION W. Zhang1*, A. Rohlmann1, T. Südhof2, M. Missler1; Neurexins constitute a family of highly polymorph transmembrane proteins for which an involvement in synaptic function has been proposed. Identified as the first, Ca++-dependent α-latrotoxin receptor, their exact physiological role has remained elusive. Here we have used a gene targeting approach to investigate this question. All three αneurexin genes were deleted by conventional knockout technology, and all possible combinations of double (DKOs), and the triple mutant mice (TKOs) have been generated. DKOs show an increased mortality (between 40% to 100% affected), while all TKOs die within several hours of birth, demonstrating that α-neurexins are essential genes for postnatal development. In newborn mutant mice, ventilation is severely impaired depending on the combination of the differentially distributed forms. Patch-clamp recordings in brainstem respiratory neurone of one day old animals revealed greatly reduced whole cell calcium currents in DKOs and TKOs, while Na +- and K +IR -currents appeared to be unaffected. Among all HVA-Ca++-currents studied (N-, P/Q-, and L-type), the N-type currents were mostly impaired, whereas radioligand experiments with ω-conotoxin GVIA could not detect differences in the number of binding sites on membranes. Compared to control mice, both the frequency and amplitude of miniature postsynaptic currents (mPSC) in respiratory neurones were severely reduced in TKOs. In addition, the amplitudes of evoked postsynaptic currents (ePSC) were reduced and the failure rate were increased. Our results strongly support the hypothesis that α-neurexins play a key role in modulation of calcium channel function. *Zentrum Physiologie und Pathophysiologie, Universität Göttingen, Humboldtallee 23, D-37073 Göttingen; 1 Zentrum Physiologie und Pathophysiologie, Universität Göttingen; 2Center for Basic Neuroscience and Howard Hughes Medical Institute, Dallas, USA

SNAP-25 AND SYNTAXIN 1 REDUCE EXOCYTOSIS OF INSULIN-CONTAINING GRANULES IN PANCREATIC β -CELLS J. Schriever-Abeln, X. Ma, L. Eliasson Background and Aims: Secretory granules within the pancreatic βcells need to be docked and primed prior to fusion and release. In many endocrine cells as well as the pancreatic β-cell SNARE-proteins have been suggested to be present in these processes. The aim of this study was to investigate the involvement of the t-SNARE proteins SNAP25 and Syntaxin 1 in the exocytotic process of primary β-cells. Material and Methods: We used capacitance measurements to investigate the temporal aspects of exocytosis on a single cell level. The standard whole-cell configuration of the patch-clamp technique was used which allows an intracellular application of the different antibodies used. Identification of the proteins was obtained by immunocytochemistry and Western Blot analysis preceded by subcellular fractionation. Results: The presence and distribution of these proteins within the βcell was investigated prior to the functional studies. Using subcellular fractionation followed by Western Blot analysis, SNAP-25 and Syntaxin 1 was found to be present mainly in the plasma membrane but also in a granular fraction. Immunostaining with antibodies against SNAP-25 and Syntaxin 1 further verified these results. To investigate the functional importance of the proteins, a train of ten 500 ms depolarising pulses from -70 to 0 mV were applied to single cells. Under control conditions the increase in membrane capacitance during the first depolarisation was 47±15 fF (n=6), corresponding to 24±8 granules using a conversion factor of 2 fF/granule. The release during the first depolarisation was almost abolished (5±1 fF; P<0.01) in the presence of an antibody against SNAP-25. This antibody also reduced the increase in membrane capacitance during the total train to 43±24 fF (n=6) compared with 176±50 fF (n=5) under control conditions (P<0.05). Exocytosis was likewise affected in presence of an antibody against Syntaxin 1 and the exocytotic response during the train was reduced by 70% (P<0.01). Conclusions: These data suggest that SNAP-25 and Syntaxin 1 are important for Ca2+-dependent exocytosis in pancreatic β-cells and that they are crucial for the priming of the insulin-containing granules prior to release. Dept. of Molecular and Cellular Physiology, Lund University, Tornavägen 10, BMC Fl 1, 22184 Lund, Sweden

O 15-3

O 15-5

VOLTAGE-GATED Ca2+ CHANNELS ARE COUPLED TO THE EXOCYTOTIC MACHINERY D. Atlas, O. Wiser, M. Trus, R. Cohen Syntaxin 1A has a pronounced inhibitory effect on the activation kinetics and current amplitude of voltage-gated Ca 2+ channels. We explored the molecular basis of syntaxin interaction with Cav1.2 (Lctype) and Cav2.2 (N-type) Ca2+channels by way of functional assays of channel gating in a Xenopus oocytes expression system. A chimera of syntaxin 1A and syntaxin 2 in which the transmembrane domain (TMD) of syntaxin 2 replaced the TMD of syntaxin 1A (Sx1-2) significantly reduced the rate of activation of Cav1.2 and Cav2.2. The major sequence differences at the TM of syntaxin isoforms are the two highly conserved Cys 271 and Cys 272 in syntaxin 1A, which correspond to Val 272 and Val 273 in syntaxin 2 TMD. Mutating either Cys in Sx1-1 (syntaxin 1A) to Val, did not affect modulation of the channel while a double mutant C271/272V was unable to regulate inward current. Transfer of these two Cys to the TM of syntaxin 2 by mutating Val 272 and Val 273 to Cys 272 and Cys 273 led to channel inhibition. When cleaved by Bot-C1, the syntaxin 1A fragments, aa 1253 and aa254-288, which includes the TMD, were both unable to inhibit current amplitude but retained the ability to modify the activation kinetics of the channel. A full-length syntaxin 1A and the integrity of the two Cys within the TMD are crucial for coordinating Ca2+ entry through the Ca v1.2 and Ca v 2.2. These results suggest that upon membrane depolarization, the gated channel signals the exocytotic machinery through syntaxin 1A TMD domain. Cleavage with Bot-C1 disrupts the syntaxin 1A/channel coupling and abolishes exocytosis. These results are consistent with the dynamic role of the channel in the Ca2+-regulated release of channel-associated vesicles (CAV).

REGULATION OF EXOCYTOSIS IN PANCREATIC ACINAR CELLS BY CAPACITATIVE CA2+ ENTRY T. Bagrij, M. Campos-Toimil, J. M. Edwardson, P. Thomas In pancreatic acinar cells muscarinic agonists stimulate both the release of Ca2+ from intracellular stores and an influx of extracellular Ca2+. The part played by Ca 2+ released from intracellular stores in the regulation of secretion is well established; however, the role of Ca2+ influx in zymogen (secretory) granule exocytosis is unclear. Using digital imaging of fura-2 fluorescence and continuous time-differential analysis [M. Campos-Toimil, J. M. Edwardson & P. Thomas (2000) J. Physiol. (Lond.) 528, 317-326], we have now shown that removal of extracellular Ca 2+ (Ca out) reduces both Ca 2+ influx and exocytosis stimulated by 1 µM acetylcholine (ACh). Remarkably, wortmannin (100 nM), an inhibitor of phosphatidylinositol 3-kinase (PI 3-K), potentiated exocytosis (~45%) in the presence of Caout, but inhibited exocytosis (~45%) in the absence of Caout. Gd 3+ (2 µM), a selective blocker of capacitative Ca2+ entry, both inhibited exocytosis and brought about the switch to a wortmannin-inhibitable form of exocytosis. In contrast, inhibition of non-capacitative Ca2+ entry by SB203580 (10 µM; a p38 MAP kinase inhibitor) had no effect on exocytosis. Interestingly, even in the presence of Caout, exocytosis induced by a supra-maximal concentration of ACh (10 µM) was potently inhibited by wortmannin (~70%). We conclude that supra-maximal doses of ACh lead to a switch in the mode of exocytosis by inhibiting capacitative Ca2+ influx, and that the mode employed at high agonist concentrations is dependent upon PI 3-K. Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge, CB2 1PD, United Kingdom

S 185 O 15-6

O 16-2

WEIBEL-PALADE BODY EXOCYTOSIS IN HUMAN ENDOTHELIAL CELLS OCCURS AT SELECTIVE CELL MEMBRANE AREAS S. W. Schneider*, R. Matzke, R. Ossig, P. Rogge, A. Niemeyer, T. Gorge, H. Oberleithner Human umbilical vein endothelial cells (HUVEC) secrete vasoactive substances such as von Willebrand factor (vWF). This molecule is stored in large (up to 3 µm) cone-like vesicles called Weibel Palade bodies (WPB). By using atomic force microscopy (AFM) we are able to visualize the apical surface topography of fixed human endothelial cells with nanometer resolution. In addition, AFM allows to measure local cell stiffness with a spatial resolution of 100 nm. In previous studies we showed that endothelial cells have a readily releasable pool of WPB. In resting cells this intracellular with a height of 140 ± 50 nm (±SEM; n = 8) and a diameter of 275 ± 85 nm (±SEM; n = 8). Humps represent areas of docked WPB. Stiffness measurements revealed that humps are characterized by decreased cell membrane stiffness of 45% compared to surrounding cell membrane. Incubation of HUVEC with the actin depolymerization agent cytochalasin D decreased overall cell stiffness and eliminated stiffness differences. Therefore, cell stiffness is mainly due to the subapical actin network. After stimulation of the cells with hyperosmolaric solutions or histamine these docked WPB immediately fuse with the plasma membrane forming large (diameter: ~500 nm) exocytotic pores and release vWF into the supernatant. Appearance of exocytotic pores matches the disappearance of humps. The data indicate that human endothelial cells have a readily releasable pool of WPB that allows the instantaneous release of vWF. These distinct areas of exocytosis are characterized by cell membrane protrusions (humps) and decreased cell membrane stiffness due to a missing actin cortical network. Lack of cortical actin at distinct membrane areas is a prerequisite for the release of prothrombotic substances into the blood stream.

THE INTERACTION BETWEEN NITRIC OXIDE AND SUPEROXIDE ANION ON NEURAL REGULATION OF PROXIMAL TUBULAR FLUID REABSORPTION IN HYPERTENSIVE RATS X.C. Wu, E.J. Johns Previously, we demonstrated that nitric oxide (NO) exerted a tonic inhibition on proximal tubular fluid reabsorption (Jva) in normotensive Wistar rats (NR) but not in the stroke prone spontaneously hypertensive rats (SHRSP) (Wu, Harris, Johns, 1999). In further studies, we have found that blockade of NO generation prevented low frequency renal nerve stimulation (RNS) from inducing an increase in Jva in NR and that it was independent of superoxide anion (O2-). The aim of the present study was to examine the roles of NO and O2- in regulating Jva under basal conditions and during RNS in SHRSP where oxidative stress exists. Male SPSHR (250-320 g) were anaesthetised with Inactin (100mg/kg ip), the left kidney was exposed and the renal nerves placed in contact with stimulation electrodes. The shrinking split-droplet technique was used to measure Jva, (x 10-4 mm3 mm-2 s-1) in proximal tubule S2 segments using at least one pair of surface nephrons / rat. Basal Jva was measured and subsequent estimations made when the renal nerves were stimulated at 0.75 Hz and 1 Hz (2ms, 15V). A second set of measurements was performed on different nephrons. Groups of 6-7 rats received intraluminally either L-NAME (l0-4 M), superoxide dismutase (SOD 10-4 M), sodium nitroprusside (SNP 10-4 M) alone or SOD plus SNP (10-4 M). Data, means±S.E.M., were subjected to ANOVA and Student’s ‘t’ test with significance taken at P<0.05. The animals were killed humanely at the end of experiment. The BP was stable (l22±2 mmHg), RNS had no effect on GFR but urine flow decreased by 13% and 18% (P<0.05) and sodium excretion by 12% and 34% at 0.75 Hz (P<0.05) and 1 Hz (P<0.01), respectively. The basal Jva was 2.06±0.16 xl0-4 mm3 mm2 s-1 and although Jva rose by 27% (P<0.05) at 0.75 Hz there was no further change at 1 Hz. LNAME had no effect on basal Jva but it increased by 34% at 0.75 Hz (P<0.01) and 24% at I Hz RNS. SOD enhanced basal Jva by 33% (P<0.001) and during RNS increased a further 21 % at 0.75 Hz (P<0.05) but not at 1Hz. SNP had no effect on basal Jva or the rise induced by RNS at 0.75 Hz but at 1 Hz Jva was increased by 45% (P<0.05). SNP plus SOD had no effect on the pattern of response to RNS, which was comparable to that observed when vehicle was given alone. These data suggested that over-production of O2- contributed to the low basal Jva in the SHRSP rats. During lower level of RNS, at 0.75 Hz, neither O2nor NO changed the magnitude of response of Jva to RNS. However, at 1 Hz RNS, O2- did appear to be involved and one possibility was that peroxinitrite was in someway modulating Jva since SNP appeared to interact with O2- to restore the responsiveness of Jva to RNS. Reference: Wu, XC, Harris, PJ, Johns EJ, Am J Physiol. 277 (46): F560F566, 1999.

Physiologisches Institut, Robert-Koch-Str. 27a, D-48149 Münster O 16-1 REGULATION OF THE EPITHELIAL SODIUM CHANNEL BY N4WBP5A, A NOVEL NEDD4/NEDD4-2-INTERACTING PROTEIN A.-A. Konstas*, L. Shearwin-Whyatt‡, A. Fotia‡, D.I. Cook†, S. Kumar‡, C. Korbmacher* The amiloride-sensitive epithelial sodium channel (ENaC) plays a critical role in fluid and electrolyte homeostasis and consists of α, β and γ subunits. The carboxyl terminus of each ENaC subunit contains a PPxY motif which is important for interaction with the WW domains of the ubiquitin-protein ligases, Nedd4 and Nedd4-2. Disruption of this interaction, as in Liddle’s syndrome where mutations delete or alter the PPxY motif of either the β or γ subunits, has been shown to result in increased ENaC activity and hypertension. Using the two-electrode voltage clamp technique and a protein surface expression assay in Xenopus oocytes, we showed that N4WBP5A, a novel Nedd4/Nedd42 binding protein, is a potential regulator of ENaC. Surface expression and amiloride-sensitive sodium currents (∆Iami) were increased by 4fold in oocytes coexpressing ENaC and N4WBP5A compared with ENaC control oocytes. The stimulatory effect of N4WBP5A on ENaC is specific, since N4WBP5A failed to stimulate either CFTR or Kir1.1a. Coexpression of N4WBP5A reduced the brefeldin A induced decline of ∆Iami. This indicates that the stimulatory effect of N4WBP5A is due to a reduction in the rate of ENaC retrieval. The lack of effect of N4WBP5A on ENaC with Liddle’s syndrome mutation suggests that the stimulatory effect of N4WBP5A requires the presence of an intact endocytotic retrieval mechanism. Stimulation of ∆Iami by N4WBP5A required sodium loading of the oocytes indicating that N4WBP5A prevents sodium feedback inhibition of ENaC possibly by interfering with the xNedd4-2 mediated regulation of ENaC. As N4WBP5A binds Nedd4-2 and appears to be associated with specific intracellular vesicles, we propose that N4WBP5A functions by regulating Nedd4/Nedd4-2 availability and trafficking. * Univ. Laboratory of Physiology, Univ. of Oxford, Parks Road, Oxford OX1 3PT, UK; ‡ Hanson Centre for Cancer Research, Institute of Medical and Veterinary Science, Frome Road, Adelaide, SA 5000, Australia; † Department of Physiology, Univ. of Sydney, Sydney, NSW 2006, Australia

Department of Physiology, Division of Medicine, University of Birmingham, BI5 2TT UK O 16-3 INTRACELLULAR Na CONCENTRATION AND Rb UPTAKE IN PROXIMAL CONVOLUTED TUBULE (PCT) CELLS AND ABUNDANCE OF Na/K-ATPase α1-SUBUNIT IN NHE3-/- MICE F.-X. Beck1, W. Neuhofer1, A. Dörge1, G. Giebisch2, T. Wang2 Proximal solute and fluid absorption is greatly reduced in mice in which the gene encoding the Na/H exchanger isoform 3 is ablated (NHE -/- ). To obtain more information on the cellular mechanisms underlying these transport defects, Na, Cl and K concentrations and cell Rb uptake − an index of Na/K-ATPase activity − were assessed in proximal convoluted tubule (PCT) cells of NHE -/- and wild-type (NHE +/+) mice using electron microprobe analysis. In addition, the relative abundance of the α1-subunit of the Na-K-ATPase in outer cortex was determined by Western blot analysis. In NHE-/- mice, PCT cell Na concentration was slightly, but significantly, lower than in NHE+/+: 13.1 ± 0.6 (64) vs. 14.9 ± 0.6 (62) mmol/kg wet wt; means ± SEM, (n). This lower intracellular Na concentration was associated with reduced Rb uptake rates (measured after a 30-s Rb infusion): 9.7 ± 0.6 (59) vs.14.8 ± 0.8 (50) mmol/kg wet wt/30 s. In contrast to Rb uptake, the abundance of the α1-subunit of the Na/K-ATPase was not different between NHE-/- and NHE+/+ mice. Intracellular Cl concentration was higher (14.2 ± 0.4 vs. 12.8 ± 0.41 mmol/kg wet wt) and intracellular K concentration unchanged (122.7 ± 2.7 vs. 121.6 ± 2.5

S 186 mmol/kg wet wt) in NHE-/- compared with NHE+/+ mice. These findings are compatible with the view that the elimination of apical NHE3 in PCT cells of NHE3-/- mice reduces apical Na entry and, due to lower cell Na concentrations, Na/K-ATPase activity. Apparently, intracellular Na concentration is not the sole factor determining the expression of Na/K-ATPase in the renal cortex of NHE3-/- mice. 1

Physiologisches Institut der Universität, Pettenkoferstr. 12, 80336 München; 2Department of Cellular and Molecular Physiology, Yale University, Medical School, New Haven, USA O 16-4 THE NA-K-2CL COTRANSPORTER FORMS HIGH MOLECULAR WEIGHT COMPLEXES IN FERRET RED BLOOD CELL MEMBRANES I. Matskevich, D.K. Apps, P.W. Flatman The Na-K-2Cl cotransporter plays an important role in regulating cell K and Cl concentrations and cell volume. It is highly expressed in ferret red cells, yet has not been clearly identified by staining gels of their membrane proteins separated by SDS gel electrophoresis. A study of rat parotid glands (Moore-Hoon ML & Turner RJ Biochem. 39, 3718-3724, 2000) found that two cotransporter molecules can be chemically cross-linked together suggesting that the protein exists in membranes as a homodimer. In this study we show that high molecular weight complexes containing the cotransporter can also be detected in ferret red cell membranes without using cross-linking. Blood was taken by cardiac puncture from anaesthetised ferrets (sodium pentobarbitone, 120 mg/kg i.p.). Membranes were prepared from washed red cells by hypotonic lysis or by freeze/thawing in isotonic buffer and were analysed by SDS gel-electrophoresis. Proteins were blotted from gels and immunodetected with T4 mouse monoclonal antibodies (Developmental Studies Hybridoma Bank, University of Iowa). The cotransporter appears on blots as a smeared band (sometimes a doublet) at 164 ± 6 kDa (mean ± s.e.m, n = 8). With prolonged wet blotting bands appear between 250 – 480 kDa, with a prominent band at 341 ± 11 kDa (n = 8). These bands are seen on blots from immunoprecipitates and whole membranes and are stable, surviving boiling in the presence of SDS and treatment of samples with reducing agents (e.g. dithiothreitol) and alkylation with iodoacetamide. The cotransporter was also detected in two bands at about 200 and 500 kDa after separation of proteins by blue native gel-electrophoresis (without SDS or reducing agents). Our results suggest that the Na-K-2Cl cotransporter forms high molecular weight complexes in ferret red blood cell membranes. The molecular weight of these complexes, 2 or 3 times that of single cotransporter polypeptide, is consistent with the cotransporter existing as dimers or trimers, though complexation with other membrane proteins cannot be excluded. This work is supported by the Wellcome Trust. Membrane Biology Group, Division of Biomedical and Clinical Laboratory Sciences, The University of Edinburgh, Edinburgh EH8 9XD, UK O 16-5 INTESTINAL NA+ AND CL- TRANSPORT PROTEIN EXPRESSION AND FUNCTION AND PARACELLULAR PERMEABILITY IN CFTR -/- MICE I. Blumenstein, H. Rossmann, C. Weinhold, P. Jacob, B. Riederer, C. Neff, M. Gregor, U. Seidler Backround: Patients and mice with a defective CFTR protein suffer from malabsorption and intestinal obstruction. Besides a lack of agoniststimulated short-circuit current (Isc) and HCO3- secretory response, a defective regulation of expression and/or regulation of a whole array of ion transport proteins have been described in CFTR-deficient GI cell lines or tissues, as well as changes in paracellular permeability, protein secretion and apoptosis. Aim and Methods: To better understand the intestinal electrolyte transport dysfunction associated with defective CFTR expression, we studied the mRNA expression of NHE3 and DRA, the major apical transport proteins involved in electroneutral Na+ and Cl- absorption, and NBC1 and NKCC1, the major proteins for basolateral anion uptake, in jejunal and proximal colonic mucosa of CFTR -/- mice and their normal littermates by a quantitative RT-PCR protocol. In parallel, we measured transepithelial Na + and Cl -, and mannitol movements in the basal state and after an increase in

intracellular cAMP in stripped upper jejunum and proximal colon of these mice by isotope flux measurements. Results: When compared to 18sRNA expression, NHE3 and DRA expression levels were slightly, and NBC1 and NKCC1 more markedly reduced in CFTR -/- compared to CFTR +/+ mucosa. When compared to the brush border membrane protein villin expression, NHE3 and DRA mRNA levels were increased and NBC1 and NKCC1 levels similar in CFTR -/- compared to +/+ mucosa. Correspondingly, intestinal NaCl absorption was not significantly different in the basal state, but inhibition by NHE3 blockade was somewhat stronger in CFTR -/- than +/+ mucosa. An increase in cellular cAMP inhibited jejunal and proximal colonic NaCl absorption only in +/+ mucosa. 3H+-mannitol flux rates were similar in +/+ and -/- mucosa and were not increased by 8-Br-cAMP. Interestingly, 8-Br-cAMP elicited an slight increase in bilateral Na+ flux rates in CFTR -/- mucosa, and serosa to mucosa Na+ flux only in +/+ mucosa, suggesting that cAMP may cause an increase in tight junctional permeability to small ions only. Conclusion: No major changes in the expression levels or function for NHE3, DRA, NBC1 and NKCC1, and no change in paracellular permeability were observed in CFTR -/intestine. However, a complete lack of cAMP-mediated inhibition of electroneutral Na+ absorption is present in CFTR -/- small and large intestine. This is likely one major pathophysiologic factor in the development of intestinal obstruction in CF patients. Dept. of Medicine, University of Tübingen, Germany O 16-6 THE SERUM AND GLUCOCORTICOID DEPENDENT KINASE SGK1 AND THE NA+/H+-EXCHANGER REGULATING FACTOR NHERF2 SYNERGIZE TO ACTIVATE THE RENAL EPITHELIAL K+-CHANNEL ROMK1 M. Palmada, H. Embark, G. Henke, Y. Feng, C. Böhmer, EJ. Weinman, S. Hebert, F. Lang, C. Yun Mineralocorticoids stimulate Na+ reabsorption via the renal epithelial Na+-channel ENaC and K+ secretion via the renal epithelial K+-channel ROMK1 in principle cells of connecting tubule and collecting duct. Following heterologous expression in Xenopus oocytes the serum and glucocorticoid sensitive kinase SGK1 has been shown to up-regulate ENAC activity by enhancing transport protein abundance in the plasma membrane. With the same method, ROMK1 appeared to be insensitive to regulation by SGK1. On the other hand, ROMK1 has been shown to colocalize with NHERF2, a protein mediating targeting and trafficking of transport proteins into the cell membrane. The present study has been performed to test, whether NHERF2 is required for regulation of ROMK1 by SGK1. Upon expression in Xenopus oocytes, ROMK1 displayed the typical current-voltage relation and pH sensitivity reported earlier. Coexpression of either NHERF2 or SGK1 with ROMK1 does not significantly enhance ROMK1 K + channel activity (1.2 ± 0.1 µA and 1.2 ± 0.2 µA, respectively, n=10). However, coexpression of NHERF2 and SGK1 together with ROMK1 leads to a five fold increase of K+ channel activity as compared to expression of ROMK1 alone (from 1.0 ± 0.1 µA to 4.7 ± 0.3 µA, n=10). This upregulation was not paralleled by alteration of voltage and pH sensitivity of the channel. In conclusion, NHERF2 and SGK1 interact to up-regulate ROMK1 channel activity. This interaction allows the integration of genomic regulators and activators of SGK1 and of NHERF2 in the control of ROMK1 activity and renal K+ excretion. Physiologisches Institut der Universität Tübingen, Gmelinstr. 5, 72076 Tübingen, Germany. O 17-1 SOLUTION STRUCTURE AND FUNCTION OF THE TANDEMINACTIVATION DOMAIN OF Kv1.4 CHANNELS R. Wissmann, W. Bildl, D. Oliver, P. Jonas, D. Bentrop, B. Fakler Fast inactivating (A-type) potassium channels channels are responsible for shaping the presynaptic action potential and thus for determining timing and strength of synaptic transmission. Kvl.4, the subunit reconstituting fast inactivation in a hippocampal presynapse, exhibits inactivation-gating prototypic for the famous ‘ball-and-chain’ mechanism. Unique among inactivating Kvα subunits, however, Kvl.4 was found to present two inactivation balls that are both able to occlude the open channel pore when present at the N-terminus of the molecule. Here we investagted the solution structure and function of

S 187 this ‘tandem-inactivation domain’ using NMR spectroscopy and patchclamp recording. Inactivation domain 1 (IDl, residues 1-38) consists of a flexible N-terminus anchored at a 5-turn helix, while ID2 (residues 40-50) is a 2.5-turn helix made up of small hydrophobic amino acids. Functional analysis showed that IDl is the ball domain occluding the channel pore, while ID2 works as a ‘docking domain’ that binds IDl to the cytoplasmic face of the channel. Deletion of ID2 slows inactivation considerably and thus largely impairs the process of cumulative inactivation underlying the channels ability of signal-integration and spike proadening. Together, the concerted action of IDl and ID2 promote rapid inactivation of Kvl.4 that is crucial for the channel’s function in presynaptic plasticity. Physiologisches Institut II, Hermann-Herder Str. 7, 79104 Freiburg, Germany O 17-2 LUMINAL K+ RECYCLING VIA KCNE2/KCNQ1 CHANNELS IS REQUIRED FOR GASTRIC ACID SECRETION D. Heitzmann, A. Schmitt-Gräff, F. Grahammer, R. Nitschke, J. Barhanin, R. Warth H+ secretion of gastric parietal cells via the H+/K+-ATPase is coupled to the uptake of K+. Recently, we have shown that KCNQ1 is the poreforming α-subunit of a luminal K+ channel enabling K+ recycling. Immunofluorescence revealed a colocalization of KCNQ1 and H+/K+ATPase in tubulovesicles of gastric parietal cells. Inhibition of KCNQ1 K+ channels by the chromanol 293B and KCNQ1 gene disruption in mice abolish H+ secretion completely. KCNE1, 2, and 3 are known to coassemble with KCNQ1 leading to distinct functional properties of the heteromultimeric channel. This study was aimed to investigate subunit composition and regulation of heterologously expressed and native KCNQ1 K + channels. In rodent stomach, KCNE1-3 are expressed. In contrast, only KCNE2 and KCNE3 are abundant in human stomach; KCNE1 transcript could not be detected by PCR techniques. In situ hybridization showed a strong labeling of parietal cells by KCNE2specific antisense but not sense probe. To examine the functional properties of the putative channel complex, COS cells were cotransfected with KCNE2 and KCNQ1. Compared to cells transfected with KCNQ1 alone, KCNE2 cotransfected cells exhibited less K+ current under control conditions. In addition, KCNQ1-typic slow activation kinetics and voltage dependence of the current was changed to an instantaneous and voltage-independent current. Interestingly, KCNE2/KCNQ1 current was activated by acidic extracellular pH and by stimulation of cAMP and IP3/Ca2+ pathways. 293B inhibited KCNE2/KCNQ1 current in transfected cells and whole cell current of freshly isolated mouse parietal cells. In conclusion, KCNE2 is the likely candidate to coassemble with KCNQ1 in the luminal membrane compartment. The channel complex is activated by cAMP and IP3/Ca2+ which represent physiological stimuli of acid production. In contrast to many other K + channels, KCNE2/KCNQ1 is activated by acidic extracellular pH. Inhibition of KCNE2/KCNQ1 by 293B derivatives could offer new perspectives for the treatment of peptic ulcer disease. Physiologisches Institut, Winterthurerstr. 190, CH-8057 Zürich O 17-3 SRC TYROSINE KINASE MODULATES CLONED M-TYPE K+ CHANNELS N. Gamper, M.S. Shapiro We discovered a novel pathway of neuronal M-type K + channel regulation by the tyrosine kinase Src. Heteromultimers of KCNQ2 and KCNQ3 channels underlie neuronal M current (Wang et al. 1998, Science 282:1890-3). In the present study, cloned KCNQ2 and KCNQ3 subunits were heterologously expressed in CHO cells to form functional heteromultimers. Co-transfection of Src with KCNQ2/KCNQ3 channels dramatically decreased whole-cell current amplitude and slowed activation kinetics. In cells transfected with KCNQ2 and KCNQ3 (control), mean amplitudes at 0 mV and +40 mV were 695 ± 78 pA (n=34) and 1170 ± 152 pA (n=21) respectively, whereas in cells cotransfected with Src, they were 150 ± 19 pA (n=43) and 277 ± 33 pA (n=41). Activation time constants at 0 mV were increased from 158 ± 11 ms (n=34) in control to 289 ± 17 ms (n=43) in Src-transfected cells. The half-activation voltage (V1/2) was shifted by 6 mV more positive by Src co-transfection. Application of the tyrosine kinase inhibitors

erbstatin (20 µM) and PP2 (100 nM) to the bath during the experiment increased current amplitudes at 0 mV in Src-transfected cells to 234 ± 18% (n=7) and 142 ± 14% (n=6) of the initial values, respectively and completely restored control activation kinetics. Erbstatin also shifted the V1/2 in Src-transfected cells by -16 mV. The inactive PP2 analogue, PP3, did not affect current amplitudes nor activation kinetics in Srctransfected cells. Individual transfection of KCNQ2 or KCNQ3 channels with or without Src reveals dramatic inhibition of KCNQ3 channels by Src whereas the activity of KCNQ2 subunits was not affected. The effects of Src transfection on KCNQ2/3 currents were mimicked by 1h incubation of CHO cells expressing KCNQ2/3 with the tyrosine phosphatase inhibitor vanadate (100 µM). Vanadate also reduced M current in cultured rat superior cervical ganglion sympathetic neurons. Cells were held at –25 mV and M current amplitude was quantified as the time-dependent deactivating current at –60 mV. M current density decreased from 0.87 ± 0.14 pA/pF (n=15) in control cells to 0.47 ± 0.12 pA/pF (n=13, P≤0.05) in vanadate-treated cells. Department of Physiology University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229 O 17-4 IDENTIFICATION OF A NOVEL CONUS PEPTIDE LIGAND FOR K+ CHANNELS M. Ferber, A. Sporning, G. Jeserich, R. DeLaCruz, M. Watkins, B. M. Olivera, H. Terlau Voltage-gated ion channels, especially K + channels determine the electrical excitability of cells. Although many Conus peptides that interact with voltage-gated Na + and Ca ++ channels have been characterized, relatively few have been identified that interact with K+ channels. Here we describe a novel Conus peptide that interacts with potassium channels. The mature peptide sequence was deduced from a Conus radiatus cDNA clone and the peptide chemically synthesized. The peptide has the same class III scaffold as the µ-conotoxins that block sodium channels and ψ-conotoxins, which are noncompetitive ACh-receptor antagonists. Despite its structural affinities, the C. radiatus peptide proved to have an entirely different pharmacological specificity: it affects the Shaker K+ channel and is therefore the defining member of a new family of Conus peptides, the κM-conotoxins. By using the Xenopus oocyte expression system and the two electrode voltage clamp technique, we demonstrate that this peptide (κMconotoxin RIIIk) despite its structural homologies to µ-conotoxins does not affect sodium channels. Instead it blocks Shaker K+ channels with an IC 50 of 1.21 ± 0.25 µM (n = 5). The Hill coefficient is ~1, suggesting that binding of a single toxin molecule is sufficient to inhibit the Shaker channel. Studies using Shaker K + channel mutants with single residue substitutions revealed that the peptide interacts with the pore region of the channel. Introduction of a negative charge at residue 427 (K427D) greatly increases the affinity of the toxin (IC50 = 109 ± 61 nM, n = 5). In contrast, the substitutions at two other key residues within the Shaker channel pore, F425 and T449, drastically reduced toxin affinity. The data are consistent with κM-conotoxin RIIIk blocking the conductance of the Shaker K + channel by interactions with the outer vestibule of the pore. Since Conus radiatus is believed to be a fish-hunting cone snail, our results suggest that the presumptive physiologically-relevant molecular target is a voltage-gated K+ channel in fish. We tested the Sha1 channel from trout, because its pore region is structurally very similar to the K427D Shaker mutant. As expected the TSha1 channel is effectively blocked by κM-conotoxin RIIIk. The affinity and binding kinetics of the block are state dependent, with an IC50 of about 20 nM for the closed state and 60 nM at 0 mV for the open state of TSha1 channels. These data demonstrate that most likely a teleost homologue of the Shaker K + channel is the natural target of κM-conotoxin RIIIk. Max-Planck-Institut für Experimentelle Medizin, AG Molekulare & Zelluläre Neuropharmakologie, Hermann-Rein-Str. 3, D-37075 Göttingen, Germany O 17-5 PROTEIN SYNTHESIS IS REQUIRED FOR K+ CURRENT ENβ 1-40 IN RAT CEREBELLAR GRANULE HANCEMENT BY Aβ NEURONES L.D. Plant, M. Ramsden, Z .Henderson, H.A. Pearson

S 188 Recent studies have described a role for amyloid beta protein (Aβ) as a modulator of ion channel activity in central neurones (Price et al.,1998; MacManus et al., 2000; Ramsden et al., 2001). Specifically, Aβ 1-40 augments the voltage-gated K+ channel current (IK+) of cerebellar granule neurones (CGN; Ramsden et al., 2001). We hypothesised that Aβ1imparts this effect via upregulation of protein synthesis and investi40 gated this notion using the whole-cell patch clamp technique. CGN were cultured as previously described (Ramsden et al., 2001). IK+ was evoked in cells held at -70mV by step depolarisations to between -60 and +70mV following a prepulse to -140mV. Peak IK+ evoked at +50mV was increased by 32±18% (n=23, p<0.05) in CGN incubated with 100nM Aβ1-40 for 24 hours, compared with reverse sequence peptide controls. This effect was due to specific augmentation of the inactivating component of IK+ (IKA), as the delayed rectifier component (Ikv) was unaltered. IKA augmentation was abolished when CGN were co-incubated with the protein synthesis inhibitor, cycloheximide (1µM). For example, peak IK+ was decreased by 15±16% (n=15). Importantly, cycloheximide treatment did not affect IKV in these cells. These data indicate that Aβ1-40 augments IKA in central neurones by a mechanism involving upregulation of protein synthesis. References: MacManus A, Ramsden M, Murray M, Henderson Z, Pearson HA, Campbell VA (2000). J. Bio .Chem. 275:4714-8; Price SA, Held B, Pearson HA (1998). Neuroreport 9:539-45; Ramsden M, Plant LD, Webster NJ, Vaughan PFT, Henderson Z, Pearson HA (2001). J. Neurochem 79:699-712 School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, United Kingdom

that endogenously released PTHrP improves post-ischemic recovery. Since PTHrP expression is regulated in an estrogen-dependent way, we analyzed the influence of PTHrP and PTHrP receptor antagonists on early post-ischemic recovery in hearts from female and male rats. Methods: Rat hearts were perfused in the Langendorff mode for 20 min, then exposed to 30 min no-flow ischemia, and reperfused for 30 min. Ventricular function was investigated as left ventricular developed pressure (LVP). To increase the concentration of PTHrP during early reperfusion, PTHrP(1-36) was added before the onset of ischemia. To inhibit the effect of locally released PTHrP, 5lle-PTHrP(136) was used to inhibit ventricular effects of PTHrP or PTHrP(7-34) was used to inhibit cardiac effects of PTHrP. Results: During the preischemic period 5lle-PTHrP reduced LVP in hearts from females by 16±9% (n=12,p<0.05) but not in hearts from male rats. PTHrP was without any effect. Post-ischemic recovery in control hearts reached 51%, 68%, and 72% of initial values after 5, 15, and 30 min, respectively. Addition of PTHrP did not further increase post-ischemic recovery. Inhibition of endogenously released PTHrP by 5lle-PTHrP reduced post-ischemic recovery to 18%, 38%, and 49%. This effect was more pronounced in female hearts than in male hearts. PTHrP(734) had a more pronounced inhibitory effect on post-ischemic recovery than 5lle-PTHrP. Conclusions: Endogenously released PTHrP improves the early post-ischemic recovery in rat hearts. This effect of PTHrP is more pronounced in hearts from female rats than in hearts from male rats. In summary, this study indicates a functional role for PTHrP released from coronary endothelial cells under energy-depleting conditions during reperfusion. In addition, our study shows that the estrogen-dependent increased expression of PTHrP in ventricles is of functional importance.

O 17-6 K.-D. Schlüter, Physiologisches Institut, Aulweg 129, 35392 Giessen HYPOXIA-INDUCED INHIBITION OF K-CHANNELS OF ALVEOLAR EPITHELIAL CELLS (A549 CELLS) C. Karle 1, T. Gehrig2, R. Wodopia2, S. Höschele2, V.A.W. Kreye3, P. Bärtsch 2, H. Mairbäurl2 In excitable cells hypoxia inhibits K channels, which causes membrane depolarization and initiates hypoxic responses. Here we studied, whether in lung alveolar epithelial cells (A549 cells) inhibition of ion transport by hypoxia might also be mediated by mechanisms that involve Kchannel modulation. A549 cells were exposed to acute hypoxia (1.5%O2) during the measurements. Membrane currents were recorded with the whole cell patch clamp technique. The activity of the Na/K pump and of Na/K/2Cl-cotransport was measured by unidirectional uptake of 86 Rb to test, whether K channel inhibition also affected ion transport activity. Hypoxia causes a 60% inhibition of total membrane current. No effect of hypoxia was seen, when cells were pretreated with TEA. TEA and iberiotoxin inhibit whole cell current of normoxic cells to a similar degree, whereas 4-aminipyridine and BaCl 2 were ineffective. Results of RT-PCR indicate the presence of mRNA encoding for maxiKchannels as well as for the voltage sensitive K-channels Kv1.2, Kv1.5 and Kv2.1. Measurements of unidirectional 86Rb uptake show that TEA (in normoxic control cells) and 4h of hypoxia inhibit 86Rb uptake by 40% and 35%, respectively; the effects seem to be additive. Increasing extracellular K to 20 mM also inhibits transport. The K channel opener 1-EBIO increases 86Rb uptake more than 2-fold in normoxic and hypoxic cells. These results indicate that K channels, most likely Ca-activated maxi K channels, are involved in the regulation of transepithelial transport, whereas typical voltage sensitive Kchannels, although expressed, seem not to be active. K channel inhibition might be one step required to initiate hypoxic inhibition of ion transport activity. Medical Clinic and Policlinic, Department of Internal Medicine, 1 Section III, Cardiology, 2Section VII, Sports Medicine, 3Department of Physiology and Pathophysiology, University of Heidelberg, 69115 Heidelberg, Germany O 18-1 ENDOGENOUSLY RELEASED PTHrP IMPROVES EARLY POST-ISCHEMIC RECOVERY IN RAT HEARTS K.-D. Schlüter, H. Degenhardt, S. Wenzel Coronary endothelial cells release Parathormon-related peptide (PTHrP) under energy-depleting conditions. Therefore, the local concentration of PTHrP increases during ischemia. The peptide hormone exerts positive chronotropic and inotropic effects and dilates vessels. Due to these effects of PTHrP on ventricular function we postulated

O 18-2 MYOCARDIAL INFARCTION IN MICE: FUNCTIONAL CHARACTERIZATION AND EXPRESSION OF CYTOKINES A. Deten, H.-G. Zimmer Large myocardial infarctions (MI) were induced in mice (20-25 g) by ligating the left coronary artery (LAD). Heart function was measured by right and left ventricular (RV and LV, resp.) catheterization in closed-chest spontaneously breathing mice after 6, and 12 h and 1, and 4 weeks. LV function was severely depressed after MI, whereas RV developed pressure (RV dev P) and RV maximal rate of rise in pressure (dP/dt max) were elevated 4 weeks after MI (Table 1). We further investigated the expression of cytokines in the infarct-area (MI-A), the border zone adjacent to the infarct (MI-B) and in the non-infarcted LV (LV) by ribonuclease protection assay. IL-6 was induced after 6 and 12 h in MI-A as well as in MI-B and LV (250-, 50- and 15fold in MI-A, MI-B and LV, resp.). This was accompanied by an increase in IL-6 receptor mRNA expression, while the expression of the signal transducing receptor component gp130 was unchanged. Also the mRNA expression of IFNγ and TNFα was upregulated 6 and 12 h after MI (for both IFNγ and TNFα 30-, 15- and 5-fold in MI-A, MIB and LV, resp.). None of these cytokines was elevated at later times after MI. Additionally, the expression of TGF-β1 increased in MI-A 3-fold after 12 h and was still significantly elevated after 4 weeks (2fold). In MI-B and LV, TGF-β1 expression increased later (2-fold at 1 and 4 weeks after MI). TGF-β2 showed a similar expression pattern, but the maximum increase occurred in MI-B (6-fold at 1 week after MI). Contrarily, TGF-β3 expression was increased later (after 1 and 4 weeks), predominantly in MI-A and MI-B (7-fold), but only slightly in the LV (2-fold). These changes in the expression pattern of TGF-βisoforms were accompanied by an increase in TGF-β receptor (TGFβR) expression. Especially TGF-βRII was elevated 2.5-fold in MI-A and MI-B after 1 and 4 weeks, while TGF-βRI only slightly increased in MI-B and MI-A (1.5-fold after 1 week). Table 1: Functional characterization (n.m. not measured; * p<0.05 v CTRL): CTRL (16) 6 h MI (4) 12 h MI (4) 1 week MI (4) 4 weeks MI (4)

HR LV dev P 435 ± 11,6 122 ± 2,1 403 ± 24,2 71 ± 5,8* 416 ± 12,6 64 ± 3,2* 393 ± 16,1 82 ± 4,1* 414 ± 21,7 99 ± 8,3*

LV dP/dt max 12.624 ± 677 4.621 ± 623* 4.498 ± 518* 5.427 ± 625* 6.867 ± 519*

RV dev P 27 ± 0,6 n.m. n.m. 33 ± 3,4 48 ± 4,9*

RV dP/dt max 2.508 ± 151 n.m. n.m. 2.752 ± 296 3.391 ± 333*

Carl-Ludwig-Institute of Physiology, Liebigstr. 27, D-04103 Leipzig

S 189 O 18-3

O 18-5

HYPOXIC PRECONDITIONING PROTECTS RIGHT BUT NOT LEFT VENTRICULAR FUNCTION FROM ISCHEMIA / REPERFUSION DAMAGE IN ISOLATED RAT HEARTS J. Forkel, X. Chen, S. Wandinger, S. Frede, P. Massoudy, R. Schulz, H. Jakob, G. Heusch Background. Chronic hypoxia may precondition the myocardium and protect from ischemia/reperfusion damage. A separate analysis of such protection for right and left ventricular function is not yet available. We therefore studied the functional recovery after ischemia and reperfusion in isolated working right and left hearts of chronically hypoxic rats. Methods. Male Wistar rats were housed in a hypoxic chamber (10.5% fraction of inspired oxygen) for 2 weeks. Control rats were housed in ambient air. Right and left blood-perfused isolated working heart preparations were used. All hearts performed pressure-volume work, then underwent warm ischemia, reperfusion and performed again pressure volume-work (15 min each). Functional right and left heart recovery was determined. Hypoxia-inducible-factor-1 (HIF-1) mRNA was measured. Results. The hematocrit was 46±1% in normoxic and 67±1% in hypoxic rats (p<0.01). Right ventricular weight was 18.9±1.0% of whole heart in normoxic rats and 21.8±1.1% in hypoxic rats (p<0.05). Expression of HIF-1 mRNA was increased 25-fold in hypoxic rat hearts. Postischemic left ventricular work of normoxic rats recovered to 69±5% of baseline. Left hearts of hypoxic rats had a recovery to 65±6% of baseline. Right hearts of normoxic rats had a recovery to 68±4% of baseline, right hearts of hypoxic rats recovered to 94±2% of baseline (p<0.05 vs. right heart recovery of normoxic rats). Conclusions. Hypoxic preconditioning enhanced postischemic right heart functional recovery. In contrast, with hypoxic preconditioning , there was no increased recovery of the left ventricle. The observed upregulation of HIF-1 mRNA in chronically hypoxic hearts indicates increased HIF-1 transcriptional activity. HIF-1 might play a critical role in the observed cardioprotection of right hearts .

ROLE OF MYOCARDIAL MAST CELLS IN RELEASE OF TNFα FROM HEARTS OF RAT AND MOUSE AFTER ISCHEMIA B.F. Becker*, S. Gilles*, L. Hültner+, J. Reil*, U. Welsch#, S. Zahler* An immediate, transient release of TNFα is observed during postischemic reperfusion of the human heart following bypass surgery. Myocardial mast cells have been proposed to store preformed TNFα and, thus, be the cellular source of this immediately released cytokine. In the present experiments we sought to compare hearts of rats, wildtype mice (with mast cells) and mice deficient in mast cells (kit w/w-v) with respect to release of TNFα into coronary effluent as opposed to transudate (interstitial fluid emerging on the epicardial surface), and immuno-histologically with respect to colocalisation of TNFα with mast cells. Measurements were performed on isolated hearts under basal flow conditions and after a 15 min period of global ischemia. Hearts of rats (n=7) released about 2-4 pmol TNF/min per gram of tissue into the interstitial space. Immediately after reperfusion, interstitial release rose to approx. 25 pmol/minxg and remained at this level for about 10min, before transiently declining to 10-15 pmol/ minxg after 30min. Coronary venous levels were always below detection. Hearts of mice were different: TNFα was present equally in transudate and in coronary effluent, total release amounting to about 10 pmol/minxg before ischemia and as much as 30 pmol/minxg after ischemia, however, with the majority now present in the venous effluent. Furthermore, there was no quantitative distinction between wild-type and kit w/w-v mice (n=5). Hearts of rats and wild-type mice possessed mast cells in the perivascular tissue. Immunoreactive TNFα colocalized to these cells, but in the mouse heart was also found diffusely in endothelial cells of some larger venous vessels. In the case of kitw/wv mice, no mast cells were present and immunoreactive staining seemed generalised in the myocardium. Localisation of TNFα in the heart differs between rats and mice, mast cells being of relatively little importance in the latter with more cytokine derived from endothelial cells, both before and after ischemia. Supported by SFB 469

Div. of Cardiothoracic Surgery, Dept. of Pathophysiology, Dept. of Physiology, Univ.-Klinikum Essen, Hufelandstr. 55, D-45122 Essen

*Department of Physiology, Schiller Str. 44, 80336 Munich; +GSF, Munich-Grosshadern; #Department of Anatomy, University of Munich, Germany O 18-6

O 18-4 ASCORBIC ACID INHIBITS ISCHEMIC PRECONDITIONING IN PIGS A. Skyschally, R. Schulz, P. Gres, M. Thielmann, G. Heusch Objective: The involvement of free oxyradicals in ischemic preconditioning in rats and rabbits has been established. Data from larger mammals which are closer to man are lacking. We have therefore tested the impact of the natural radical scavenger ascorbic acid on ischemic preconditioning in pigs. Methods: In 29 anesthetized pigs the LAD coronary artery was perfused from an extracorporal circuit. Subendocardial blood flow (radioactive microspheres) and myocardial infarct size (TTC-staining) were determined. Results: Ischemic preconditioning (IP) by 10 min ischemia followed by 15 min reperfusion reduced infarct size after 90 min severe ischemia and 120 min reperfusion (I/R). The pretreatment with ascorbic acid (A; i.v.; 30 min before I or IP, resp.; 2g bolus; maintenance dose 25mg/min up to the end of I) had no effect on infarct size per se (A+I/R), but abolished the infarct size reduction by ischemic preconditioning (A+IP+I/R). Infarct size Subendocardial blood [% area at risk] flow [ml/min/g] Group I/R (n=8) 25.5 ± 3.8 0.048 ± 0.01 IP+I/R (n=9) 5.2 ± 1.7 * 0.057 ± 0.01 A+I/R (n=3) 29.0 ± 4.7 0.044 ± 0.03 A+IP+I/R (n=9) 19.1 ± 4.0 # 0.045 ± 0.03 # Mean±SEM; * p<0.05 vs. I/R; p<0.05 vs. IP+I/R

Conclusion: Scavenging free oxyradicals with ascorbic acid prevents the beneficial effect of ischemic preconditioning on infarct size in pigs. Therefore, free oxyradicals are also involved in the signal transduction of ischemic preconditioning in pigs. Their role as triggers or mediators remains to be established. Abteilung für Pathophysiologie, Universitätsklinikum Essen, Hufelandstraße 55, 45122 Essen

CARDIAC FUNCTION AND GENE EXPRESSION IN MYOGLOBIN KNOCKOUT MICE SUBJECTED TO HYPOXIC STRESS G. Schlieper, J.-H. Kim, A. Molojawiy, A. Gödecke, J. Schrader Recently generated myoglobin knockout (myo -/-) mice show no obvious phenotype but revealed increases in hemoglobin, coronary flow, and cardiac capillary density as compensatory mechanisms. The aim of this study was to investigate whether myo -/- mice show a different gene expression pattern compared to wildtype controls (WT) and whether differences in phenotype arise from hypoxic stress. Myo -/- and WT mice were exposed to hypoxia (10% O 2) for 2 weeks. Thereafter functional parameters (heart rate, blood pressure) were measured and cardiac gene expression profile of 2352 genes was determined using cDNA arrays from Clontech (n=4-5). Real time PCR was performed on selected genes. Hypoxia reduced arterial pO2 by 50% in WT and myo -/- mice. Hematocrit increased by 65% in both groups. Hypoxia did not significantly change heart rate, systolic and diastolic pressure. WT and myo -/- mice were phenotypically not different during hypoxia. Cardiac gene expression of myo -/- mice differed significantly from WT mice in 15 genes under normoxic conditions (e.g. keratinocyte lipid binding protein: +420%; heat shock protein 86: -50%; cytochrome c oxidase Vb: +30%). After 2 weeks of hypoxia 53 genes were found to be significantly changed in WT hearts (e.g. procollagen 1 alpha subunits 1 and 2: +80%; glyceraldehyde-3phosphate dehydrogenase: +45%; lactate dehydrogenase 2 B subunit: –30%; S100 calcium binding protein A1: -55%). However the expression profile of myo -/- mice stressed by hypoxia was similar to the WT response. HIF 1 alpha was unchanged during hypoxia in WT and myo -/- mice. Differential gene expression was verified by real time PCR. Our data demonstrate changes in cardiac gene expression during hypoxia that reflect important structural and metabolic adaptations. Myo -/- mice are functionally and genetically well adapted even during hypoxia. Cardiovascular Physiology, Universitätsstr. 1, 40225 Düsseldorf

S 190 O 19-1

O 19-3

CONCENTRATIONS OF ANTIEPILEPTIC DRUGS IN THE EXTRACELLULAR SPACE OF FOCAL NEOCORTICAL AREAS, CEREBROSPINAL FLUID, AND IN BLOOD SERUM OF EPILEPTIC PATIENTS (EPILEPSY SURGERY; MICRODIALYSIS) B. Rambeck, U.H. Jürgens, H. Straub*, H.W. Pannek, F. Behne, D. Kuhlmann, E.-J. Speckmann The pharmacological treatment of epilepsies is commenly known to be not satisfactory. This may be due not only to minor efficacy of antiepileptic drugs (AED) but also to pharmocokinetic problems (c.f. multi drug transporters). The aim of the present investigation was to compare the concentration of AED in the extracellular space of the brain in vivo with those in blood and in cerebrospinal fluid (CSF). Extracellular fluid was obtained from patients (n=6) during surgical resection of epileptic foci using microdialysis. The microdialysis probe (cut off 20.000 kD; flow 1 µl/min) was placed into the cortex which had to be removed and fixed by a pressure foot. An equilibration period of 35 min was allowed to obtain basal drug levels. Sampling was performed in 10 min fractions. Parallel to the micodialysis blood and CSF were taken. Samples were analyzed by HPLC with UV detection. The intraoperative microdialysis was performed in patients treated with carbamazepine (n=2), lamotrigine (n=2) and oxcarbamazepine (n=2). The concentrations of carbamazepine, lamotrigine and 10-OHcarbamazepine in the in vivo dialysate were in the same order of magnitude as those of dialysates of serum and CSF as well as of the ultrafiltrates of both. The measured concentrations of the in vivo dialysates were 40 – 50% of those of the free, i.e. not protein bound, drugs found by direct analysis of the serum. The results point to an equilibrium of concentrations of free drugs in the three compartements mentioned.

TRANSIENT HYPEREXCITABILITY OF CA1 PYRAMIDAL CELLS IN RAT HIPPOCAMPUS AFTER CLOSED HEAD INJURY D. Griesemer, A.E.M. Mautes+, B. Neumcke Traumatic brain injury frequently elicits epileptic discharges originating in the hippocampus. To study this hyperexcitability on the cellular level, we analysed the effects of closed head injury (CHI) on neurones in the CA1 field of the rat hippocampus. CHI was induced under ether anaesthesia by a calibrated weight drop on the left skull of young rats (p11-p15). At different times after CHI, ipsilateral hippocampal slices were prepared and CA1 pyramidal cells and interneurones were studied under current clamp in the whole-cell patch configuration. Bicucullin was added to the bath to block GABA-mediated synaptic transmission. We found a dramatic increase in the frequency of spontaneous action potentials of CA1 pyramidal cells (control: 1.67 ± 0.95 Hz, n=7, mean ± SEM) after CHI compared to sham-operated animals. This hyperexcitability was observed as early as 2 h after CHI (AP frequency after CHI: 2.62 ± 0.54 Hz, n=11; sham: 0.57 ± 0.39 Hz, n=7), was maximal at 24 h (CHI: 3.00 ± 0.49 Hz, n=19; sham: 0.31 ± 0.16 Hz, n=6) and disappeared after 3 d (CHI: 1.57 ± 0.33 Hz; n=26; sham: 1.23 ± 0.25 Hz, n=21). CA1 interneurones of the stratum radiatum responded by a much weaker increase of AP frequency after CHI. Resting membrane potentials of CA1 pyramidal cells (-56.5 ± 1.0 mV, n=7) and interneurones (-57.6 ± 1.0 mV, n=7) were hardly affected by CHI. Our results show that CHI has a stimulatory and ether anaesthesia a depressing effect on the excitability of CA1 pyramidal cells. The strong hyperexcitability of CA1 pyramidal cells after CHI is cell-type specific and transient in contrast to the persistent seizure susceptibility of neurones in the dentate gyrus after traumatic brain injury (G. Golarai et al: J Neurosci 21, 8523 (2001)).

Institut für Physiologie, Robert-Koch-Str. 27a, 48149 Münster, Germany

Physiologisches Institut, + Neurochirurgisches Forschungslaboratorium, Universität des Saarlandes, D-66421 Homburg O 19-4

O 19-2 SYSTEMIC TREATMENT WITH THE PYROGENS IL-6 AND LPS INDUCES NUCLEAR STAT3 TRANSLOCATION PREDOMINANTLY IN THE RAT BRAIN VASCULAR ORGAN OF THE LAMINA TERMINALIS E.-M. Harré, T. Hübschle, M. Kueth, U. Pehl, J. Roth, R. Gerstberger Interleukin-6 (IL-6) represents an endogenous pyrogen and is known to mediate fever responses induced by the exogenous pyrogen lipopolysaccharide (LPS). IL-6 is thought to act on the brain at sites which lack a blood-brain barrier, the circumventricular organs (CVOs). Cells that are activated by IL-6 respond with nuclear translocation of the ”signal transducer and activator of transcription 3” (STAT3) molecule and can be detected using immunohistochemistry. Therefore we investigated whether the LPS-induced release of IL-6 into the systemic circulation or rats was accompanied by nuclear STAT3 translocation within the sensory CVOs, the area postrema, the vascular organ of the lamina terminalis (OVLT) and the subfornical organ. In physiological experiments, systemic administration of LPS (100µg/kg i.p.) induced a biphasic febrile response which lasted about 8 h and started to develop 2 h after LPS injection. At the same time plasma IL-6 slightly rose 1 h after LPS treatment, then showed a strong increase within the following hours and declined at the end of the febrile response. In histological experiments nuclear STAT3 translocation in response to LPS was predominantly observed within the OVLT 2 h after LPS treatment. In order to investigate whether this effect was truly mediated by IL-6, the cytokine itself was systemically applied (40µg/kg i.p.). Indeed an almost identical pattern of nuclear STAT3 translocation was observed, however, already 1 h after IL-6 treatment. Our observations represent the first molecular evidence for an IL-6-induced STAT3-mediated genomic activation of OVLT neurons and support the proposed role of this brain area as a sensory structure for humoral signals created by the activated immune system and resulting in the generation of fever. Veterinary-Physiology, Justus-Liebig-University Giessen, Frankfurter Str. 100, D-35392 Giessen, Germany

ION TRAP MASS SPECTROSCOPY SHOWS THAT THE NA + CHANNEL BLOCKER QYNAD EXISTS IN THE CSF OF MS PATIENTS BUT, IS ABSENT IN THE CSF OF ALS PATIENTS P. Aulkemeyer1, R. Rüdel1, H. Tumani2, H. Brinkmeier1 Ion trap mass spectroscopy allows one to determine mass, amino acid sequence and concentration of small peptides dissolved in body fluids with high specificity at picomolar resolution. An enormous advantage of the method is that no extensive purification of the test fluids is required. We have used this method to determine more exactly the CSF concentration of QYNAD, a pentapeptide with Na+ channel blocking properties that we had discovered in samples of CSF from patients with inflammatory demyelinating nerve diseases such as multiple sclerosis or Guillain-Barré syndrome (Brinkmeier et al., Nature Med. 6:808811, 2000). Using the much more cumbersome HPLC method we had determined the CSF concentrations in eight MS patients between 9 and 34 µM. In 8 control patients the QYNAD concentration was determined to 3.2 ±1 µM (Brinkmeier l.c.). Ion trap determinations with the CSFs from eight previously not investigated MS patients yielded a very similar range of 11 to 34 µM. Re-analysis of the previously HPLC-analysed control CSFs showed that with the more accurate ion trap method no QYNAD could be detected. The existence of QYNAD in the CSF could, therefore, be specific for inflammatory demyelinating nerve diseases. To test whether QYNAD is existent in the degenerative nerve disorder of amyotrophic lateral sclerosis (ALS), we investigated the CSFs from five ALS patients with the ion trap. The results were inconsistent, four times 0 and once 195 µM. More experiments are required to determine the specificity of QYNAD in the CSF of neurological patients. Supported by DFG, Br-1139/4. 1 Dept.of General Physiology, University of Ulm, D-89069 Ulm, 2Dept of Neurology, University Hospital, University of Ulm, Oberer Eselsberg, D-89081 Ulm

O 19-5 NEUROTROPHIN RECEPTOR BLOCKADE INDUCES DIFFERENTIATION IN RETINOBLASTOMA CELLS N. Wagner1, K.D. Wagner1, H. Scholz 1, M. Sefton 2, A. RodriguezTébar2, R. Grantyn 2

S 191 Purpose. To clarify the expression of neurotrophins and their receptors in retinoblastoma (Rb) cells, to elucidate their potential role in the proliferation of neuroectodermal tumor cells, and to establish conditions for Rb cell differentiation. Methods. The Rb-derived cell line Y-79 was grown in serum-free suspension or monolayer culture. Proliferating and differentiated cells were isolated and submitted to semiquantitative reverse transcription-polymerase chain reaction (RTPCR) analysis, immunostaining, and flow cytometry. The proliferation rate of the cells was estimated by 5-bromo-2’-deoxyuridine (BrdU) incorporation, and the effects of neurotrophins and laminin on BrdUincorporation, process outgrowth, or immunostaining were determined. Results. In contrast to previously studied normal retinal precursor cells, Y-79 cells not only express nerve growth factor (NGF), brainderived neurotrophic factor (BDNF), and neurotrophin-3 (NT-3), but also the corresponding high affinity receptors TrkA, TrkB, and TrkC, and the low affinity p75 receptor. Proliferation was stimulated by exogenous and endogenous neurotrophin receptor ligands. Inhibition of protein kinase phosphorylation with K252a blocked proliferation and promoted differentiation. The effect of K252a on differentiation was enhanced by the addition of soluble laminin. After 9 days of combined treatment, the fraction of differentiated cells amounted to 30%, differentiation being characterized by improved attachment, neurite outgrowth, expression of NF-68, and a loss of glial fibrillary acidic protein (GFAP) and parvalbumin immunoreactivity. These changes were accompanied by a downregulation of TrkB and TrkC, but not TrkA or p75. Differentiated cells were isolated and further grown in the absence of K252a. However, despite the high level of TrkA expression in differentiated cells, the addition of NGF had no effect on their survival. Conclusions. A mitogenic action of exogenous and / or endogenous neurotrophins may contribute to retinal tumor growth.

elastic muscle protein titin. Here we studied functional properties of the PEVK-domain of cardiac-muscle titin by using a hybrid experimental approach. First we measured the force-extension behaviour of cloned cardiac PEVK-titin by single-molecule atomic force microscopy (AFM). The results then were compared to the extensibility of PEVKtitin measured by immunoelectron microscopy in intact rabbit cardiac sarcomeres. We found that the data obtained in AFM measurements could be readily applied to explain PEVK extensibility in situ. The analysis revealed that cardiac PEVK acts as an entropic spring with the properties of a random coil exhibiting mechanical conformations of different flexibility. Since in situ, titin is in close proximity to thin filaments, we also studied whether the PEVK-domain may interact with actin filaments. Interaction was indeed seen in the in-vitro motility assay, in which recombinant cardiac-PEVK constructs slowed down the sliding velocity of actin filaments over myosin. The PEVK effect was partially suppressed in the presence of physiological concentrations of Ca2+. Co-sedimentation assays confirmed a weak Ca2+-modulated actin-binding propensity of the cardiac PEVK-domain. Furthermore, we demonstrate that, in cardiac myofibrils, the actin-PEVK interaction gives rise to a viscous force component opposing filament sliding. Thus, the PEVK-domain contributes not only to the extensibility of cardiac myocytes, but also affects contractile properties. 1

University of Heidelberg, Institute of Physiology and Pathophysiology, Im Neuenheimer Feld 326, D-69120 Heidelberg; 2Dept. of Physiology and Biophysics, Mayo Clinic Rochester, MN, U.S.A.; 3Max-Planck Institute of Medical Research, Heidelberg, Germany; 4Max-PlanckInstitute of Molecular Physiology, Dortmund, Germany O 20-2

1

Institute of Physiology, Medical Faculty (Charité), Humboldt-University, Berlin, Germany; 2Department of Developmental Neurobiology, Ramon y Cajal Institute for Neurobiology, Consejo Superior de Investigaciones, Madrid, Spain O 19-6 SPATIAL PATTERNS OF NEURONAL ACTIVITY IN GLIOMAINVADED NEOCORTEX R. Köhling1, W. Paulus2, W. Senner2, H. Straub1, E.-J. Speckmann1 Gliomas often cause epileptic seizures as one of their first symptoms. It remains largely unknown which functional changes of neuronal activity are responsible for this, and to which extent the degree of tumour invasion correlates to the changes in neuronal excitabiltiy. To address the question whether the spread of activity in neuronal networks is altered by neoplastic glial cells, in a first pilot study, optical imaging was used to detect spatial patterns of stimulus-evoked responses in neocortical slices. Gliomas were induced in adult Wistar rats (n=6) by stereotactic intracortical implantation of a C6 tumour cell line suspension (10 µl; cell concentration: 2x105/µl) transfected with EBFP or lacZ. Gliomas were left to develop for 14-21 days. After this period, animals were sacrificed, and coronal neocortical slices (500 µm) were made. Evoked neuronal activity (single electical stimuli, pial surface) was monitored using the voltage-sensitive dye RH 795. Age-matched animals (n=6) served as controls. Implantation of tumour cell suspension led to development of intracortical gliomas in all cases, with both solid tumours and a surrounding 200-300 µm wide zone of dispersed invasion. Single electrical stimuli of pial surface resulted in homogenous, primarily supragranular activation in control tissue. By contrast, in glioma-invaded slices, network activity was either reduced, or increased, in the latter case extending to both supra- and infragranular layers. These results show that glioma infiltration causes substantial alterations in network activity patterns. Supported by DFG KO 1779/4-1. Institut für Physiologie, Bereich Neurophysiologie, Universität Münster, Robert-Koch-Str. 27a, 48149 Münster, Germany

TITIN EXPRESSION IN ADULT AND NEONATAL RAT FAST AND SLOW TWITCH MUSCLES G. Mutungi Previously, we have shown that the tension responses to a ramp stretch, in resting (=relaxed) adult (Mutungi & Ranatunga, 1996) and neonatal (Mutungi & Ranatunga, 2000) rat skeletal muscle fibres, consist of a viscous, a visco-elastic and an elastic component. In adult rats, all three tension components show characteristic fast/slow fibre type differences (Mutungi & Ranatunga, 1996) that are absent in the neonate (Mutungi & Ranatunga, 2000). From the behaviour of the viscoelastic component under various experimental conditions, we suggested that it arises from characteristic perturbations of the gap-filament and that the fast/slow fibre type differences may result from variations in the lengths of the I-band region of the various titin isoforms. Although, fast and slow mammalian muscles are known to express different titin isoforms (Labeit & Kolmerer, 1995); it is not clear whether the various fibre types, within a muscle, express the same or different titin isoforms. The aim of this study was to investigate the titin isoforms expressed in adult and neonatal rat, fast (extensor digitorum longus, edl) and slow (soleus) twitch muscles. Immediately after isolation, the muscles were snap-frozen and pulverised to a fine powder in liquid nitrogen, solubilised in SDS-containing buffer and finally analysed using 2-4% polyacrylamide/SDS gels. Peptide maps from 12 adult and 12 neonatal rats revealed that the adult slow and to some extent fast muscles had at least three titin bands. However, only two titin bands were seen in the neonatal muscles and the slow migrating band seen in the adult muscles was missing. Taken together with the myosin isoforms present in these muscles, these results suggest that the various fibre types may each express a fibre-type specific titin isoform. References: Labeit, S. & Kolmerer, B. (1995). Science 270, 293-296; Mutungi, G. & Ranatunga, K.W. (1996). Journal of Physiology 496, 827-837; Mutungi, G. & Ranatunga, K.W. (2000). Journal of Physiology 525.P, 86-87. I thank The Wellcome Trust for financial support. Department of Physiology, University of Bristol, Bristol BS8 1TD, UK O 20-3

O 20-1 STRETCH AND ATTACH: THE PEVK DOMAIN OF TITIN IS AN ENTROPIC SPRING WITH ACTIN-BINDING PROPERTIES W.A. Linke1,2, M. Kulke 1, H. Li2, S. Fujita-Becker 3, C. Neagoe1, D.J. Manstein3, M. Gautel4, J.M. Fernandez2 The PEVK-domain is a unique sequence contained within the giant

TITIN ISOFORM SWITCH IN ISCHEMIC HUMAN HEART DISEASE C. Neagoe1, M. Kulke1, F. del Monte2, R. Hajjar2, W.A. Linke1 Ischemia-induced cardiomyopathy usually is associated with elevated left ventricular end-diastolic pressure (LVEDP), which follows from increased myocardial stiffness due to increased collagen expression.

S 192 Whereas collagen is a main contributor to left ventricular wall stiffness, the giant sarcomere protein titin is another. Interestingly, human cardiac muscle co-expresses, at the level of the sarcomere, two main titin isoforms of different length, a longer N2BA isoform and a shorter N2B variant. In this study, we compared the N2BA:N2B titin isoform expression ratio in left ventricular samples of normal human hearts (n=8) to that of ischemic human hearts from coronary artery disease (CAD) patients (n=8). By using unconventional gel electrophoresis and Western blotting, we demonstrate that CAD hearts contain higher relative amounts of N2BA titin than normal hearts. The titin isoform shift correlated with the appearance of ischemia markers, i.e., complexes containing cardiac troponin I. Passive tension was measured in isolated cardiac myofibrils obtained from human hearts with known LVEDP. Myofibrils from two CAD hearts with high LVEDP (32 and 25 mm Hg) showed a reduced passive-tension response to stretch, compared to myofibrils from two human hearts with normal LVEDP (8-12 mm Hg). Immunofluorescence microscopy of tissue sections from both normal and diseased human hearts, using antibodies to titin and collagen, showed increased expression of collagen and confirmed elevated N2BAtitin expression. We conclude that in ischemia-induced cardiomyopathy, expression of a more compliant titin isoform lowers the stiffness of myofibrils, whereas collagen accumulation around myocytes stiffens the diseased tissue. Concerted modifications of titin and collagen expression in diseased myocardium is likely to reduce the ability of the heart to respond to higher diastolic stretch with increased active force generation. 1

University of Heidelberg, Institute of Physiology and Pathophysiology, Im Neuenheimer Feld 326, D-69120 Heidelberg, Germany; 2CVRC, Cardiology Division, Massachusetts General Hospital, Boston, U.S.A O 20-4 CALCINEURIN-NFATC1 SIGNAL TRANSDUCTION DURING FAST-TO-SLOW TRANSFORMATION IS TRIGGERED BY CALCIUM TRANSIENTS IN ELECTROSTIMULATED MYOCYTES H.-P. Kubis, N. Hanke, R. Scheibe, J.D. Meißner, G. Gros The calcineurin-NFATc1 signal transduction is involved in mediating the switch from fast to slow fiber type gene expression in skeletal muscle cells. Specifically, the expression of the slow myosin heavy chain I (MHC I) gene is positively correlated with the activity of this signal transduction pathway. From T-cells it is known that the calcineurin-NFAT pathway is activated by a long-lasting elevation of [Ca2+]i . In skeletal muscle, it has not been shown which type of calcium signal is necessary to activate the calcineurin-NFATc1 cascade. We electrostimulated primary skeletal muscle cells grown on microcarriers for different time intervals to activate the calcineurin-NFATc1 system and to induce fast-to-slow transformation. The effectiveness of the stimulation procedure with respect to fiber transformation was verified by Northern blot analysis of MHCI and fast myosin heavy chain IId (MHC IId) expression. Activity of the calcineurin-NFATc1 pathway was established by NFATc1 immunofluorescence, which is found inside the nuclei when the system is activated. Resting [Ca 2+]i was measured using the Fura-2 fluorescence technique. Electrostimulation of myocytes for one or two weeks induced an increase of MHC I and a decrease of MHC IId protein expression and led to an increase in resting [Ca 2+] i by ~50 % compared to control cells. A stimulation period of only 24 hours (15 min stimulation at 1 Hz followed by 30 min pause; continuous repetition of this cycle) also induced a distinct fiber transformation at the MHC I mRNA level and revealed a full activation of the calcineurin-NFATc1 pathway, but this was not associated with an increase in resting [Ca2+]i. Similarly, a single stimulation train of 20 minutes led to a nearly complete nuclear import of NFATc1 but did not influence resting [Ca2+]i. We conclude that the calcineurin signal transduction pathway in these muscle cells is activated by a 20 min series of short calcium transients, as they are associated with excitation-contraction coupling, rather than by a continuous elevation of resting calcium concentration. Zentrum Physiologie, Medizinische Hochschule, D-30623 Hannover

Slowly activating L-type calcium currents (ICa) were recorded using the Two-Micro-Electrode Voltage Clamp (2-MVC, Friedrich et al., 1999, J.Physiol. 517) in three different mouse strains: C57/SV129 control mice, mdx mice (an animal model for Duchenne Muscular Dystrophy, DMD) and transgenically engineered mini-dystrophin containing mice which have undergone gene therapy for DMD by introducing a socalled mini-dystrophin as substitute for membrane bound full length dystrophin (e.g. Phelps et al., 1995, Hum.Mol.Gen. 4). The activation and inactivation properties of ICa were examined in an age dependent manner for animals from two to 18 months of age. Mdx mice had smaller peak ICa amplitudes increasing during the age of the animals and accounted for about 30 % of ICa in young animals (~ -23 µA/ cm²), 50 % in the mediate age group (~ -35 µA/cm²) and about 80 % in aged mdx mice (~ -50 µA/cm²) compared to controls of same age. Examining the minimum peak ICa in all age groups a second population of mdx fibres became apparent showing nearly normal minimum peak ICa values in aged mdx mice probably reflecting a growing percentage of ‘revertant’ fibres. Time-to-peak (TTP) was prolonged in mdx fibres compared to controls. Mini-dystrophin fibres always showed similar activation kinetics, peak ICa values and TTP as in controls of the same age. Voltage dependent inactivation and channel closure kinetics were, however, not greatly altered in mdx mice and thus comparable in all three genotypes. Intracellular calcium concentrations as judged from fura-2 ratio measurements were not different in young animals of all three genotypes where the discrepancy of peak ICa was greatest. We conclude, that cytoskeletal components such as dystrophin may have a marked influence of L-type calcium channels, i.e. a retarded maturation as judged from the decreased peak ICa and prolonged TTP. Inactivation kinetics do not seem to be altered by lack of dystrophin. Interestingly, calcium channel properties seem to be fully restored in mini-dystrophin containing fibres, indicating that mini-dystrophin might be a potential candidate for effective gene therapy in DMD. Medical Biophysics, Institute of Physiology & Pathophysiology, INF 326, Ruprecht-Karls-University, Heidelberg, Germany

O 20-6 DYNAMIC FATIGUE ASSESSMENT DURING REPETITIVE ISOTONIC AND QUASI-ISOMETRIC CONTRACTIONS OF RAT SKELETAL MUSCLE IN VITRO P. Vedsted, A.H. Larsen*, K. Madsen*, G. Sjøgaard The most commonly used fatigue variable is the decline in force, but dynamic fatigue variables may be more valid especially in the study of shortening contractions. We studied fatigue development following isotonic (isot.) and quasi-isometric (q-isom.) contractions in rat soleus (SOL) and extensor digitorum longus (EDL) muscles in vitro, using the same stimulation protocol (60 Hz, 400 ms/s) for 100s in SOL and 60s in EDL. Fatigue was quantified as the relative decline in shortening (D), shortening velocity (dD/dt), and relaxation (-dD/dt) during the isot. contractions and correspondingly for the q-isom. contractions as the relative decline in peak force (Fpeak), peak rate of force development (dFpeak/dt), and relaxation (-dFpeak/dt). For both contraction modes the decline in work per contraction could be assessed. Test contractions (60 Hz, 1.5 s) performed before and after fatigue were analysed for relative decline of maximal force development (Fmax) and maximal rate of force development (dFmax/dt). Results are summarized in the table as post-fatigue values in percentage of pre-fatigue: EDL (isot.) EDL SOL (isot.) SOL (q-isom.) (q-isom.) Fmax 30.4±4.1%a 33.3±3.6% 58.5±6.4% a, 64.4±7.0% dFmax/dt 19.1±4.3% a,b 22.3±3.2%b 46.3±9.9% a,b 52.3±8.5% b D or Fpeak 0.7±0.6% a 18.3±2.6% 17.0±3.1% a 35.6±5.8% dD/dt or dFpeak/dt 6.1±1.3% a 19.9±3.6% 25.8±5.2% a 39.5±7.5% -dD/dt or -dFpeak/dt 3.6±0.9% a 8.0±1.5%c 10.7±2.2% a 13.5±2.2%c Work per contraction 0.7±0.6% a 3.0±0.8%c 17.0±3.1% 11.8±3.6%c Values are mean±SD. a significantly lower than quasi-isometric, b significantly lower than Fmax, c significantly lower than Fpeak (p<0.05).

O 20-5 CALCIUM CURRENTS IN SINGLE MUSCLE FIBRES OF MDX MICE AND TRANSGENIC MINI-DYSTROPHIN MICE O. Friedrich, J.M. Gillis, J.S. Chamberlain, R.H.A. Fink

Test contractions declined to a lower level in Fmax and dFmax/dt after isot. vs. q-isom. fatiguing contractions, and the dynamic fatigue variable dFmax/dt declined to a significant lower level than Fmax. When comparing fatigue development during the two fatiguing contraction

S 193 modes, the respective fatigue variables again decreased more rapidly and to lower levels during isot. vs. q-isom. contractions. During the qisom. fatiguing contractions the relative decline in Fpeak significantly underestimated fatigue development compared to the decrease in the dynamic variables -dFpeak/dt and ”work per contraction”, while this was not the case during isot. contractions. Conclusions: Fatigue development was significantly larger during isot. vs. q-isom. contractions, in concert with expected differences in the Fenn-effect. The use of force as the only experimental fatigue variable may in a contraction mode dependent manner underestimate the degree of fatigue development, neglecting the fatigue effect on time and length dimensions. National Institute of Occupational Health, Lersø Parkallé 105, 2100 Copenhagen Ø, Denmark; * Institute of Sport Science and Clinical Biomechanics, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark

(13.6±1.6 %, n=9) in ventricular cells but neither in atrial (n=8) nor in pacemaker-like cardiomyocytes (n=13). By contrast, intracellular dialysis with the catalytic subunit of PKA (7 µM) enhanced I f in pacemaker (11.4±2.2 %, n=5) and atrial cells (21.7±3.35 %, n=10), whereas no effect was observed in ventricular cells (n=10). In line with these findings 8-Br-cAMP exerted no additional upregulation of If in early stage atrial (n=10) and pacemaker (n=5) cells pretreated with the catalytic subunit of PKA. This was in clear contrast to late stages of development, where in the atrial and pacemaker-like cardiomyocytes only weak stimulation of If by the catalytic subunit of PKA but stronger enhancement after subsequent application of 8-Br-cAMP by 7.9±1.8 % (n=8) and 30.1±7.2 % (n=6), respectively, was observed. Thus, differential modulation of If is seen during embryonic development. The underlying mechanism is being further investigated. Institute of Neurophysiology, University of Cologne, Robert Kochstr. 39, D-50931 Cologne, Germany, +Department of Cardiac Surgery, University of Bonn

O 21-1 PROTEIN KINASE A PHOSPHORYLATION POTENTIATES VR1 RESPONSES TO NOXIOUS HEAT IN HEK293T CELLS P.K. Rathee, C. Distler, O. Obreja, C. Nau, G.K. Wang, S.Y. Wang, M. Kress The vanilloid receptor 1 (VR1) is one of the heat transducers in nociceptive neurons contributing to inflammatory pain and hyperalgesia. Several inflammatory mediators sensitize nociceptors to noxious heat by activating G-protein coupled receptors and the cAMP/protein kinase A (PKA) cascade. Here we investigated the role of cAMP/PKA on the sensitivity of VR1 to noxious heat and the possible involvement of A kinase anchoring proteins (AKAP). Whole cell patch clamp recordings were performed in human embryonic kidney cells (HEK293t) cells transiently transfected with rat wild-type VR1 or VR1 where PKA phosphorylation sites (T144, T370, S502) were mutated using sitedirected mutagenesis. Cells were exposed to 5 s heat stimuli at 1 min intervals and forskolin (FSK) was used as a conditioning intermittent stimulus. Exposure to FSK (10 µM for 1 min) resulted in a 4-fold increase of heat responses from 345 pA to 1597 pA in cells expressing wild type VR1. This potentiation was significantly reduced in the presence of the specific PKA inhibitor PKI14-22 or after addition of the StHt31 peptide to the intracellular solution, which inhibits PKA AKAP interaction. The negative control peptide StHt31-P was ineffective. HEK293t cells expressing the VR1 mutants still exhibited heat responses, however, there was a significant reduction in FSK induced potentiation of heat activated currents. RT-PCR and indirect immunocytochemistry showed the expression of PKA subunits as well as AKAPs in HEK293t cells. These results suggest that VR1 is phosphorylated at the predicted PKA consensus sites and that phosphorylation at these sites results in potentiation of heat responses. This requires the functional anchoring of the PKA via an AKAP in close vicinity to the VR1 channel protein. This work was supported by DFG (SFB 353/A10) and Wilhelm-SanderStiftung Dept. Physiol. & Exp. Pathophysiol., Universitaetsstr. 17, D-91054 Erlangen, Germany O 21-2 MODULATION OF THE HYPERPOLARIZATION ACTIVATED NON-SELECTIVE CATION CURRENT IN MURINE EMBRYONIC CARDIOMYOCYTES DURING DEVELOPMENT Z.J. Lu, JB. Zhang, +W. Roell, P. Sasse, J. Hescheler, B.K. Fleischmann The hyperpolarization activated non-selective cation current (I f ) appears involved in the generation of the spontaneous electrical activity of pacemaker cells in heart. If is regulated by the autonomous nervous system and has been shown to be directly stimulated by cAMP. Our group has recently found that If is enhanced by protein kinase A (PKA)mediated phosphorylation in early stage embryonic stem cell-derived cardiomyocytes. Aim of the present study was to investigate the modulation of If in early (E9.5-E11.5) and late stage (E16.5-E18.5) embryonic cardiomyocytes. For this purpose murine hearts were harvested at different time points of development and the hormonal modulation of If assessed with the whole cell voltage clamp technique. Early stage murine cardiomyocytes are spontaneously beating and express I f at a high density (6.4±0.6 pA/pF, n=44) whereas at later stages If is downregulated (1.9±0.3 pA/pF, n=20). At the early stage extracellular application of 8-Br-cAMP (200-400µM) stimulated If

O 21-3 UP- AND DOWN-REGULATION OF AN INSECT NEURONAL α-BUNGAROTOXIN-RESISTANT NICOTINIC RECEPTOR SUBTYPE BY TWO DISTINCT PROTEIN KINASES C R. Courjaret, F. Grolleau, B. Lapied Compared to their vertebrate counterparts, the intracellular regulation of insect neuronal nicotinic receptors (nAChRs) is quite unknown. However, in a recent work we began to identify phosphorylation/ dephosphorylation process of nAChR1, a subtype of α-bungarotoxinresistant nAChRs expressed in Dorsal Unpaired Median (DUM) neurones isolated from the terminal abdominal ganglia of the cockroach (Periplaneta americana) (Courjaret and Lapied, 2001; Mol. Pharmacol. 60: 80-91). Using whole-cell patch-clamp technique adapted on shortterm cultured DUM neurones, we revealed that pressure application of nicotine (10 mM, 100 ms in duration), performed in the presence of 500 nM α-bungarotoxin, activated nAChR1 at steady-state holding potentials ranging from –90 mV to –30 mV. This nAChR1, sensitive to d-tubocurarine (10 µM) was 1) regulated via a cAMP/protein kinase A cascade and 2) directly and indirectly modulated by a CaMKII via an inhibition of PP1/2A activity. In the present study, we performed additional experiments to determine if nAChR1 was also regulated by other intracellular messenger pathways involving protein kinase C (PKC). Application of the diacylglycerol analog DiC8 (50 µM) produced a reduction of the current amplitude (49.4 ± 4.8 %, mean ± S.E.M., n=11), whereas the phorbol ester PMA (1 µM) increased the nicotinic response (101.6 ± 6.0 %, n=18). Both effects were suppressed by the PKC inhibitor, chelerythrine (20 µM). The effect of PMA was strongly reduced when the [Ca2+]o was decreased from 5 mM (control condition) to 1 mM. By contrast, when [Ca2+]i was increased with caffeine (5 mM), PMA reduced the current amplitude by 49.6 + 8.8 % (n=6) and the effect of DiC8 previously observed was potentiated. Under high [Ca2+]i (100 µM in the pipette), the current amplitude was strongly reduced (70.2 ± 4.1 %, n=9) and no additional effect of DiC8 was observed. Together, these results suggested that two PKCs, differing in their [Ca2+]i sensitivity, were involved. This was confirmed by bath application of muscarine. Muscarine (1µM) increased the current amplitude by 116.6 + 12.1 % (n=4). This increase was less important with higher concentration of muscarine (100 µM). These effects inhibited by pirenzepine (1 µM), the phospholipase C (PLC) inhibitor U73122 (10 µM) and heparine (50 µM) might reflect the small and high elevation of [Ca2+]i produced by 1 µM and 100 µM muscarine (⊗[Ca 2+]i 32.8 + 8.2 nM, n=4 and 736.1 + 22.1 nM, n=3, respectively). We demonstrate here that two PKCs up- and down-regulate nAChR1. Université d’Angers Laboratoire de Neurophysiologie (RCIM) UPRES EA 26472, Bd Lavoisier Angers, France O 21-4 INHIBITION OF SMOOTH MUSCLE MYOSIN LIGHT CHAIN KINASE (MLCK) ACTIVITY AND CONTRACTION BY P21 ACTIVATED PROTEIN KINASE1 (PAK1) A. Wirth, M. Schroeter, E. Manser, P. de Lanerolle, G. Pfitzer The monomeric GTPase Rac is activated by reactive oxygen species (ROS) in vascular smooth muscle. Effectors of Rac are PAKs, a pro-

S 194 tein kinases familiy consisting of 3 mammalian isoforms, which have multiple effects on cytoskeletal organization in non-muscle cells. Here, we investigated whether PAK1 affects smooth muscle contraction directly in a Ca2+-independent manner. In triton skinned carotid arteries, PAK1 significantly inhibited submaximal Ca2+-activated force. The mechanism of inhibition was further elucidated in taenia coli. After eliciting a control submaximal (pCa 6.79) and maximal (pCa 4.35) contraction (F max), triton skinned taenia coli was incubated in relaxing solution containing 10 nM okadaic acid with recombinant PAK1 (72 µg/ml) or buffer for 1 hr (time control). Force elicited at pCa 6.79 after incubation with PAK1 and in the time controls was 31 ±10% and 62 ±8% of F max respectively (p<0.05). Steady state myosin regulatory light chain (MLC20) phosphorylation at pCa 6.79 was 41 ±13% in control and 21 ±6% in PAK1 treated strips (p<0.05). Heat denatured PAK1 did not inhibit force. Inhibition of force was reversible upon washing out PAK1 from the strips. Treatment with PAK1 in the presence of [32P]ATP exhibited an increase in 32P-incorporation into MLCK but not into caldesmon immunoprecipitated from the strips. The rate of contraction and MLC20 phosphorylation elicited at pCa 7.3 was inhibited in strips in which myosin phosphatase activity was completely inhibited by microcystin-LR. As under this condition, these rates depend soley on the activity of MLCK, this result suggests that phosphorylation of MLCK by PAK1 is associated with inhibition of the enzyme. In conclusion, PAK1 inhibits contraction at constant submaximal [Ca2+] in vascular and intestinal smooth muscle by inhibition of the activity of MLCK. Institut für Vegetative Physiologie, Universität Köln, Robert-KochStr. 39, 50931 Köln

from the membrane-bound form by proteolytic cleavage of the Cterminal domain. Since intracellular phosphorylation of ACE might be involved in the regulation of the cleavage process, we determined whether the cytoplasmic tail of ACE is phosphorylated and whether this process regulates secretion. Methods & Results: Immunoprecipitation of ACE (180 kDa) from 32P-labelled endothelial cells revealed that ACE is phosphorylated. This phosphorylation was not observed in endothelial cells over-expressing a mutant form of ACE (ACE?S, all five cytoplasmic serine residues replaced by alanine). The cytoplasmic tail of ACE contains a consensus sequence for CK2, and CK2 co-precipitated with ACE from endothelial cells. Moreover, CK2 phosphorylated both ACE as a peptide corresponding to the cytoplasmic tail in vitro. Mutation of serine 1270 within the CK2 consensus sequence almost abolished ACE phosphorylation. In ACE overexpressing endothelial cells, ACE was mostly localised to the plasma membrane. However, no ACE was detected in the plasma membrane of ACE∆S-overexpressing cells, although a precursor ACE (170 kDa) was prominent in the endoplasmic reticulum and the cell supernatant contained substantial amounts of the soluble protein (175 kDa). A correlation between ACE-phosphorylation and secretion was confirmed in endothelial cells treated with the specific CK2-inhibitor, 5,6-dichloro-1-(beta-D-ribofuranosyl), which time-dependently decreased the phosphorylation of ACE and increased its secretion. Conclusion: These results clearly indicate that the CK2-mediated phosphorylation of ACE regulates its retention in the plasma membrane. Moreover, the finding that ACE is an active signal transduction molecule may account for some of the beneficial effects of ACE inhibitors. Institut für Kardiovaskuläre Physiologie und Institut für Biochemie II, Klinikum der J.W.G.-Universität, D-60590 Frankfurt am Main

O 21-5 ROLE OF CALCIUM-DEPENDENT PROTEIN KINASES IN THE CHOLINERGIC STIMULATION OF GASTRIC ACID SECRETION M. Fährmann, M. Kaufhold, U. Seidler Introduction: Calcium is a relevant mediator of acid secretion, however, the precise mechanism of calcium activation is unknown. We examined two calcium-dependent protein kinases, the protein kinase C (PKC) α-isoform, the only calcium-dependent PKC expressed in rabbit parietal cells, and the Ca2+/calmodulin-dependent protein kinase II (CaMKII). The latter is found in the cytosol, the apical domain, and localized on H+-K +-ATPase containing gastric tubulovesicles. Activation of the M3 muscarinic acetylcholine receptor initiatially stimulates in a calcium-dependent fashion the fusion of the tubulovesicle with the secretory membrane to position the proton pump or H +-K+-ATPase for acid production. Aims: We analyzed the intracellular localization and interaction of both PKC-α and CaMKII before and after carbachol-stimulated H+ secretion. Methods: PKC-α and CaMKII were monitored by phosphorylation activity or by immunodetection including a phospho-specific anti-CaMKII antibody against activated CaMKII. The acid secretory response of isolated and cultured rabbit gastric parietal cells was determined by [ 14C]-aminopyrine. Kinases were intracellular modulated by specific kinase agents (TPA, Gö 6976, Ro 318220, KN-62). Results and discussion: Both PKC-α and CaMKII were redistributed to the secretory menbrane after carbachol stimulation of H+ secretion. The use of divers pharmacological agents revealed the central role of CaMKII in acid secretion. In contrast to this PKC-α inhibited about one fourth of carbachol-evoked acid formation. We observed in an in vitro-model a direct phosphorylation and inhibition of CaMKII by PKC-α. Conclusions: The main mediator of muscarinic receptor stimulation is CaMKII. Tanslocated CaMKII is inhibited by calcium-dependent PKC-α, and, therefore, translocation of CaMKII appears to be a prerequisite for down-regulation by co-translocated PKC-α. Institut für Zoophysiologie, Hindenburgplatz 55, D-48143 Münster, Germany

O 22-1 CLAUDIN-2 FORMS CATION-SELECTIVE PARACELLULAR PORES IN TIGHT JUNCTIONS OF EPITHELIAL CELLS S. Amasheh, N. Meiri, A.H. Gitter, T. Schöneberg, J. Mankertz, J.D. Schulzke, M. Fromm Introduction: Claudins are recently detected transmembranal proteins forming the tight junctions which seal epithelial cells. Understanding of the function of these proteins is largely unknown. We analyzed function of claudins in 2 epithelia of different tight junctional permeability, strains C7 and C11 of MDCK cells, employing electrophysiological, molecular, and biochemical techniques. Results: Monolayers of MDCK C7 cells feature a relatively high transepithelial resistance Rt, whereas C11 cells show R t of a leaky epithelium. The genuine expression of occludin and claudin-1, -2, and -3 was detected by immunoblotting. The most obvious difference was observed for claudin-2, which was markedly expressed in C11 but hardly detected in C7 cells. To examine the contribution of claudin-2 to the functional barrier, stable transfection of C7 cells with claudin-2 was performed (clone C7-cld2). This approach resulted in localization of claudin-2 in the tight junction strand region. C7 cells transfected with vector alone did not reveal detectable claudin-2 expression by means of confocal fluorescence microscopy. Paracellular resistance, determined from Ca2+switch impedance measurements, and cation permeability, obtained from NaCl dilution potentials, was measured in C7-cld2 and in controls transfected with vector alone, C7-vec and C11-vec. The originally tight C7 cells became leaky for cations after stable transfection with claudin-2. 3H-mannitol flux was not altered in clone C7-cld2. Hence, claudin-2 forms a paracellular channel for monovalent cations (i.e. Na+, K+, and H +) but not for molecules >182 Da. Conclusion: This study demonstrates, for the first time, that the presence of a single tight junction protein can regulate in a wide range the paracellular cation permeability of epithelia. Assembly of claudin-2 within the tight junction acts as a cation-selective paracellular pore. Dept. of Clinical Physiology, Dept. of Gastroenterology, and Dept. of Pharmacology, Benjamin Franklin University Hospital, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany

O 21-6 CK2 REGULATES THE PHOSPHORYLATION AND SECRETION OF THE ANGIOTENSIN CONVERTING ENZYME (ACE) IN ENDOTHELIAL CELLS K. Kohlstedt, W. Müller-Esterl, R. Busse, I. Fleming Background: Soluble angiotensin converting enzyme (ACE) is derived

O 22-2 BICARBONATE AND H+ - TRANSPORT OF ISOLATED AIRWAY SUBMUCOSAL GLANDS M.J. Hug1, A. Waltz1, H. Oberleithner1, R.J. Bridges2

S 195 The mechanisms by which airway submucosal glands secrete electrolytes and fluid are poorly understood. Results on a airway serous cell line indicate that these cells transport either Cl- or HCO3- ions depending on the mode of stimulation. The aim of the present study was to investigate the pH regulatory mechanisms of serous cells of freshly isolated submucosal glands (SMG). Porcine SMG were dissected out of pig tracheas obtained from a local slaughterhouse. Single glands were transferred into the chamber of an inverted microscope, immobilized by holding pipettes and the serous cells loaded with the fluorescent pH probe BCECF. The fluorescence ratio of the emission after excitation at 488nm and 436nm respectively was used to estimate cytosolic pH (pHi). Resting pHi of SMG cells in the absence of HCO3-/CO2 was 7.1 ± 0.18 (n=26). Addition of a solution buffered with HCO3-/CO2 transiently acidified the cells by 0.18 ± 0.03 (n=18). pHi rapidly recovered to a slightly more alkaline value than baseline pHi. Removal of the HCO3-/ CO2 buffer strongly alkalinized SMG cells by 0.2 ± 0.03 (n=18). To challenge pHi regulatory mechanisms we exposed the cells to 20 mmol/ l NH 4+ in the absence and presence of HCO3-/CO2. In both cases we observed only a slight increase in pH i followed by a pronounced acidification indicative for a high rate of NH4+ transport into the cytosol. Washout of NH4+ further acidified the cells. Realkalinization could only be observed in the presence of Na +. This effect was inhibited by application of the specific NHE1 blocker (HOE694) with an apparent IC50 of 20 µmol/l. Full recovery of pHi in the presence of HOE694 was observed when the cells were bathed in HCO3-/CO2 solution. Increasing the extracellular HCO 3 - concentration or decreasing the bath Cl concentration alkalinized porcine SMGs. The rate of alkalinization could be enhanced by the addition of forskolin whereas carbachol decreased the alkalinization rate. We conclude that SMG cells possess HCO 3- dependent and independent pH i regulatory mechanisms that require the presence of extracellular Na+. HCO3- transport is modulated by agonists increasing either cAMP or cytosolic Ca2+. Supported by IMF HU 11 01 03. 1

Department of Physiology, University of Münster, Robert-Koch-Str. 27a, D-48149 Münster, Germany; 2Department of Cell Biology and Physiology, University of Pittsburgh, Pittsburgh PA 15261, USA

O 22-3 THE CFTR BINDING ADAPTER PROTEIN CAP70 INTERACTS WITH THE APICAL CL-/HCO3- EXCHANGER DRA IN RABBIT SMALL INTESTINAL MUCOSA H. Rossmann, G. Lanprecht, P. Jacob, C. Neff, S. Baisch, J. Biber, U. Seidler, W. Fieh Introduction and Aim: CAP70 has been shown to be an adapter protein coupling characteristic C-terminal amino acid sequences of transport and channel proteins, e.g. CFTR (cystic fibrosis transmembrane regulator, Wang et al., 2000, Cell). CFTR and the electroneutral Cl-/HCO3-exchanger DRA (down regulated in adenonma) are thought to play a major role in duodenal HCO3- secretion. We hypothesized that CAP70 mediates the formation of macromolecular complexes in the apical membrane of duodenal enterocytes, containing one or more CFTR and DRA molecule(s). Methods and Results: Basolateral and apical membranes were isolated from rabbit duodenal and ileal mucosa, apical membranes from rabbit colonic mucosa. Western blot experiments using a specific, polyclonal anti-DRA antibody localized DRA in the apical membrane fraction of all studied tissues (Jacob et al., Gastroenterology, in press). Preincubation of the anti-DRA antibody with the recombinant DRA protein, initially used for immunization, blocked the DRA signal completely. Simultaneously a strong 70 kD band appeared in duodenum and ileum, which was restricted to the brush border membrane fraction and was missing in the colon. Deletion of the Cterminal four amino acids (-ETKF) of the recombinant DRA protein and repetition of the described overlay experiment resulted in the disappearance of the 70 kD band suggesting a specific protein-protein interaction between DRA and a DRA-binding protein. Incubation of a parallel blot with an anti-CAP70 antibody revealed a band, which displayed the size and the expression pattern expected for the DRA binding protein. In contrast to the situation in duodenum and ileum, neither CAP70 mRNA and protein expression nor the described overlay band was observed in rabbit colon. This data suggested to us that CAP70 represents the DRA-binding protein observed by the overlay experiments. The different expression pattern of CAP70 and DRA raised the question, whether different DRA variants are expressed in

small and large intestine. 5' and 3' RACE were carried out using several primers deduced from the known part of the rabbit coding sequence. Cloning and sequencing of the PCR products showed sequence identity of several clones from duodenal, ileal, and colonic mucosa. Summary and Conclusion: As already known for CFTR, the DRA protein interacts with CAP70 in the brush border membrane of duodenal and ileal mucosa. The assembly of such macromolecular complexes may result in coordinated activity or facilitate regulatory processes. In the colon, however, no CAP70 expression was observed, while DRA expression is high. Though the data suggest a different situation in small and large intestine, the DRA molecule displays sequence identity throughout the gut. Abt. Innere Medizin I, Zentrallabor, Universität Heidelberg, Bergheimer Straße 58, 69115 Heidelberg, Deutschland

O 22-4 EFFECT OF K+ CHANNEL ACTIVATION/BLOCKADE ON cAMPcGMP AND Ca2+ DEPENDENT STIMULATION OF ISC AND HCO3SECRETORY RESPONSE IN RAT DUODENAL MUCOSA I. Blumenstein, M. Gregor, U. Seidler Backround: We have previously observed that the relationship of HCO3- to Cl- in the CFTR-dependent duodenal anion secretory response varies strongly between different secretagogues. We have also found that the isoform 1 of the Na+ HCO3--cotransporter (NBC1) is a major supply pathway during duodenal HCO3- secretion. This transporter is electrogenic and could therefore be influenced by changes in the basolateral membrane potential. Aim: We investigated the effect of K+ channel activation/K+ channel blockade on the HCO3- (∆JHCO3-) and Isc (∆ Isc) secretroy response to different secretagogues. Methods: Short circuit current (Isc) and HCO 3- secretory rate was assessed in isolated and muscle-stripped rat duodenal mucosa in conventional Ussing-chambers. Basolateral membrane potential was hyperpolarized by the K+ channel opener EBIO. Reversely, the K+ channel inhibitors CTX and Hoe293B were used to prevent a potential agonist-induced hyperpolarization. Results: The ∆JHCO3- /∆Isc ratio was highest with cGMP-dependent agonists, followed by 8-Br-cAMP, EBIO, and carbachol, and was higher for all agonists at lower agonist concentrations. The chromanol Hoe 293b, an inhibitor of the KCNQ1 K+-channels, was found to partially inhibit the secretory response to cGMP, cAMP and Ca2+-dependent agonists, had a stronger inhibitory effect with increasing agonist-concentration, and a stronger inhibitory effect on ∆Isc than on ∆J HCO3-. Charybdotoxin partially inhibited carbachol- and EBIO-induced anion secretion only. When acetazolamide was present to inhibit cellular CO 2 hydration (thus making HCO3- secretion dependent on Na+HCO3- cotransport), EBIO did not elicit a HCO 3- secretory response any more, demonstrating that Na+HCO3- cotransport is indeed inhibited by K+-channel activation. Conclusion: During stimulation of rat duodenal anion secretion, high agonist-concentrations result in the activation of Hoe 293B inhibitable K+ channels, and Ca2+-dependent agonists stimulate CTxsensitive K + channels as well. The resultant hyperpolarization of the basolateral membrane potential favours Cl - secretion over HCO 3 secretion. One of the potential mechanisms for this is a hyperpolarization-induced inhibition of basolateral Na + HCO 3 cotransport. Thus, differences in the activation of basolateral K + channel activation by different agonists are one likely to explain their different ratio of HCO3 - to Cl - in the intestinal anion secretory response. 1st Med. Dept, University Hospital Schnarrenberg, Otfried-Müller-Str. 10, D-72076 Tübingen

O 22-5 INFLUENCE OF TRANSPORT VIA RAT ORGANIC CATION TRANSPORTER ON CYTOSOLIC PH M. Dieter1, I. Setiawan1, C. Boehmer1, H. Koepsell2, F. Lang1 The organic cation transporter OCT2 mediates the Na+-independent electrogenic transport of a wide variety of organic cations including tetraethylammonium (TEA). Conflicting data and ideas have been reported in the past on the involvement of H+ in the transport function

S 196 of OCT2. The present study has been performed to test whether the transport of TEA influences cytosolic pH (pHi). To this end rOCT2 has been expressed in Xenopus oocytes and ion selective electrodes and dual electrode voltage clamp applied to determine TEA induced current (ITEA) and change of cytosolic pH (pHi) at an extracellular pH (pHo) of 6.0 and 8.0. In OTC2 expressing oocytes not exposed to TEA pHi averaged 7.21 ± 0.02 at pHo 6.0 and 7.23 ± 0.02 at pHo 8.0 (n=10) and the potential difference across the cell membrane -29 mV at pH 6.0 and -42 ± 3 mV at pH 8.0. ITEA approached 18 ± 1 nA at pHo 6.0 and 33 ± 4 nA at pHo 8.0 (n=15). ITEA is paralleled by depolarization of the cell membrane (by -10 ± 1 mV at pHo 6.0 and -16 ± 1 mV at pHo 8.0) and by intracellular acidification (by 0.018 ± 0.002 and 0.019 ± 0.002 pHi/min at pHo 6.0 and 8.0, respectively, n=10). A depolarization by inhibition of K+ channels with 5 mM Ba2+ (by 24 ± 1 mV, n=5) had only a slight effect on pHi (by 0.04 ± 0.01 pHi/min, n=5). In water injected oocytes TEA did not produce a significant ITEA. The TEA-induced acidification in rOCT2 injected oocytes suggests but does not conclusively prove that TEA transport into the cell through rOCT2 is paralleled by translocation of H+ into the cell or by exchange of a buffer anion thus leading to cytosolic acidification. Department of Physiology, Univ. of Tübingen 1, and Department of Anatomy, Univ. of Würzburg, Germany2

O 22-6 STRUCTURE AND FUNCTION OF A HUMAN PROTONCOUPLED AMINO ACID TRANSPORTER CLONED FROM THE INTESTINAL CELL LINE CACO-2 V. Ganapathy, Z. Chen, W. Huang, D. Thwaites We have isolated a human proton-coupled amino acid transporter clone from a Caco-2 cell cDNA library using a probe specific for the recently cloned rat lysosomal proton-coupled amino acid transporter rPAT1. The human clone (hPAT1) is 5,585 bp-long with a 1431 bplong coding region (with the termination codon). The predicted protein consists of 476 amino acids and exhibits 85% identity in amino acid sequence with rPAT1. Among various human tissues examined by Northern blot, the hPAT1-specific mRNA transcript is expressed most predominantly in the intestinal tract. When expressed heterologously in mammalian cells, hPAT1 induces the transport of the model amino acid substrate alpha-methylaminoisobutyrate (MeAIB). The cDNAinduced transport is sodium-independent, but is energized by an inwardly directed proton gradient (extracellular pH 6.0). The transport of MeAIB is 6-fold higher in cDNA-transfected cells than in vector-transfected cells. The cDNA-specific transport activity is markedly influenced by extracellular pH, maximal transport occurring at pH 5.0. Subsequent functional studies of this clone were therefore carried out at pH 5.0 in the absence of sodium. Among various L-amino acids tested, hPAT1 interacts with glycine, alanine, proline, alpha-aminoisobutyrate, gamma-aminobutyrate, and MeAIB as evidenced from competition experiments with radiolabeled MeAIB as a transport substrate. However, hPAT1 also recognizes the D-isomers of alanine and proline. With serine and cysteine, although the L-isomers do not interact with the transporter to any significant extent, the corresponding D-isomers are recognized by the transporter. The ability of hPAT1 to transport Dserine is demonstrable directly using radiolabeled D-serine. The transporter exhibits a Michaelis constant of 2.3 ± 0.2 mM for glycine as determined from direct transport measurements. The affinity of other amino acids was then tested by competition experiments. The inhibition constant for MeAIB, alpha-aminoisobutyrate, and gammaaminobutyrate is 1.5 ± 0.2, 2.6 ± 0.7, and 2.1 ± 0.2 mM, respectively. The inhibition constant for proline and alanine is 2.0 ± 0.6 and 1.7 ± 0.2 mM, respectively and these values do not vary significantly between L- and D-isomers of these amino acids. However, with cysteine and serine, the D-isomers show 6- to 8-fold higher affinity for the transporter than the corresponding L-isomers. These functional characteristics of the cloned hPAT1 closely resemble those that have been described previously for the proton-coupled amino acid transport system in Caco-2 cells. These data suggest that the cloned hPAT1 is responsible for the proton-coupled amino acid transport expressed in the plasma membrane of the Caco-2 cells. Department of Biochemistry and Molecular Biology, Medical College of Georgia, Augusta, GA 30912, USA

O 23-1 DIABETES INHIBITS STORE-OPERATED CALCIUM INFLUX IN RAT RETINAL ARTERIOLES THROUGH A PKC-DEPENDENT MECHANISM T.M. Curtis, E. Major, C.N. Scholfield In diabetes, early hyperglycemia leads to a deterioration in retinal microvascular function and there is now strong evidence that the persistent upregulation of PKCβ might be responsible. We have previously shown that retinal arterioles rely upon a novel dihydropyridine sensitive store-operated Ca2+ influx (SOC) pathway1 and that Ca2+ signalling in response to vasoconstrictors is blunted in retinal vessels from diabetic rats2. The present work aimed to assess the role SOC entry in this latter phenomenon and the effects of PKC antagonists. Sprague-Dawley rats were (200-230g) given 60mg kg-1 streptozotocin I.P. and killed 4-6 weeks later. Animals with blood glucose concentrations >15 mM were taken as diabetic (mean glucose 29 ± 2.2 mM, s.e.m). Injected animals with glucoses <7 mM were used as non-diabetic controls (mean glucose 5.5 ± 0.4 mM, s.e.m). Cytosolic Ca2+ was estimated from freshly isolated rat retinal arterioles using fura-2 microfluorimetry as previously described1. In vessels from diabetic animals, 25 mM D-glucose was present through the dispersion and experiment. SOC influx was assessed by depleting Ca2+ stores with Ca2+ free/cyclopiazonic acid over 10 min and subsequently measuring the rise in Ca2+ over the initial 10 s on adding 2 mM or 10 mM Ca2+ solution. At least 4 rats were used per treatment group. Comparisons used the unpaired t-test. SOC influx was attenuated in vessels from diabetic rats: in non-diabetic vessels the addition of 2 and 10 mM Ca2+ induced a mean rate of rise in Ca 2+ of 7.4 and 17 nM s -1 (n=17), respectively, compared with 3.2 and 9.5 nM s-1 (n=18) in vessels from diabetic animals (2 mM Ca2+, P=0.007; 10 mM Ca2+ P=0.01). In vessels from non-diabetics 25 mM D-mannitol had no effect on SOC entry, consistent with a non-osmotic mechanism (P=0.8 and 0.98 for 2 and 10 mM Ca2+ respectively; n=8). Diabetic animals injected daily with insulin (protamine-zinc, 5 units) for 2 weeks had a mean blood glucose of 3.5 ± 1 mM (s.e.m) and SOC influx similar to non-diabetic controls (2 mM Ca2+, P=0.48; 10 mM Ca2+, P=0.2; n=10). The reduced SOC entry in diabetic vessels was reversed by 2 h exposure to 100 nM staurosporine (SP), a non-specific PKC antagonist (P values were 0.0006, and 0.009 for 2 & 10 mM Ca2+, respectively, paired comparison with and without SP in the same animals; n=14). The specific PKCβ antagonist LY379196 (100 nM) also reversed the poor Ca 2+ influx although its action was less efficacious (P=0.02 & P=0.04 using the same experimental design as for SP above). These results demonstrate that SOC influx is inhibited in retinal arterioles from animals having sustained elevated blood glucose and that PKCβ appears to play a major role in mediating this effect. Decreased SOC influx may contribute to the demise of microvascular smooth muscle/pericytes and hence the retinal haemodynamic changes in the early stages of diabetes mellitus. 1 Curtis & Scholfield (2001) J Physiol 532, 609-623. 2 Curtis & Scholfield (1999) J Physiol 515P, 6P. We thank the Wellcome Trust and British Heart Foundation for financial support. Ophthalmology/School of Medicine, Institute of Clinical Science, Royal Victoria Hospital, Belfast, BT12 6BA and Smooth Muscle Group/ School of Medicine, Physiology, 97 Lisburn Rd, Belfast, BT9 7BL O 23-2 CRITICAL DETERMINANTS OF Ca2+-DEPENDENT REGULATION OF E-TYPE Ca2+ CHANNELS ARE LOCATED IN AN ARGININERICH REGION IN THE CYTOSOLIC II-III LINKER J. Lérôy, A. Pereverzev, R. Vajna, N. Qin, J. Hescheler, G. Pfitzer, C. Malecot, T. Schneider, U. Klöckner The entry of Ca2+ through voltage dependent Ca2+ channels is an important step in signal transductions cascades triggering a variety of intracellular processes ranging from secretion of neurotransmitter to muscle contraction. In contrast to L- and P/Q-type calcium channels, where calmodulin is the intracellular Ca2+-sensor which mediates the Ca 2+-responsiveness of channel activity, E-type calcium channel activity was regarded to be Ca2+ insensitive. Here we present evidence that elevation of [Ca2+]i can drastically modifies the biophysical properties of inward currents induced by expression of Cav2.3d (α1Ed) in HEK 293 cells. The current density is increased, the inactivation time course slower and the recovery from short term inactivation is accelerated compared to recording conditions which prevent a rise in [Ca2+]i by either dialysing the cells with 10 mM BAPTA or exchanging extra-

S 197 cellular Ca2+ by Ba2+. This behaviour was unchanged after mutating homologous regions known to be involved in the Ca2+ dependent modulation of L- and P/Q type calcium channels. However, when testing the in vivo expressed splice variant Cav2.3e which differs structurally from Ca v2.3d by lacking a 19 aa long insertion in the intracellular II-III linker encoding part of an arginine rich region, a reduced sensitivity of the channel activity towards elevation of [Ca2+]i could be detected. Our results demonstrate that a novel functional region is involved in the Ca2+ dependent feed back regulation of the activity of E-type calcium channels, putting forward the hypothesis that Ca2+ influx through these channels might be controlled by a novel molecular mechanism.

nisoldipine was 0.13 µM at a holding-potential of –80 mV. In embryonic CaV1.2 +/+ cardiomyocytes, IBa was blocked by nisoldipine with an IC 50-value of 0.1 µM. It is possible that the calcium current of cardiomyocytes at day 12.5 p.c. is mediated by the CaV1.3 gene. Further experiments are needed.

Institute of Physiology, University of Cologne, Robert-Koch-Str. 39, D-50931 Köln, Germany

CaT-L MEDIATED Ca2+ ENTRY IN COMPARISON TO STOREOPERATED Ca2+ INFLUX M. Bödding, U. Wissenbach, V. Flockerzi The recently cloned human Ca 2+ transport like (CaT-L) channel was investigated in whole-cell patch-clamp experiments on transiently transfected HEK-293 cells. CaT-L is a member of the transient receptor potential (trp) family and comprises unique biophysical properties of the Ca2+-release-activated Ca2+ current (ICRAC) in RBL cells such as the inward rectification, positive reversal potential and high Ca 2+ selectivity. However, a fundamental difference between Ca2+ currents mediated by CaT-L and CRAC-pores is the activation mechanism: Depletion of inositol-1,4,5-trisphosphate (IP3)-sensitive Ca 2+ stores by IP3, the sarco/endoplasmic reticulum Ca 2+-ATPase (SERCA) inhibitor thapsigargin or the Ca 2+ ionophore ionomycin activated ICRAC but not CaT-L. The gating of CaT-L was highly dependent on the cytosolic free Ca 2+ concentration ([Ca2+]i). Lowering the [Ca2+]i resulted in Ca 2+ influx and the intracellular EGTA concentration correlated with current amplitude. Therefore, CaT-L functions as a Ca2+ sensing pore and not as a store-operated channel.

O 23-3 EFFECTS OF AUXILIARY SUBUNITS ON CHARGE MOVEMENT OF THE EXPRESSED α1G SUBUNIT OF T-TYPE CALCIUM CHANNEL L. Lacinova1, N. Klugbauer, F. Hofmann The charge movement was measured in α1G T-type calcium channel expressed in HEK293 cells alone or together with the α2δ-2a or γ5 auxiliary subunits. The α2δ-2a subunit was previously shown to facilitate transition of the expressed α1G channel into an inactivated state while γ5 subunit supported both activation and inactivation of the channel. The aim of this study was to test, if both auxiliary subunit affect the gating current observed during activation and inactivation of the α1G channel. Ion current through the expressed channel was blocked by 1 mM La3+. The onset of charge movement was detected between 70 and -60 mV. The amount of charge moved increased slowly and saturated above +60 mV. For all three channels, the voltage dependence of on- and off-charge movement could be fitted by a single Boltzmann function. For the on-charge, a half-maximum was reached at +12.9±1.4 (n=25) mV, 12.1 ±1.4 (n=21) mV and +11.3±2.4 (n=16) mV for α1G, α1G + α2δ-2a and α1G + γ5 channels, respectively. The offcharge reached half-maximum at +12.3±0.7 mV, 11.2±1.5 mV and +7.3±1.9 mV for α1G, α1G + α2δ-2a and α1G + γ5 channels, respectively. The shift caused by γ5-coexpression was significant (p<0.01). The time course of both on- and off-charge movements could be fitted by a single exponential. Coexpression of both auxiliary subunits significantly accelerated the decay of the on-charge movement and α2δ-2a significantly accelerated the decay of the off-charge movement. These findings support the hypothesis that both auxiliary subunits interact with the α1G subunit of the T-type calcium channel. Institute of Pharmacology and Toxicology TUM, Biedersteinerstr. 29, 80802 München, Germany; 1present address: Institute for Molecular Physiology and Genetics, Vlarska 5, 833 06 Bratislava, Slovakia O 23-4 IS THE CaV1.3-TYPE CALCIUM CHANNEL IMPORTANT IN THE DEVELOPING MURINE HEART? A. Welling, S. Lefrank, C. Seisenberger, F. Hofmann, N. Klugbauer Voltage-gated L-type calcium channels play a crucial role for hypertension and arrhythmias. In the cardiovascular system the most abundant isoform is formed by the CaV1.2-gene. Additionally CaV1.1 and CaV1.3 were identified in different amounts by RT-PCR analysis. The differential contribution of the different L-type channels is unclear, because of the absence of selective blockers. Therefore we deleted the CaV1.2-gene in a mouse line (CaV1.2 -/-). No viable mice were obtained, but embryos survived till day 13.5 p.c. From day 9.5 p.c the embryonic hearts were beating regularly, therefore it is possible that another Ltype channel is important in the early developing heart. According to its current kinetic Ca V1.3 is a likely candidate. To characterize its biophysical and pharmacological properties we have cloned CaV1.3 from embryonic murine CaV1.2 -/- heart tissue. Functional channels could be expressed in a heterologous HEK293 expression system together with ß3 and α2δ−1 subunits and measured with the whole-cell patch-clamp method. Ba2+- inward currents (IBa) were maximal between –10 and 0 mV with maximum amplitudes of 10.3 ± 1.9 pA/pF (n=15) with 5 mM Ba2+ as the charge carrier. Inactivation was slowely with an apparent t½ of 500 ms. The steady-state inactivation constant was –26 ± 0.3 mV (n=3). The IBa was increased by the dihydropyridine (DHP) agonist Bay K 8644 and blocked by nisoldipine. The IC 50-value for

Institut für Pharmakologie und Toxikologie der Technischen Universität, Biedersteiner Str. 29, 80802 München, Germany O 23-5

Institut für Pharmakologie und Toxikologie, Universität des Saarlandes, 66421 Homburg, Germany O 23-6 PALMITATE POTENTIATES GLUCOSE-INDUCED INSULIN SECRETION VIA A KATP-CHANNEL INDEPENDENT MECHANISM C.S. Olofsson, A. Salehi, C. Holm, P. Rorsman Background and aims: It is well established that acute exposure of pancreatic islets to free fatty acids leads to stimulation of insulin release. We have correlated the effects of palmitate on insulin secretion to changes of the cytoplasmic free Ca2+ ([Ca2+]i) and whole-cell Ca2+ currents, with the goal of identifying the underlying mechanisms. Material and methods: Collagenase-isolated intact mouse pancreatic islets or dispersed B-cells were used. Changes in [Ca2+]i were recorded in fura-2 loaded islets using microfluorimetry. Insulin secretion was assessed by radioimmunoassay. Currents were measured applying the perforated patch whole-cell configuration. Palmitate was applied at a total concentration of 1 or 0.5 mM, dissolved with 1% fatty acid free bovine serum albumin. Results: In islets exposed to 15 mM glucose, [Ca2+]i oscillated between a plateau level and discrete peaks. The frequency of these oscillations (peak-to-peak) was 1.4±0.26 min1 . Addition of 1 mM palmitate elicited a sustained increase in [Ca2+]i by 24% (based on the calculated time average; n=13; p<0.01). The frequency of the oscillations increased to 2.5±0.67 min-1 (+61%; p<0.001). Palmitate remained capable of increasing [Ca2+]i in islets depolarised by 30 mM K+ in the presence of 500 µM diazoxide (27%; n=8; p<0.001) and in islets exposed to 400 (µM of the ATP-regulated potassium channel (KATP-channel) inhibitor tolbutamide (26%; p<0.05; 7 out of 11 islets). Current-voltage relationships for Ca2+ currents (-60 to + 60 mV) were recorded before and after addition of 1 mM palmitate. The peak current at 0 mV decreased from -128±27 pA under control conditions to -71±21 pA 4 min after inclusion of the fat (5 experiments; p<0.05). In a 6th experiment currents increased. When the concentration of palmitate was reduced to 0.5 mM Ca2+ currents did instead increase (n=6). Palmitate (1 mM) elevated insulin secretion from 1.15±0.11 to 2.24±0.14 ng/islet/h in the presence of glucose alone (15 mM). In islets depolarised by 15 mM glucose and 400 µM tolbutamide, palmitate was capable of increasing insulin release from 2.05±0.09 to 3.44±0.19 ng/islet/h. In the presence of glucose (15 mM), diazoxide (500 µM) and K + (30 mM) palmitate elevated insulin secretion from 2.40±0.14 to 3.67±0.23 ng/islet/h. Note that the absolute increase was equal under the different conditions (n=8). Conclusions: Our results suggest that the stimulatory action of palmitate on glucose-

S 198 induced insulin secretion is not solely mediated via closure of KATPchannels but also by an effect on the voltage-gated Ca2+-channels. Dept of Physiological Sciences, Lund University, Tornav 10, BMC F11, 221 84 Lund, Sweden O 24-1 ATP ANTAGONIZES THROMBIN-INDUCED BARRIER FAILURE IN HUMAN ENDOTHELIAL CELLS D. Gündüz, F. Hirche*, M. Schäfer, G. Pfitzer*, H. M. Piper, T. Noll Failure of endothelial barrier function induced by thrombin is due to Ca2+-mediated intracellular mechanisms triggering of activation of contractile machinery and disturbance of cell-cell adhesion. In contrast, extracellular ATP, which also stimulates intracellular Ca2+ signaling mechanisms, inactivates the contractile apparatus and stabilizes cellcell adhesion. Here the hypothesis was studied, whether ATP can antagonize thrombin-provoked hyperpermeability. Changes in intracellular Ca2+ (Fura-2, Ca i), activation of contractile apparatus (phosphorylation of myosin light chains; MLC~P), isometric tension (cell coated collagen gels, IT), cell adhesion (localization of cadherins, immunofluorescence microscopy) and permeability (albumin flux, P) of cultured monolayers of human umbilical vein endothelial cells (HUVEC) were studied. Thrombin (0.2 U/ml) increased Cai from basal level of 78±8 to 570±63 nM, MLC~P from 53±7 % to 153±18 % (Maximum MLC~P=200 %, since MLC are phosphorylated at two sites), IT from 170±23 to 367±57 µN, P from 3,1±0.4 to 11±1xl0-6 cm/s (mean±SD, n=5, P<0,05 for all changes) and caused cadherin delocalization at the cell-cell borders. ATP (10 µM), which also increased Cai 7-fold, reduced MLC~P to 21±5 %, IT to 89±7 µN, P to 0.8±0.2x10-6 cm/s (P<0.05 compared to control for all parameters), and had no effect on cadherin localization. Addition of ATP (10 µM) to thrombin stimulated cells (0.2 U/ml for 5 min) caused a complete reduction of thrombin-induced MLC~P, IT, P, and restored cadherin localization but did not affect the Ca2+ rise. Pretreatment of the cells for 5 min with ATP abolished the thrombin effects on MLC~P, IT, P and cadherin localization. Conclusions: ATP antagonizes and prevents thrombin induced barrier failure in HUVEC. The antagonism relies upon opposite effects of ATP and thrombin on the contractile machinery and cell-cell adhesion, but not on cytosolic Ca2+ control.

of 2.3±1nmol/min as compared to intact laminin I (1.1±0.5nmol/ min). Upon elevated shear stress Elastase is activated by EC, which in turn results in a degradation of laminin to fragments amongst them E8. In contrast to P1 and intact laminin, this proteolytic fragment serves then as an inhibitor of endothelial proliferation and induces apoptosis and oxygen radical production. These results are indicating that during shear stress the extracellular matrix is altered post synthesis which results in an induction of signaling cascades which in total might be part of a negative feed back loop to limit shear stress induced proliferation and differentiation of endothelial cells. Institute of Physiology, LMU, Schillerstr. 44, 80336 Munich O 24-3

Physiologisches Institut, Justus-Liebig-University Giessen, Aulweg 129, D-35392 Giessen; *Insitut für Vegetative Physiologie, Universität zu Köln, Robert Koch Straße 39, 50931 Köln

REGULATION OF NAD(P)H OXIDASE SUBUNIT EXPRESSION BY ARTERIAL SHEAR STRESS IN HUMAN ENDOTHELIAL CELLS N. Duerrschmidt, W. Goettsch, H. Morawietz Objective of study: The impact of biomechanical forces on the generation of oxidative stress in endothelial cells is not well understood. An NAD(P)H oxidase complex has been identified as a major source of endothelial superoxide anion generation. Previous investigations described an increased superoxide anion formation in response to shear stress. In this study, we analyzed the effect of laminar shear stress on expression of NAD(P)H oxidase subunits in endothelial cells. Methods and Results Primary cultures of human umbilical vein endothelial cells (HUVEC) were exposed to shear stress (1, 5, 10, 15, 30 dyn/cm2) in a cone-and-plate viscometer for up to 24 h. Expression of NAD(P)H oxidase subunits was determined by standard-calibrated competitive RT-PCR (in % of internal control without shear stress). The NAD(P)H oxidase subunit p67phox mRNA is 1.4-fold upregulated by arterial shear stress of 30 dyn/cm2 after 24 h. However, the p67phox protein expression is not affected by arterial shear stress. Arterial shear stress has no effect on the mRNA expression of NAD(P)H oxidase subunit p22phox. In contrast, mRNA expression of subunits gp91phox and p47phox is downregulated by chronic laminar arterial shear stress to 47% and 74% of control without shear stress, respectively. The protein kinase C inhibitor RO-31-8220 (1 µM) and the tyrosine kinase inhibitor herbimycin A (1 µM) have no effect on shear stress-dependent regulation of NAD(P)H oxidase subunits. Conclusions The upregulation of superoxide anion formation in response to arterial shear stress seems not to be mediated on the level of NAD(P)H oxidase subunit expression in human endothelial cells.

O 24-2

Institute of Pathophysiology, Martin Luther University HalleWittenberg, Magdeburger Str. 18, D-06097 Halle, Germany.

SHEAR STRESS INDUCES GENERATION OF LAMININ FRAGMENT E8 BY ACTIVATION OF ELASTASE AND PREVENTS ENDOTHELIAL CELL PROLIFERATION T. Gloe, S. Zahler, H.-Y. Sohn, U. Pohl Activated metallo-proteases generate matrix-fragments that may decisively modulate cellular functions such as proliferation and differentiation of endothelial cells. We investigated whether shear stress contributes to the release of proteases from EC and whether proteolytic fragments of laminin influences their proliferation. Porcine aortic endothelial cells (PAEC) were subjected to shear stress (20dyn/ cm² for 2h, n=8). Subsequently the conditioned media were tested for proteolytic activity of Elastase in a specific enzyme assay. Western analysis of the extracellular matrix proteins were performed to identify proteolytic laminin fragments. For proliferation assays growth-arrested HUVEC were plated on cover-slips precoated either with collagen I, fibronectin, laminin I or with one of its proteolytic fragments P1 (Pepsin) and E8 (Elastase). After 24h, 48h, 72h and 96h bFGF induced cellular proliferation was measured with the MTT reduction assay. In parallel experiments apoptosis assays were performed by FACS analysis as cell surface binding of annexin V. Finally O2- production was quantified using the cytochrome assay. After shear stress an enhanced Elastase activity of approx. 14±0.8mU/ml was found in conditioned media from shear stress cells as compared to static controls (5±1mU/ml, n=8). This went along with an increased occurrence of the laminin fragment E8 within the matrix. Seeding EC on E8 resulted in a significantly reduced proliferation (3-fold increase in cell number after 96h compared to 6-fold on laminin, p<0.05, n=4/8), while P1 had no effect. In parallel to these findings EC growing on E8 exhibited a 2fold higher apoptosis rate (n=4, p<0.05) and an increased O2- production

O 24-4 SUSTAINED HYPERCAPNIC INDUCED VASODILATION REQUIRES INCREASED NO PRODUCTION VIA SHEAR STRESS A. Heintz, A. Deussen The hypercapnic induced vasodilatation of coronary vessels is a widely known reaction. However, the underlying mechanisms that mediate this response remain unclear. Results published on this topic remain controversial due to different experimental set ups and species. This study was designed to investigate the contribution of NO to the hypercapnic flow response by comparing constant flow and constant pressure perfused isolated working hearts. Methods: Hearts were perfused with Krebs-Henseleit-buffer either under constant pressure (54 mmHg, group I ) or constant flow (10 ml/min, group II). After 30 min nitroprusside (100 µmol/l) was infused or hearts were perfused under hypercapnic conditions (pO2 = 588 mmHg, pCO2 = 59 mmHg, pH = 7.16) for 10 minutes, respectively. L-NAME (100 µmol/l) was used to inhibit the endothelial NO-synthase. Results: Group I: The flow response to hypercapnia and nitroprusside, respectively, was biphasic with an early fast (max. 60 s) and a consecutive persisting rise in flow (mean flow increase after 10 min 21.6 ± 12.9 (SD) %). LNAME specificially abolished the flow increase during the delayed hypercapnic flow response (mean flow increase after 10 min –0.12 ± 8.55 %). Group II: On average perfusion pressure decreased during the first two minutes of hypercapnic perfusion. Thereafter, however, perfusion pressure returned to control.The nitroprusside induced flow increase was constant over 10 min. Conclusions: Under constant

S 199 pressure the hypercapnic flow response in guinea pig heart is biphasic. The delayed hypercapnic flow response largely depends on intact NO production. In contrast, under constant flow perfusion there was no consecutive persisting decrease in coronary resistance. The results are in agreement with the hypothesis that the delayed resistance decrease requires a shear stress induced increase of NO production. Institute of Physiology, Medical Faculty, Fetscherstr.74, TU Dresden, 01307 Dresden O 24-5 cAMP-INDUCED INCREASE IN MACROMOLECULE PERMEABILITY IN MICROVASCULAR CORONARY ENDOTHELIAL CELLS IS DUE TO DISINTEGRATION OF CELLADHESION BUT NOT ACTIVATION OF THE CONTRACTILE APPARATUS K. Bindewald, D. Gündüz, D. Klingenberg, H. M. Piper, T. Noll In microvascular coronary endothelial monolayers activation of adenylate cyclase causes failure of endothelial barrier function. The effector mechanisms downstream of the cAMP-PKA signal transduction pathway causing barrier failure are largely unknown. In principle, loss of barrier function may result from activation of the contractile machinery and/or disintegration of endothelial cell-adhesion structures. Here the effect of the cAMP-PKA signalling pathway on macromolecule permeability (albumin flux across endothelial monolayers) the contractile activation (phosphorylation of the regulatory myosin light chains, MLC∼P) and cell-cell adhesion (localization of cadherins; immunofluorescence microscopy) were analysed in cultured monolayers of microvascular coronary endothelial monolayers. Under basal conditions, MLC was phosphorylated by 82 ± 9 %, albumin permeability was 5.8 ± 0.7 x 10-6 cm/s (mean ± SD, n = 5, P < 0,05) and cadherin positive immunofluorescence staining was localized at cell-cell contact sites. Activation of adenylate cyclase by forskolin (5 µM) caused a reduction of MLC~P to 13 ± 2 % (P<0.05 compared to control) within 20 min. After the same period of time, however, permeability was increased to 214 ± 23% (P<0.05) and cadherin-positive immunofluorescence staining could not be detected at the cell-cell borders. As indicated by western blot analysis, the cellular cadherin content did not change in presence of forskolin. Inhibition of adenylate cyclase by SQ22586 (20 µM for 30 min) as well as PKA by H89 (20 µM for 30 min) prevented the effect of forskolin on MLC~P, permeability, and cadherin delocalisation. Conclusion: The data of the present study show that in microvascular coronary endothelial cells cAMP has two effects: it stabilizes the barrier by reducing the activity of the contractile machinery and provokes a destabilization by disturbing the cell-cell adhesion. Disturbance of the cell-cell adhesion structures prevails and is the cause of barrier failure induced by cAMP. Physiologisches Institut, Justus-Liebig-Universität Giessen, 35392 Giessen O 24-6 RECALCIFICATION BUT NOT CA2+-DEPLETION STIMULATE TYROSINE PHOSPHORYLATION OF THE VE-CADHERIN/ CATENIN COMPLEX H.-J. Schnittler, J. Seebach, F. Luo, P. Dieterich Calcium dependent vascular endothelial (VE-) cadherin/catenin complexes are the backbone of endothelial adherence junctions that are crucially involved in regulation of barrier function and cell-cell adhesion. VE-cadherin is linked to the actin filament cytoskeleton via catenins. While calcium-depletion causes a dissociation of VE-cadherin and catenins from the junctions, recalcification leads to the reformation of calcium-dependent cell adhesion. VE-cadherin-mediated cell adhesion is modified by tyrosine phosphorylation, but it is unclear to whether tyrosine phosphorylation increases or decreases cell adhesion and barrier function. We addressed this important question in human-umbilicalvein endothelial cells by calcium depletion and recalcification experiments combined with transendothelial resistance (TER) measurements. Tyrosine phosphorylation of VE-cadherin and catenins was determined by Western Blotting employing phosphotyrosine antibodies of immunoprecipitated VE-cadherin/catenin complexes. Calcium depletion ([Ca 2+] < 10-7M) caused VE-cadherin to disappear from the junctions (within minutes) accompanied with a rapid and dramatic decrease in barrier function. We found that disappearance of

VE-cadherin under calcium depletion occurred completely independent from tyrosine phosphorylation. In contrast, recalcification ([Ca2+] =1.8 mM) caused a reappearance of VE-cadherin at the junctions (within 5 minutes) and was accompanied by a transient tyrosine phosphorylation of VE-cadherin, β- and γ-catenin but not α-catenin. Tyrosine phosphorylation occurred within 5 minutes after recalcification, returned to baseline levels within 120 minutes and was accompanied with a complete reestablished barrier function. Recalcification-dependent increase in barrier function occurred in a two phases: a quick initial increase was observed within 20-30 minutes followed by a plateau (around 20 minutes) and a subsequent complete establishment of barrier function within further 30 minutes. Genistein (100µM), a tyrosine kinase inhibitor, only partly blocked reappearance of VE-cadherin/catenin complex at the junctions but blocked the establishment of barrier function and cell adhesion largely. The data show that tyrosine phosphorylation promote the development of functional competent junctions in endothelial cells. Institute of Physiology TU-Dresden, Fiedlerstrasse 42, D-01307 Dresden, Germany, e-mail: [email protected] O 25-1 DIRECT INFLUENCE OF CYTOKINES ON HYPOTHALAMIC VASOPRESSINERGIC NEURONS OF RATS IN VITRO AND IN VIVO J. Roth1, H.A. Braun2, T. Hübschle1, U. Pehl1, K. Voigt2, R. Gerstberger1 In response to systemic injection of bacterial lipopolysaccharide (LPS), cytokines such as IL-1 and IL-6 can be detected within the hypothalamus. In this brain area the paraventricular and supraoptic nuclei (PVN, SON) are located which are responsible for production and release of the antidiuretic hormone (ADH). The electrical activity of the PVN and SON neurons is controlled by a number of physiological stimuli including rising temperature which results in increased impulse frequences of these neurons. Cytokines, on the other hand, exert inhibitory effects on the impulse patterns of PVN and SON neurons. Since the appearance of cytokines within the hypothalamus results in fever, we wondered if there are interactions between cytokines and the febrile increase of body temperature on the control of activity of PVN / SON neurons. Electrophysiological data indicate that the inhibitory effect of IL-1 on the electrical activity of PVN / SON neurons in brain slices is reversed to a pronounced stimulation, if the temperature of the slice preparation is increased from 35°C to 40°C (febrile temperature). In vivo we observe that IL-1-induced fever is accompanied by a significant rise of ADH-levels in plasma, a finding which correlates well to the electrophysiological data. A direct influence of cytokines on hypothalamic PVN and SON is further demonstrated by a pronounced translocation of cytokine-specific transcription factors (STATs) from the cytoplasm into the nuclei of SON and PVN neurons. These results clearly show that there are neuromodulatory effects of cytokines in neuroendocrine hypothalamic nuclei. The investigation of interactions between cytokines and other signal parameters which control the activity of PVN / SON neurons can provide novel evidence for the mechanisms responsible for the induction of brain-controlled sickness symptoms. 1

Institut für Veterinär-Physiology der Justus-Liebig Universität Giessen, Frankfurter Strasse 100, D-35392 Giessen and 2Institut für Normale und Pathologische Physiologie der Philipps-Universität Marburg, Deutschhausstrasse 2, 35033 Marburg O 25-2

SYMPATHETIC MODULATION OF VASCULAR TONE DEPENDS ON α1-ADRENERGIC RECEPTOR GENE EXPRESSION A. Gericke, K.D. Wagner, H.M. Stauss The frequency of sympathetic-mediated Mayer waves is shifted to lower frequencies in hypertension. A compensatory downregulation of vascular α1-adrenergic receptors may contribute to the lower frequency of Mayer waves in hypertensive patients. Therefore, the purpose of the present study was to investigate if the α1-adrenergic receptor density determines the time course of vascular smooth muscle cell (VSMC) contraction in response to α1-adrenergic receptor stimulation. VSMC were isolated from the aorta of 3-week-old Sprague Dawley rats by collagenase dispersion. After one passage, cells were doubletransfected with genes encoding the α1b-adrenergic (n=8) or the α1d-

S 200 adrenergic receptor (n=8) and a green fluorescent protein (GFP). GFP was used to identify successfully transfected cells. VSMC, transfected with the GFP gene only, served as controls (n=8). Transfection was accomplished using the DMRIE-C transfection system. Subsequent stimulation with the α1-adrenergic agonist, phenylephrine, caused dosedependent VSMC contractions that were almost twice as fast in cells, transfected with either of the two α1-adrenergic receptor subtypes (1.17±0.15 %/s, α 1b and 1.24±0.13 %/s, α1d) compared to controls, transfected with GFP alone (0.72±0.07 %/s, p<0.05). The maximum change in cell diameter in response to the α1-adrenergic receptor stimulation was not altered in cells transfected with α1-adrenergic receptors (4.7±0.6 %, α1b; 5.4±1.2 %, α1d; 4.3±0.8 %, controls). These data suggest that the time course of catecholamine-induced contraction of VCMS depends on the α1-adrenergic receptor density. A reduced vascular α1-adrenergic receptor gene expression may explain the lower frequency of Mayer waves in hypertension. Johannes-Müller-Institut für Physiologie, Humboldt-Universität (Charité), Tucholskystraße 2, 10117 Berlin O 25-3 DISTINCT ELECTROPHYSIOLOGICAL CHARACTERISTICS OF NODOSE GANGLION NEURONS WITH CARDIAC AXONS P. Linz, K.F. Hilgers, P.W. Reeh, R. Veelken Nodose ganglion neurons with sensoric afferents responding to mechanical stimuli like volume and pressure changes are involved in the regulation of the autonomous nervous system. Nodose ganglion cells with cardiac afferents are peculiar in that these afferents dot not only sense changes in cardiac filling pressure but are also exposed to strong mechanical stimuli due to regular heart beats. Hence, we wanted to test the hypothesis that nodose ganglion neurons with cardiac afferents exhibit distinct electrophysiological properties as compared to nodose ganglion neurons with putative projections from other organs. The lipophilic fluorescent dye DiI ( Molecular Probes, Oregon) was applyed into the pericardial sac of rats to label cells with cardiac axons. The dye was allowed to diffuse retrogradely along the afferent cardiac nerves to the cell-bodies in the nodose ganglion. Four days after the labeling procedure, the ganglia were dissected and neurons from the nodose ganglion plated on coverslips. Cardiac neurons were identified by DiI fluorescence under laser light. Whole cell patch recordings were obtained from the cells and current-voltage relationships established using a ramp protocol (voltage increases from -100 mV to +50 mV over 4s). Cells were characterised under control conditions and after exposure to hypoosmotic mannitol solution for 5 minutes to induce mechanical stress. This protocol was also used after the stretch activated channel blocker gadolinium (20µM) was added to the medium. Cardiac cells exhibited a lower baseline current at -80 mV and the mechanically induced inward currents was significantly lower as compared to the currents in non-cardiac cells (control: -153±29 pA and hyposomotic stress: -203±29 pA versus control: -230±22* pA and hyposomotic stress: -377±44# pA; *, # p<0,05 cardiac [n=12] versus non-cardiac [n=26], mean+SEM ). Responses to mechanical stress could be abolished in both groups of cells by gadolinium. DiI itself did not influence the electrophysiological characteristics of either cardiac or non-cardiac nodose ganglion cells. Hence, nodose ganglion neurons with cardiac afferents were less sensitive to mechanical stimuli as compared to nodose ganglion neurons with putative sensoric afferents from other organs. Dr. P. Linz, IV Medizinische Klinik und Institut fur Physiologie und experimentelle Pathophysiologie, Universitat Erlangen-Nürnberg, Universitatsstraße 17, 91054 Erlangen

sodium channel) family located on afferent vagal fibers. These channels can be blocked by benzamil. In 18 anesthetized male Sprague-Dawleyrats we did volume loading (saline i.v.; 5% body weight over 15 minutes) to increase LVEDP. Blood pressure (BP) and heart rate (HR) were measured intraarterially. Integrated SRNA was recorded with an implanted bipolar stainless steel electrode. The animals were instrumented with an intrapericardial catheter to administer drugs to the epicardial site. In control rats (n=6) normal saline was injected intrapericardially (50 µl ipc. ) before volume loading. Experimental groups (n=6, respectively) were given 50 µl of a benzamil solution (0.1 or 1.0mM ipc.) before volume loading. Volume loading lowered SRNA by -62±5% in controls. Benzamil (ipc.) significantly impaired this mechanosensitive reflex response in a dose dependent manner ( 44±4%[0.1mM]; -34±4%[1.0mM] {n=6 each}). Hence, the reflex response to volume loading was blunted significantly by ipc. benzamil. Since benzamil is a specific inhibitor of mechanosensitive sodium channels (DEG/ENaC) we conclude that mechanotransduction in the afferent limb of the volume-sensitive cardio-renal reflex could be mediated at least in part by mechanosensitive ion channels of the DEG/ENaC-family located on vagal afferent nerve fibers. 1

Department of Nephrology, Med. Clinic IV, J.W.Goethe-University Frankfurt/M., Germany; 2Department of Nephrology, Friedrich-Alexander-University, Erlangen, Germany O 25-5 IMPAIRED GLUCOSE-INDUCED INSULIN SECRETION IN α2ADRENOCEPTOR KNOCK OUT MICE M. Peterhoff, A. Sieg, C.M. Chao, M. Brede*, L. Hein*, S. Ullrich Insulin secretion is regulated by stimulators and inhibitors. Adrenaline, an important physiological inhibitor of insulin secretion, is known to inhibit adenylyl cyclase and to hyperpolarize the cells. The α2adrenoceptor (AR) subtypes mediating these effects are unknown. Animals deficient in α2-AR subtypes were generated and the effects of adrenaline on insulin secretion and membrane potential examined. In both, α2A- and α2C-AR knock out (KO) mice islets, adrenaline (1 µM) inhibited secretion by 76 % and 100 %, respectively. Surprisingly, glucose (16.7 mM)-induced secretion in the absence and presence of 1µM forskolin was largely impaired in α2A-AR KO islets. However, when cAMP levels were increased dramatically by the combination of forskolin (5 µM) and IBMX (100 µM), secretion was stimulated 3-fold (9-fold in wt islets). When islets where cultured overnight in the presence of pertussis toxin (100 ng/ml), adrenaline-mediated inhibition was abolished and glucose-induced secretion improved (about 4fold in both KO and wt islets). In α2A/α2C-AR double KO mice islets adrenaline did not inhibit secretion. Membrane potential was measured using the amphotericin B permeabilized patch clamp method. At 2.8 mM glucose, membrane potential of α2A-AR KO cells was -55 ± 2 mV (n =17). Glucose depolarized to -35 ± 2 mV (n =18) and induced action potentials in all cells examined. Similar results were obtained with α2C-AR KO cells. Adrenaline (1 and 10 µM) hyperpolarized α2A- but not α2C-AR KO cells. These data suggest that two isoforms of α2- ARs, α2A and α2C, are expressed in insulin secreting cells. Glucoseinduced insulin secretion was impaired in α2A-AR KO animals through a mechanism which involves pertussis toxin sensitive G proteins. Institut für Neurophysiologie, Universität zu Köln, Robert-Koch-Str. 39, D-50931 Köln, Germany and *Institut für Pharmakologie/ Toxikologie, Universität Würzburg, Versbacher Straße 9, D-97078 Würzburg, Germany O 25-6

O 25-4 A PUTATIVE ROLE OF EPITHELIAL SODIUM CHANNELS (ENAC) IN THE AFFERENT LIMB OF THE VOLUME SENSITIVE CARDIO-RENAL REFLEX IN RATS T.Ditting1, O.Jung1, P. Linz2, KF. Hilgers2, H. Geiger1, R. Veelken2 Afferent signals of the volume sensitive cardio-renal reflex are mediated via afferent fibers of the vagus nerve. Increase of left ventricular end diastolic pressure (LVEDP) by volume loading decreases sympathetic renal nerve activity (SRNA) without significant changes in blood pressure. We hypothezised that this reflex response is mediated by stretch-sensitive ion channels of the DEG/ENaC (degenerin / epithelial

NICOTINIC ACETYLCHOLINE RECEPTOR (nAChR) MODULATION OF MESENTERIC AFFERENT DISCHARGE FROM THE ANAESTHETISED RAT JEJUNUM W. Jiang, A.K. Kirkup, M.E. Kreis, D. Grundy Acetylcholine released from neurones within the enteric nervous system may activate extrinsic afferent fibres and thus be a mediator of ”cross-talk” between the enteric and extrinsic innervation of the gastrointestinal tract. Nicotinic receptor agonists stimulate jejunal mesenteric afferents directly (Jiang et al., 2000, Gastroenteology, 120. A178). However, as multiple nicotinic acetylcholine receptor (nAChRs) subtypes are expressed in both dorsal root ganglion and nodose ganglion neurones (Genzen et al., 2001, J. Neurophysiol 86,

S 201 1773-82, Zhuo et al., 1997, Prog. Neurobiol. 52, 79-107), the nAChRs involved in modulating sensory transduction from the gut to the brain have yet to be characterised. The aim of this study was to investigate the functional contributions of homomeric alpha7 nAChRs to activation of extrinsic afferents supplying the rat jejunum in vivo and to determine any differential effects on intestinal motor activity and afferent sensitivity. Extracellular recordings of mesenteric jejunal afferent discharges were performed in pentobarbitone sodium (60mg kg-1, i.p.) anaethetized male Sprague Dawley rats (300-400g). Responses to systemic administration of the nicotinic agonist N,N-Dimethyl-N’phenyl-piperazinium iodide (DMPP) at doses of 100 and 300 µgkg-1 were quantified pre- and post-treatment with methyllacaconitine (MLA, α7 nAChR, 1mgkg -1) followed by mecamylamine (a non-selective neuronal nAChR antagonist, 1 mgkg-1). Data are expressed as mean±SEM from 4-7 animals and analysed using appropriate Student’s t tests corrected for multiple comparisons. DMPP provoked an intense dosedependent increase in afferent discharge with onset < 2.5s and duration < 15s, accompanied by a prolonged reduction in the luminal pressure, responses that were consistent in vehicle-treated animals. In contrast, MLA attenuated the magnitude of the afferent response to 100 µgkg1 DMPP from 119±9 spikes s-1 to 85±3 spikes s-1 (P<0.05 N=6) but had no significant effect on the afferent response to the higher dose of DMPP (from143±9 to 156±12 spikes s -1). Subsequent administration of mecamylamine almost completely abolished the response to both low (11± 3 spikes s-1) and high dose DMPP (4±2 spikes s-1). MLA had no effect on the magnitude of the falls in intestinal pressure in response to both low and high doses of DMPP while subsequent administration of mecamylamine completely abolished the inhibitory motor responses to DMPP. In conclusion, alpha7-nAChR contribute to the mesenteric afferent response to DMPP but not the intestinal motor inhibition arising from activation of enteric inhibitory neural pathways to the muscle. However, the major component of the mesenteric afferent response and the entire motor response is sensitive to mecamylamine implicating other neuronal nicotinic receptor subunits in sensory signal transduction from the intestine. 1

Department of Biomedical Science, University of Sheffield, Alfred Denny Building, Western Bank, Sheffield, S10, 2TN, UK. 2Department of General Surgery, School of Medicine, University of Tuebingen. Germany. O 26-1 BLOCKADE OF IP3-RECEPTORS MODULATES Ca 2+ SPARKS IN MYOCYTES FROM THE GUINEA PIG VAS DEFERENS J.G. McGeown, D.E. Kean, C. White The release of Ca2+ from the sarcoplasmic reticulum can give rise to localised intracellular Ca2+ transients. These signals may be generated by the opening either of inositol 1,4,5-triphosphate (IP3) receptors, giving rise to events known as Ca2+ ‘puffs’, or of ryanodine receptors, producing Ca 2+ ‘sparks’. The purpose of the present study was to characterise the properties of elementary Ca 2+ release events in myocytes isolated from the guinea pig vas deferens. Intracellular calcium concentration ([Ca2+]i) was monitored confocally (BioRad) using cells loaded with fluo-4AM (5 µM). Loaded cells were continuously superfused with physiological salt solution at 37oC. The fluorescence of a single scanline (0.2 µm width) was recorded at a frequency of 500 Hz. Data are expressed as F/F0 were F is the measured fluorescence intensity and F0 the resting intensity. Spontaneous focal increases in fluorescence with the properties of Ca2+ sparks [1] were observed under control conditions (Table 1). 2-Aminoethoxydiphenyl borate (2APB, 100 µM), an IP3-receptor blocker [2], altered the frequency, kinetics and spread of these events (Table 1). This suggests that IP3R activity can affect spontaneous release via ryanodine receptors, perhaps by affecting SR Ca2+ loading.

Control n=71 2APB n=91

Amplitude (∆F/F0) 1.9 ± 0.1 1.7 ± 0.1 P > 0.05

Frequency (Hz) 0.9 ± 0.2 1.4 ± 0.1 P < 0.05

FTHM (ms) 42.4 ± 5.6 66.9 ± 7.0 P < 0.01

FWHM (µm) 2.3 ± 0.3 3.2 ± 0.3 P < 0.05

Table 1. Values are expressed as means ± SEM, n represents the number of Ca2+ events. A t-test was used to assess statistical significance. FWHM = full width of Ca2+ transient at the half maximal amplitude. FTHM = full timecourse of Ca2+ transient at the half maximal amplitude. 1. Jaggar JH, Porter VA, Lederer WJ, Nelson MT. (2000) Am. J. Physiol.

278: C235-C256; 2. Maruyama T, Kanaji T, Nakade S, Kanno T, Mikoshiba K. (1997) J. Biochem. 122:498-505. Funded by the Wellcome Trust. Smooth Muscle Group, Dept of Physiology, 97 Lisburn Road, Queen’s University of Belfast, BT9 7BL. U.K. O 26-2 ROLE OF RHO-KINASE IN CONTROL OF FORCE, CALCIUM AND MLC 20 RELATIONSHIP IN URETERIC SMOOTH MUSCLE S. Shabir, T. Burdyga In smooth muscle, recent work has shown that the Rho pathway by decreasing myosin phosphatase activity is functionally important in tonic smooth muscle. The aim of our study was to investigate its importance in the ureter. As recent work has indicated large differences in the signalling pathway of guinea pig and rat ureteric smooth muscle, we have also determined whether its effects on [Ca2+]I and contraction differed between theses two species. To determine the input and activity of the Rho pathway to force production in the ureter, a specific inhibitor of Rho-associated kinase, Y27632 was used, and its effects on force, calcium, and phosphorylation of myosin light chains were measured in rat and guinea pig ureter strips. The strips were stimulated electrically to obtain phasic contractions, or with carbachol (100uM). Concentrations of Y27632 ranging from 0.01 to l0uM were added to obtain dose response curves for the inhibition of force and calcium by these different modes of stimulation. All data are means ± SEM. Application of 1uM Y27632 to rat ureter completely inhibited the tonic component of force upon carbachol stimulation, while only reducing calcium to 80±4.1% (n=10). The guinea pig ureter shows no response to carbachol stimulation. The data for electrical stimulation showed that the force in rat ureter was inhibited with an IC50 of 0.3uM (n=5), and the rate of relaxation of force was enhanced 1.5 times that seen in the control. Measurement of MLC phosphorylation showed that the dephosphorylation of myosin light chains was enhanced in the presence of Y27632. Under control conditions when calcium had fallen to 50% of peak, the phosphorylation was also around 50% of its peak. In the presence of Y27632 however, the phosphorylation had returned to basal levels. The inhibitory effects of Y27632 on the rate of relaxation and MLC could be reversed by the MLC phosphatase inhibitor calyculin A (1uM, n=3). In contrast, the guinea pig data showed no significant decrease in force or calcium with Y27632 up to 1uM, and only a small decrease using l0uM (n=3). These data clearly show that Rho modulation of force and calcium is more important in rat ureteric smooth muscle than in guinea pig. Physiological Laboratories, University of Liverpool, Crown Street, Liverpool L69 3BX O 26-3 MYOGENIC TONE IN RESISTANCE ARTERIES IS INCREASED AFTER TRANSFECTION OF VASCULAR SMOOTH MUSCLE CELLS WITH SPHINGOSINE KINASE DUE TO ACTIVATION OF THE RHOA/RHO KINASE PATHWAY St.-S. Bolz1, L. Vogel1, D. Sollinger1, R. Derwand1, S. Pitson2, U. Pohl1 The RhoA/Rho kinase pathway is an important modulator of microvascular tone. It was tested whether endogenous sphingosine-1phosphate (S1P) stimulates this pathway in resistance arteries (RA) and whether this lipid mediator affects resting tone and myogenic responses (MR). RA were kept in organ culture and incubated for 19-22h with plasmids coding for the S1P-generating enzyme sphingosine kinase (wild type Sphk: wtSphk) or its inactive mutant hSK-G82D. Dominant active RhoA (L63RhoA) or an inactive RhoA mutant (N19RhoA) or inactive Rho kinase (KD1A) were coexpressed with wtSphk to assess the involvement of the RhoA/Rho kinase pathway in mediating Sphk effects. GFP (green fluorescent protein)-transfected RA, that showed vascular responses virtually identical to untransfected RA, served as controls. Resting tone (rT) was significantly enhanced in RA expressing wtSphk (-23±3% of max. diameter vs. -10±1% in RAGFP, p<0.01) and L63RhoA (-22±2%) but almost abolished in hSK-G82D-transfected RA (-2%). Coexpression of N19RhoA or KD1A with wtSphk abolished development of rT. MR (initiated by pressure steps from 45 to 110mmHg) were enhanced after transfection of wtSphk and L63RhoA

S 202 (-154±14% and -92±12% of initial distension vs. -61±8% in RAGFP, p<0.01) but almost completely inhibited after coexpression of N19RhoA or KD1A or transfection with hSK-G82D. The ability of endogenous Sphk to activate the RhoA/Rho kinase pathway via its product S1P implicates a possible important role for the Sphk as a modulator of microvascular tone and myogenic reactivity. 1

Institute of Physiology, LMU, Schillerstrasse 44, 80336 Munich, Germany and 2Department of Medicine, University of Adelaide, Australia

N/M; P<0.01). Conclusions: The switch towards IK1 expression seems to be important in promoting excessive VSMC proliferation. The blockade of IK1 in proliferating neointimal VSMC could represent a new therapeutic action of clotrimazole to prevent restenosis after angioplasty. R. Köhler, Universitätsklinikum Benjamin Franklin, Hindenburgdamm 30, 12200 Berlin, Germany O 26-6

O 26-4 DESYNCHRONIZING EFFECT OF THE ENDOTHELIUM ON INTRACELLULAR CA2+ CONCENTRATION DYNAMICS IN VASCULAR SMOOTH MUSCLE CELLS OF RAT MESENTERIC ARTERIES M. Sell, W. Boldt, F. Markwardt The kinetics of the intracellular Ca2+ concentration ([Ca2+]i) of vascular smooth muscle cells (vsmcs) in small rat mesenteric arteries was investigated by confocal laser scanning microscopy using the fluorescent Ca2+ indicator fluo-3 AM. 1 µM noradrenaline (NA) induced randomly distributed transient elevations of [Ca2+]i in several single vsmcs which were weakly temporally coupled. Higher NA concentrations of 3 or 10 µM, however, induced strongly synchronized [Ca2+]i oscillations in vsmcs. In preparations with intact endothelium, the synchronization of [Ca2+]i signals was attenuated by acetylcholine but augmented by the NO synthase antagonist L-NAME, pointing to a desynchronizing effect of the endothelium already under basal conditions. In preparations with or without intact endothelium sodium nitroprusside (SNP) as well as the gap-junction uncoupler heptanol reversibly desynchronised the [Ca2+]i transients. The effect of ACh but not that of SNP was influenced by L-NAME. Propagated intracellular [Ca 2+]i waves had a velocity of 25 µm/s. Because the phase shift of [Ca2+]i oscillations between single vsmcs were maximally 2 s and independent on the distance of up to 90 µm between individual cells we consider intercellular [Ca2+]i waves to be too slow to account for the synchronization of [Ca2+]i oscillations. We conclude that the coupling of [Ca2+]i signals in vascular smooth muscle cells is not constant but highly regulated by NA and by endothelium derived NO. Oscillations of vessel contraction at high sympathetic tone may be induced by synchronisation of [Ca2+]i transients of distinct vsmcs whereas endothelium derived NO inhibits vasomotion by desynchronizing [Ca2+]i transients of single vsmcs. Julius-Bernstein-Institut für Physiologie, Martin-Luther-Universität Halle-Wittenberg, Magdeburger Str. 6, 06097 Halle (Saale), Germany

THROMBIN STIMULATES THE EXPRESSION OF TISSUE FACTOR VIA A RAC1-DEPENDENT PATHWAY IN VASCULAR SMOOTH MUSCLE CELLS A. Görlach, O. Herkert, R. Busse, J. Hess Thrombin is a potent activator for vascular smooth muscle cells and has been shown to induce the expression of tissue factor (TF) in these cells. TF is the main regulator of the extrinsic coagulation pathway finally leading to the generation of thrombin. However, the signaling pathways involved in this potential thrombogenic cycle are not well understood. Reactive oxygen species (ROS) have been implicated as signalling molecules in smooth muscle cells and are involved in the regulation of TF. The GTPase Rac plays an essential role in activating the ROS-generating enzyme NADPH oxidase. We therefore investigated whether Rac is involved in signaling pathways modulating the expression of TF. Human aortic or pulmonal smooth muscle cells were treated with a Clostridium difficile toxin B (TxB) which inhibits Rac but also Cdc42 and Rho. This treatment significantly inhibited thrombin-induced TF mRNA expression. To specifically investigate the involvement of Rac1, cells were transfected with expression vectors containing either wildtype Rac1 (Rac-WT), dominant-negative RacT17N or constitutive active RacG12V. Thrombin-stimulated TF mRNA expression was enhanced in RacT17N. In addition, TF-dependent procoagulant activity was lowered in RacT17N-transfected cells compared to cells expressing Rac-WT or RacG12V. Thrombin also stimulated the phosphorylation of p38 MAP kinase, ERK1/2 and protein kinase B/Akt. Treatment with TxB as well as expression of RacT17N prevented thrombin-induced phosphorylation of p38 MAP kinase, protein kinase B/Akt as well as p21-activated kinase (PAK), but not of ERK1/2. Consistently, thrombin-induced TF mRNA expression was attenuated by inhibitors of p38 MAP kinase as well as by wortmannin, but not by inhibitors of ERK1/2. These findings indicate that thrombin stimulates a Rac-dependent pathway leading to the activation of p38 MAP kinase and protein kinase B/Akt and subsequently to stimulating TF expression and TF-dependent procoagulant activity. Thus, the small GTPase Rac1 may play an essential role in promoting the thrombogenic cycle in smooth muscle cells. Experimentelle Kinderkardiologie, Deutsches Herzzentrum München, Klinik an der TU München, Lazarettstr. 36, D-80636 München

O 26-5 2+

+

INHIBITION OF THE IK1 CA -ACTIVATED K CHANNEL BY CLOTRIMAZOLE AS A NEW THERAPEUTIC STRATEGY FOR RESTENOSIS R. Köhler, I. Eichler, M. Kneifel, R. Real, , I. Grgic, J. Wibawa, S. Brakemeier, K. Borner, HD Orzechowski, J. Hoyer Objective: Angioplasty stimulates proliferation and migration of vascular smooth muscle cells (VSMC) leading to neointimal thickening and restenosis. Here we investigated whether alterations in hyperpolarizing Ca2+ activated K+ channels (KCa) are involved in neointimal cell proliferation after angioplasty. Methods and Results: Function and expression of KCa in neointimal VSMC were investigated by combined single-cell RT-PCR and patch-clamp techniques in situ in intact vessel slices of rat carotid arteries at 2 and 6 weeks after balloon catheter injury (BCI) and in mature freshly isolated carotid VSMC. In single-cell RT-PCR experiments, expression of large conductance KCa gene rSlo was detected in 87 % of normal VSMC samples. rSlo expression was significantly reduced in neointimal VSMC samples at 2 weeks after BCI (36% rSlo +-samples, P<0.01) and was almost absent in neointimal VSMC at 6 weeks after BCI (3% rSlo+-samples, P<0.001). In contrast to mature VSMC, neointimal VSMC at 2 weeks frequently expressed the intermediate-conductance KCa (63% rIK1+-samples, P<0.01). At 6 weeks after BCI, rIK1 expression was much less often detected (16% rIK1+-samples). Changes in K Ca gene expression pattern were paralleled by corresponding alterations in hyperpolarizing K Ca currents. Treatment of the rats for 2 weeks after BCI with the IK1 blocker clotrimazole (CLT 120 mg/kg/d) resulted in a pronounced reduction of neointimal/medial ratio (CLT 0.30 ± 0.05 N/M vs controls 0.57 ± 0.07

O 27-1 SHAPING DENDRITIC CALCIUM KINETICS: THE ROLE OF ENDOGENOUS BUFFERS AND EXTRUSION MECHANISMS H. Schmidt1, K.M. Stiefel1, P. Racay2, B. Schwaller2, J. Eilers1 Ca 2+ transients in dendrites and spines of central neurons represent short-lived and highly compartmentalized signals. They serve a variety of neuronal functions that are likely to be determined by their spatial and temporal pattern. We used high-resolution confocal microfluorometry to quantify how Ca2+ buffers, diffusion and extrusion mechanisms shape climbing fiber-mediated Ca2+ transients in spiny dendrites of cerebellar Purkinje cells. In wild-type (WT) mice, activation of the afferent climbing fiber induced Ca2+ transients with peak amplitudes of ~300 nM and biphasic decays with time constants of 30 and 400 ms. In Parvalbumin-deficient (PV-/-) mice the peak amplitudes were not different from the WT while the decay was slightly prolonged (by ~ 40 ms). In Calbindin/Parvalbumin double-knockout mice (CB/PV-/-) the peak amplitudes were about two-fold larger than in the WT and the decay kinetics were substantially accelerated but still biphasic (~30 and 170 ms). In combination with a numerical simulation of dendritic Ca2+ extrusion mechanisms we conclud that i) CB acts as a fast buffer that limits the peak amplitude of Ca2+ signals while representing a relevant Ca2+ source during the decay of Ca2+ transients; ii) PV acts as a slow buffer that does not affect the peak amplitude but accelerates the decay due to delayed Ca2+ buffering; iii) dendritic extru-

S 203 sion mechanisms are about two-fold more efficient than somatic extrusion mechanisms and, as in the soma, exhibit a non-linear dependence on [Ca2+]i. Our results demonstrate a dynamic interplay between buffers and extrusion mechanisms in shaping the waveform of Ca2+ signals. The kinetic profile of PV and CB as well as the non-linear activity of dendritic Ca2+ pumps seems to be fine tuned to effectively modulate physiological Ca2+ signals. 1

Department of Neurophysiology, Max-Planck-Institute for Brain Research, 60528 Frankfurt, Germany, and 2Institute of Histology and General Embryology, University of Fribourg, 1705 Fribourg, Switzerland

release and causes volume decrease by net efflux of water. It seems to be unlikely that a [Ca2+]i - dependent Cl - - efflux participates in gap junctional uncoupling. Release of anions should take place in parallel with K+ release because of the restraint of electroneutrality. It is proposed that a K+ release from granulosa cells is a prerequisite to apoptotic cell death. Institute of Biophysics, University Hanover, Herrenhaeuser Strasse 2, D-30419 Hanover, Germany O 27-4

O 27-2 CALCIUM RELEASE AND CALCIUM ENTRY IN MAMMALIAN ASTROCYTES J.W. Deitmer, R. Zur Nieden, H.P. Schneider, A. Beck Calcium signalling in glial cells has received increasing attention recently due to its potential role in cellular information processing in nervous systems. The most prevalent pathway is via G-protein-coupled receptors generating IP 3 resulting in calcium release from the endoplasmic reticulum (ER). Depleted ER stores then initiate the storeoperated calcium entry (SOCE) to refill with calcium. Mitochondria take up calcium with lower affinity, but large capacity. We have studied the relative contribution of ER and mitochondria to store, release and induce entry of calcium in cultured rat cerebellar astrocytes and in hippocampal astrocytes from acute rat and mouse brain slices. Identification of astrocytes in acute brain slices was achieved online with the calcium responses in low-potassium saline (Dallwig & Deitmer, 2000, Cell Calcium 28, 247-259), and with immunocytochemical methods. We have employed metabotropic receptor agonists like ATP and phenylephrine (α1-adrenergic receptor agonist), mitochondrial inhibitors such as carbonyl cyanides and oligomycin, and the inhibitor of the endoplasmic reticulum calcium pump cyclopiazonic acid (CPA) to evaluate the contribution of ER and mitochondria for intracellular calcium signalling in astrocytes. The rate and the amplitude of the cytosolic calcium change were compared in the presence and absence of external calcium and CPA, when intra- and extracellular compartments were calcium-depleted. Astrocytes show both spontaneous and evoked calcium transients and oscillations, which depend upon intracellular calcium release; they can be maintained in calcium-free saline for a short while, before they stop and only resume with the extracellular calcium readded. Our results suggest some functional interaction between the different calcium pathways, whereby the SOCE seems to be modulated by previous calcium signalling. In conclusion, calcium signalling in astrocytes involves a complex set of interacting calcium pathways, of which calcium release and voltage-independent calcium entry are the key regulated processes. FB Biologie, Univ. Kaiserslautern, Postfach 3049, D-67653 Kaiserslautern O 27-3 K+ - EFFLUX MEDIATES LOSS OF CELL VOLUME ACCOMPANIED BY AN INCREASE OF CYTOPLASMIC FREE CALCIUM AND GAP JUNCTIONAL UNCOUPLING IN GRANULOSA CELLS A. Ngezahayo, B. Altmann, H.-A. Kolb Maturation of the ovule is linked to the functioning of the surrounding granulosa cells. Gap junctional coupling of granulosa cells is considered as prerequisite for a synchronised response to hormones like gonadotropins and estradiols. For the first time we applied the dual whole-cell technique to pairs of cultured granulosa cells to analyse the physiological regulation of gap junctional coupling. After establishing the dual whole-cell configuration the cells shrink and morphological deterioration occur which were accompanied by cell-cell uncoupling. In parallel, a sustained increase of cytoplasmic free calcium ([Ca2+]i) was recorded by use of Fura-II. Cell shrinkage could not be prevented by an appropriate adjustment of the osmotic pressure of pipette solution. Addition of 5 mM ATP did not suppress the recorded electrical uncoupling, but a detailed analysis showed that also the K+-efflux and anion release had to be blocked. Replacement of 140 mM KC1 in the pipette solution by 40 mM CsCl and 100 mM K-gluconate prevented an increase of [Ca2+]i, the observed morphological deterioration like cell blebs and stabilised gap junctional coupling. The data indicate that gap junctional uncoupling is mediated by an sustained increase of [Ca2+]i which promotes K+

INVOLVEMENT RHO GTPASES IN THE CALCIUM SIGNALING INDUCED BY ACTIVATION OF MUSCARINIC RECEPTORS IN RBL 2H3-HM1 MAST CELLS E. Aneiros, N. Djouder, K. Aktories, A. Cavalié Previous studies suggested an important role of Rho GTPases in the calcium mobilization of immune cells. Using a Fura-2 imaging system, we examined the calcium signaling elicited by activation of muscarinic (M1) and high affinity IgE receptors (FcεRI) in RBL 2H3-hm1 cells and used the toxin B of Clostridium difficile to inhibit Rho, Rac and Cdc42. In cells primed with anti TNP IgE (0.3 µg/ml), the activation of the FcεRI pathway with TNP-OVA (50 ng/ml) evoked elevations of the intracellular calcium concentration with plateaus lasting more than 10 min in the presence of extracellular calcium (1.8 mM). The calcium plateaus induced through the M1 pathway with carbachol (1 mM) were regularly shorter than 3 min. Carbachol had no effect on calcium mobilization when it was applied during calcium plateaus evoked with TNP-OVA. In the absence of extracellular calcium, carbachol had also no effect on calcium release when the calcium stores were previously depleted with TNP-OVA. Furthermore, the calcium release induced via M1 and FcεRI receptors seemed not to be affected in cells treated with toxin B (40 ng/ml, 2 h). By contrast, toxin B disrupted the calcium plateaus induced through the FcεRI pathway but not when the stimulation was via M1 receptors. Typically, cells treated with toxin B responded to FcεRI receptor stimulation with oscillations of the intracellular calcium concentration for periods longer than 5 min. The calcium oscillations of cells treated with toxin B disappeared when calcium was removed from the extracellular solution. These data suggest that M1 and FcεRI pathways do not linked calcium entry to store depletion in the same way. Additionally, Rho GTPases appear not no to be essential for the calcium mobilization induced through the M1 pathway in RBL 2H3-hm1 cells. Pharmakologie und Toxikologie, Universität des Saarlandes, Homburg; Pharmakologie und Toxikologie, Universität Freiburg, Freiburg, Germany O 27-5 CELL SIGNALING IN THE HYPERTONIC ACTIVATION OF RAT HEPATOCYTE Na+ CONDUCTANCE H. Olsen, K. Bierhals, C.-T. Lin, H. Waldmann, F. Wehner The activation of an amiloride-sensitive Na + conductance is one of the key elements in the regulatory volume increase of rat hepatocytes. It could be shown recently that this process is mediated by protein kinase C (PKC). Here, the actual isoform of PKC employed was specified and some of the mechanisms further upstream to PKC were defined. This was done in primary cultures and by use of cableanalysis, intracellular recordings with low Na+ pulses, and Western blots. The activation of Na+ conductance was found to be Ca++ independent and, under isotonic conditions, it could be mimicked by indolactam V indicative that n-type isoforms of PKC mediate the process. Because the nPKCs expressed in rat hepatocytes are δ and ε, and ε did not respond to hypertonicity, this rendered PKCδ the most likely isoform mediating the activation of Na + conductance. This could be confirmed with immunoblots and, under isotonic conditions, by use of a specific activator of PKCδ. The hypertonic activation of Na + conductance was inhibited by genistein, pertussis toxin, and wortmannin. These findings led to a working hypothesis in which hypertonicity activates phospholipase Cγ (PLCγ), likely by means of a disassembly of actin filaments. This activation is modified by tyrosine kinases and via Gi and/or Go proteins and this modulation most likely involves PI3-kinase. PLCγ then stimulates PKCδ by production of diacylglycerol (DAG). The activation of Na+ conductance was also blocked by the microtubule-disruptor nocodazole which may sug-

S 204 gest that the exocytotic insertion of channels into the plasma membrane plays an additional (constitutive) role in the activation process. Max-Planck-Institut für molekulare Physiologie, Abteilung Epithelphysiologie, Otto-Hahn-Straße 11, 44227 Dortmund, Germany O 27-6 THE INTERACTION OF 14-3-3 WITH RGS7 IS DYNAMICALLY REGULATED BY EXTRACELLULAR SIGNALS, AND CONTROLS RGS7 ACTIVITY IN VIVO M. Köttgen*, T. Benzing*, M. Johnson, B. Schermer, H. Zentgraf, E. Kim, G. Walz Regulator of G protein Signaling (RGS) proteins accelerate the intrinsic GTP hydrolysis of activated Gαi and Gαq subunits. The adaptor protein 14-3-3 binds to a subset of RGS proteins, and inhibits their GTPase accelerating activity in vitro. The interaction between 14-3-3 and RGS7 requires phosphorylation of serine 434 of RGS7. We demonstrate now that phosphorylation of this serine residue is dynamically regulated by extracellular signals. Cellular stimulation by TNF-α or G protein coupled receptor activation transiently decreased the phosphorylation of serine 434 of RGS7, thereby abrogating the inhibitory interaction with 14-3-3. To demonstrate that the interaction serves as a molecular switch to control RGS7 function in vivo we investigated the effect of 14-3-3 on RGS-mediated deactivation kinetics of G protein-coupled inwardly rectifying K channels (GIRKs). We co-expressed the muscarinic acetylcholine receptor (M2), GIRK 1/2 and RGS proteins in Xenopus oocytes and studied RGS protein function using standard double electrode voltage clamp. Injection of recombinant 14-3-3 expressed in bacteria, inhibited the function of wild-type RGS7, but had no effect on either RGS7P436R or RGS4, two proteins that fail to bind 143-3. Our findings demonstrate for the first time that extracellular signals can modulate the activity of RGS proteins by regulating their interaction with 14-3-3. * These authors contributed equally to this work. Universitätsklinik Freiburg, Medizinische Klinik (Abt. IV), Hugstetterstr. 55, 79106 Freiburg

O 28-1 TNFalpha INHIBITS RENIN TRANSCRIPTION BOTH IN VIVO AND IN VITRO V. Todorov, M. Müller, A. Kurtz Background/Objectives: Renin, produced in renal juxtaglomerular (JG) cells, is a fundamental regulator of blood pressure. Accumulating evidence suggests that cytokines may influence directly renin production in the JG cells. TNFalpha, which is one of the key mediators in immunity and inflammation, is known to participate in the control of vascular proliferation and contraction, and hence in the pathogenesis of cardiovascular diseases such as hypertension and atherosclerosis. Therefore we have tested the effect of TNFalpha on renal renin transcription in As4.1 cells, which represent transformed mouse JG cells, and in native mouse JG cells in culture. Renin gene expression was also determined in genetically engineered mice, lacking the gene for TNFalpha (TNFalpha knockout mice). Results: TNFalpha inhibited renin gene expression via an inhibition of the transcriptional activity, targeting the proximal 4.1 kb of the renin promoter. TNFalpha was also found to attenuate the forskolin stimulated renin gene expression in primary cultures of mouse JG cells. Mice lacking the TNFalpha gene were shown to have almost 3-fold higher basal renal renin mRNA abundance relative to the control strain. The TNFalpha knockout mice and the wild type control mice displayed no differences in the magnitude of stimulation of renin gene expression by low-salt diet (10 days, 0.02% w/w) plus ACE-inhibitor (3 days ramipril, 10 mg/kg/day) treatment. However, high-salt diet (10 days, 4% w/w), which inhibited renal renin gene expression about 2-fold in wild type mice, did not affect the abundance of renal renin mRNA in TNFalpha knockout mice. Conclusions: Our data suggest that TNFalpha inhibits renin gene expression in vitro and in vivo. TNFalpha seems to be involved in the regulation of renal renin production by dietary salt intake. Institute of Physiology, Regensburg University, D-93040 Regensburg, Germany

O 28-2 Β-INDEPENDENT MECHANISM EVIDENCE FOR A NOVEL TGF-Β OF FIBRONECTIN PRODUCTION IN MESANGIAL CELLS OVEREXPRESSING GLUCOSE TRANSPORTERS C. Weigert, K. Brodbeck, M. Huber, R. Lehmann, U. Friess, S. Facchin, S. Aulwurm, C.W. Heilig, H.U. Häring, E.D. Schleicher Recent experimental work indicates that the hyperglycemia-induced increase in mesangial matrix production, a hallmark in the development of diabetic nephropathy, is mediated by an increased expression of glucose transporter 1 (GLUT1). Mesangial cells stably transfected with the human GLUT1 (GT1 cells) mimick the effect of hyperglycemia on the production of the extracellular matrix proteins, particularly fibronectin, when cultured under normoglycemic conditions. In the present study we investigated the molecular mechanism of the increased fibronectin production in GT1 cells. In contrast to normal mesangial cells the increased fibronectin production was not mediated by the prosclerotic cytokine TGF-β1. We found markedly increased nuclear content in Jun proteins leading to enhanced DNA binding activity of AP-1. Addition of the AP-1 inhibitor curcumin reduced fibronectin production dose-dependently. Moreover, inhibition of classical protein kinase C isoforms prevented the activation of AP-1 and the enhanced fibronectin production. In contrast to mesangial cells exposed to high glucose no activation of the hexosamine biosynthetic pathway, the p38 or the ERK1/2 MAPK pathway and no increase in TGF-β1 synthesis could be detected, which can be explained by the absence of oxidative stress in the GT1 cells. Our data indicate that increased glucose uptake and metabolism induces PKC-dependent AP1 activation which is sufficient for an enhanced fibronectin production, but not for an increased TGF-β1 expression. Department of Internal Medicine, Division of Endocrinology, Metabolism and Pathobiochemistry, University of Tübingen, OtfriedMüller-Straße 10, D-72076 Tübingen, Germany O 28-3 RAC1 MODULATES THE EXPRESSION OF PLASMINOGEN ACTIVATOR INHIBITOR-1 A. Görlach, T. Kietzmann Plasminogen activator inhibitor 1 (PAI-1) is the primary physiological inhibitor of plasminogen activation. PAI-1 levels are elevated in patients with coronary heart disease and may play an important role in the development of atherothrombosis. Increased PAI-1 expression can also directly influence vessel wall remodelling. A number of stimuli such as hypoxia and TGF-ß have been shown to activate PAI-1 expression. However, the signaling pathways regulating the expression of PAI-1 are not completely understood. TGF-ß has been shown to modulate gene expression via an NADPH oxidase. The GTPase Rac1 is critically required for activation of the NADPH oxidase. We therefore investigated whether Rac1 is involved in modulating the expression of PAI-1. Human hepatoma HepG2 cells were transfected with expression vectors containing either wildtype Rac1 (Rac-WT), dominant-negative RacT17N or constitutive active RacG12V. Northern and Western blot analyses demonstrated that in the presence of RacT17N, PAI-1 levels were enhanced compared to cells transfected with Rac-WT whereas expression of RacG12V decreased PAI-1 levels. Consistently, cotransfection of luciferase (Luc) constructs containing the hypoxia- and TGF-ß- responsive human PAI-1 promoter resulted in decreased Luc activity in the presence of RacG12V. MAP kinases and protein kinase B/Akt have been shown to act downstream of the NADPH oxidase and Rac1 to regulate gene expression. Transfection of RacT17N decreased phosphorylation of ERK1/2 and protein kinase B/Akt but enhanced phosphorylation of p38 MAP kinase whereas expression of RacG12V stimulated phosphorylation of ERK1/2 and protein kinase B/Akt. Cotransfection of the PAI-1 Luc gene construct with a vector encoding ERK decreased Luc activity whereas cotransfection of vectors expressing MKK3 or MKK6, the upstream kinases of the p38 MAP kinase, enhanced PAI-1 Luc activity. Treatment with wortmannin, an inhibitor of phosphoinositol-3 kinase, an upstream kinase of protein kinase B/Akt, on the other hand, elevated PAI-1 Luc activity. These findings demonstrate that dominant-negative Rac1 promotes phosphorylation of p38 MAP kinase but inhibits phosphorylation of ERK1/2 and protein kinase B/Akt. Thus, Rac1dependent modulation of the activity of these kinases may be involved in the pathogenesis of diseases associated with enhanced PAI1 levels.

S 205 Experimentelle Kinderkardiologie, Deutsches Herzzentrum München, Klinik an der TU München, Lazarettstr. 36, D-80636 München

O 28-4 β -HYDROXYEFFECT OF BILE DUCT LIGATION ON 11β STEROID DEHYDROGENASE TYPE 2 EXPRESSION IN THE KIDNEY AND COLON H.C. Thiesson, B.L. Jensen , O. Skøtt Late stages of liver failure are characterised by accumulation of sodium and water, and this cannot be completely explained by increased aldosterone concentrations. The aim of this study was to clarify whether reduced activity of 11β -hydroxysteroid dehydrogenase type 2 (11βHSD2) in biliary cirrhosis would result in apparent mineralocorticoid excess with sodium retention and potassium loss. Aldosterone stimulates transepithelial sodium transport in the kidney and colon, and 11β-HSD2 provides selective access of aldosterone to the mineralocorticoid receptor by inactivating glucocorticoids. In male Wistar rats, we established a model of decompensated biliary cirrhosis (ascites volume 28 ± 10 ml) by bile duct ligation (BDL). Total RNA was extracted from colon and kidney tissue using the RNeasy Midi Kit (Qiagen). Expression of renin and 11β-HSD2 were measured with a Ribonuclease protection assay. In the kidney expression of cortical 11β-HSD2 mRNA decreased after 8 weeks BDL (834 ± 70 cpm/total RNA, p<0.05, n=10), compared to the 8 weeks sham group (1838 ± 105 cpm/total RNA, n=12). In colon, BDL induced an up-regulation (7620 ± 1117 cpm/ total RNA, n<0.01) of 11β-HSD2 mRNA expressions after 8 weeks compared to the 8 weeks sham group (2082 ± 375 cpm/total RNA). Western blot analysis confirmed lower protein levels of 11β-HSD2 in the BDL group. Water excretions were similar in the two groups, while sodium excretion decreased in the BDL group (290 ± 82 µmol Na+ /24 h/100 g, p<0.05) compared to the sham group (558 ± 48 µmol Na+ /24 h/100 g). Plasma potassium decreased significantly in the 8 weeks BDL group (5.1± 0.04 mmol/1, p<0.05) compared to the sham group (5.6 ± 0.04 mmol/1). No significant change could be detected in p-renin-, paldosterone concentration and renal cortical renin mRNA expression between the groups. Conclusion: Downregulation of 11β-HSD2 in renal cortex might explain the apparent mineralocorticoid excess in bile duct ligated rats. Dept of Physiology and Pharmacology, University of Southern Denmark, Winsløwparken 213, DK-5000, Odense C, Denmark

O 28-5 ASYMMETRIC INTERACTIONS OF CORTICOSTERONE WITH THE ORGANIC CATION TRANSPORTER OCT2 C. Volk, V. Gorboulev, T. Budiman, C. Popp, G. Nagel, H. Koepsell Members of the OCT1 transporter family play a crucial role in homeostasis of organic cations. Since the first organic cation transporter (OCT1) was cloned in our labarotory in 1994, many additional members including organic anion transporters have been cloned. The subfamily of organic cation transporters consists of three subtypes (OCT1-3) which had been cloned from various species. These proteins are able to translocate a variety of organic cations, and are sensitive to several structurally different inhibitors which are not transported themselves. Inhibition of OCTs by corticosterone has been reported previously, but the mechanism of inhibiton is not understood. In this study we investigated this inhibition in detail using Xenopus laevis oocytes as expression system. First, uptake measurements with radiolabelled substrates were performed. All OCT subtypes tested were strongly inhibited by corticosterone, but with different affinities (IC50 values between 1µM (hOCT3) and 150µM (rOCT1). As corticosterone can diffuse easily through the membrane, it cannot be discriminated if the corticosterone binding site is located on the inner or outer face of the transporter. Therefore, electrophysiological measurements including two electrode voltage clamp (TEVC) and giant patch clamp were employed in oocytes expressing rOCT2. Interestingly, in TEVC measurements, different IC50 values were observed, depending on the duration of corticosterone preincubation (65 µM without, 4.8 µM after

10 min preincubation). This indicates that corticosterone inhibits on the outside with a comparatively low affinity, but diffuses into the cell and can inhibit on the inside with a tenfold higher affinity. Inside-out giant patch measurements, which allow a direct control of the cytoplasmic side yielded IC 50 values which correspond to the values obtained after corticosterone preincubation (2 µM). Using different substrate concentrations (100µM – 10mM choline) we found that the corticosterone inhibition is not competitive, in contrast to another high affinity inhibitor, tetrabutylammonium. As corticosterone can inhibit from both sides, we hypothesize that the binding site for corticosterone is located within the substrate binding pocket that changes from outside to inside conformation during the transport cycle. Competition between choline and corticosterone may not be observed because the respective binding domains within the binding pocket may overlap only partially. Supported by Deutsche Forschungsgemeinschaft (SFB 487/A4). Institut für Anatomie und Zellbiologie, Bayerische Julius-MaximilliansUniversität, Koellikerstr.6, 97070 Würzburg

O 28-6 ROLE OF PUTATIVE PKC-DEPENDENT PHOSPHORYLATION SITES IN THE REGULATION OF THE RAT ORGANIC CATION TRANSPORTER rOCT1 G. Ciarimboli, V. Gorboulev, H. Koepsell, E. Schlatter The amino acid structure of rOCT1 contains several potential PKCdependent phosphorylation sites (PKC-sites), which may be involved in transport regulation. Previously, we reported phosphorylation and an increase of the substrate affinity of rOCT1 by PKC-activation (J Am Soc Nephrol 11: 1216-1224, 2000). In the present work, we compared the functional properties of rOCT1 after site-directed mutagenesis of PKC-sites (rOCT1x) with wild-type rOCT1 (rOCT1wt). The transporters were stably transfected in HEK293 cells and their function was examined fluorimetrically with the cation 4-(4-(dimethylamino)styril)-methylpyridinium iodide (ASP+, 1 µM) as substrate. Activation of PKC with sn-1,2-dioctanoyl glycerol (1µM) and of PKA with forskolin (1µM) significantly stimulated rOCT1wt–mediated ASP + accumulation by 95.3±37.4% (n=12) and 60.9±20.9% (n=13), respectively, but had no influence on rOCT1 x –mediated ASP + accumulation. Inhibition of p561ck tyrosine kinase by aminogenestein (10 µM) led to a reduction of ASP+-uptake in both rOCT1x and rOCT1wt (74.3±3.5%, n=11 and 54.5±8.7%, n=11). The affinities of rOCT1x for TEA+ and for quinine were significantly lower than those of rOCT1wt (IC50 values: 362 µM vs 86 µM for TEA+ and 2.1 µM vs 0.65 µM for quinine, respectively). In conclusion, we found that 1) the presence of PKC-sites is important for the regulation of rOCT1wt transport activity; 2) an endogenous activation of the p561c tyrosine kinase is present in both rOCT1x and rOCT1wt and 3) mutations of PKC-sites lead also to changes in the substrate affinity of the transporter. These findings support the hypothesis that direct phosphorylation of the rOCT1 protein regulates the transport capacity probably due to an increased affinity to the substrates. Universitätsklinikum Münster, Medizinische Klinik und Poliklinik D, Experimentelle Nephrologie, Domagkstr 3a, 48149 Münster, Germany

O 29-1 UNCONTROLLED INSULIN SECRETION FROM PANCREATIC ß-CELL ADENOMA DOES NOT INVOLVE DEFECTS IN ATPSENSITIVE POTASSIUM CHANNELS K.E. Cosgrove 1, P.D. Barnes1, R. O’Brien1, A. Natarajan 1, A.T. Lee1, R.M. Shepherd 1, A.-M. Gonzalez 1, K. Hussain 2 , P.E. Clayton 3, A. Aynsley-Green2, K.J. Lindley2, J. Rahier4, B. Glaser5, M.J. Dunne1 Hyperinsulinism in Infancy (HI) is the most common cause of recurrent or persistent hypoglycaemia in early childhood. In children under the age of one year the disease (usually) manifests as either diffuse abnormalities of pancreatic ß-cell function (Di-HI), or focal adenomatous hyperplasia of ß-cells (Fo-HI). Di-HI is caused by defects

S 206 in the KATP channel genes ABCC8 (SUR1) or KCNJ11 (Kir6.2). Fo-HI is a condition of unregulated insulin release arising from the somatic loss of maternal heterozygosity resulting in the expression of paternallyderived mutation(s) in SUR1 or Kir6.2, and the loss of genes that are normally responsible for controlled cell growth and proliferation, including H19 and p57 kip2 (CDKN1C). By contrast, little is known about the molecular pathogenesis of ß-cell adenomas. We have isolated tissue from 4 patients with adenoma (AD) following surgery and from seven patients with Fo-HI (with permission). AD subjects were aged 1 month, 8.5 years, 14 years and 79 years at the time of surgery and we have used patch-clamp techniques, immunofluorescence, RT-PCR and genetic studies to document the pathogenesis of this condition with reference to Fo-HI. In all 4 patients with AD we consistently found KATP channels in ß-cells when recordings were made using the inside-out patch configuration. The expression of SUR1 and Kir6.2 was also confirmed by RT-PCR. By contrast, peak K ATP channel values were reduced to zero in 3 patients with Fo-HI and to an average of <9% of control values in 4 others; AD 13±3pA (n=36, 4 patients) vs. 25.5±1.5pA (n=269, 56 donors) in control human ß-cells, and 1.7± 0.3pA (n=20, 4/7 patient) in Fo-Hi patients. In AD ß-cells KATP channels responded to ATP (0.5mM), diazoxide (0.2mM) and tolbutamide (0.25mM) in the normal manner, and were also activated by internallyapplied ADP and GDP (n=20). In the case of the 8.5 year old AD patient, we were also able to show that in comparison to patients with Fo-HI, genetic analysis of the isolated adenoma revealed no loss of heterozygosity for chromosome 11p15 (ABCC8, Exon 31 R1274R polymorphism) and expression of the cyclin-dependent kinase inhibitor p57kip2 in both the adenoma and non-adenoma portions of the pancreas. Summary: These data document that hyperinsulinism as a consequence of adenoma has a different aetiology to that observed in HI which is due to abnormalities in the ß-cell KATP channel and ß-cell proliferation. 1 Dept Biomedical Science, Sheffield University, Sheffield, UK; 2Institute of Child Health, University College London, London, UK; 3Manchester Children’s Hospital, Manchester University, Manchester, UK; 4 Université Catholique de Louvain, Brussels, Belgium; 5Department of Endocrinology, Hadassah University, Jerusalem, Israel

O 29-3 ATP HYDROLYSIS-INDEPENDENT MECHANISM OF K ATP CHANNEL ACTIVATION DIFFERS AMONGST CHANNEL OPENER U. Russ, U. Lange, C. Löffler-Walz, A. Hambrock, U. Quast Smooth muscle KATP channels are activated in the presence of MgATP by several vasodilator compounds, the KATP channel openers. There is growing evidence that the openers activate the channel by enhancing ATP hydrolysis by SUR, the regulatory subunit of KATP channels to which the openers bind. However, at high concentrations, openers activate the channel also in the absence of MgATP. Here we describe the interaction of openers with the channel in the absence of Mg2+. The low affinity binding of openers in the absence of Mg2+ was measured by inhibition of [3H]glibenclamide binding to SUR2B(Y1206S), a mutant with high affinity for glibenclamide. Inhibition by P1075 and rilmakalim occurred with K i values approximately 200 fold higher than the high affinity Ki in the presence of MgATP. Diazoxide binding was unaffected by the absence of MgATP and only very poor binding was detected for minoxidil sulphate and nicorandil up to 300 µM. Electrophysiological experiments on Kir6.2 / SUR2B channels were performed in inside-out patches with a cytoplasmic solution without Mg 2+ (+ 5 mM EDTA). In the presence of 0.1 mM ATP, P1075 opened the channels from 1 to 100 µM in a concentration-dependent manner. In the beginning, efficiency of P1075 decreased exponentially with ~ 25 % min-1 and, after ~ 20 min, reached a stable level of about 20 % of the current in the absence of ATP. 10 µM rilmakalim was less effective than 10 µM P1075 and showed a decreased effect at 100 µM. Diazoxide [300 µM] and minoxidil sulfate [300 µM] produced no or very small channel activation; nicorandil at 100 µM or 1 mM, however, was effective in half of the patches. This results show that KATP channel openers differ in their ability to activate the channel without stimulating ATP hydrolysis. Department of Pharmacology, Medical Faculty, University of Tübingen, Wilhelmstr. 56, D-72074 Tübingen O 29-4

O 29-2 NOVEL SPLICE VARIANTS OF SULFONYLUREA RECEPTOR SUR1 A. Hambrock 1, U. Russ 1, A. Piehl1, R. Preisig-Müller 2, P. Hanley2, J. Ray2, J. Daut2, U. Quast1, C. Derst2 ATP-sensitive potassium channels (K ATP channels) are gated by nucleotides and link the metabolic state of a cell to its excitability. They are composed of pore-forming Kir6.x subunits and sulfonylurea receptors (SURs) as regulatory subunits. SURs are members of the ATP-binding cassette proteins with two nucleotide binding folds (NBFs) and binding sites for the blocking sulfonyl-ureas like glibenclamide and for the openers. The channel formed by SUR1 and Kir6.2 couples insulin secretion from the pancreatic ß-cell to plasma glucose levels and is also the dominant KATP channel in many neurons. In this study we report the identification of four novel splice variants of guinea-pig SUR1 by RT-PCR and their functional characterization. Three of these splice variants originate from alternative splicing of the region coding for NBF1 and lack exon 17 (SUR1∆17), 19 (SUR1∆19) or both (SUR1∆17∆19). The fourth (SUR1C) is a C-terminal SUR1-fragment formed by exons 31-39 containing the last two transmembrane segments and the carboxy terminus of SUR1. RT-PCR analysis showed that these splice variants were expressed in various guinea-pig tissues, with strong expression of SUR1C in cardiomyocytes. Confocal microscopy using EGFP-tagged SUR or Kir6.x did not provide any evidence for involvement of these splice variants in the mitochondrial K ATP channel, the molecular composition of which is still under discussion. Only SUR1 and SUR1∆17 showed high affinity binding of glibenclamide with no significant differences in the pharmacological properties and the MgATP dependence of glibenclamide binding. In addition, only SUR1 and SUR1∆17 formed functional KATP channels upon coexpression with Kir6.2 and showed similar sensitivity to block by ATP. The physiological role of the four splice variants remains unknown. 1

Institute for Pharmacology, University of Tübingen, D-72074 Tübingen; 2 Institute for Normal and Pathological Physiology, University of Marburg, D-35037 Marburg, Germany

Mg2+ SENSITISES KATP CHANNELS TO INHIBITION BY DIDS IN A SUR-DEPENDENT MANNER L. Gojkovic Bukarica, U. Russ, A. Hambrock, U. Quast ATP-dependent K + (KATP ) channels are composed of pore-forming subunits (Kir6.2) and sulphonylurea receptors (SUR1 or SUR2). Two isoforms of SUR2 exist, SUR2A and SUR2B, that differ in the last carboxyterminal exon. SUR1 has only 66%-67% identity with SUR2, but the last carboxyl terminal exon of SUR2B resembles that of SUR1. The stilbene DIDS (4-acetamido-4’-isothiocyanatostilbene-2,2’disulfonic acid) recognises ATP binding sites. In the absence of Mg2+, DIDS (>70 µM) inhibits pancreatic β-cell (SUR1) and cardiac (SUR2A) KATP channels by interacting with a site near the ATP binding site of the Kir6.2 subunit (Proks et al. 2001, Br J Pharmacol 132:973). In this study, we have investigated the effects of Mg2+ and of the SUR subunit on the inhibition by DIDS of recombinant KATP channels expressed in HEK cells. In inside-out patches and in the presence of 0.7 mM Mg2+, DIDS irreversibly inhibited the channels formed by Kir6.2 + SUR1/2A/2B with IC50 values ≈9/24/2.6 µM and Hill coefficients of 1.0/2.7/1.7, respectively. In the absence of Mg2+, DIDS (<10 µM) was ineffective. At 100 µM, DIDS inhibited SUR1/Kir6.2 completely and the SUR2-containing channels by ≈40 %. These experiments show that Mg2+, at physiological concentrations, increases the sensitivity of the KATP channel towards DIDS in a manner depending on the SUR subtype. The greatest increase in sensitivity is found for SUR2B (70x), followed by SUR2A (7.5x) and SUR1 (4x). LGB is supported by the Alexander von Humboldt-Stiftung. Department of Pharmacology, Medical Faculty, University of Tübingen, Wilhelmstr. 56, D-72074 Tübingen O 29-5 ADRENALINE AS POTENT INHIBITOR OF INSULIN SECRETION IN SUR-1 KO MICE M. Buchenau, A. Sieg, J. Su, L. Aguilar-Bryan**, J. Bryan*, S. Ullrich Human β cells, which are deficient in functional KATP channels, secrete high levels of insulin in the presence of hypoglycemia as a consequence

S 207 permanent depolarisation. In contrast, reduced and regulated secretion was described in sulfonylurea receptor knock out (SUR1 KO) mice islets. We analysed the role of adrenaline, a potent physiological inhibitor of insulin secretion, in these mice. In freshly isolated islets, adrenaline (0.1 and 1 µM) inhibited insulin secretion induced by glucose (16.7 mM), and forskolin (10 µM) plus IBMX (100 µM), by 86% and 95%, respectively. Glucose alone, which did not stimulate secretion from freshly isolated KO islets, augmented secretion 2.6-fold in cultured islets. In comparison, in WT islets secretion was stimulated 5.7-fold by glucose. When islets were treated with pertussis toxin (PTX) during culture, glucose-induced secretion was further increased in wt and KO islets. However, insulin content, which was 37 % lower in KO than in wt islets, decreased during culture at 11 mM glucose, in the absence and presence of PTX, by 53 % and 67 % in wt islets and by 50 % and 79 % in KO islets, respectively. PTX treatment blocked inhibition of secretion by adrenaline in WT islets. In KO islets, however, the block was only partial (57%). Standard whole cell patch-clamping of Sur1KO β-cells revealed continuous action potentials in both low and high glucose that were inhibited by adrenaline (1 µM) and this hyperpolarized the β-cells from -37 ± 1 mV to –57 ± 1.5 mV (n = 35). The α 2 adrenoceptor antagonists, α-yohimbine (10 µM), pertussis toxin (PTX, 100 ng/ml) blocked the adrenaline-induced hyperpolarisation as did 100 µM BaCl 2 , while other K + channel inhibitors, TEA (10 mM), apamin (200 nM), dendrotoxin (40 nM) and E-4031 (400 nM) did not. We conclude that the hyperpolarizing effect of adrenaline is not mediated by KATP channels and this and other inhibitory pathways may be involved in the reduced in vivo insulin response in these mice relative to neonates with familial hyperinsulinism. Institut für Neurophysiologie, Universität zu Köln, Robert-Koch-Str. 39, D-50931 Köln, Germany; *Dep. of Mol. and Cell. Biol. and **Dep. of Endocrinology, Baylor College of Medicine, Houston, Texas 77030, USA

release of Ca2+ from the sarcoplasmic reticulum (SR) which is triggered by Ca2+ influx through voltage operated L-type Ca2+ channels during the action potential. Although much is known about the subcellular properties of ventricular myocyte ec-coupling, the mechanisms that shape the spatio-temporal appearance of Ca2+ signals in atrial myocytes are less clear. We investigated the microscopic details of Ca2+ signalling in rat atrial myocytes using electrical field stimulation of fluo4-loaded cells and simultaneous monitoring of subcellular Ca2+ events by high speed (120 frames/s) confocal imaging. During steady-state stimulation at 20 or 37°C in 1 mM Ca 2+, the majority of cells (59%; n =120) displayed Ca2+ signals that were confined to 1-5 µm of the periphery of the myocytes. Only 10% of all cells tested exhibited homogeneous Ca 2+ transients, with the remainder exhibiting various degrees of inhomogeneities during ec-coupling. In cells exhibiting inhomogeneous Ca 2+ signals, increases in the Ca 2+ influx (by acutely elevating extracellular Ca 2+), loading of the SR Ca2+ content (by temporarily halting electrical pacing) and sensitization of ryanodine receptors (by sub millimolar caffeine concentrations) all resulted in a homogeneous Ca2+ signals and a substantial (up to 6 fold) increase in the inotropic response of the atrial myocytes. We furthermore investigated the role of mitochondrial Ca2+ uptake and SERCA pump activity in restricting Ca 2+ signals to the periphery of atrial myocytes. Inhibition of both mitochondria (by antimycin/oligomycin) and SERCAs (by cyclopiazonic acid) provoked an inward propagation of the initial peripheral Ca2+ response resulting in homogeneous Ca2+ signals. This inward propagation of the Ca 2+ signal was always accompanied by a substantial increase in the inotropic response and very often also by the occurrence of Ca2+ signal alternans. From these data, we conclude that atrial myocyte ec-coupling is characterised by the existence of a functional ‘firewall’ for Ca2+ transients, that under normal conditions restricts ec-coupling-induced Ca2+ signals to the periphery of the cell. All processes contributing to positive or negative feedback in Ca2+induced Ca2+ release determine the functional state of the firewall. The existence of such a firewall is another example how cells modulate their physiological output by sole alteration of the spatial profile of the underlying Ca2+ signals.

O 29-6 EFFECT OF AN ATP-REGULATED-K+-CHANNEL OPENER ON ISCHEMIA-INDUCED DAMAGE AND LONG-TERM POTENTIATION IN RAT HIPPOCAMPUS IN VITRO U.H. Schröder, S. Busse, L. Schild, J. Breder, K.G. Reymann The hippocampus is highly susceptible to hypoxic/ischemic injury. The sustained influx of Ca2+ and Na+ and the efflux of K+ during such events eventually lead to pronounced neuronal damage, especially in the CA1 region. Many ion channels have been reported to be involved in these ion movements but their role in ischemia-induced neurodegeneration is still not completely clear. Using acutely isolated hippocampal slices from adult and organotypic hippocampal slice cultures (OSC) from juvenile rats we investigated if ATP-regulated-K+channel opening affects neuronal recovery after hypoxia/hypoglycemia and long-term-potentiation. Recovery of the synaptically evoked population spike in the CA1 region in acutely isolated slices and propidium iodide staining after 24 h in OSC were taken as measures of neuronal viability. The novel ATP-regulated-K+-channel opener Y26763 does not only protect CA1 neurons when present before but also when present after an insult. Since Y-26763 does not alter mitochondrial function the drug might be specific for plasma membrane K+-channels and its protective effect might not involve an action on mitochondrial K+-channels. K+-channel opening only slightly affected long-term potentiation induced by a weak tetanus (four times two paired pulses) in slices from adult rats, indicating that it may not severely alter physiological responses. In conclusion we suggest that the activation of ATP-regulated-K + -channel is critical for neurodegeneration. *

Leibniz Institute for Neurobiology, POB 1860, Brenneckestraße 6, 39008 Magdeburg, Germany and FAN GmbH, Leipziger Str. 44, 39120 Magdeburg, Germany

O 30-1 MECHANISMS CONTRIBUTING TO THE ‘FIREWALL’ IN ATRIAL MYOCYTE EC-COUPLING L. Mackenzie, M.D. Bootman, M.J. Berridge, P. Lipp. A principal step in cardiac excitation contraction (ec-) coupling is the

Laboratory of Molecular Signalling, The Babraham Institue, Cambridge, CB24AT, UK

O 30-2 POSITIVE INOTROPIC EFFECT OF ADENOVIRUS-MEDIATED EXPRESSION OF THE HUMAN ATRIAL MYOSIN LIGHT CHAIN 1 IN RABBIT VENTRICULAR CARDIOMYOCYTES H. Kögler, G. Hasenfuss, I. Morano In the human heart two isoforms of the myosin light chain 1 (LC1) are expressed in a developmentally-regulated and chamber-specific manner, the atrial (ALC1) and ventricular (VLC1) isoform. In the fetal heart, ALC1 is also expressed in the ventricles, but in the ventricles of normal human hearts it is completely replaced by VLC1 during early postnatal development. During myocardial hypertrophy, however, ALC1 is reexpressed in the ventricles. This replacement may represent a molecular adaptation process to compensate an increased workload.We used adenovirus-mediated gene transfer to express the human ALC1 (hALC1) in rabbit ventricular cardiomyocytes. Cells transfected with an adenovirus encoding for β-galactosidase served as controls. Immunofluorescence microscopy 48h post transfection revealed abundant expression of hALC1, colocalized with myosin heavy chain, in virtually 100% of the cells, while control cells exhibited only a faint background signal. We measured fractional shortening of electrically stimulated myocytes using a video edge-detection system. At a [Ca2+]o of 2.0 mM, the amplitude of fractional shortening in hALC1 cells was enhanced by 39%, compared to controls (p<0.001), and the maximum speed of both shortening and relengthening was increased (p=0.003 and p=0.01, respectively). The shortening-[Ca2+]o relationship between 1.0 and 3.0 mM [Ca 2+]o was shifted upward in hALC1 cells, indicating enhanced Ca2+ responsiveness. In conclusion, the human atrial myosin light chain 1, when expressed in rabbit cardiomyocytes, exerts a strong positive inotropic effect, presumably due to direct modulation of actin-myosin interaction, without any obvious deleterious effects, especially on diastolic function. This favorable profile makes hALC1 a promising candidate for gene therapeutic approaches to treat the contractile dysfunction in human heart failure.

S 208 Georg-August-University Göttingen, Department of Cardiology and Pneumology, Robert-Koch-Str. 40, D-37075 Göttingen

whereas the low content of myofibrils in the SA node may be the result of diminished translation of α-MHC mRNA into protein and/or diminished organisation of myofilaments into myofibrils.

O 30-3

School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK

STRETCH-ACTIVATED CURRENTS IN HUMAN ATRIAL MYOCYTES A. Kamkin1, I. Kiseleva1, K.D. Wagner1, J. Bohm2, H. Theres2, J. Günther1, H. Scholz1 Stretch activated currents (I SAC), which contribute to extra-action potentials in isolated ventricular myocytes were demonstrated in healthy rat (Zeng, Bett & Sachs, 2000:[1]) and healthy rat and guinea pig hearts (Kamkin, Kiseleva & Isenberg, 2000:[2]). Stretch sensitivity of cells increases with age, hypertension and hypertrophy [2]. In this study, we analyzed ISAC in human atrial myocytes. From the late currents (IL), the stretch-activated difference current (ISAC) was calculated. The bath solution contained (mM): 150 NaCl, 5.4 CsCl, 1.8 CaCl2, 1.2 MgCl2, 20 glucose, 5 HEPES (pH 7.4) and the pipette solution contained (mM) 140 CsCl, 5 Na 2 ATP, 5.5 MgCl 2 , 10 HEPES (pH 7.2). The holding potential was -45 mV. 5 µM EGTA adjusted pCa to 7.0. BAPTA electrode solution was complemented with 5 mM BAPTA. We applied axial stretch. An increase of length by 4 and 6 µm produced growing respective sarcomere lengthening from 1.74±0.02 to 1.91± 0.03, and to 2.00±0.02 µm. The amplitude of IL reversibly increased with the amplitude of stretch. At –45 mV, growing stretches by 2, 3, 4, and 6 µm induced I SAC of -54±18, -148±20, -272±22, and -613±22 pA, respectively. ISAC were suppressed by 5 µM Gd3+. ISAC were insensitive to substitution of Cl- ions by aspartate ions. Stretch suppressed ICa-L. This reduction of I Ca-L was absent when cell dialysis with 5mM BAPTA preceded the stretch, whilst I SAC remained unmodified. At -45 mV, stretch by 3 µm produced ISAC of –148±20 pA without BAPTA and -140±10 pA at 5 mM BAPTA. Linear voltage-dependence, E rev, Gd3+ sensitivity, and Cl- insensitivity suggest that ISAC flows through nonselective cation SACs. Effects of stretch on I SAC in Ca2+-free bathing solution were similar to those at 1.8 mM Ca2+. Therefore, preferentially Na+ ions are suggested to flow through these SACs. We conclude that axial stretch of human atrial myocytes induced significant I SAC comparable to that we described in human ventricular myocytes [2]. This current could be of pathophysiological relevance because of its large intensity compared to currents obtained from healthy myocardium of other species. 1

Institute of Physiology (Charité), Tucholskystr.2, 10117 Berlin, and Clinic of Cardiac Surgery (Charité), Schumannstr. 20-21, 10098 Berlin, Germany.

2

O 30-5 INCREASED MITOCHONDRIAL IRON IN HYPERTROPHIED VENTRICULAR MYOCYTES OF ADULT SPONTANEOUSLY HYPERTENSIVE RATS F. Rudolf, M. Schultz, M. F. Gallitelli The mechanisms through which mitochondria succeed in the longterm balance of energy demand and energy supply in the hypertrophied non-failing heart are not known. According to the idea of a Camediated regulation of ATP production, an increase in mitochondrial Ca2+ would activate enzymes of oxidative phosphorylation and eventually enhance ATP formation. Up to now, however, such a long-term increase of mitochondrial calcium could not be demonstrated in the hypertrophied heart. The mitochondrial metabolism is also regulated by the import of iron into the mitochondria ([Fe]m): reducing [Fe] m impairs, while increasing [Fe]m activates the metabolic activity of the cell. With voltage-clamp techniques and electron probe microanalysis we have analyzed the Ca- and Fe-load of cytosol and mitochondria in quickly-frozen ventricular myocytes of spontaneously hypertensive rats (SHR) and their age-matched controls (WKY). The differences in mitochondrial calcium-load ([Ca]m) between hypertrophied and control myocytes were not significant. In central mitochondria [Ca]m was 0.62±0.10 mmol/kg dw (dry weight) in SHR (or 340±0.05µM) and 0.63± 0.13 mmol/kg dw (or 350±0.07 µM) in WKY. Measurements of [Fe] and [S] revealed for the first time that the [Fe]m and [S]m in SHR myocytes were significantly higher than in controls. In subsarcolemmal mitochondria [Fe]m was 6.5±0.1 mmol/kg dw in SHR (n=95) and 6.0±0.1 in WKY (n=100, p<0.05). In central mitochondria [Fe] was 6.74±0.1 (n=161) in SHR and 6.32±0.1 mmol/kg dw (n=162) in WKY (p<0.01). Quantitative images of [Fe] in mitochondria and cytosol (at 16 nm spaial resolution) confirmed the data of point analysis. Although we cannot exclude that calcium is involved in the long-lasting adjustment of the balance between energy demand and energy supply in the hypertrophied heart, the higher [Fe] in mitochondria of SHR let us suggest that, in the hypertrophied myocytes, the regulation of electron transport and the synthesis of enzymes involved in the citric cycle such as mitochondrial aconitase may be mediated by Fe. Supported by DFG and BMBF to M.F.G. Julius-Bernstein-Institute of Physiology, University of Halle, Magdeburger Str. 6, D-06097 Halle/Salle

O 30-4 DIFFERENT DISTRIBUTION AND CELLULAR LOCALISATION OF CONNEXIN43 AND α -MYOSIN HEAVY CHAIN mRNA IN THE SINOATRIAL NODE H. Dobrzynski, H. Musa, T.W. Claydon, S.R. Coppen, E. Dupont, N.J. Severs, M. Peckham, R. Billeter, M.R. Boyett The sinoatrial (SA) node is a specialised tissue in the heart: it lacks the major gap junction connexin of the heart (Cx43) and has a low content of well organised myofibrils. In this study we first confirmed the lack of Cx43 and the low content of well organised myofibrils in the SA node by immunohistochemistry using suitable antibodies. Next, we localised mRNAs for Cx43 and á-myosin heavy chain (α-MHC; major constituent of the thick myofilament) in tissue sections through SA node and the surrounding atrial muscle of the crista terminalis using in situ hybridisation; 1400 bp (Cx43) and 300 bp (α-MHC) antisense and sense riboprobes were used. Distinct hybridisation signals were seen using the antisense probes, but no signal was seen in the negative controls (using the sense probes). Similar results were obtained from sections through three rat (Cx43) and three rabbit (α-MHC) hearts. Cx43 mRNA signal was absent from the SA node (the same region that lacks Cx43 protein signal; confirmed by immunohistochemistry) but was present in the atrial muscle of the crista terminalis. α-MHC mRNA signal was present throughout the SA node and the atrial muscle of the crista terminalis. The cellular localisation of Cx43 and α-MHC mRNAs differed: if present, Cx43 mRNA signal was located in a ring surrounding the nucleus, corresponding to the rough endoplasmic reticulum, whereas α-MHC mRNA signal was located throughout the cytoplasm presumably on free ribosomes. In conclusion, these results suggest that the lack of Cx43 protein in the SA node is the result of a lack of transcription,

O 30-6 TRANSCRIPTION FACTOR AP-1 AS MEDIATOR OF αADRENERGIC GROWTH STIMULATION IN VENTRICULAR CARDIOMYOCYTES G. Taimor, A. Schlieper, H.M. Piper α-Adrenergic stimulation of cardiomyocytes results in myocardial hypertrophy. Concurrently with this hypertrophic growth response induction of immediate early genes is found. It remains to be elucidated if induction of these genes only appears in parallel or is directly involved in the development of cardiac growth. Some of the immediated early genes, e.g. fos and jun, code for subunits of the transcription factor AP-1. The aim of this study was to analyse the involvement of this transcription factor in α-adrenergic growth stimulation of isolated adult cardiomyocytes of rat. Activity of AP-1 in nuclear extracts from cardiomyocytes was determined in retardation assays testing for AP-1 binding activity to its consensus binding site TRE. The αadrenoceptoragonist phenylephrine (PE, 10 µM) increased AP-1 binding activity in cardiomyocytes within 30 minutes. Preincubation of cardiomyocytes with TRE-decoy-oligonucleotides inhibited this response. To determine the rate of protein synthesis as parameter for hypertrophic growth incorporation of 14C-phenylalanin during 24 h was measured. Incubation of cardiomyocytes with PE increased the rate of protein synthesis to 161.4 ± 23.9 % (n=4, p<0.05 vs. control). This induction of hypertrophic growth was abolished when AP-1 binding activity was inhibited by preincubation of cardiomyocytes with TRE-decoy-oligonucleotides (122.3 ± 13.6 %, n=4, n.s. vs. control). In conclusion, intracellular scavenging of AP-1 binding activity inhib-

S 209 its α-adrenergic induced growth stimulation in isolated adult cardiomyocytes. These results indicate importance of this transcription factor during the development of cardiac hypertrophy. Physiologisches Institut, Aulweg 129, 35392 Gießen, Germany O 31-1 CORONARY CONSTRICTION DURING STIMULATION OF THE PERIAQUEDUCTAL GREY MATTER - A MECHANISM UNDERLYING NON-CARDIAC CHEST PAIN? M.I. Blomfield, S. Egginton, T.A. Lovick Some 50% of patients with chest pain have no obvious cardiac abnormality (Mayou, R 1998, J Psychosom Res 44, 53-70). Psychiatric evaluation of these patients reveals a high incidence of anxiety states such as panic disorder (Bass, C & Wade, C 1984 Psychol Med 14, 5161). Panic attacks are characterised by periods of intense irrational fear and autonomic arousal, often with chest pain. In anaesthetised rats the autonomic components of panic-like responses evoked by stimulation of panic circuitry in the dorsal periaqueductal grey matter (dPAG) include an increase in the duration of the ST component of the ECG (Lovick, TA and Drew, S 1998 J Physiol 511 P, 112P) that may reflect myocardial ischaemia.47 urethane-anaesthetised (1.5g 100g-1 i.p) male 300-350g Wistar rats were instrumented to record blood pressure, heart rate, ECG at V4 and tracheal air flow. Rectal temperature was maintained at 36.5-37.5°C. Electrical stimulation in the dPAG (100µA, 1ms, 80Hz for 10s) produced a pressor response (76.5±4.0 mm Hg), tachycardia (36.5±5.2 beats min -1), hyperpnoea and a 18.54±1.8% increase in the interval between the peaks of the S and T waves of the ECG. The PAG-evoked increase in ST duration was not changed significantly (25.1 ±6.6 % v. 22.4±6.6 %, n = 4) during paralysis (pancuronium bromide 1mg Kg-1 i.v.) and artificial ventilation. In 31 rats fluorescent microspheres were used to measure blood flow in the heart, kidneys and right m. extensor digitorum longus (Deveci, D and Egginton, S Exp Physiol 84 1999, 615-30). During PAG stimulation renal vascular resistance increased by 966.7±199.8% and the resistance of the hindlimb either increased 118.8±46.6%, n=6 or decreased -554.2±6.8%, n=25). In most rats (n=21) coronary resistance did not change significantly (+18.8±45.3%, p= 0.27, paired t-test) but in 10 rats an increase in coronary vascular resistance occurred which correlated significantly with the increase in ST interval (R 2=0.86).These data identify a significant population of rats in which vasoconstriction within the coronary circulation occurs in response to activation of the dPAG. This phenomenon may underlie the episodes of chest pain experienced by patients with panic disorder but angiographically normal coronary arteries. Supported by the Wellcome Trust Dept. of Physiology, University of Birmingham, Birmingham B15 2TT, U.K O 31-2 EXPERIMENTAL EVIDENCE FOR ALTERED BACKGROUND ACTIVITY IN DORSAL HORN NEURONES AS A CAUSE OF AT-LEVEL PAIN IN PARAPLEGIC PATIENTS C. Scheifer, U. Hoheisel*, P. Trudrung, T. Unger**, S. Mense Patients with complete spinal cord injury (SCI) often suffer from chronic pain which is perceived around the segment of the lesion. However, little is known about the mechanisms responsible for chronic SCI pain. The present study tests the hypothesis that chronic SCI pain is due to pathological neuronal activity in the spinal segment just rostral to the SCI. In rats with complete chronic SCI, the impulse activity of single dorsal horn neurones was recorded close to the lesion. The experiments were performed on anaesthetised adult SD rats. In 8 rats the spinal cord was transected under deep anaesthesia 2 to 6 weeks prior to the final experiments. 6 rats served as a sham-operated control group (laminectomy only) and 6 rats underwent no surgical interventions (2nd control). Compared with both controls, in SCI animals the spontaneous neuronal activity had a significantly higher level. Some neurones discharged at 390 to 1580 impulses per minute. About 10 % of all neurones recorded in SCI animals exhibited an abnormal type of background activity which was not present in both controls. The neurones showed a high frequency discharge without long silent periods or discharges of a bursting nature with long periods of silence alternating with phases of increased activity. The increase in impulse activity and change in discharge pattern was associated with abnormal behaviour

(autotomy and excessive grooming) indicative of dysaesthesia or spontaneous pain. The behaviour was directed to dermatomes close to the site of injury. The results support the hypothesis that chronic SCI pain is due to an abnormal neuronal activity at the spinal level. If the data are also valid for patients, the results mean that a small percentage of dorsal horn neurones - namely the highly active ones - are responsible for the spontaneous pain of SCI. Supported by the Medical Faculty of the University Heidelberg. Institut für Anatomie und Zellbiologie, Universität Heidelberg, Im Neuenheimer Feld 307, D-69120 Heidelberg, *Institut fur Pharmakologie, Universität Kiel, Hospitalstraße 4, D-24105 Kiel, and **Institut fur Pharmakologie und Toxikologie, Charite, Humboldt Universität, Dorotheenstraße 94, 10117 Berlin O 31-3 ESTABLISHMENT OF C-FIBRE EVOKED FIELD POTENTIALS IN LAMINA I/II OF THE SPINAL DORSAL HORN IN MICE AND THE INFLUENCE OF B1-RECEPTOR AGONISTS IN INFLAMMATORY PAIN S.Steinfatt, C.Schott, H.Seller Bradykinin is known to be a mediator involved in nociception. Two different receptors had been identified: the B2-receptor, permanently expressed, and the B1-receptor whose expression appears to be correlated with the induction of pain. We investigated the involvement of B1-receptors in the spinal cord of mice during inflammatory pain. First, we established a model of measuring C-fibre-evoked field potentials in the spinal cord of mice analogous to that in the rat. The lumbosacral segment of the spinal cord of adult mice (p42-70) with its dorsal roots was removed under deep ether anasthesia. On a vibratome, 2mm-thick transverse slices with at least one dorsal root attached were cut. The slice was placed in a recording chamber superfused with Krebs solution equilibrated with 95% O2 and 5% CO2 at 27°C. Postsynaptic extracellularly evoked potentials were recorded in the substantia gelatinosa (lamina I/II) of the spinal cord in response to electrical stimulation (2 times of threshold, pulsewidth 0.5ms, 30s intervals). Capsaicin reduced C-fibre-evoked potentials (1 µM; n=5; 52±1% of control). DAMGO as a µ-opiate agonist also induced a reduction of evoked field potentials (1µM; n=6; 38±1% of control). The latter can be inhibited by the µ-opiate antagonist Naloxon (10µM; n=6; 85±2% of control). These results are comparable to those obtained in the rat. Three hours after generating acute inflammatory pain by injection of 10µL carrageenan (2%; 20mg/ml) in the left hindpaw, the spinal cord was removed. The B1-receptor agonist Lys-(Des-Arg)Bradykinin produced a depression of C-fibre-evoked potentials on the ipsilateral side (10µM; n=6; 54.1±1% of control), which is not the case in control mice (10µM; n=5; 104±5% of control). This indicates that B1 receptors contribute to the transmission of inflammatory pain in the spinal cord of mice. Institute of Physiology and Pathophysiology, University of Heidelberg, D-69120 Heidelberg, Germany O 31-4 LTD-LIKE MODULATION OF HUMAN TRIGEMINAL NOCICEPTION A. Schorr, J. Ellrich Low-frequency stimulation (LFS) of Aδ-fibers causes sustained depression of synaptic transmission. This so-called long-term depression (LTD) has been demonstrated in many different areas of the central nervous system in animal experiments. Here we have identified a robust depression of the nociceptive blink reflex by electric LFS of trigeminal afferents in man. In 8 healthy volunteers (5 female, 3 male; 25 to 35 yrs) neuronal excitability of sensory trigeminal neurons of the spinal trigeminal nucleus was detected by applying the nociceptive blink reflex. The electromyographic activity (EMG) of both orbicularis oculi muscles was recorded by surface electrodes (1001000 Hz). Electric rectangular pulses (500 µs duration) were applied by a custom-made concentric electrode (central cathode 0.5 mm diameter, ring anode 30 mm diameter) to the left and right forehead in four volunteers each. Each volunteer participated in two experimental sessions on different days. In each session 11 blocks of 10 painful electric test stimuli each (0.2 Hz) were applied every eight minutes. In the LFS session 1200 painful electric conditioning stimuli were ap-

S 210 plied with a frequency of 1 Hz via the same electrode between the 3rd and 4 th test stimuli blocks. In the control session no conditioning stimuli were applied. Under control conditions integrals and onset latencies of ipsi- and contralateral reflex responses remained statistically unchanged. After LFS integrals of ipsi- and contralateral reflexes significantly decreased by about 53% on an average. Onset latencies significantly increased by 5% and 8% on ipsi- and contralateral side. This depression was stable at least within one hour after the end of LFS. Prestimulus baseline EMG activity under control and LFS conditions was not statistically different. This is the first report on LTDlike depression of orofacial nociceptive processing by low-frequency stimulation of trigeminal afferents in man. The results would suggest the involvement of LTD in long-lasting therapeutic effects of counterstimulation in chronic pain states.

0.01) suppressed 2 hours after morphine-sulfate administration as indicated by a decrease of the pain ratings by about 45 % and a decrease of the pain-related cerebral activity by about 50%. Both effects were highly correlated. In spite of the marked morphine-induced analgesia, there was no effect on sedation as shown by all sedation indicators. Neither the auditory evoked potentials, the spontaneous EEG, the reaction time nor the subjective mood of the subjects were significantly altered. These results distinguish the analgesic properties of morphine as very powerful, even compared to other strong analgesics, while inducing no relevant sedation under pain. In contrast to other studies which found sedative effects in pain-free healthy subjects after morphine-sulfate administration we conclude that there is a relevant arousal function of the phasic pain which surmounts the morphine induced sedation.

Department of Experimental and Clinical Pharmacology and Toxicology, University of Erlangen-Nuremberg, Emil-Fischer-Center, Fahrstraße 17, D-91054 Erlangen

Institute of Physiology, University Hospital Eppendorf, Martinistr. 52, 20246 Hamburg O 32-1

O 31-5 A DISSOCIATIVE CHANGE IN THE EFFICACY OF SUPRASPINAL VERSUS SPINAL MORPHINE IN THE NEUROPATHIC RAT A. Pertovaara*, H. Wei The efficacy of spinally versus supraspinally administered morphine was studied in rats with a spinal nerve ligation-induced neuropathy. Behavioral assessment indicated that the effect of intrathecally (i.t.) administered morphine on pain-related responses was attenuated when compared with controls. In contrast, the pain attenuating effect of morphine in the periaqueductal gray was enhanced in neuropathic animals when compared with controls. The effect of systemically administered morphine on pain-related behavior of neuropathic rats was in the same range as in controls or decreased, depending on the test. Microinjection of lidocaine or MK-801, an NMDA receptor antagonist, into the rostroventromedial medulla partly restored the pain attenuating effect of i.t. administered morphine in neuropathic animals. Electrophysiological recordings of nociceptive wide-dynamic range (WDR) neurons in the deep spinal dorsal horn corresponded to a large extent with behavioural results. The inhibitory effect of spinally administered morphine on WDR neuron responses was attenuated in neuropathic animals, whereas the inhibitory effect induced by morphine in the periaqueductal gray was enhanced following neuropathy. Paradoxically, systemic administration of morphine in neuropathic rats did not suppress the responses of WDR neurons at a dose that was behaviourally effective in neuropathic rats and that suppressed spinal neuronal responses in control animals. Lidocaine block of the rostroventromedial medulla did not modulate the inhibitory effect of systemically administered morphine on WDR neurons. The results indicate that the neuropathy-induced modulation of morphine action varied depending on the site of administration: at the spinal cord level morphine effect was reduced and supraspinally enhanced. The reduction of morphine effect at the spinal level may have been partly due to descending pathways.

MOLECULAR IDENTIFICATION OF A VOLTAGE-GATED SODIUM CHANNEL INVOLVED IN BDNF-EVOKED NEURONAL EXCITATION R. Blum, K.W. Kafitz, A. Konnerth Brain-derived neurotrophic factor (BDNF) belongs to a family of neurotrophins that mediate differentiation and survival of neurons. More recent evidence indicates that BDNF also acts as a modulator of activity-dependent neuronal plasticity. In a recent study, BDNF was shown to be a potent neuro-excitant that mediates a very rapid neuronal depolarisation via the tyrosine kinase receptor Trk-B (Kafitz et al., Nature 401 (1999): 918-921). The molecular mechanism underlying neurotrophin-evoked rapid neuronal excitation remained unknown. Here we demonstrate the molecular reconstitution of BDNF-mediated neuronal excitation. In HEK-293 cells, stimulation of recombinant Trk-B receptors by BDNF activated the co-expressed sodium channel Na v1.9, a TTX-insensitive member of the family of voltage gated sodium channels. As found for BDNF-evoked inward currents in hippocampal neurons and SH-SY5Y neuroblastoma cells, the reconstituted current was insensitive to TTX, however highly sensitive to the sodium channel blocker saxitoxin. Control experiments showed that the homologous channel Nav1.7 was not able to replace the function of Na v1.9 in BDNF-evoked excitation. Vector-mediated antisensemRNA-expression targeted against either Trk-B or Nav1.9 transcripts inhibited the intrinsic BDNF-evoked excitation of SH-SY5Y cells. Our results demonstrate for the first time that neurotrophin ligands, bound to a receptor tyrosine kinase, induce a fast cellular depolarization through a TTX-insensitive voltage-gated sodium channel. This indicates a novel function of these channels in ligand-mediated neuronal signalling. Supported by the DFG. Physiologisches Institut, Ludwig-Maximilians-Universität, 80802 München O 32-2

Dept.Physiol., Inst. Biomed., Univ.Turku, FIN-20520 Turku & Dept.Physiol., Inst. Biomed., Univ. Helsinki, POB 63, FIN-00014 Helsinki, Finland O 31-6 SEDATION VERSUS ANTINOCICEPTION: CORTICAL SITES OF MORPHINE ACTION E. Scharein, M. Quante, R. Zimmermann, B. Bromm The antinociceptive and sedative effects of a low dose of morphinesulfate (10 mg i.v.) were determined using a standardized experimental pain model (intra-cutaneously administered electrical pulses) in a randomized placebo-controlled repeated measurement design. Seven healthy subjects participated in this study. Five blocks (1 block = 80 stimuli) of painful stimuli were applied, one before and 4 after morphine-sulfate or saline injection. The blocks covered a period of 3 hours. Analgesia was assessed by changes in the subjective pain ratings and by the pain-induced changes in the electroencephalogram (EEG), i.e. pain-related cerebral potentials. Sedation was determined by measuring changes in spontaneous EEG and auditory evoked potentials (AEP), reaction times and mood, as rated on five bipolar mood scales. In all subjects the pain related measurements were significantly (p <

TWO TYPES OF BDNF-INDUCED CALCIUM SIGNALS IN NEURONS AND GLIAL CELLS IN ACUTE BRAIN SLICES C.R. Rose, B. Pichler, A. Konnerth Neurotrophins are secreted proteins that regulate the differentiation and survival of different neuronal populations. They also play an essential role in the modulation of activity-dependent neural plasticity. Furthermore, we could recently show that neurotrophins rapidly excite central neurons by activation of a TrkB-receptor-gated Na+ conductance. The resulting depolarisation leads to the opening of voltage-gated Ca2+ channels and to intracellular neuronal calcium transients. BDNFinduced calcium transients are also observed in cultured astrocytes (see Pichler et al., this issue). Here, we studied the mechanisms of neurotrophin-induced calcium transients in neurons and glial cells in acutely isolated brain slices. Hippocampal or cerebellar slices were prepared from juvenile Wistar-rats (postnatal days 9-15) and loaded with the calcium indicator Fura-PE3-AM. Calcium imaging was performed using two-photon laser-scanning microscopy. Areas and plane of focus were chosen such that both glial cells and neurons could be imaged at the same time. The neurotrophin BDNF (50 ng/ml, 10-50 ms) was applied by a picospritzer coupled to micropipettes. Application of BDNF evoked prominent calcium transients in glial cells and neurons of the hippocampus and cerebellum. The time course of neighbouring

S 211 neuronal and glial calcium transients, however, differed significantly. The response observed in neurons (CA1 pyramidal cells, granule cells of the dentate gyrus and cerebellar Purkinje neurons) peaked within 200 ms after the BDNF application and decayed monoexponentially. In contrast, BDNF-evoked calcium transients in neighbouring glial cells (hippocampal astrocytes and cerebellar Bergmann-glial cells, respectively), were only observed after a delay of 1-2 s and displayed complex decay kinetics. Glial responses were unaltered by TTX and by glutamate-receptor antagonists (CNQX, APV, MCPG), demonstrating the absence of any contamination by secondary glutamatergic signals. Our results demonstrate that the neurotrophin BDNF directly elicits calcium signalling in different types of glial cells. They also suggest that neurotrophins activate glial cells by a mechanism that differs from that observed in neurons.

the presence of the CB1 antagonist SR 141716A, the Gi/0-protein inhibitor pertussis toxine, the adenylyl cyclase (AC) activator forskolin and Sp-cAMPS, a membrane-permeable activator of the protein kinase A (PKA). SR 141716A, pertussis toxin and SPcAMP, but not forskolin, reversed the effect of WIN 55,212-2. The results suggest that CB1 activation clearly inhibits synaptic transmission in the LA through involvement of Gi/0-proteins and PKA, but not necessarily through inhibition of the AC.

Physiological Institute, Ludwig-Maximilians-Universität, D-80802 Munich,

EVIDENCE FOR P2X 7 PURINOCEPTORS IN OPTIC NERVE GLIA IN SITU G. James, A.M. Butt We have previously shown that white matter glia in situ respond to application of ATP with an increase in intracellular calcium concentration ([Ca2+]i), and that this response is mediated by a variety of P2Y and P2X receptor subtypes (James & Butt, 2001). We suggested that the low-affinity component of the ATP response could have been mediated by the pore-forming subtype P2X7, which is known to have important roles in pathophysiology. In this study we investigate directly whether white matter glia in situ possess the P2X7 subtype in intact isolated optic nerves of post-natal day 10-30 rats. Rats were humanely killed by CO 2 narcosis according to UK Home Office regulations and optic nerves were dissected free. Intact isolated optic nerves were then either calcium imaged as described previously (James & Butt, 2001) or they were put through a standard Western blotting procedure. Application of ATP caused a concentration-dependent increase in average [Ca2+]i for an area of nerve, with an initial response at 10 nM ATP and a maximal response at 10 mM ATP ([Ca2+]i increase = 123 ± 10 µM), with an EC 50 of 25 µM ATP. The response (1 mM ATP) was potentiated by removal of Na+ and Mg2+ from the extracellular milieu by 160 ± 20% and 46 ± 11% respectively, consistent with the effects of these ions on P2X 7 receptors. The P2X 7 receptor agonist BzATP caused a concentration dependent increase in [Ca2+]i, maximal response at 1 mM BzATP ([Ca2+]i increase = 43 ± 2 µM; EC50 = 32 µM ATP), and this could be blocked (100 %) at 10 µM by the P2X7 specific antagonist, periodate-oxidised ATP (oATP, 100 µM). Western blotting confirmed expression of P2X 7 receptor protein in the optic nerve. Incubation of optic nerves with YO-PRO-1 (40 µM) showed that glial cells possessed pore-forming purinoceptor P2X subtypes. In summary, the results demonstrate expression of P2X 7 receptors have a role in optic nerve glia in situ, and we suggest that these and other purinoceptors may have important roles in the pathophysiology of white matter. G. James & A.M. Butt (2001), Cell Calcium 30: 251-259. GJ is supported by Guy’s and St. Thomas’s Charitable Foundation.

O 32-3 SUBSTANCE P DEPRESSES EVOKED EXCITATORY POSTSYNAPTIC CURRENTS IN THE RAT NUCLEUS ACCUMBENS IN VITRO S.B. Kombian, K. V.V. Ananthalakshmi, P. Narayanan, W.C. Matowe The nucleus accumbens (NAc) is a forebrain structure that forms part of the neural circuit which mediates complex emotions including drug seeking behaviour. It is believed to serve as an interface between the limbic and motor systems where emotional events (limbic origin) are converted into action (motor output). Biochemical and behavioural studies implicate dopamine, glutamate and GABA in these behaviours. Similar studies indicate some that peptides present in the NAc mimic the actions of dopamine while others antagonize them. The cellular mechanism(s) by which one such peptide, Substance P, may alter accumbens function is not known. Using parasaggital forebrain slices containing the NAc, and whole cell and nystatin-perforated patch recording techniques, we tested the hypothesis that Substance P would depress excitatory synaptic transmission in the NAc in a manner similar to dopamine. Bath application of Substance P for 3-5 minutes caused a reversible depression of evoked, pharmacologically isolated (50 µM picrotoxin) excitatory postsynaptic currents (EPSC). The synaptic depressant effect was concentration dependent (0.1-5 µM) with a maximal depression of 46.7 ± 1.7% (n=3) at 1.0 µM. Beyond this concentration, the effect tended to decrease. The synaptic depressant effect at 1.0 µM was not accompanied by changes in the input resistance (determined around the resting membrane potential) of the recorded neurons (237 ± 114 MΣ vs 232 ± 95 MΣ, p>0.05, n=3). Furthermore, the steady-state current-voltage (I-V) curves in control and in the presence of 1.0 µM Substance P were superimposable. However, the effect of Substance P was routinely accompanied by increases in the paired pulse ratio ranging from 6-17% of control. Taken together, these findings suggest that Substance P depresses evoked EPSC in the NAc by a presynaptic mechanism, possibly through an action to decrease glutamate release. Supported by KFAS-980709.

Max-Planck-Institute of Psychiatry, Kraepelinstrasse 2, 80804 München O 32-5

Centre for Neuroscience Research, King’s College London, Guy’s Hospital, London, SE1 1UL

Faculty of Pharmacy, Kuwait University, Safat 13110, Kuwait O 32-6 O 32-4 THE EFFECT OF THE CANNABINOID RECEPTOR-1 (CB1) AGONIST WIN 55,212-2 ON SYNAPTIC TRANSMISSION IN THE MOUSE LATERAL AMYGDALA S. Azad, W. Zieglgänsberger, G. Rammes Cannabinoids exhibit a variety of central effects including influences of emotional states. The lateral amygdala (LA) which is known to be involved in emotional processes, containes high levels of CB1. We therefore investigated the effects of CB1 activation on synaptic transmission in this area using whole-cell patch-clamp recordings and extracellularly recorded population spikes (PS). In an in vitro slice preparation of 6-8 weeks old mice synaptic responses were evoked in the LA by stimulating the external capsule. The CB1 agonist WIN 55,212-2 (2.5-5µM) reduced the PS amplitude, excitatory postsynaptic currents (EPSC), pure AMPA receptor-mediated EPSCs, GABAergic PSCs and EPSCs of interneurons to 51.6±5.8%, 71.6±10.7%, 52.4±4.3%, 50.5±10.2% and 49.2±12.6%, respectively. WIN 55,212-2 also increased paired-pulse fascilitation and reduced miniature EPSC frequency, indicating a presynaptic mechanism. To further elucidate the CB1mediated mechanisms, we investigated the effects of WIN 55,212-2 in

CARBACHOL-INDUCED NETWORK OSCILLATIONS IN THE INTACT CEREBRAL CORTEX OF THE NEWBORN RAT Heiko J. Luhmann , Werner Kilb The cholinergic system has a modulatory action in the adult cortex and induces oscillatory network activity and facilitates synaptic plasticity. In conventional brain slices and in a novel in vitro preparation of the intact cerebral cortex we studied carbachol (CCh)-induced network oscillations during the early postnatal period. Multi-site extracellular recordings revealed CCh-induced transient beta oscillations (duration 4.6±0.2 s, amplitude 123±7.4µV, frequency of 17.7 ± 0.5 Hz), which propagated uniformly at 0.5-1.5 mm/s over the whole cortex and which were reversibly blocked by TTX and atropine. This activity was not blocked by bicuculline methiodide or gabazine, but reversibly abolished by kynurenic acid, indicating that activation of glutamate receptors, but not GABA-A receptors was required. CNQX caused a significant decrease in the power of the Fourier frequency spectrum of the CCh-induced oscillations and CPP or MK-801 completely blocked the activity, indicating an essential role of NMDA receptors and a contribution of AMP AIkainate receptors to the CChinduced oscillations. Oscillations were synchronized between sites sepa-

S 212 rated horizontally by ≈1 mm and for delays of 2-8 ms. Whole-cell recordings from visually identified and biocytin-labelled pyramidal neurons in neonatal rat cortical slices demonstrated that CCh causes an increase in spontaneous synaptic activity and a CPP-sensitive membrane depolarization. Extracellular biocytin tracing in the intact cerebral cortex revealed long-range patchy connections of up to 2.7 mm in lateral extent. This axonal network in neonatal cortex may form the substrate for the generation, propagation and synchronization of CCh-induced oscillations. Our data indicate, that activation of muscarinic receptors in neonatal cerebral cortex elicits action-potential dependent oscillations, which are mediated by a long-range network of intracortical connections coupled via NMDA receptors. Institute of Neurophysiology, University of Düsseldorf, POB 101007, 0-40001 Düsseldorf, Germany. [email protected] O 33-1 THE WILMS’ TUMOR GENE WT1 MEDIATES RETINAL GANGLION CELL SURVIVAL THROUGH ACTIVATION OF THE POU-DOMAIN FACTOR POU4F2 K.D. Wagner1, N. Wagner1, V. Vidal2, G. Schley1, D. Wilhelm3, C. Englert3, A. Schedl2, H. Scholz1 The Wilms’ tumor gene Wt1 encodes a zinc finger transcription factor which is required for the formation of the kidney and mesothelial tissues. Mutational inactivation of Wt1 is responsible for a subset of pediatric renal carcinoma (Wilms’ tumor). Wilms’ tumor patients are at increased risk for developing ocular disorders such as aniridia and optic nerve hypoplasia. In view of these clinical observations, we examined whether Wt1 is necessary for the normal development of the vertebrate retina. Wt1 transcripts were initially detected by in situ mRNA hybridization in the neural retina of 12-day old wild-type (Wt1 +/+) mouse embryos and, at later developmental stages, in the presumptive retinal ganglion cell layer. Wt1 null mutant mice (Wt1-/-) at E12 had markedly thinner retinas than their age-matched wild-type counterparts due to impaired proliferation of retinal progenitor cells. At E18, the thinner ganglion cell layers of the Wt1 -/- retinas corresponded with a 4-fold increased number of apoptotic cells and a failure of the optic nerve fibers to grow normally. Strikingly, the Wt1/mutants had optic nerves that never reached more than some 100 µm beyond optic disc level. Expression of the POU-domain transcription factor Pou4f2, which is essential for retinal ganglion cell survival, was lost in the mutant retinas. Stable expression of Wt1 increased Pou4f2 mRNA more than 8-fold (by real-time RT-PCR) in cultured cells. Furthermore, co-transfection of a Wt1 plasmid activated a 3.8 kilobase construct which contained the regulatory sequences of the human POU4F2 gene, approximately 4-fold. These retinal malformations, including the lack of Pou4f2, were rescued by transgenic expression of a yeast artificial chromosome (YAC) carrying the human WT1 gene in Wt1-/- embryos. In conclusion, the Wilms’ tumor transcription factor, acting through transcriptional activation of Pou4f2, is a key regulator of retinal development with impact on progenitor proliferation and retinal ganglion cell survival. These findings demonstrate for the first time that Wt1 has a critical role in neuronal development. 1 Johannes-Müller-Institut für Physiologie, Humboldt-Universität, Charité, Berlin; 2University of Newcastle, Human Molecular Genetics Unit, 3 Forschungszentrum Karlsruhe, Institut für Toxikologie und Genetik

with functional pili, e.g. PAO-I and PAK, induced significant apoptosis as shown by the infection-induced breakdown of phosphatidylserine asymmetry as well as mitochondrial and morphological alterations typical for apoptosis. After 3h of infection PAO-I and PAK induced increased FITC-Annexin staining when compared to uninfected control cells (35.1±7.9% and 39.1±11% versus 6±1.4%). In contrast, the pilus-deficient P. aeruginosa mutants PAK ∆pilA and PAK ∆pilA∆all failed to induce apoptosis in Chang cells. After 3h of infection the FITC-Annexin positive cells amounted to 7.4±4.7% and 6.6±3.5%. The pilus-deficient mutants also failed to induce significant mitochondrial alterations such as mitochondrial depolarisation, synthesis of reactive oxygen intermediates and release of cytochrome c, as observed for PAO-I and PAK. However, the pilus-deficient mutants still induced a small upregulation of CD95 within 90 min (12 and 13% compared to 59 and 63% for PAO-I and PAK, respectively). We conclude that functional pili are essential for P. aeruginosa-induced apoptosis in human epithelial cells. The small upregulation of CD95 induced by the pilus-deficient mutants is not sufficient to trigger the apoptotic signalling cascade within the infected epithelial cells. Pili will faciliate intimate adherence of the bacterium to the host cells allowing type III secretion system dependent delivery of cytotoxins into the host cells, which finally triggers apoptosis. Physiologie I, Universität Tübingen, Gmelinstrasse 5, 72076 Tübingen, Germany; *Institut für Molekularbiologie, Hufelandstraße 55, 45122 Essen, Germany O 33-3 INTRINSIC P-GLYCOPROTEIN EXPRESSION IN MULTICELLULAR PROSTATE TUMOR SPHEROIDS IS REGULATED BY REACTIVE OXYGEN SPECIES M. Wartenberg, F. C. Ling, M. Schallenberg, A. T. Bäumer, K. Petrat, J. Hescheler, H. Sauer Intrinsic expression of the multidrug resistance (MDR) transporter Pglycoprotein (Pgp) may be regulated by reactive oxygen species (ROS). A transient expression of Pgp was observed during the growth of multicellular tumor spheroids. Maximum Pgp expression occurred in tumor spheroids with a high percentage of quiescent, Ki-67-negative cells, elevated glutathione levels, increased expression of the cyclindependent kinase inhibitors p27Kip1 and p21WAF-1 as well as reduced ROS levels and minor activity of the mitogen-activated kinase (MAPK) members c-jun terminal kinase (JNK), extracellular signal-regulated kinase ERK1,2 and p38 MAPK. Raising intracellular ROS by depletion of glutathione with buthionine sulfoximine (BSO) or glutamine starvation resulted in downregulation of Pgp and p27 Kip1, whereas ERK1,2 and JNK were activated. Downregulation of Pgp was furthermore observed with low concentrations of hydrogen peroxide and epidermal growth factor (EGF), indicating that ROS may regulate Pgp expression. The downregulation of Pgp following BSO treatment was abolished by agents interfering with receptor tyrosine kinase signaling pathways, i.e. the protein kinase C (PKC) inhibitors bisindolylmaleimide I (BIM-1) and Ro-31-8220, the p21ras farnesyl protein transferase (FPT) inhibitor III, the c-Raf inhibitor ZM 336372 and PD98059 which inhibits ERK1,2 activation. ROS involved as second messengers in receptor tyrosine kinase signaling pathways may act as negative regulators of Pgp expression. Department of Neurophysiology, University of Cologne, Robert-KochStr. 39, D-50931 Cologne, Germany

O 33-2 PSEUDOMONAS AERUGINOSA INDUCED APOPTOSIS IS DEPENDENT ON PILUS-MEDIATED ADHESION OF THE BACTERIA TO HUMAN HOST CELLS V. Jendrossek, I. Müller, E. Gulbins* P. aeruginosa strains are gram negative opportunistic pathogens associated with severe infections in immunocompromised hosts and in patients with cystic fibrosis. Apoptosis of lung epithelial cells has been shown to be essential for host defense of mice against nasal infections with P. aeruginosa. Induction of apoptosis in infected cells was dependent on type III secretion system dependent delivery of cytotoxic proteins from the bacterium to the infected host cells as well as on a functional CD95-receptor/CD95-ligand system of the epithelial cell. The purpose of the present study was to define the contribution of bacterial pili to the apoptotic response of human conjunctiva epithelial Chang cells. Our results demonstrate that only P. aeruginosa strains

O 33-4 3D-VISUALISATION OF HIF-1α AND ITS INTERACTING COFACTORS IN HYPOXIC HEPG2 CELLS U. Berchner-Pfannschmidt, C. Wotzlaw, I. Stolze, J. Fandrey, T.Kietzmann, H. Acker Erythropoietin gene expression is regulated by the transcriptional complex hypoxia inducible factor 1 (HIF-1) which consists of the constitutive HIF-1β- and the O2-labile HIF-1α-subunit. Under normoxic conditions HIF-1α is hydroxylated at proline residues and recognized by the von Hippel-Lindau tumor suppressor protein (pVHL). pVHL is part of an E3 ubiquitin ligase which tranfers ubiquitin residues to HIF1α tagging HIF-1α for proteasomal degradation. Under hypoxia HIF1α is stabilized, accumulates and translocates into the nucleus. After dimerisation with the β-subunit and complexing with hepatocyte nuclear

S 213 factor-4 α (HNF-4 á) and p300 HIF-1á binds to hypoxia response gene elements (Masson et al. EMBO 2001). Methods: 2 Photon-laser scanning microscopy (2P-LSM) and computer assisted deconvolution methods were applied to immunostained or transfected HepG2 cells in order to visualize HIF-1α and interacting nuclear or cytoplasmic components under anoxic/hypoxic or reoxygenation conditions. Results: Nuclear domains like nucleoli and PML-bodies appeared as hollow or dense spheres respectively whereas HIF-1α as well as p300 and HNF-4α formed speckle like structures. Under anoxic/ hypoxic conditions HIF-1α speckles colocalised with p300 and HNF-4α significantly, less with PML-bodies and not with nucleoli. Under hypoxia or after reoxygenation HIF-1α was found to be colocalized with pVHL close to the perinuclear endoplasmatic reticulum which was hypothesized to house reactive O2-species generated in hot spots of Fenton reaction affecting the iron state in this cellular compartment (Kietzmann et al NIPS 2001). Conclusion: High resolution 2P-LSM detects new sites of interaction within the O2-sensing pathway by visualising hypoxia induced HIF-1α trafficking.

markedly decreased INFγ formation. At 500 mosmol/l the INFγ formation was completely abolished. Hypertonicity similarly reversed the INFγ production following stimulation with phorbolester PMA and Ca++ ionophore. Thus, the inhibitory effect of osmotic cell shrinkage must be downstream of PKC activation and Ca++. INFγ formation was not significantly inhibited, when osmolarity was increased after antigen stimulation, indicating that INFγ translation was not impaired by osmotic cell shrinkage. On the other hand pretreatment with hyperosmolarity impaired INFγ production. CMV-antigen further triggered the formation of CD69, a C-type lectin similarly used as a marker for lymphocyte activation, an effect again reversed by hyperosmolarity. On the other hand, osmotic cell shrinkage enhanced TNFα production. In conclusion, exposure to hypertonic extracellular fluid blunts the stimulatory action of antigen exposure on INFγ and CD69, an effect maintained even following correction of extracellular osmolarity. Department of Physiology, University of Tübingen, Gmelinstr. 5, D 72076 Tübingen

Max-Planck-Institut für molekulare Physiologie, Otto-Hahn-Str.11, 44227 Dortmund/FRG, Germany O 34-1 O 33-5 MECHANISMS OF ESTROGEN RECEPTOR INDEPENDENT ACTION OF TAMOXIFEN IN HEPG2 CELLS S. Brandt, J. Grote, H. Heller, K.-D. Schuster Tamoxifen is a potent agent for the treatment of breast cancer. It might prove useful in the therapy of hepatocellular carcinoma. As previously shown (Brandt et al. 2001, Pflügers Arch. 441, R 207), the exposition of the estrogen receptor negative human hepatoblastoma cell line HepG2 to 2.10-5 M tamoxifen reduces telomerase activity and cell viability. The latter effect is at least in part due to apoptotic cell death. The findings raised the questions: (i) Is apoptotic cell death caused by the downregulation of telomerase activity? (ii) Does the tamoxifen-induced depression of protein kinase C (PKC) activity mediate the decrease of telomerase activity? Inhibition of telomerase activity was induced by the transfection of the oligonucleotide (TTAGGG)3. PKC activity was inhibited by bisindolylmaleimide I. We quantified telomerase activity using TRAP assay. Cell viability was assessed by performing the MTT-test. Apoptotic nuclei were detected by fluorescence microscopy after DNA staining. Furthermore, telomerase activity and viability of HepG2 cells were determined treating the cells with tamoxifen or with tamoxifen and α-tocopherol. αtocopherol has been demonstrated to antagonise tamoxifen-induced inhibition of PKC activity in the estrogen receptor α-negative breast carcinoma cell line MDA-MB-231 (Gundimeda et al. 1996, J. Biol. Chem. 271, 13504-13514). Both (TTAGGG)3 and bisindolylmaleimide I (24 h incubation) caused a reduction of telomerase activity, preceding a decrease of cell viability and an occurrence of apoptotic nuclei, αtocopherol slowed down the tamoxifen-induced depression of telomerase activity and cell viability. We conclude that the occurrence of apoptotic cell death in HepG2 cells during tamoxifen treatment can be attributed to the reduction of telomerase activity which may be due to the downregulation of PKC activity. Department of Physiology, University of Bonn, Nussallee 11, 53115 Bonn, Germany O 33-6 INHIBITION OF INTERFERON GAMMA EXPRESSION BY OSMOTIC SHRINKAGE OF PERIPHERAL BLOOD LYMPHOCYTES K.S. Lang, S. Fillon , H.-G. Rammensee Hypertonic environment, as it prevails in renal medulla impairs the defence against pathogens and thus facilitates the development of infection. The present experiments have been performed to test whether or not hypertonicity influences activation of lymphocytes. To this end, peripheral blood lymphocytes (PBL) have been activated by cytomegalovirus (CMV) antigen (HLA-A2 restricted epitope NLVPMVATV) to produce interferon gamma (INFγ) at varying extracellular osmolarity. Stimulation (for 6 h) with the CMV antigen increases the production of INFγ from less than 1 % of B- and T-blood lymphocytes to 93 % of the T lymphocytes and to 90 % of the Blymphocytes. Hypertonicity during exposure to the CMV antigen

IMPAIRED RENAL NA+ RETENTION IN THE SGK1-KNOCKOUT MOUSE P. Wulff, V. Vallon, D.Y. Huang, H. Völkl, K. Richter, K. Klingel, J. Loffing, G. Kauselmann, F. Lang, D. Kuhl The serum and glucocorticoid dependent kinase sgk1 is a mineralocorticoid stimulated gene and up-regulates renal epithelial Na+ channel ENaC heterologously expressed in oocytes. Thus, sgk1 is considered to mediate the mineralocorticoid regulation of ENaC activity and antinatriuresis. To address the in vivo functional significance of sgk1 in renal Na+ transport, a knockout mouse (sgk1 -/-) was generated. Sgk1 -/- do not display any gross functional or histological abnormalities. On normal NaCl diet (2.5 g Na+/kg) growth, blood pressure, GFR and urinary excretion of fluid and electrolyte are not significantly different between sgk1 -/- and respective littermate wild type mice (sgk1 +/ +). However, enhanced plasma aldosterone concentration in sgk1-/(mean±SE: 1.2±0.2 vs. 0.6±0.1 ng/ml, n=8-9) suggests that adequate Na + excretion requires an increase of aldosterone release. Despite hyperaldosteronism, the plasma K+ concentration is higher in sgk1-/vs. +/+ (4.7±0.2 vs. 4.1±0.2 mM, n = 13-14). Upon reduced NaCl intake (0.15 or 0 g Na +/kg) sgk1-/- decrease renal NaCl excretion which, however, remains enhanced compared with sgk1 +/+ over the 8 day observation period. Consistent with a concurrent net body Na + loss in sgk1 -/-, the body weight increase observed in sgk1 +/+ was reversed in sgk1-/-. Further studies at day 3 after switching to low NaCl diet reveal that this Na+ λοss occurs in sgk1 -/- inspite of lower blood pressure (79±2 vs. 99±5 mmHg) and GFR (172±26 vs. 273±27 µl/min, n=5-6), and higher plasma aldosterone concentration (6.0±0.5 vs. 1.7±0.2 ng/ml) vs. sgk1 +/+. Micropuncture experiments confirm a decline of single nephron GFR in sgk1-/- mice and disclose a marked increase of fractional Na + reabsorption in the proximal tubule. Electrophysiological analysis of isolated perfused cortical collecting duct (CCD) reveals a smaller amiloride sensitive transepithelial potential difference in sgk1 -/- vs. +/+ (-10±1 vs. –20±1 mV, n = 12-16). In accordance, blinded immunohistochemical analysis reveals that αENaC is present in the apical membrane of distal tubule to CCD in sgk1-/-, the abundance, however, is significantly reduced vs. sgk1 +/+. Thus, while sgk1 appears to be dispensable under normal dietary conditions, it is required for adequate up-regulation of Na + reabsorption in mineralocorticoid-sensitive segments under dietary Na+ restriction. Center for Mol. Neurobiology, Univ. Hamburg, 20246 Hamburg; Dept. of Pharmacology, Physiology, Mol. Pathology, Univ. Tübingen; Dept. of Physiology, Univ. Innsbruck; Dept. of Anatomy, Univ. Zürich O 34-2 ATTENUATED GLOMERULAR HYPERFILTRATION IN DIABETIC DOPAMINE D3-RECEPTOR KNOCK-OUT MICE B. Mühlbauer, M. Tappe, H.-G. Lemaire, G. Gross, G. Luippold Dopamine D3-receptors have been shown to be involved in the physiological regulation of renal hemodynamics in rats and mice. The purpose of the study was to further evaluate the role of D3-receptors in regulating glomerular hyperfiltration in mice with diabetes mellitus. Methods: Dopamine D3-receptor knock-out (KO) and wild-type (WT)

S 214 mice were injected with 120 mg/kg streptozotocin (i.p.) which induced a modest hyperglycemia not requiring insulin substitution. Fourteen days after onset of diabetes mellitus (DM), kidney function was assessed in pentobarbital-anesthesia during infusion of Ringer solution. Age-matched non-diabetic animals served as controls (CON). Blood pressure and heart rate were monitored continuously and glomerular filtration rate (GFR) was assessed by renal clearance of 3[H]-inulin. Values are given as means ± SEM (each group n=10). Results: Blood glucose levels were around 470 mg/dl in DM animals and around 200 mg/dl in CON groups. Urinary flow rate was significantly increased in DM compared to CON animals. GFR did not significantly differ when comparing CON-WT and CON-KO (0.46 ± 0.04 vs. 0.52 ± 0.03 ml/ min per g kidney wet weight, respectively). In DM-WT mice a significant glomerular hyperfiltration was observed while in DM-KO mice GFR was only slightly increased (0.83 ± 0.04 vs. 0.65 ± 0.05 ml/min per g kw, respectively). Mean arterial blood pressure and heart rate did not differ among the groups. Conclusions: These data support the hypothesis that dopamine D 3-receptors are significantly involved in the regulation of glomerular hyperfiltration in experimental diabetes mellitus and that dopamine D3-receptor antagonists may be a therapeutic approach in the treatment of diabetic nephropathy. Department of Pharmacology, Faculty of Medicine, Eberhard-Karls University, D-72074 Tubingen, Germany.

O 34-3 PERFUSATE COMPOSITION MODULATES ENDOTHELIAL NITRIC OXIDE HOMEOSTASIS IN RAT JUXTAMEDULLARY NEPHRONS Janos Pittner, Mats Wolgast and A. Erik G. Persson In cell-free perfused juxtamedullary nephron or whole kidney preparations the total perfusate flow has been observed to be higher than in the corresponding, blood perfused preparations. The first attempts to explain this were concerned with possible rheological and other physical differences between blood and cell-free perfusates. With the discovery of the role of nitric oxide in setting basal vascular tone and the observation that circulating hemoglobin acts as a natural luminal scavenger of endothelial nitric oxide, another plausible explanation has emerged. Cell-free perfusion may not provide a proper intraluminal nitric oxide scavenging and may result in a relative endothelial overproduction and a following vasodilation. On the other hand albumin, a common ingredient of cell-free perfusion fluids, has been reported to have the capacity to scavenge nitric oxide. To investigate the magnitude of this, we used the isolated perfused juxtamedullary nephron preparation perfused with Krebs-Ringer-bicarbonate buffer with 0-20% bovine serum albumin added. We loaded the afferent arteriolar endothelium with the NO-sensitive fluorophore DAF-2 DA and measured steady state fluorescence from single endothelial cells and mid-afferent arteriolar diameters during perfusion with different albumin concentrations at 100 mmHg perfusion pressure. The intensity of the fluorescence decreased during stepwise elevations of the albumin concentration (0%, 1131±47; 1%, 1044±51; 5%, 999±43; 10%, 891±47; 20%, 826±39 arbitrary fluorescence units) and vasoconstriction ensued (0%, 19.14±0.53; 1%, 18.72±0.48; 5%, 17.75±0.62; 10%, 15.37±0.58; 20%, 14.54±0.52 µm mid-afferent arteriolar diameter, n=5, mean±SD) during these maneuvers. We conclude that afferent arteriolar endothelial nitric oxide homeostasis relies on the scavenging capacity of the perfusate and a lack of this scavenging leads to a “predilated” state of the perfused vessels, hence the observed “abnormally” high perfusate flows. Department of Medical Cell Biology, Uppsala University, Uppsala, Sweden

O 34-4 CONTRIBUTION OF POTASSIUM TO ANGIOTENSIN-INDEPENDENT STIMULATION OF ALDOSTERONE RELEASE INDUCED BY REDUCED TOTAL BODY SODIUM S. Hartzendorf, E. Seeliger, H. W. Reinhardt The importance of the components of the renin-angiotensin-aldos-

terone-system (RAAS) for regulation of total body sodium (TBS) has previously been demonstrated by balance studies in freely moving chronically instru-mented Beagle dogs. A 10% reduction of TBS caused an extreme stimula-tion of aldosterone (Aldo) release via the reninangiotensin II-axis (RA). Intriguingly, a comparable but RA-independent stimulation of Aldo-relea-se was observed following RAAS-interruption by ACE-inhibition (ACEI). Reduction of TBS induced an increase in plasma potassium concentration (PP), that may be involved in this RA-independent Aldo-release. This study aims at determining the role of PP in both RA-dependent and RA-independent Aldo-release. Two groups of Beagles were fed different diets: either high potassium intake (HPI: 3.5 mmol K per kg BW per day) or low potassium intake (LPI: 1.5 mmol K per kg BW per day). Both groups were studied in 2 protocols. 1) time control with unchanged TBS (CON), 2) 10% reduction of TBS (?TBS) by peritoneal dialysis, on study day 1, the RAAS was left intact, on day 2, it was interrupted by ACEI. CON: LPI reduced PP and Aldo-release significantly as compared to HPI. ?TBS: LPI completely prevented the increase in PP, as it was induced by TBS-reduction during HPI. With intact RAAS (day 1), LPI reduced RA-dependent stimulation of Aldo-release by ~ 30%. With ACEI (day 2), LPI markedly reduced but did not completely suppress RA-independent Aldo-release. In addition, LPI prevented a retention of potassium, as it was induced by TBS-reduction during HPI. Conclusion: With unchanged TBS, Aldo-release is partially controlled by PP. Three mechanisms mediate the stimulation of Aldo-release induced by TBSreduction: 1) stimulation of renin release (RA-dependent Aldo-release), 2) RA-independent Aldo-release mediated by increase in PP and 3) RA-independent Aldo-release mediated by a yet unknown factor. Inst. für Physiologie, Charité Berlin, Tucholskystr. 2, 10117, Germany

O 34-5 BENEFICIAL EFFECTS OF THE VASOPEPTIDASE INHIBITOR OMAPATRILAT ON NEUROGENIC VOLUME RETENTION IN RATS WITH EXPERIMENTAL CONGESTIVE HEART FAILURE (CHF) B. Klanke, K.F. Hilgers, T. Ditting, R.E. Schmieder, R. Veelken We previously reported that angiotensin II is mandatory for the volume retention in CHF rats due to increased and dysregulated renal sympathetic nerve activity. We tested the hypothesis that the vasopeptidase inhibitor omapatrilat, inhibiting angiotensin converting enzyme (ACE) and neutral endopeptidase (NEP), reverses neurogenic volume retention after an acute volume load in rats with CHF. To assess the neural control of volume homeostasis, 21 days after the induction of congestive heart failure (CHF) by ligature of the intraventricular coronary artery, rats underwent volume expansion (0.9% NaCl; 10% body weight over 30 min) to decrease renal sympathetic nerve activity. Untreated animals, rats with renal denervation or rats pretreated with a non-hypotensive dose of the vasopeptidase inhibitor omapatrilat (250 nmol / animal) were studied. An additional group of animals received a non-hypotensive dose of the ACE-inhibitor captopril (250 nmol / animal) before administering the volume load. Experiments were performed in chronically instrumented rats. Excretion of the volume load and the administered sodium, glomerular filtration rate and renal plasma flow as well as blood pressure and heart rate were assessed. In some animals activity of renal sympathetic nerve activity was measured with bipolar electrodes. Untreated CHF rats excreted only 74,9±3 % of the administered volume load (untreated control rats without CHF: 106±5 %)(*). In heart failure rats pretreated with omapatrilat (250 nmol / animal iv) or in rats after renal denervation these values were 106,5±6 %(*) and 106±8 %(*) respectively. Rats receiving captopril exhibited a volume load excretion of 101.6±6 %(*) (*p<0,05 versus untreated CHF rats, mean ± SEM). Decreases of renal sympathetic nerve activity by volume expansion were Acute volume retention in rats with congestive heart failure is due to effects of an increased renal sympathetic nerve activity. Impaired volume excretion is reversed by low dose pretreatment with omapatrilat (NEP and ACE inhibitor) and to the same extent by captopril (ACE inhibitor). This findings underline the importance of the renin-angiotensin system for volume regulation in CHF. Dr. B. Klanke, IV Medizinische Klinik, Universität Erlangen-Nürnberg, Universitätsstraße 17, 91054 Erlangen

S 215 O 34-6

O 35-2

REABSORPTION OF myo-INOSITOL IN RAT PROXIMAL TUBULES AND LOOPS OF HENLE S. Silbernagl, K. Voelker, W.H. Dantzler* Fractional excretion of myo-Inositol (mIn), a renal organic osmolyte, is 1-2 %. To determine the localization and kinetics of tubular mIn reabsorption as well as the specificity of the carrier type(s) involved, we microinfused and -perfused (10 nl/min) tubule segments at the cortical and papillary surface of of rat kidney in vivo et situ. The microinfusate/-perfusate contained 3H-mIn or 3H-D-glucose (Gluc) and 14 C-inulin. The fractional 3 H-reabsorption (FR) between (→) the microinfusion site (early proximal [EP], late proximal [LP], early distal tubule [ED], papillary loop of Henle [LH] or vasa recta [VR]) and the final urine (U) was determined. Moreover, segments of PCTs were microperfused (EP → LP). Results (± SEM): FR of 3H-mIn (10-5 M) was 95 ± 1.4 % (EP → U), 96 ± 0.8 % (LP → U) and 3.7 ± 5.2 % (ED → U), and was blocked by 10-4 M phloridzin. Addition of 5 · 10-2 M D-glucose resulted in a FR of 3H-mIn (10-5 M) of 83 ± 2.5 % (LP → U), whereas addition of 5 · 10 -2 M D-fructose, D-mannose, or L-fucose had no effect. At 5 · 10-2 M mIn, FR of mIn was 16 ± 2.2 % (EP → U) and 14 ± 5.0 % (LP → U). In microperfusion experiments (EP → LP), FR of 3 H-mIn was 56 ± 4.0 % (at 10-5 M), 9 ± 2.3 % (at 5 · 10-2 M), and 36 ± 2.2 % (at 10-5 M + 5 · 10-2 M D-glucose). In further experiments (EP → LP), FR of 3H-Gluc was 74 ± 2.8 % (at 10-5 M), 16 ± 2.5 % (at 5 · 102 M), and 73 ± 4.7 % (at 10-5 M + 5 · 10-2 M mIn). When 3H-mIn (9 · 105 M) was microinfused into LH or VR, the fractional 3H-recovery in the ipsilateral urine was 56 ± 4.6 % and 27 ± 4.5 %, respectively. The latter value was reduced to 12 ± 2.2 % by 10-4 mol/1 phloridzin. Conclusions: (i) 3H-mIn is reabsorbed in the PCT and in short loops of Henle by (a) saturable, phloridzin-sensitive transporters). The latter tubule segment has the higher reabsorption capacity. (ii) 3H-mIn is also reabsorbed in the ascending limb of long loops of Henle, and can enter the tubular lumen from the papillary vasa recta. The latter process is phloridzinsensitive. (iii) Along the PCT and the loop of Henle, 3H-mIn is reabsorbed by a carrier different from the glucose carriers SGLT1 and 2. Whereas most of our data suggest that mIn is reabsorbed by the SMIT carrier(canine), some of them are at variance with the specificity of canine SMIT expressed in Xenopus oocytes.

SIMILAR CHLORIDE CHANNELS IN THE BASOLATERAL MEMBRANES OF THE EARLY AND LATE DISTAL TUBULE OF THE MOUSE J Teulon, S Lourdel, P Marvao, M Paulais We investigated Cl - channels in the basolateral membranes of the distal convoluted tubule (DCT), discriminating between early (DCTe) and late (DCTl) portions. The DCTs were microdissected from the kidneys of male mice after collagenase treatment. The pipette and bath solutions contained (in mM) : 140 NaCl, 4.8 KCl, 1.2 MgCl2, 1 CaCl2, 10 glucose, 10 HEPES (pH 7.4). Using the cell-attached and excised configurations, we found one channel of 8-10 pS (Table 1) in 50 % (n = 64) of DCTe and 63 % (n = 15) of DCTl tubules. Upon excision, most patches lost channel activity rapidly but it was possible to measure conductive properties in some cases. The PNa/ PCl ratio was estimated using a bath NaCl concentration of 14 mM : we found that the channel was Cl- selective (Table 1).

Depts. of Physiol., Univ. of Wuerzburg & *Univ. of Arizona, Tucson, AZ

O 35-1 BARTTER SYNDROME WITH SENSORINEURAL DEAFNESS DUE TO DISTURBED ACTIVATION OF CLC-K CHLORIDE CHANNELS BY BARTTIN S. Waldegger, N. Jeck, P. Barth, M. Peters, H. Vitzthum, K. Wolf, A. Kurtz, M. Konrad, H. W. Seyberth The term Bartter syndrome encompasses a heterogeneous group of autosomal recessive salt-losing nephropathies which are caused by disturbed transepithelial sodium chloride reabsorption in the distal nephron. Mutations have been identified in the NKCC2 (Na +-K +-2Cl --) cotransporter and ROMK potassium channel, which cooperate in the process of apical sodium chloride uptake, and ClC-Kb chloride channels, which mediate basolateral chloride release. Recently, mutations in barttin, a protein not related to any known ion transporter or channel, were described in BSND, a variant of Bartter syndrome associated with sensorineural deafness. Here we show that barttin functions as an activator of ClC-K chloride channels. Expression of barttin together with ClC-K in Xenopus oocytes increased ClC-K current amplitude, changed ClC-K biophysical properties, and enhanced ClC-K abundance in the cell membrane. Co-immunoprecipitation revealed direct interaction of barttin with ClC-K. We performed in situ hybridization on rat kidney slices and RT-PCR analysis on microdissected nephron segments to prove co-expression of barttin, ClC-K1 and ClC-K2 along the distal nephron. Functional analysis of BSND-associated point mutations revealed impaired ClC-K activation by barttin. The results demonstrate regulation of a CLC chloride channel by an accessory protein and indicate that ClC-K activation by barttin is required for adequate tubular salt reabsorption and inner ear endolymph-production. Department of Pediatrics, University-Hospital of Marburg, Deutschhausstr. 12, D-35033 Marburg, Germany

DCT

G (pS) Er (mV) G (pS) PNa /PCl Excised Cell attached Cell attached Excised Early 9.4 ± 0.3 (23) 2.0 ± 2.2 (23) 10.7 ± 1.1 (6) 0.08 ± 0.04 (9) Late 8.1 ± 0.6 (13) 0.1 ± 1.6 (13) 9.6 ± 0.6 (7) 0.08 ± 0.02 (7) Using the same bath solution, a sub-conductance level with an amplitude 0.53 ± 0.07 as great as the fully-open amplitude (n = 6) was frequently observed, and showed Cl- selectivity. The Cl- channel blocker DPC, at 10-5, 10-4 and 10-3 M, reduced the Po to 85 ± 8 % (n = 3), 57 ± 9 % (n = 4) and 20 ± 5 % (n = 3) of control, respectively. Although further studies are needed to characterize regulation properties, we suggest that this basolateral channel, might participate in the NaCl absorption across the DCT, in relation with the apical Na-Cl cotransporter. INSERM U 426, Faculté Xavier Bichat, Paris, France

O 35-3 INTERACTION OF PLASMA MEMBRANE VACUOLAR H +ATPase IN THE KIDNEY WITH THE Cl--CHANNELS CFTR AND ClC-5 C.A. Wagner1, U. Lükewille1, S. Breton2, D. Brown2, A.S. Yu3, B. Grubb4, R. Boucher4, G. Giebisch1, J.P. Geibel1,5 Vacuolar H +-ATPases (V-H +-ATPase) are the major H +-extrusion proteins beside Na+/H+-exchangers in the kidney and are important for regulation of acid-base homeostasis. Plasma membrane H+-ATPase function has been shown to be Cl--dependent in vesicle studies and in isolated tubule preparations. The molecular identity of the Cl - channel(s) has remained elusive to date. Two Cl- -channels have been implicated in interactions with V-H+-ATPases, CFTR and ClC-5. CFTR is highly abundant in intercalated cells (IC) in the cortical collecting duct (CCD). ClC-5 has been shown to colocalize with V-H+-ATPases in endocytotic vesicles in the proximal tubule (PT) and has also been found in the IC in the CCD. Mutations in ClC-5 cause Dent-s disease characterized by defective PT protein endocytosis and subsequent proteinuria and hypercalciuria. CFTR deficient mice showed an almost complete loss of V-H+-ATPase activity in IC in isolated CCD. V-H+ATPase activity, however, was not Cl- -dependent in these cells and could not be inhibited by Cl- -channel blockers. V-H+-ATPase expression levels and localization were not altered in CFTR knock-out mice. Coimmunoprecipitation demonstrated a possible close interaction between the B1 V-H+-ATPase subunit and CFTR. In contrast, in mice with highly reduced ClC-5 expression levels no apparent difference in plasma membrane V-H+-ATPase activity in the PT or IC in the CCD were found. Also, V-H +-ATPase expression levels and localization were unaltered. These results demonstrate that V-H+-ATPase activity is Cl- -dependent in the PT but not in the CCD. ClC-5 is not the Cl- channel necessary for PT plasma membrane V-H +-ATPase function. CFTR, in contrast, is involved in regulation of V-H+-ATPase activity in IC and may directly interact. Departments of 1Cellular and Molecular Physiology and 5Surgery, Yale University, School of Medicine, 333 Cedar Street, New Haven, CT, USA, 2Program in Membrane Biology, Mass Gen Hospital, Boston, MA, USA, 3Department of Medicine, Brigham´s and Women´s Hospital and Harvard Medical School, Boston, MA, USA 4Cystic Fibrosis/ Pulmonary Research and Clinical Treatment Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

S 216 O 35-6

O 35-4 -

ALTERATION OF CL SWELLING-ACTIVATED CHANNELS BY CFTR KNOCK-OUT IN PRIMARY CULTURES OF RENAL TUBULES H. Barrière, R. Belfodil, M. Tauc, C. Poujeol, P. Poujeol Objective: several studies demonstrated that cells expressing mutated CFTR protein exhibit altered channel regulation. Transgenic homozygous CFTR-/- mice presenting an inactivation of gene encoding CFTR were used to study alterations of different Cl- channels in renal cells. Method: primary cultures of proximal convoluted (PCT), distal convoluted (DCT) and cortical collecting tubules (CCT) were obtained from both wild type (CFTR +/+) and CFTR knock-out (CFTR-/-) mice . Cl- conductances were characterised using the whole cell clamp technique in 7 day-old cultures. Results: in CFTR+/+ mice, a cAMP activated Cl- conductance was measured in cultured DCT and CCT but not in PCT cells. As expected, cultured DCT and CCT from CFTR-/- mice did not exhibit cAMP-activated Cl - conductances. Ca++-activated Clcurrents induced by ionomycin were recorded in all cultured segments from both CFTR+/+ and CFTR-/- mice. Swelling-activated Cl- conductances were also studied. In cultured PCT, DCT and CCT from CFTR+/+ mice, a hypo-osmotic shock (30%) induced outwardly rectifying and Cl- currents which were strongly inhibited by hyper-osmotic solution, 0.1 mM NPPB, 1 mM DIDS and 1 mM DPC in three segments. Extracellular adenosine or ATP (10 µM) mimicked the swelling Cl - conductance A1 receptor antagonist DPCPX (10µM) and ectoATPase inhibitor blocked the effects of ATP, adenosine or hypotonic shock. Moreover, the application of adenosine or the hypotonic shock produce a Ca 2+ influx which is responsible for the activation of the swelling Cl - conductance. In CFTR -/- mice, the hypotonic shock did not induce Cl- currents in all segments. However, the perfusion of ATP, adenosine or the application of ionomycin restored swelling-activated Cl- conductances. Conclusion: In the kidney of CFTR +/+ mice, swelling Cl- channels are controlled by A1 receptors and Ca++ influx across the cell membrane. The inactivation of the CFTR gene suppresses this swelling-activated Cl- conductances This could be due to the absence of adenosine production resulting from the loss of ATP release.

L- AND P/Q-TYPE VOLTAGE DEPENDENT CALCIUM CHANNELS ARE CO-EXPRESSED IN RENAL VASCULAR SMOOTH MUSCLE CELLS D. Andreasen, U.G. Friis, B.L. Jensen, P.B. Hansen, O. Skøtt We have previously shown that both L-type and P/Q-type Ca2+ channels are expressed in renal vascular smooth muscle cells, and that both channel types are important for contraction of renal resistance vessels (1, 2). In this work we wanted to examine whether the two channel types were co-expressed, as opposed to expression in separate populations of smooth muscle cells. Using patch clamp, we measured calcium currents in a culture of immortalized rat aortic smooth muscle cells (A7r5, American Tissure Type). These cells did have calcium currents, which could be partially blocked by applying either the Ltype blocker calciseptine (30.6% ± 2.8 SE, n=5), or the P/Q-type blocker ω-Agatoxin IVA (22.1% ± 4.2 SE, n=4). Both A7r5 cells and freshly isolated renal vascular smooth muscle cells (rVSMC) were used for immunocytochemistry. The cells were labeled with anti-α1C antibody (Alomone Labs.), and this labeling visualized using goat-anti rabbit IgG antibody conjugated to the Alexa 488 fluorophore (Molecular Probes). Subsequently, the same cells were labeled with anti-α 1A antibody (Alomone Labs.) conjugated to the Alexa 586 fluorophore (Molecular probes). This confirmed that both L- and P/Q-type channels are expressed in the same cell. It further revealed that both channel types are located in clusters in the smooth muscle cells. In individual cells there seems to be regions with higher densities of both calcium channels. It also appears that there is a high density of both channels in the vicinity of the nucleus. We can thus conclude that the cultured A7r5 cells and freshly isolated renal vascular smooth muscle cells express both L- and P/Q-type channels. PB Hansen, BL Jensen, D Andreasen, UG Friis, O Skøtt. Circ Res, 87: 896 - 902 (2000). PB Hansen, BL Jensen, D Andreasen, O Skøtt.. Circ Res, 89: 630-638 (2001). Ditte Andreasen, Physiology and Pharmacology, University of Southern Denmark, Winsloewparken 21, 3., DK5000Odense, Denmark

UMR-CNRS 6548, Université de Nice Sophia Antipolis, Parc Valrose, 06108 Nice cedex, France

O 36-1 O 35-5 FUNCTIONAL ANALYSIS OF ROMK MUTATIONS: CORRELATION WITH THE CLINICAL PHENOTYPE M. Peters, N. Jeck, S. Weber, K.P. Schlingmann, U. Schulte, S. Waldegger, H.W. Seyberth, M. Konrad Mutations in the renal potassium channel ROMK lead to a lifethreatening salt-losing tubular disorder with antenatal onset due to defective salt reabsorption in the medullary thick ascending limb of Henle‘s loop. Clinical characteristics include fetal polyuria leading to polyhydramnios and preterm delivery. Postnatally, affected patients show severe salt-wasting and marked hypercalciuria with consecutive nephrocalcinosis. Eleven naturally occurring ROMK mutations were studied in Xenopus oocytes and mammalian cells by immunohistochemical and electrophysiological methods revealing different underlying pathomechanisms. Beside five mutations which lead to decreased potassium conductance, three premature stop-mutations and three mutations with impaired trafficking to the cell surface were identified. As already shown for mutations in the CFTR, inframe stopand trafficking-mutations are targets for further pharmacological intervention in order to restore at least residual K+ conductance. The clinical relevance of such a residual K+ conductance is reflected by the mild phenotype and late clinical onset of a patient harboring a homozygous ROMK mutation which shows a significant residual function in electrophysiological recordings. This indicates that even partial restoration of the ROMK mutations might effectively alleviate the disease phenotype. Sponsored by DFG and the Kempkes Stiftung, Marburg Department of Pediatrics, Philipps-University Marburg, Deutschhausstr. 12, 35037 Marburg, Germany

DIAZOXIDE AND ISCHEMIC PRECONDITIONING DIFFERENTLY ON THE MITOCHONDRIA: IS MITOCHONDRIAL KATP CHANNEL A RED HERRING? K.H.H. Lim, S.A. Javadov, S.J. Clarke, M.-S. Suleiman, A.P. Halestrap Objective. Recent data have cast doubt on whether the mitochondrial ATP sensitive potassium channel (mitoK ATP) is the end-effector of ischemic preconditioning (IP). This study reappraises the role of the mitoK ATP and specific mitoK ATP modulators, diazoxide and 5hydroxydecanoate (5HD), in preconditioning by measuring mitochondrial matrix volume and respiration, which are regulated by mitoK ATP. Methods. Langendorff perfused Wistar rat hearts (oxygenated Krebs Henseleit buffer, 37°C, pH 7.4) were either treated with 50µM diazoxide or IP (5’5’5’5) ± 100 or 300µM 5HD before 30 mins global isothermic ischemia and 30 mins reperfusion. Mitochondria were rapidly isolated for determination of matrix volume (3H2O and [14C]-sucrose) and respiration (Clarke oxygen electrode). Results. IP and diazoxide treated hearts showed better postischemic functional recoveries vs controls (n=9,7,7 respectively; rate pressure product mean±SEM: 97±6 and 66±l3 vs 4l± 7% of baseline values respectively; ANOVA p<0.05). Mitochondria matrix volumes were increased after IP and diazoxide treatment vs controls (0.87±0.09 and l.03±0.07 vs 0.65±0.05µl/mg respectively, p<0.05 both), whilst IP increased and diazoxide decreased mitochondrial respiration (State 3 oxidation of 5 µM 2oxoglutarate: 474±36 and 29l±21 vs 37l±29 nmol/mg/min in controls respectively, both p<0.05). 5HD had variable effects on these changes. Conclusions. The results suggest that (l) ischemic and chemical preconditioning act differently on mitochondria, (2) the mitoKATP may not be the end effector mechanism, and (3) diazoxide and 5HD may not be acting specifically on the mitoKATP as previously proposed. Bristol Heart Institute, University of Bristol, Bristol BS2 8HW, UK

S 217 O 36-2

O 36-4

UPREGULATION OF CARDIAC ENDOTHELIAL NITRIC OXIDE SYNTHASE BY ANGIOTENSIN CONVERTING ENZYME INHIBITION IN HEART FAILURE PATIENTS H. Morawietz, S. Rohrbach, U. Rueckschloss, E. Schellenberger, K. Hakim, H.-R. Zerkowski, G. Kojda, D. Darmer, J. Holtz Objective of study ACE inhibitors are considered as gold standard in heart failure therapy. The knowledge on the underlying mechanisms remains incomplete. Therefore we investigated the impact of ACE inhibition on the cardiac NO system in heart failure patients. Methods and Results The mRNA expression (amol/µg RNA, standard calibrated competitive RT-PCR) of endothelial constitutive NO synthase (ecNOS) was significantly higher (22.5±4.8) in right atrial myocardium of patients with ACE inhibitor treatment before elective coronary artery bypass grafting or valve surgery, than in atria of patients without this therapy (8.9±0.7, P<0.0001). Furthermore, ACE inhibitor treatment increased atrial ecNOS protein expression from 9.0±0.7 relative units (RU) to 12.0±0.9 RU (P<0.05) and myocardial ecNOS activity from 17.6±1.3 to 23.7±1.1 pmol/mg protein/min (P<0.001). This increased ecNOS expression was specific for ACE inhibitor therapy because no significant differences in clinical characteristics (including NYHA classification and hemodynamic parameters) and concomitant therapy between these groups of patients were observed. In addition, we found a trend of decreased ecNOS mRNA expression in failing right atrial and left ventricular myocardium of patients without ACE inhibitor treatment, but a similar upregulation of ecNOS mRNA in failing right atrium and left ventricular myocardium of heart transplantation patients with ACE inhibitor therapy before surgery compared with NYHA I or nonfailing donor hearts. Conclusions These data suggests a new mechanism by which ACE inhibitors impact on cardiac NO-dependent pathways. This effect might contribute to the beneficial effects of ACE inhibition in the treatment of heart failure.

KATP CHANNEL-INDEPENDENT TARGETS OF DIAZOXIDE AND 5-HYDROXY DECANOATE IN THE HEART P.J. Hanley, M. Löffler, M. Mickel, J. Daut Diazoxide and 5-hydroxydecanoate (5-HD; C10:0) are generally assumed to target specifically mitochondrial ATP-sensitive K + (K ATP) channels. Here we have elucidated KATP channel-independent targets of diazoxide and 5-HD in the heart. Using submitochondrial particles isolated from pig heart, we found that diazoxide (10 − 100µmol/L) dose-dependently decreased the activity of complex II, whereas it did not inhibit complexes I, III and IV. Pinacidil (23 − 230µmol/L), a nonselective KATP channel opener, did not inhibit complex II, however, it selectively inhibited complex I (NADH dehydrogenase). These direct inhibitory effects of diazoxide and pinacidil cannot be explained by activation of mitochondrial K ATP channels. Furthermore, application of either diazoxide (100µmol/L) or pinacidil (100µmol/L) did not decrease mitochondrial membrane potential in isolated guinea pig ventricular myocytes. We tested whether the medium-chain fatty acid 5HD is ‘activated’ via acyl-CoA synthetase (EC 6.2.1.3), an enzyme present both on the outer mitochondrial membrane, as well as in the matrix. Using electrospray ionization mass spectrometry and analytical HPLC, we showed that 5-HD-CoA (5-hydroxydecanoyl-CoA) is synthesized from 5-HD via acyl-CoA synthetase. 5-HD-CoA may be the active form of 5-HD, serving as substrate for acyl-CoA dehydrogenase (β-oxidation) and/or exerting some other cellular action. In conclusion, we have identified KATP channel-independent targets of 5HD, diazoxide and pinacidil. Our findings suggest that pharmacological preconditioning may involve inhibition of respiratory chain complexes.

Institute of Pathophysiology, Martin Luther University HalleWittenberg, Magdeburger Str. 18, D-06097 Halle, Germany

Institut für Normale und Pathologische Physiologie der Universität Marburg, Deutschhausstr. 2, D-35037 Marburg O 36-5

ANALYSIS OF THE CARDIAC PROTEOME OF MYOGLOBIN DEFICIENT MICE REVEALS REDUCED EXPRESSION OF PROTEINS INVOLVED IN FATTY ACID BETA-OXIDATION T. Laussmann, C. Fingas, S. Metzger, J. Schrader In order to investigate compensatory changes on the proteome level in hearts of myoglobin deficient mice (myo -/-), myocardial tissue was analyzed by large - format (24 x 20 cm) two dimensional gel electrophoresis (2D-PAGE, n=12 myo -/- vs. n=12 wildtype (wt) mice) combined with mass spectrometry (ESI-MS/MS). A total number of 450 protein species was recorded of which 22 were significantly altered in expression (t-test: p<0.005). Besides alterations like: [Mn] superoxide-dismutase (-69% in Mb-KO in relation to wt), voltage-gated anion channel (-60%), alpha (b)-crystallin (-38%), heat shock protein 70 (+162%) and others, myo -/- mice showed consistently decreased expression of proteins involved in fatty acid metabolism: electron transfer flavoprotein dehydrogenase (-90%), short-chain acyl-CoA dehydrogenase (-52%), isovaleryl-CoA dehydrogenase (-48%), enoyl-CoA hydratase (2 isoforms, -58%, -33%) and apolipoprotein AI (-72%). Additionally, considerable amounts of fragments of the mitochondrial trifuncional enzyme (TE) were detected in myo -/- mice, indicating an increased degradation of this protein. The TE plays a central role in the beta - oxidation of long - chain fatty acids. Western blot analysis (ECL) of the TE showed that the native protein is indeed reduced by 28% in myo -/- mice (n=9, p<0.05). In contrast, the amounts of the glycolytic enzyme GAPDH increased drastically in myo -/- mice (+474%). Taken together, these observations suggest a metabolic shift from fatty acid to glucose utilization when myoglobin is lacking. This would positively affect the oxygen balance in myo -/- hearts since the oxidation of glucose consumes less oxygen per mole produced ATP compared to fatty acid oxidation. Thus, the presence of myoglobin is essential for effective beta -oxidation of fatty acids in the heart.

IMPROVEMENT OF LEFT VENTRICULAR FUNCTION AFTER CELLULAR CARDIOMYOPLASTY W. Roell, Z.J. Lu, Y. Xia, #K. Tiemann, M. Breitbach, E. Kolossov, R. Stehle, G. Pfitzer, #B. Lüderitz, J. Hescheler, +A. Welz, B.K. Fleischmann The potential improvement of heart function after cellular cardiomyoplasty is currently discussed. Aim of the present study was to investigate the fate and function of transplanted cardiomyocytes at the cellular level and in the beating heart. To allow the in vivo identification and analysis of transplanted cells, murine embryonic cardiomyocytes (E13.5-14.5) expressing the enhanced green fluorescent protein (EGFP) under control of the cardiac α-actin promoter were injected (105 cells diluted in 5 µl volume) into the free left ventricular wall of wild type mice of the same strain. Two weeks after operation the hearts were enzymatically digested and ion channel expression assessed in the EGFP positive cells. The transplanted cardiomyocytes displayed typical ventricular-like action potentials and intact β-adrenergic stimulation similar to the native counterparts. Analysis of the expression of the L-type Ca2+ current showed normal biophysical characteristics and similar current density in transplanted (15,4±1.8 pA/pF, n=3) and native cardiomyocytes (15.7±3.1 pA/pF, n=5). Next, force of contraction was analyzed in skinned (Triton X) muscle strips prepared from cryoinfarcted areas with and without transplanted cells. A significant increase of force development was observed after cellular cardiomyoplasty compared to sham operated animals (221±52 µN, n=5 vs. 37±27 µN, n=5), confirming engraftment and physiological integrity of transplanted cells. These data suggest a possible improvement of left ventricular function, which was evaluated 1 and 14 days postoperatively by echocardiography using a 15-20 MHz linear scanner. A prominent improvement of the regional wall motion in the cryoinfarcted areas of animals with cellular cardiomyoplasty compared to the group without treatment was observed. The improved wall motion scores were accompanied by a clear increase of the ventricular ejection fraction (44±13%, n=5) vs. control animals (23±12 %, n=5). We conclude, that cellular cardiomyoplasty improves contractility, leads to enhanced left ventricular function and most likely is responsible for their better survival.

Institute of Cardiovascular Physiology, Universitätsstr. 1, 40225 Düsseldorf

Inst. of Physiology, Univ. of Cologne, D-50931 Cologne, +Department of Cardiac Surgery and #Cardiology, University of Bonn

O 36-3

S 218 O 36-6 INVOLVEMENT OF REACTIVE OXYGEN SPECIES IN CARDIOTROPHIN-1 (CT-1)- INDUCED CARDIOMYOGENESIS OF MURINE EMBRYONIC STEM CELLS H. Sauer, W. Neukirchen, G. Rahimi, J. Hescheler, M. Wartenberg Cardiotrophin-1 (CT-1) is a novel cytokine which is involved in the growth and survival of cardiac cells. In the present study we demonstrate that treatment of embryoid bodies grown from pluripotent embryonic stem (ES) cells with CT-1 significantly stimulated cardiomyogenesis and increased nuclear expression of the proliferation marker Ki-67. The increase in Ki-67 expression was inhibited upon pretreatment with free radical scavengers, indicating a role for the intracellular redox state in the signaling cascade. CT-1 treatment of cardiac cells raised intracellular reactive oxygen species (ROS), which were presumably generated by an NADPH oxidase since ROS generation was downregulated upon preincubation with the NADPH oxidase inhibitor diphenylen iodonium chloride (DPI). CT-1 activated nuclear factorκB (NF-κB), and induced phosphorylation of the Janus kinase signal transducer-2 (JAK-2), as well as the signal transducer and activator of transcription-3 (STAT-3). JAK-2 and STAT-3 phosphorylation could be inhibited by pretreatment with the JAK-2 antagonist AG490, the free radical scavenger vitamin E, the NADPH oxidase inhibitor DPI as well as by the phosphatidylinositol 3 kinase (PI 3-kinase) inhibitor LY294002. It is concluded that CT-1 stimulates the proliferation of ES cell-derived cardiomyocytes by a JAK-STAT pathway that involves ROS as signaling molecules in the signal transduction cascade. Institut für Neurophysiologie, Universität zu Köln, Robert-Koch-Str. 39, D-50931 Köln, Germany O 37-1 QUANTAL AND ANATOMICAL PROPERTIES OF LAYER 4 LAYER 2/3 SYNAPSES IN RAT BARREL CORTEX D. Feldmeyer1, R.A.Silver2, J.Lübke3, B.Sakmann Layer 4 of the barrel cortex receives sensory information from the periphery via the thalamus and distributes it to other cortical layers via the axons of spiny stellate cells. In this study we have investigated functional and anatomical properties of synapses made by layer 4 spiny neurones onto layer 2/3 pyramidal cells. Whole-cell voltage recordings were made at 35-37 °C from 37 pairs of synaptically coupled layer 4 spiny neurones and layer 2/3 pyramidal cells in brain slices of barrel cortex from P17-P23. Single action potentials in the layer 4 neurones evoked unitary glutamate receptor mediated EPSPs in the L2/3 pyramidal cell with a peak amplitude of 0.66±0.1 mV (S.E.M.) in 2 mM calcium. To impose conditions with different release probabilities, solutions with a different Ca/Mg ratio were superfused. Quantal parameters were estimated using multiple-probability fluctuation analysis. Preliminary analysis of 5 cell pairs gave a mean quantal size of 0.16±0.01 mV (at -70mV) and 4.8±0.6 (range 3 - 6) functional release sites. In 2 mM calcium the mean release probability across sites was 0.80±0.04. Anatomical analysis of the same biocytin filled pairs showed the number of putative synaptic contacts varied between 4-5, with an average of 4.8±0.2 (n = 5). Synaptic contacts were exclusively made within pyramidal cells of that cortical column in which the spiny cell was located and contacts were made predominantly on the basal dendritic branches of the pyramidal cell. The fact that the mean number of anatomical contacts and functional release sites was similar suggests that each contact behaves in an all-or-non manner. Supported by the Wellcome Trust and SFB 505 (Grant C6) 1

Abteilung Zellphysiologie, Max-Planck Institut für Medizinische Forschung, Heidelberg, Germany; 2Dept. Physiology, UCL, London, United Kingdom; 3 Anatomisches Institut der Albert-LudwigsUniversität, Freiburg, Germany O 37-2 SYNAPTIC FACILITATION BY NON-LINEAR SUMMATION OF CALCIUM CONCENTRATION IN THE MICRODOMAIN F. Felmy, E. Neher , R. Schneggenburger Facilitation of transmitter release is caused by Ca2+ concentration ([Ca]i) remaining from a previous stimulus. [Ca]i could act either by binding to a high affinity Ca2+ site (“facilitation-site model”), and/or by increasing the microdomain [Ca] i at the Ca2+ sensors for vesicle fusion (“Ca2+

summation”). To distinguish between these possibilities, we performed paired pre- and postsynaptic recordings at the calyx of Held synapse, with presynaptic Ca2+ uncaging (1-1.5 mM DM-nitrophen, 70-100 µM fura-4F or fura-2FF) and detection of transmitter release from excitatory postsynaptic currents (EPSC). Weak pre-depolarizations of the presynaptic terminal (range –34 to –23 mV, 84 ms) led to strong facilitation of transmitter release (359±114 % of control; n = 12 cells), as tested by action-potential (AP) like stimuli. However, weak pre-depolarizations did not change the Ca 2+ dependence of transmitter release in Ca2+ uncaging experiments (F.F., E.N. & R.S. 2001; Pflügers Arch. 441, R178), contrary to the expectations of the facilitation-site model. We next determined the Ca 2+ sensitivity of facilitation by pre-elevating [Ca]i by Ca2+ uncaging, and probing release with AP-like stimuli. At 0.776 ± 0.14 µM [Ca]i, facilitation was 212 ± 85 % of control. This data was analyzed by calculating the inferred waveform of microdomain [Ca]i for individual EPSCs after AP-like stimuli, according to a kinetic model for Ca2+ binding and transmitter release (R.S.& E.N. 2000, Nature 406). When the measured, preelevated [Ca]i was added linearly to the inferred microdomain [Ca]i, about one-half of the observed facilitation could be explained. Since we find no evidence in favor of a separate facilitation site (see above), we hypothesize that the remaining half of facilitation is caused by supralinear summation of [Ca]i, possibly caused by Ca2+ buffer saturation. Max-Planck-Institut für biophysikalische Chemie, Membranbiophysik, 37077 Göttingen, Germany

O 37-3 MITOCHONDRIAL CONTROL OF SHORT-TERM SYNAPTIC DEPRESSION AT THE RAT CALYX OF HELD B. Billups, I.D. Forsythe Mitochondria are present at high concentrations in presynaptic terminals. In addition to providing metabolic support for exocytosis, their ability to buffer cytoplasmic calcium may contribute to regulating transmitter release. We investigated this possibility at the calyx of Held. The calyx of Held is a glutamatergic presynaptic terminal in the auditory brainstem. Its large size is suitable for whole-terminal patchclamping (Forsythe et al., 1998, Neuron 20, 797-807) and measurement of intracellular calcium concentration using fluorescent dyes. We have recently shown that Rhod-2 can be used to image presynaptic mitochondrial calcium transients (Billups and Forsythe, 2001, J.Physiol 531P, p110) and have demonstrated that mitochondrial calcium uptake plays a role in shaping the presynaptic cytoplasmic calcium transient. To investigate whether mitochondrial calcium buffering has any effect on neurotransmitter release we simultaneously patch-clamped the postsynaptic neurone in the medial nucleus of the trapezoid body (MNTB) and the calyx of Held presynaptic terminal. Neurotransmitter release was evoked by presynaptic depolarization from –80 to 0mV for 1ms, repeated at 200Hz for 100ms. The postsynaptic glutamatergic response, mediated by AMPA receptors, was recorded in the presence of NMDA receptor antagonists. 50µM cyclothiazide and 3mM kynurenate were also added to reduce AMPA receptor desensitisation and saturation. Under these conditions the postsynaptic response is a reflection of the amount to neurotransmitter released upon stimulation of the terminal. The magnitude of the post-synaptic current (EPSC) reduces dramatically throughout the duration of the 200Hz train. This synaptic depression is probably due to the depletion of vesicles from the readily releasable pool in the presynaptic terminal. The rate of recovery from synaptic depression was assessed by a test stimulus at various times after the train of stimuli. In control conditions the magnitude of a test EPSC elicited 500ms after the depleting train recovered to 81±4% (mean±SEM, n=3) of the initial EPSC amplitude. Mitochondrial calcium sequestration was then inhibited by the addition of 25µM rotenone and 1µM FCCP in the presence of 5µg/ml oligomycin. Under these conditions the magnitude of the test EPSC only recovered to 40±5% (n=3, p<0.01) following the depleting train. The slowing of recovery from depression was reversed by the addition of 1mM EGTA to the presynaptic terminal, implicating presynaptic calcium as a mediator of this effect. Ru360, a specific inhibitor of the mitochondrial calcium uniporter had the same effect on recovery from synaptic depression (recovery to only 51±6%, n=3) as rotenone and FCCP. These data suggest that mitochondrial calcium buffering plays an important role in controlling the rate of recovery from synaptic depression at a mammalian central synapse. Supported by the Wellcome Trust.

S 219 Department of Cell Physiology and Pharmacology, University of Leicester, Leicester LE1 9HN, UK O 37-4 CAPACITANCE MEASUREMENTS FROM A CORTICAL PRESYNAPTIC ELEMENT S. Hallermann, C. Pawlu, P. Jonas, M. Heckmann Improved slice techniques and enhanced optical resolution allow patchclamp recordings from mossy fiber boutons (MFBs) forming en passant synapses on CA3 pyramidal neurons in acute hippocampal slices of rats (Geiger and Jonas, 2000, Neuron 28:927-39). To gain a presynaptic measure for the transmitter release at this synapse we studied the capacitance of MFBs. Square wave and sinusoidal (5 kHz) voltage commands were used. A two compartment model was fitted to the current transients in response to square pulse stimulations. This gave estimates of 1.68 ± 0.17 pF for the capacitance of the MFB and 2.17 ± 0.35 pF for a second compartment connected by a resistance of 231 ± 47 MΩ (n = 19 presynaptic elements; 22-24°C). The size of the first component was roughly consistent with the physical size of the MFBs. Capacitance increases were detected with a sinusoidal voltage command before and after short depolarisations to 0 mV to open presynaptic Ca2+ channels. ∆Cm increased with increasing pulse duration and approached saturation with 30-ms pulses. With a ∆Cm value of about 1 fF with 1-ms depolarisations and an expected capacitance increase of 30 aF per vesicle (assuming 1 µF m-2 specific capacitance and a vesicle diameter of 30 nm) our data indicate that on average 30 vesicles are released per action potential. This is in reasonable agreement with the likely number of release sites and the function of MFBs. Supported by DFG SFB 391 and SFB 505 Physiologisches Institut der Universität Freiburg, Hermann-HerderStr. 7, D-79104 Freiburg, Germany O 37-5 α 1D L-TYPE CA2+ CHANNELS CONTROL EXOCYTOSIS AT THE AFFERENT SYNAPSE OF COCHLEAR INNER HAIR CELLS A. Brandt, J. Striessnig, T. Moser Ca 2+ dependent exocytosis of glutamate mediates synaptic transmission at the inner hair cell (IHC) afferent synapse. Here, we studied the role of L-type Ca2+ channels for synaptic vesicle fusion in IHCs using patch-clamp membrane capacitance measurements. The availability of L-type Ca2+ channels was pharmacologically and genetically manipulated. First, IHCs were stimulated by short depolarizations, which selectively recruit the readily releasable pool of vesicles at the active zone (RRP), in the increasing extracellular presence of the Dihydropyridine (DHP) antagonists nifedipine or isradipine (up to 10µM). The antagonists reduced both Ca2+-influx and release of the RRP down to 15% of control. Exocytosis and Ca2+ influx were inhibited to the same degree over the whole range of blockade. The DHP agonist BayK 8644 increased both Ca2+ influx and fusion of RRP in a similarly linear fashion. Phasic and tonic components of depolarization-induced exocytosis were nearly abolished in IHCs from mice lacking the 1D subunit of Ca2+ channels. However, intracellular release of Ca2+ from cage compounds was still sufficient to trigger robust exocytosis in α1D deficient IHCs. Therefore, exocytosis was compromised mainly by the lack of the stimulating Ca2+ influx through α1D L-type Ca2+ channels. Together, the results demonstrate the large impact of L-type Ca 2+ channels on secretion control at the inner hair cell afferent synapse. The linear relationship between the reduction of Ca2+ influx and exocytosis by DHP antagonists suggests that the Ca2+ concentration at each release site is controlled by very few L-type channels. Tobias Moser, HNO-Klinik, Universität Göttingen, Robert-Koch-Strasse 40 , 37075 Göttingen, Germany O 37-6 TEMPORAL INTEGRATION OF EPSPS IS MODIFIED BY INTERPLAY BETWEEN GIRK CURRENT AND IH C. Alzheimer, T. Takigawa Trains of brief iontophoretic glutamate pulses were delivered onto the apical dendrites of CA1 pyramidal cells at variable frequencies (3 - 100 Hz) to examine how the activation of a G protein-activated, inwardly

rectifying K + (GIRK) conductance alters the postsynaptic processing of repetitive excitatory input. Application of the GIRK channel agonist, baclofen (20 µM), reduced the amplitude of individual glutamateevoked postsynaptic potentials (GPSPs) and attenuated summation of GPSPs so that higher stimulus intensities were required to fire the cell. Notably, GIRK channel activation not only decreased GPSPs, but also suppressed the subsequent afterhyperpolarization (AHP), which arises from a transient deactivation of the hyperpolarization-activated cation current (I h). Voltage-clamp recordings ruled out a direct modulatory action of baclofen on Ih. GIRK channel activation alone does account for AHP suppression, firstly because, with smaller GPSP amplitudes, less I h channels are deactivated resulting in a diminished AHP, and secondly because, owing to its progressive increase in the hyperpolarizing direction, the GIRK conductance shunts a large portion of the remaining AHP. We provide experimental evidence that the suppression of the I h -dependent AHP by GIRK channel activation bears particular significance on the processing of repetitive excitatory inputs at frequencies where the deactivation kinetics of I h exert a prominent depressing effect. In functional terms, activation of GIRK current not only produces a time-independent mitigation of incoming excitatory input, which results directly from the opening of an instantaneous K + conductance, but might also cause a time-dependent redistribution of synaptic weight within a stimulus train, which we link to an interplay with the deactivation of I h. Supported by the DFG (SFB 391, A9). Physiology, University of Munich, Pettenkoferstr.12, D-80336 Munich, Germany O 38-1 CONTROL OF [Cl-]i IN RETINAL ON BIPOLAR CELLS D. Billups, D. Attwell A non-uniform distribution of Cl- transporters, with Cl--accumulating NKCC1 at the dendrites and Cl --extruding KCC2 at the synaptic terminals, has been proposed to generate a gradient of [Cl-]i along ON bipolar cells, allowing GABA to generate a depolarization at the dendrites and a hyperpolarization at the synaptic terminals (Vardi et al., 2000, J. Neurosci. 20, 7657). We used the gramicidin perforated patch technique to measure the reversal potential of GABAA/C receptor mediated currents in ON bipolar cells in rat retinal slices. At a holding potential of -40mV, ECl was only slightly (~ 4mV) more positive at the dendrites than at the synaptic terminal. Altering the membrane potential produced a compensatory shift of ECl with a half time of around 10sec. In the steady state ECl changed by about half the change of membrane potential: around a holding potential of –45mV ECl was similar to the membrane potential, at potentials negative to -50mV ECl was positive to the membrane potential (suggesting control of [Cl]i by NKCC1), while positive to -40mV ECl was more negative than the membrane potential (suggesting control of E Cl by KCC2). Raising [K+]o from 2 to 6mM displaced ECl positive, presumably by inhibiting KCC2 and stimulating NKCC1. These data suggest that the observed transporter distribution in ON bipolar cells has only a small effect on the reversal potential of GABA-evoked currents, and that after synaptic input alters the cell membrane potential adaptation of [Cl -]i will lead to an adaptation of lateral inhibitory signals generated in bipolar cells at the outer and inner plexiform layers. Dept Physiology, University College London, Gower St., London WC1E 6BT, Great Britain O 38-2 EMERGING PRINCIPLES OF INHIBITORY SYNAPTIC SIGNALING IN HIPPOCAMPAL INTERNEURON NETWORKS M. Bartos*, I. Vida+, A. Meyer++, H. Monyer++, J.R.P. Geiger*, P. Jonas* Gamma oscillations (30-100 Hz) are postulated to be important for temporal encoding of information in the brain. Experimental and theoretical studies indicated that networks of mutually connected GABAergic interneurons are involved in the generation of coherent oscillatory activities, where the oscillatory period is a function of the decay time constant of the inhibitory postsynaptic conductance change. However, the functional properties of inhibitory transmission between synaptically connected interneurons in the hippocampus are largely unknown. We have used transgenic mice which express the enhanced green fluorescent protein (EGFP) as a marker for parvalbumin-

S 220 expressing interneurons to record from synaptically connected basket cells (BCs) at 34°C in the dentate gyrus (DG), the CA3 and CA1 hippocampal subfields. We demonstrate, that unitary IPSCs at BC-BC synapses show several similarities in the different hippocampal subregions: (1) IPSCs had rapid decay time constants with slight differences among subregions (DG 2.0 ± 0.2 ms N = 7, CA3 1.1 ± 0.2 ms N = 5, CA1 1.8 ± 0.2 ms N = 5). (2) Paired-pulse depression (PPD) of IPSCs showed slight differences in the extent of PPD among regions (DG 24 ± 4 %, CA3 37 ± 4 %, CA1 39 ± 2 %). (3) IPSCs were largely blocked by 20 µM bicucculine, indicating that they were mediated by GABAA receptors. (4) Unitary IPSCs showed a significantly faster decay time constant at BC-BC synapses than at BC-principal cell synapses, independent of the hippocampal subregions. We conclude that rapid IPSCs are a general functional principle for mutually connected interneuron networks in the hippocampus for the generation of coherent oscillatory activity. *Physiologisches Institut der Universität Freiburg, D-79104 Freiburg, + Anatomisches Institut der Universität Freiburg, D-79104 Freiburg, ++ Institut für klinische Neurobiologie der Universität Heidelberg, D69120 Heidelberg

in which [Cl-]i was maintained by the pipette solution. We excluded receptor desensitization as the cause for the decrement of [Ca2+]i elevations by alternating GABA and glycine applications. Blockade of GABAA receptors, but not of Ca2+ channels prevented the decrement. The decrement could be reversed by sustained cell depolarization. Measurement of a Cl- gradient between soma and dendrites in whole cell recording which provides an assay for net Cl- outward transport (Jarolimek et al., J. Neursoci. 1999, 15: 4695; Kelsch et al., J. Neurosci. 2001, 21: 8339) revealed that transport became active only after the neurons had become unresponsive to GABA in fluorimetric measurement. Activation of L-type Ca 2+ channels by an agonist, however, allowed us to measure Ca 2+ release in neurons which had transport activity. We conclude that the change from depolarizing to hyperpolarizing inhibition is passive in the first phase. [Cl-]i equilibrates according to the increasing negativity of the membrane potential. The pathway for Cl- exchange is provided by voltage and/or Ca2+ dependent Cl- channels in the first few days and by Cl- channels activated by endogenous GABA as soon as GABAergic synapses develop. As the last step hyperpolarizing inhibition is established by K+-Cl- co-transport. Supported by the Deutsche Forschungsgemeinschaft. Institut für Physiologie und Pathophysiologie, Universität Heidelberg, Im Neuenheimer Feld 326, D-69120 Heidelberg, Germany

O 38-3 PHARMACOLOGICAL HETEROGENEITY OF GABA A RECEPTORS IN RAT CEREBELLAR GRANULE CELLS W. Hevers The inhibitory neurotransmitter γ-aminobutyric acid (GABA) mediates most effects through GABA-gated Cl--channels (GABAAR) composed out of five of 19 known subunits (α1-6, β1-3, γ1-3, δ, ε and ρ13, π, θ). Such heteropentameric combination allows for numerous receptor subtypes of which cerebellar granule cells contain a unique subset. To pharmacologically characterise the native receptor subtypes that assemble in these cells from the subunits α1, α6, β2, β3, γ2 and δ, we used whole cell recordings in rat cerebellar slices. A developmental shift of the GABA sensitivity (half-maximal activation, EC50) from 92 µM at postnatal days 8 to 14 (P8-P14) to 12 µM in older animals (P29-P48) was observed. It was accompanied by an increased sensitivity towards the α6-specific inhibitor furosemide, supporting an increased expression of α6 subunits as molecular basis for the observed developmental changes. Cells from animals younger than P15 showed a pharmacology which is in accordance with either α1β2/3γ2 or α6β2/ 3(γ2/δ) containing receptors. The first showed an α1 typical diazepam potentiation and no effects of furosemide and the β-carboline DMCM (methyl-6,7-dimethoxy-4-ethyl-β-carboline-3-carboxylate), the latter no diazepam potentiation but the α6 typical DMCM potentiation and furosemide inhibition. In contrast, cells from animals aged P15 to P48 must be divided into two different groups: one showing both α1 typical diazepam potentiation as well as α6 typical furosemide and DMCM effects, the other in which both DMCM and diazepam effects were absent. Still, these cells were potentiated by loreclezole, suggested to share the same binding motif with DMCM. Accordingly, the pharmacology in older animals (P15-P48) suggests the concurrent presence of α1- and α6-containing receptors in 50 % of granule cells, whereas the remaining 50 % of cells reveal hitherto unknown pharmacological characteristics of cerebellar receptor-subunit combinations in situ. Clinical Research Group, Department of Psychiatry, University of Mainz, Untere Zahlbacher Str. 8, 55131 Mainz O 38-4 MECHANISMS RESPONSIBLE FOR THE DEVELOPMENTAL CHANGE OF GABA RESPONSES U. Misgeld, S. Titz, M. Hans, W. Kelsch, A. Lewen , D. Swandulla In developing neuronal networks GABA responses change from depolarizing to hyperpolarizing. We studied mechanisms involved in this developmental change by fluorimetric measurements and whole cell recording from cul-tured embryonic midbrain neurons. Neurons cultured for 6-10 days respon-ded to exogenous application of GABA (50µM, 15s), muscimol (10µM) or glycine (0.5mM) with a reversible increase in [Ca2+]i. At day >24 neurons were unresponsive. Pharmacological blockade of GABAA receptors revealed synchronous bursts already at day 12. During repetitive applications of GABA the response declined in fluorimetric measurements, but not in whole cell recording

O 38-5 DEPOLARIZING GLYCINE RESPONSES IN CAJAL-RETZIUS CELLS OF NEONATAL RAT CORTEX W. Kilb, M. Ikeda, K. Uchida, A. Okabe, A. Fukuda, H.J. Luhmann Cajal-Retzius cells (CRc) are among the first neurons appearing during corticogenesis and play an essential role in cortical lamination. In order to elucidate the role of glycine-receptors during cortical development we investigated whether CRc express glycine receptors and examined the properties of glycine-induced responses. For this purpose whole-cell and gramicidin-perforated patch-clamp recordings and microfluorimetric measurements of intracellular Cl- concentration ([Cl -] i ) using the Cl - sensitive dye 6-methoxy-N-ethylquinolinium (MEQ) were performed on visually identified CRc in tangential slices of neonatal rat cortex (postnatal days 0-3). Bath application of the glycinergic agonists glycine, β-alanine and taurine induced an inward current with maximal peak amplitudes of 364±22.4 pA (n=19), 217±148.2 pA (n=9) and 66.2±52.7 pA (n=11) and with an EC 50 of 0.54 mM, 1.62 mM and 2.41 mM. The membrane responses induced by 1 mM glycine were inhibited by the glycinergic antagonists strychnine (30 µM) and phenylbenzene-ω-phosphono-α-amino acid (100 µM), while the GABAA antagonist bicuculline (100 µM) and the NMDA antagonist (±)-2-amino-5-phosphonopentatonic acid (60 µM) were without effect. The reversal potential of glycine-induced currents revealed a strong [Cl-]i dependency and stimulation of glycine receptors with taurine induced an increase in MEQ-fluorescence, corresponding to a [Cl-]i decrease and thus reflecting Cl- efflux. Perforated-patch recordings from CR cells demonstrated that glycine depolarized the membrane by 19.8±6.3 mV (n=7), which in 4 of these cells was sufficient to elicit action potentials. Glycine-induced currents reversed under perforated-patch conditions at -42.7±12.1 mV (n=7), corresponding to a [Cl-]i of about 28 mM. In summary, these results suggest that CRc of neonatal rat cerebral cortex express functional strychnine-sensitive glycine receptors that mediate depolarizing membrane responses via Cl- efflux. Supported by DFG grants Lu 375/3-3 and 446JAP-111/1/00 to HJL and JSPS-DFG Cooperative Research Grant #13210065 to AF. Institute of Neurophysiology, Heinrich-Heine-University Düsseldorf, POB 101007, D-40001 Düsseldorf, Germany O 38-6 PRESYNAPTIC INHIBITION VIA GLUTAMATERGIC AUTORECEPTORS IS PREVENTED BY SUPPRESSING CL- CURRENT THROUGH THE ACTIVATED GLUTAMATE TRANSPORTER J. Dudel, M. Schramm Micromolar concentrations of glutamate (glu) as well as hundred times higher ones of aspartate (asp) applied through the perfused macropatch electrode to crayfish neuromuscular junctions inhibit depolarisation pulse elicited release of quanta of glu (Parnas et al., 1996, Europ. J. Neurosci. 8, 116-126; Dudel & Schramm, 2001, Pflügers Arch. 441,

S 221 Suppl. 6, P 18-3). In addition, they lengthen the decay of quantal excitatory postsynaptic currents, binding to the glu transporter and competing with the removal of released glu. Surprisingly, pharmacological block of glu transporters also blocks the presynaptic inhibition by glu or asp. Activated glu transporters are known to open an associated Cl- channel, generating a Cl- current that is much larger than the electrogenic pump current (Otis & Jahr, 1998, J. Neurosci. 18, 70997110). A Cl - current should hyperpolarize the nerve terminal, the higher driving potential for Na+ assisting in the glu transport. Hyperpolarization would also inhibit release. To test whether this was the basis of release inhibition via autoreceptors, the glu transport associated Cl- current was suppressed by replacing Cl- in the electrode perfusate by propionate. Replacement caused a transient phase of increased spontaneous and elicited release presumably due to some depolarisation of the terminal. For testing the effects of glu or asp, the pulse amplitude had to be reduced, therefore. With the control, Cl- containing perfusate, added glu or asp reduced the rate of elicited quantal release per pulse, m, to less than half. When, in the same electrode position, the perfusate was switched to Cl--free, release inhibition by glu or asp was abolished; in most terminals m even increased significantly, unmasking a facilitation of release by the transmitter seen sometimes under other conditions (Schramm & Dudel, 1997, Neurosci. Lett. 234, 31-34). On switching back to Cl- containing perfusate glu and asp inhibited release again. It appears that in the crayfish synapse autoinhibition of release employs Cl- currents associated with glu transport. Supported by the Deutsche Forschungsgemeinschaft, SFB 391. Institut für Physiologie der LMU, Biedersteinerstr. 29, D-80802 München O 39-1 PROBLEM SOLVING INSTRUCTION IN PHYSIOLOGY (PSIP): A 2 YEAR (4 SEMESTER) COURSE SYSTEM IMPLEMENTED UNDER CURRENT LEGAL BOUNDARY CONDITIONS FOR MEDICAL EDUCATION IN GERMANY H. Schmid-Schönbein, A. Lückhoff, Chr.Fahlke, K. Mottaghy, M. Baumann The preclinical medical training in the Federal Republic of Germany is characterised by boundary conditions unknown elsewhere, in that 1. the secondary education in the natural sciences is largely missing and 2. large numbers of first year students are admitted to the publically financed faculties with small number of permanent teaching positions. Our program aims at resolute remedies for the lack of scientific basis during the 1st semester, while a priori focussing on the comprehension of the foundations of systemic physiology the 4th semester. Local admission regulations restricted to the autumn make it possible to synchronise instruction on a yearly scheme. PSIP is based on early confrontation with highly abstract concepts, challenging the ability of students to obtain advanced scientific information. Besides the compulsary training in physics, chemistry and biology (still under the responsibility of the basic science faculty) our department confronts the students in their 1st semester with a lecture series on propedeutic physiology (physics of living systems) and on cell biology taught in cooperation with biologists, cytologists and biochemists. In the former a voluntary, in the latter a compulsary written examination is prerequisite for the admission to laboratory courses and seminars starting in the 2nd and extending to the 4th semester. Already during the first half of the 2nd semester, PSIP focuses on firm establishment of cellular and molecular physiology, requiring compulsary seminars on electrophysiology and transepithelial transport. The second half of the 2nd semester presents fundamental aspects of specific organ functions (cardiovascular, pulmonary, intestine, renal and peripheral nervous system). A tightly blocked course (extending from early October to late November) in the 3 rd semester teaches systemic physiology by frontal lectures, laboratory course and seminars with oral student presentations. Interdisciplinary, clinical and advanced system physiological instruction is offered in the 3rd and 4th semester, provoking patient associated „problem solving“. In cooperation with many clinical departments, detailed high-tech-information is presented to deepen anatomical, physiological and pathophysiological competence of students. PSIP is accompanied by various types of computer based and free examinations (see contribution BAUMANN et al.), but is culminating in an obligatory essay test (late in the 4th semester) in which each student is confronted with the task of answering two out of three medically relevant physiological questions and justify the answers (max. 5 lines) by concise theory (25 lines).

Department of Physiology, RWTH Aachen, Pauwelsstrasse 30, D 52074 Aachen, [email protected] O 39-2 TEACHING PRINCIPLES OF NEURONAL EXCITABILITY IN VIRTUAL COMPUTER LABORATORIES (CLABS-NEURON) H.A. Braun, H. Schneider, B. Wollweber, K. Voigt The understanding of dynamical biological functions can essentially be improved with interactive computer animations and simulations where the user can immediately see the effects of changed control parameters and can directly follow the time-dependencies of the dynamic processes. This is the aim of ”cLabs-Neuron”, which provides computer animations and simulations for teaching basic neurophysiology (see www.clabs.de). The programs illustrate the functional membrane properties (bilipid layer and ion channels) and also explain their electrical equivalents (membrane capacitance and conductance). Experiments can be done, for example, with voltage recordings across a simple RC-circuit (preselectable resistance and capacitance) in response to a variable number of current pulses of user-defined amplitude, duration and delay. Another module can be used to examine how the membrane potential changes with alterations of ionic conductances (potentiometer sliders). The user also can control the gating of voltage dependent K+- and Na+- channels (one and two gates). Single channel currents can be recorded in a virtual patchclamp laboratory where stimulus and also membrane parameters (e.g. holding and command potential, ionic concentrations, maximum conductances) can be changed. In parallel it is illustrated how singlechannel currents through a variable number of ion channels are summing up to whole-cell currents. A series of computer animations explain the concepts of voltage-clamp recordings. Current- and voltage-clamp experiments can be done in a virtual laboratory with different types of neurons. The neurons are simulated by Hodgkin-Huxley-type algorithms including simplified versions of pacemaker-neurons, subthreshold oscillators or slow-wave bursters. Modifications of the neurons kinetics are allowed and the parameters can be saved so that the user can develop his/her ”personal” neurons. Supported by BM&T, Heidelberg/ Marburg; DAQ-Solutions, Lohra; TransMIT, Giessen. Institute of Physiology, Deutschhausstr. 2, D-35037 Marburg, Germany

O 39-3 A 9-WEEK INTEGRATIVE AND PROBLEM BASED COURSE ”PATHOMECHANISMS” FOR 3RD YEAR MEDICAL STUDENTS A. Deussen, G. Baretton, O. Tiebel, I. Nitsche, A. Heintz Classical teaching contents and methods do not well conform with the rapidly changing concepts in modern medicine. Moreover, the ability of self-guided problem solving is only poorly represented in the traditional curriculum. The major goal of the present curriculum reform was to better prepare medical students for these tasks. In addition, it was aimed to improve the transmission of concepts of structurefunction relationships during pathological conditions and to teach students a critical use of cause-effect concepts. To achieve these goals elements of problem based learning were combined with elements of classical teaching. In this hybrid learning/teaching model the subjects pathology, pathophysiology, pathobiochemistry and clinical chemistry commit to focus their respective lectures on weekly defined topics. In addition to 12 h lectures per week small group tutorials (8-10 students per group) 3 times per week (6 h) use assays of clinical cases. Practical courses of pathology and clinical chemistry add to the structured curriculum while a considerable fraction of time was set aside to permit self-guided learning by the students. The curriculum reform was accompanied by extensive, independent evaluations on the student and the tutor side. Students and tutors have overwhelmingly welcomed the new curriculum. Students have positively remarked on the contents, the motivational strength and the high information density of the curriculum. The time devoted to self-guided studies significantly increased. Furthermore, students felt that their ability to deal with new contents improved. The rating of the assays of the clinical cases was similar by students and tutors. The tutors estimated to have had an important increase in interdisciplinary knowledge while preparing for and facilitating the tutorials. The results of the course examination (oral) was acceptable. Results in the federal examination (multiple choice) did not differ from previous examinations. Thus, the

S 222 restructured curriculum permits in a well-timed, but staff demanding course learning of important contents of pathomechanisms. The new curriculum offers motivational strength to both, students and tutors. Institut für Physiologie, Medizinische Fakultät Carl Gustav Carus, TU Dresden, Fetscherstr. 74, 01307 Dresden. O 39-4 A DIET ENRICHED IN FAT DOES NOT CHANGE THE EXPRESSION OF RESISTIN V. Hulstroem, K. Shultze, J. Vinten Recently a new adipocyte hormone, resistin, has been identified. Elevated resistin level has been shown to cause insulin resistance in mice. To investigate the possible role of resistin for dietary induced insulin resistance, we studied the expression of resitin mRNA in fat tissue from lean rats rendered insulin resistant by a high fat diet. Twelve male Wistar rats (200-250g) fed a diet with 60% lard (HF) gained 42g and 12 control rats fed a standard chow with 5% fat (CT) gained 142g in weight during 7 weeks. Fasting levels of glucose and insulin were not different between groups. Thirty minutes after an intraperitoneal injection of a mixture of insulin (100 mU/7kg) and glucose (2g/kg), the insulin levels were elevated but not different between groups, whereas the glucose levels in HF rats were significantly elevated compared to CT. Fat pad weight was lower in HF rats. Interestingly, the mean fat cell diameter did not differ between groups. The expression of resistin and leptin mRNA was quantified by means of a specific 32P-labeled probe in Northern Blots of total RNA isolated from fat fads. Leptin mRNA expression was markedly reduced in the insulin resitant HF rats. Furthermore we observed an unaltered resistin mRNA expression. We conclude that in the present lean fat fed rats it is unlikely that resistin mediates insulin resistance. Department of Physiology, University of Copenhagen, The Panum Institute, Blegdamsvej 3, DK-2200 Copenhagen N

O 39-6 BRAIN-GUT AXIS IN GASTROPROTECTION BY LEPTIN AND CHOLECYSTOKININ AGAINST ISCHEMIA-REPERFUSION (I/R) AND ETHANOL INJURY T. Brzozowski, PC. Konturek, S. Kwiecieñ, D. Drozdowicz, A. Ptak, S.J. Konturek, E.G. Hahn Objectives: Leptin, a product of ob gene controlling food intake, has recently been detected in the stomach and implicated in gastroprotection against various noxious agents but it is unknown whether centrally applied leptin influences I/R-induced gastric lesions. Methods: We compared the effects of leptin and CCK-8 applied i.c.v. or i.p. on gastric mucosal lesions induced in rats by I/R or 75% ethanol with or without A) vagotomy, B) suppression of NO-synthase with LNNA (20 mg/kg i.p.), C) capsaicin (CAP)-induced denervation of sensory nerves (125 mg/kg s.c.) and D) inhibition of CGRP receptors with CGRP8-37 (100 µg/kg i.p.) applied with or without leptin or CCK8 applied i.c.v. Area of gastric lesions was measured by planimetry and the gastric blood flow (GBF) was assessed by H2-gas clearance technique and plasma leptin and gastrin levels were determined by RIA. Results: Leptin (0.1-20 µg/kg i.p. or 100-625 ng/kg i.c.v.) dosedependently attenuated 75% ethanol and I/R-induced gastric lesions and these protective effects were similar to that obtained with CCK-8 and accompanied by a significant increase in GBF and plasma leptin and gastrin levels. CCK-8 enhanced plasma leptin levels but failed to affect plasma gastrin levels. These protective effects were completely abolished by vagotomy and significantly attenuated by sensory denervation with capsaicin or by CGRP8-37. L-NNA reduced significantly the protective and hyperemic effects of CCK but not those of leptin while CAP-denervation counteracted leptin-induced protection and hyperemia but attenuated significantly those of CCK. Conclusion: Central leptin exerts a potent gastroprotective activity against I/R-and ethanol-induced gastric lesions and this may depend upon vagal activity and sensory nerves and involves hyperemia mediated by NO. Department of Physiology, Jagiellonian Univ. Sch. Med., 31-531 Cracow, 16 Grzegorzecka Str., POLAND

O 39-5 K+ AND Cl- CHANNELS MEDIATE SHRINKAGE DURING STIMULATION OF ACID SECRETION IN CULTURED PARIETAL CELLS. O. Bachmann, A. Heinzmann, A. Mack, M. Gregor, U. Seidler Background & Aims: We have recently shown that in isolated and cultured parietal cells, a rapid initial cell shrinkage occurs upon stimulation of acid secretion, followed by a regulatory volume increase (RVI), which is mainly mediated by Na+/H+ and Cl-/HCO3- exchange. The factors leading to the initial cell shrinkage are unknown. We hypothesized that either activation of basolateral K+ channels and/or of the apical K+ and Cl- conductances involved in acid secretion cause this cell shrinkage via KCl and volume loss. Methods: Volume changes in isolated and cultured rabbit parietal cells were monitored by confocal measurement of the changes in calcein concentration in the cytoplasm. Acid secretion was quantified using the 14C aminopyrine uptake method. Results: Inhibition of basolateral K + channels with barium led to a rapid cell swelling, indicating that Ba2+ sensitive K+ channels are also open in resting parietal cells. The chromanol 293b, which inhibits the presumably apically located K+ channel KvLQTl, did by itself not induce cell swelling, but reduced the forskolin-induced cell shrinkage to a higher extent (69%) than the carbachol-induced shrinkage (44%). Charybdotoxine (ChTX), which has been shown to inhibit Ca2+ sensitive K + channels, equally did not induce cell swelling, blocked the carbachol-induced cell volume decrease by 77%, and had no effect on forskolin-induced shrinkage. The Cl -channel blocker NPPB also had no effect on cell volume in resting cells, but almost completely inhibited stimulation-associated cell shrinkage. Inhibition of acid secretion by 293b, ChTX and NPPB followed the same pattern as inhibition of the initial cell volume loss after stimulation. Conclusion: Cell shrinkage associated with acid secretion is dependent on the activation of K+ and Cl- channels by the respective secretagogues. Recently published data suggest that these channels may be at least in part located apically. K+, Cl- and water secretion into the secretory canaliculi are thus one likely mechanism of stimulation-associated cell shrinkage in cultured parietal cells. 1st Med. Dept., University Hospital Schnarrenberg, Otfried-MüllerStr. 10, 72076 Tübingen

P 01-1 MECHANISMS OF ClC-1 DYSFUNCTION INDUCED BY DOMINANT MYOTONIA CONGENITA MUTATIONS A.M. Ryan1, F-F. Wu2, E.P. Hoffman2, C. Fahlke3, R. Rüdel1 Myotonia congenita is an inherited disorder of skeletal myofibre excitability due to mutations in CLCN1, the gene encoding the human skeletal muscle chloride channel hClC-1. The study of dominant (Thomsen) mutations has provided valuable insight into channel structure and function. In this work we sought to characterise electrophysiologically a number of new (T310M, L283F, F428S) and recently described (P480T, A218T) dominant mutations using a mammalian expression system. Novel mutations were identified by screening all 23 exons of CLCN1 in 88 unrelated myotonia patients. All mutations were introduced into the expression vector pRc/CMV-hClC1 by site-directed mutagenesis. Mutant channels were transiently expressed in tsA201 cells. Standard whole-cell recordings were performed 24-36h later. Compared to WT, deactivation of L283F and T310M was less complete. L283F shifted the voltage-activation curve to more positive potentials under standard conditions. This shift was exaggerated at a physiological intracellular Cl- concentration. T310M (which only became symptomatic during pregnancy) and P480T also exhibited a rightward shift in the voltage-activation curve but only when studied at physiological [Cl]i. P480T additionally displayed an altered anion permeability sequence: I->NO3->Cl->Br-. A218T closely resembled WT with a minimal shift in voltage activation at physiological [Cl] i. The properties of F428S were indistinguishable from those of WT but channel expression was significantly reduced. To conclude, the number of mutations identified in Thomsen’s disease and the functional alterations they induce are increased. The study of ClC-1 mutations at a physiological [Cl]i is important for understanding channel dysfunction and the mammalian system provides a faithful tool for the study of these mutations. 1

Abt. für Allgemeine Physiologie, Universität Ulm, 89069 Ulm, Germany, 2 Dept. of Human Genetics, University of Pittsburgh, Pittsburgh, PA,USA, 3Institute für Physiologie, RWTH Aachen, Pauwelsstr. 30, 52057 Aachen, Germany

S 223 P 01-2

P 01-4

THE MYOTONIA CONGENITA MUTATION A331T CONFERS A NOVEL HYPERPOLARIZATION-ACTIVATED GATE TO THE MUSCLE CHLORIDE CHANNEL CLC-1 M. Warnstedt1, S. Chen2, L. Tranebjærg2,3, Ch. Fahlke1 Myotonia congenita is an inherited disorder of skeletal muscle excitability leading to a delayed relaxation following muscle contraction. We have identified a novel mutation found in a patient with recessive generalized myotonia (Becker). This mutation predicts the substitution of an alanine by threonine at position 331 located between the sixth and the seventh transmembrane domain of the human muscle chloride channel, hClC-1. Introduction of the A331T mutation into hClC-1 confers a novel slow gate that activates upon membrane hyperpolarization and closes at positive potentials. This novel gate interacts with fast opening and closing transitions that are common to WT and mutant channels. Under conditions at which the slow gate is not activated, i.e. a holding potential of 0 mV, the typical fast depolarization-induced activation gating of WT hClC-1 was only slightly affected by the mutation. In contrast, A331T hClC-1 channels with an open slow gate display an altered voltage dependence of open probability. These novel gating features of mutant channels produce a decreased open probability at –85 mV, the normal muscle resting potential, leading to a reduced resting chloride conductance of affected muscle fibers. The A331T mutation causes an unprecedented alteration of ClC-1 gating and reveals novel processes defining transitions between open and closed states in ClC chloride channels.

ANION PERMEATION IN HUMAN ClC4 CHANNELS S. Hebeisen, Ch. Fahlke ClC-4 and ClC-5 are two mammalian isoforms of the ClC family of voltage-gated chloride channels with unique ion conduction and gating properties. A characteristic feature of both isoforms is the pronounced rectification, with very small current amplitudes in the negative range and prominent rapidly activating current responses upon depolarization. Neither the underlying mechanisms nor the physiological impact of this rectification are currently understood. We studied pore properties of human ClC-4 channels using whole-cell recordings on tsA201 cells transfected with pRcCMV-hClC-4. Non-stationary variance analysis demonstrated that the unitary conductance is very small, i.e. 1.6 pS at +140 mV under symmetrical SCN- and 1.3 pS for external Cl - and internal I -. The single channel conductance is pronouncedly voltage-dependent with smallest values at negative potentials that increase with membrane depolarization. While the absolute open probability changes only little with voltage, the rectification of the single channel amplitude fully accounts for the observed macroscopic rectification. This result demonstrates that ClC-4 channels are open over the whole voltage range and capable to mediate anion flux under physiological conditions. Pore properties of hClC-4 differ from other ClC channels not only in the voltage dependence of conductance, but also in anion selectivity and block by certain anions. The relative permeability for external anions decreases in an order of SCN- > NO3> N3- > I- > F- > Cl- > Br- > HCO3- > CH3COO- > MES > Glutamate. Large and polyatomic anions do not block Cl- currents through hClC-4. The distinct features of anion permeation through hClC-4 as compared to other anion channels can be explained by assuming differences in the size of the pore narrowing and different affinities of the binding sites within the ionic pore.

1 Institut für Physiologie, RWTH Aachen, 2Department of Medical Genetics, University Hospital of Tromsoe, 3Institute of Medical Biochemistry and Genetics, University of Copenhagen and Department of Audiology, Bispebjerg Hospital, Copenhagen

Institut für Physiologie, RWTH Aachen, Pauwelsstr. 30, 52074 Aachen

P 01-3 P 01-5 ISOFORM-SPECIFIC PORE PROPERTIES OF CLC CHANNELS S. Hebeisen, S. Voswinkel, B. Poser, Ch. Fahlke The ClC-family is a large gene family encoding voltage-gated anion channels. The various ClC isoforms do not share a high degree of sequence identity and display a wide variety of functional properties. Nevertheless, it appears reasonable to assume that basic mechanisms underlying ion conduction and selectivity are conserved in all ClC isoforms. To identify these mechanisms and their molecular determinants, we studied possible molecular determinants of isoformspecific variations of two ClC isoforms, ClC-1 and ClC-4. Both isoforms select between ions by selective binding to several low-affinity anionselective binding sites with a lyotropic binding selectivity. The reason for the observed differences in anion selectivity of ClC-1 and ClC-4 is the absolute value of the interaction energy between binding site and anion. In ClC-1, anions are bound with higher affinity; the rate-limiting step in anion permeation is the dissociation of anions from one of these sites. Large and polyatomic anions block anion current and are less permeant. In ClC-4, not the dissociation, but the association of permeant anions to one of these binding sites is rate-limiting, resulting in a better permeability of anions that bind with higher affinity. We recently identified a highly conserved eight amino acid stretch (P1 region) within the D3-D5 region that lines the most narrow part of the ClC-1 pore. Cysteine substitutions of the P1 region of hClC-4 also affect anion permeation, but with a periodicity that is distinct from the results of ClC-1. The two positively charged residues, that project into the pore of ClC-1, are not accessible in ClC-4. This changed conformation of the P1 region will cause a decreased interaction energy with anions and is therefore a likely reason for the different anion selectivity of ClC-4. The functional evaluation of several chimeras between the two isoforms indicated important allosteric interactions between various parts of the ClC channel protein. For example, while the exchange of the whole D3-D5 region from hClC-4 to hClC-1 caused little changes of channel gating and permeation, the transplantation of roughly half of this region has dramatic effects on both functional properties. Such allosteric interaction may be responsible for a distinct conformation of highly conserved domains in ClC-1 and ClC-4. Institut für Physiologie, RWTH Aachen, Pauwelsstr. 30, 52074 Aachen

SELECTIVITY MODULATION OF RECONSTITUTED ICln L.Garavaglia, M. Ritter, J. Fürst, M. Jakab, S. Rodighiero, C. Bertocchi, R. Manfredi, J. Provenzano, C. Bazzini, G. Botta’, G. Meyer, M. Paulmichl Bilayer reconstitution experiments (macrobilayer and tip-dip) confirmed that the protein ICln is an ion channel with characteristics resembling those of the swelling activated channel RVDC (Regulatory Volume Decrease Channel) (Fürst et al., Pfluegers Arch. 2000; 440: 100,155). In diphytanoylphosphatidylcholine (DiphPC) bilayers the selectivity (PK/PCl) of reconstituted ICln was 4.18±0.66 (n=32) in the presence of 10/150 mM KCl solution (pH 8), larger than the one observed for RVDC in the cells. As observed for several proteins whose activity is affected by modifications in the lipid environment, also the ICln P K/P Cl is strongly influenced by bilayer physicochemical characteristics. In the presence of phosphatidylcholines with saturated linear acyl side chains of different increasing length (DLPC: dilauroylphosphatidylcholine, DMPC: dimyristoilphosphatidylcholine, DPPC: dipalmitoylphosphatidylcholine, DSPC: distearoylphosphatydilcholine), the ICln PK/PCl was shifted toward K+ or Cl respectively by an increase or a decrease in the acyl side chain length. In addition, the permeability of ICln was significantly lowered by the presence of Ca2+ in every condition tested. The observable activity of the channel was higher if the protein was reconstituted in bilayers constituted of the shorter lipids. If reconstituted in heart polar lipid extract bilayers, in asymmetrical (10/150 mM) KCl solutions at pH 7, the PK/PCl of ICln was 3.75±0.65 (n=10) and was shifted to 1.32±0.23 (n=7) in presence of 2 mM Ca 2+ . When the pH of the solutions was reduced from 7 to 6 in presence of 2 mM Ca2+ or 2 mM Mg 2+, the ICln engendered current turned chloride selective (PK/PCl = 0.47±0.07 (n=6); P K/P Cl = 0.85±0.15 (n=6) with Ca 2+ and Mg 2+ respectively). In conclusion, the selectivity of reconstituted ICln channel is dependent on different factors, like lipid composition of the membrane, pH and presence of Ca 2+ and Mg 2+ in the solutions facing the bilayer. Dipartimento di Fisiologia e Biochimica Generali, Sezione di Fisiologia molecolare e dei Trasporti, Universita’ degli Studi di Milano, Via Celoria 26, I-20133 Milan, Italy

S 224 P 01-6

P 01-8

HYPOTONICITY INDUCED CYTOSOL TO MEMBRANE SHIFT OF THE ICln ION CHANNEL AND ACCELERATION OF REGULATORY VOLUME DECREASE CURRENTS BY REPEATED CELL SWELLING C. Burtscher, M. Jakab, M. Gschwentner, A. Ravasio, A. Laich, S. Chwatal, J. Schirmer, S. Eichmüller, J. Fürst, S. Dossena, M. Ritter, M. Paulmichl The ICln protein is essential for cell volume regulation and functions as an ion channel with features resembling those of regulatory volume decrease currents (RVDC). We investigated the cytosol vs. cell membrane distribution of ICln under isotonic and hypotonic conditions by Western blotting in NIH 3T3 fibroblasts and MDCK cells. The bulk of ICln is located in the cytoplasm, but a substantial fraction resides in the cell membrane. Within 15 min after cell swelling the cytosolic fraction decreases whereas the membrane fraction increases significantly. SCN- quenching of MEQ fluorescence was performed to estimate the anion permeability of 3T3 fibroblasts. Under hypotonic conditions the rate of quenching is markedly enhanced. Addition of purified ICln protein to the extracellular solution enhances the quenching rate under isotonic conditions similar to cell swelling. To determine the electrophysiological responses of repeated hypotonic challenges, measurements of the cell membrane potential and whole cell RVDC were performed (MDCK cells and 3T3 cells). MDCK cells depolarize by 0.26±0.05 mV/sec (n=6) and by 0.63±0.12 mV/sec (n=5) upon first swelling (omission of 70 and 150 mosM, resp.) and by 0.54±0.13 mV/ sec (n=6) and by 0.79±0.11 mV/sec (n=5) following second swelling (omission of 70 and 150 mosM, resp.). Similarly in 3T3 fibroblasts RVDC develops more rapidly if cells are swollen for a second time: 100 sec after onset of cell swelling the current measured at +40 mV is 0.17±0.03 nA (n=13) and 0.40±0.07 nA (n=13) upon first and second swelling, resp. The maximum currents and the reversal potentials were not significantly different under either condition. In conclusion cell swelling fosters the transposition of ICln from the cytosol to the membrane where it can act as ion channel. Repeated swelling of cells leads to an accelerated activation of RVDC probably due to enhanced abundance of ICln.

THE ROLE OF ICln IN EXERCISE INDUCED ASTHMA S.Eichmüller, S. Hofer, A. Schmarda, E Scandella., J. Fürst, M. Jakab, C. Sironi, M. Ritter, M Paulmichl The aim of the project is to investigate the involvement of the chloride channel ICln in the pathogenesis of exercise induced asthma (EIA). EIA describes the transient airway obstruction after vigorous exercise in cold and dry air, which can also be provoked by inhalation of hyperosmolar aerosols. The mechanisms whereby hyperosmolarity of the airway surface liquid leads to airway narrowing remains unclear. By using the whole-cell patch-clamp method, we could show that in selected patients suffering from EIA, the swelling dependent chloride channel ICln is more active compared with that of healthy control individuals. We assume that these hyperreactive chloride channels could be one possible cause for the hyperresponsiveness of the bronchial system in asthma patients, since the blockage of these channels by cromolyns shows positive effects in asthma therapy. CLNSIA is the human gene coding for ICln. It consists of 6 exons and is located on chromosome 11q 13.5-14.1, which has also been identified by means of positional cloning in asthma patients. The target of our study is to compare cDNA as well as genomic DNA from selected asthmatics and normal individuals. We isolated total RNA and genomic DNA from blood or fibroblasts in primary cultures of children that suffer from EIA and normal individuals. With the PCR technique we amplified each exon and 700 nucleotides of the promoter of the CLNS1A gene. By using M13-tailed primers for the PCR we could sequence each exon directly from the PCR-product. We identified a mutation near the exon4-intron4 boundary, which occurs in nearly 50% of the analysed asthmatics but only in about 20% of normal individuals. Our next target is to figure out, whether this mutation has any effect on the quantity or quality of the expressed mRNA of ICln. Department of Physiology, Fritz Pregl Str. 3, University of A-6020 Innsbruck, Austria

P 01-9 Institute of Physiology, Fritz Pregl Str. 3, University of A-6020 Innsbruck, Austria

PROBING THE CELL SWELLING INDUCED CYTOSOL TO MEMBRANE SHIFT OF THE ICLN ION CHANNEL PROTEIN BY FLUORESCENCE RESONANCE ENERGY TRANSFER (FRET) A. Ravasio, M. Ritter, S. Chwatal, M. Jakab, S. Eichmüller, C. Burtscher, J. Fürst, M. Paulmichl The ICln protein, which is essential for cell volume regulation, has been shown to form ion channels and the current elicited by purified ICln reconstituted in lipid bilayers has remarkable features in common with the anion current induced after cell swelling. Under isotonic conditions the bulk of the ICln protein in native cells is located in the cytosol and only a minor fraction is associated with the cell membrane. We investigated the transposition of ICln into the cell membrane using fluorescence resonance energy transfer (FRET) between the cell membrane and ICln. To this end NIH 3T3 cells were cotransfected with plasmids encoding a fusion protein of cyano fluorescent protein and ICln (CFP-ICln) as donor and a yellow fluorescent protein bearing a farnesylation motive (YFP-Mem) as acceptor. YFP-Mem is selectively sorted to the cell membrane. Analysis was performed by fluorescence imaging and signals were corrected for CFP-YFP crosstalk and normalized to plasmid expression efficiency. Clear FRET from CFP-ICln to YFP-Mem can be detected in the cells under isotonic conditions. Upon cell swelling (reduction of 70 mosM) FRET signals increased after 20 min by 35±9% (n=12) at extracellular pH of 7.4. At pH 6.4 FRET decreased after 10 min by -6.7±0.4% (n=11) but still increased after 20 min by 10.7±0.5% (n=11). This is in agreement with results obtained by Western blot analysis showing a significant increase in the membrane to cytosol ratio of ICln protein upon cell swelling. Thus cell swelling fosters the transposition of ICln from the cytosol to the cell membrane.

THE REGULATORY VOLUME DECREASE ANION CURRENT IN EMBRYONIC CELLS OF THE NEMATODE CAENORHABDITIS ELEGANS M. Jakab, M. Gschwentner, J. Fürst, J. Rudzki, J.G. Danzl, C. Burtscher, M Ritter., M. Paulmichl In Caenorhabditis elegans we have recently identified two splice variants of the ICln ion channel protein, termed IClnNl and IClnN2. The latter one is characterised by an additional string of 20 AA’s (encoded by exon 2a) adjacent to the inner mouth of the putative channel pore and we have shown that these AA’s are responsible for inactivation of ICln channels reconstituted in lipid bilayers (Fürst et al., JBC 2001, in press). To investigate the role of the two splice variants for the swelling activated regulatory volume decrease current (RVDC) in native cells, RVDC was measured in freshly prepared C. elegans embryonic cells as well as in NIH 3T3 fibroblasts transfected with IClnNl. Singlecell RT-PCR confirmed the presence of both transcripts in the investigated embryonic cells; in these cells hypotonic stimulation (reduction of extracellular osmolality by 100 mosmol/kg for 15.7±2.7 min; n=6) was followed by the activation of an inwardly rectifying, noninactivating anion current (-0.34±0.08 nA at -120 mV, 0.08±0.02 nA at +100 mV; n=6) sensitive to the chloride channel blocker DIDS. Inward rectification and lack of inactivation are in contrast to RVDC found in virtually all other cells investigated so far, in which hypotonicity elicits outwardly rectifying and inactivating currents. In NIH 3T3 fibroblasts transfected with an IRES vector containing the sequences coding for IClnNl and GFP, hypotonic cell swelling leads to the development of an outwardly rectifying current with a markedly higher amplitude compared to untransfected control cells. In contrast to control cells, RVDC in fibroblasts overexpressing IClnNl were reduced by intracellular application of a peptide composed of the first 10 of the 20 additional AA’s of IClnN2 which carries 5 consecutive positively charged AA’s. Together with our previous findings this points to the ability of the 5 positively charged AA’s of IClnN2 exon 2a to act as a blocking particle for RVDC and that exon 2a may be responsible for the unique properties of RVDC in C. elegans embryonic cells.

Institute of Physiology, Fritz Pregl Str. 3, University of A-6020 Innsbruck, Austria

Department of Physiology, Fritz Pregl Str. 3, University of A-6020 Innsbruck, Austria

P 01-7

S 225 P 01-10

P 01-12

CELL VOLUME- AND REDOX-SENSITIVE INWARDLY RECTIFYING ANION CHANNELS IN PLASMODIUM FALCIPARUMPARASITIZED HUMAN ERYTHROCYTES E. Shumilina, C. Duranton, V. Tanneur, G. Henke, P. Kremsner, F. Lang, S.M. Huber Infection of human erythrocytes by the malaria parasite Plasmodium falciparum activates endogenous inwardly and outwardly rectifying anion channels in the erythrocyte membrane. These channels which differ in their voltage-dependence and pharmacology probably supply the parasite with nutrients and dispose of waste products from the host cytosol. The latter diminishes the volume stress and, therefore, prevents swelling and hemolysis of the host. To study whether the infection-induced channels are directly regulated by cell volume, we applied the continuous whole-cell recording of infected and control cells during cell shrinkage and swelling. Iso-ionic increase of bath osmolarity from 300 to 450 mosM (NaCl in bath and pipette) decreased inward current from Ginward = 9.4 ± 1.8 nS to 4.6 ± 1.2 nS (n = 11) within 3 min of incubation. Iso-ionic re-decrease of bath osmolarity to 200 mosM induced complete recovery and further activation of the inward current (to Ginward = 11.8 ± 1.7 nS). The current fraction that inactivated/ activated exhibited the current-voltage relationship of the inwardly rectifying current and activated time-dependently at negative voltage. The reversal potential of this volume-sensitive current fraction changed with Cl- equilibrium potential to -43 ± 7 mV (n = 7) when in unpaired recordings Cl- of the pipette solution was reduced to 10 mM indicating anion selectivity. Applying the reducing agent (DTE, 100 µM) in the bath induced a run-down of the inwardly rectifying current of infected cells while oxidation (t-BHP, 1 mM) of non-infected cells activated the inward current suggesting redox sensitivity of the volume-sensitive inwardly rectifying current. In sharp contrast, non-infected, untreated controls did not express inwardly and outwardly rectifying anion channels and cell swelling of control cells did not induce inwardly rectifying anion channels. In conclusion, infection of human erythrocytes by Plasmodium falciparum activates inwardly rectifying anion channels. Upon activation these channels are under the control of cell volume and, therefore, may adjust the volume of the host to the needs of the parasite.

BLINDNESS AND INFERTILITY DUE TO AN EPITHELIAL BARRIER DEFECT IN CLC-2 KNOCKOUT MICE A. A. Zdebik, S. M. Stobrawa, T. J. Jentsch ClC-2 knockout mice are blind and infertile, but do not show any other obvious phenotype (Bösl et al., EMBO 2001) in spite of its wide distribution. ClC-2 currents can be recorded from both Leydig and Sertoli cells, and ClC-2 staining was observed at characteristic patches in Sertoli cells forming the blood-testis barrier. Both ClC-3 and ClC-7 knockout animals also show degeneration of the neural retina during adolescence (Stobrawa et al., Neuron 2001, Kornak et al., Cell 2001). As RPE function may be lost secondary to degeneration of the photoreceptors in ClC-2 ko retinae, ClC-3 and ClC-7 ko retinae were examined in comparison. Eyes were cut at the ora serrata, the neural retina was detached from the RPE and the sclera-RPE sandwich was mounted in an Ussing chamber with an opening of 1 mm 2 . Transepithelial voltage was measured in zero-current mode. Intermittent injection of 0.1 µA allowed the determination of the transepithelial resistance Rte. The equivalent short circuit current I sc was calculated as Vte/Rte, after subtraction of the chamber resistance. In ClC-2 ko retinae, the outer nuclear layer (i.e., the photoreceptor nuclei) reached half-thickness around P15, but a considerable fraction of photoreceptors was still present at P20. In contrast, in ClC-3 ko animals most nuclei disappeared between P15 and P20. Degeneration in ClC-7 ko animals followed a similar time-course, and in both ClC-3 and -7 ko animals few nuclei were left after 4 weeks whereas in ClC-2 ko animals, the loss was more protracted. Ussing chamber experiments performed on ClC-2 wt and ko animals at P10, P12, P15, P24, P36 and with adult eyes showed that transport activity did not develop until P24 in wt retinae. In older wild type animals, a lumen-positive transepithelial potential developed fully until P36 but was most robust in adult animals (Vte=8.0±0.6 mV). In ClC-2 ko retinae, the transepithelial resistance remained low, and transepithelial voltage remained below 0.5 mV. Surprisingly, transepithelial transport remained intact in ClC-3 ko mice until P80 despite the more rapid loss of photoreceptors. In ClC-7 ko mice that survived P30, transport was normal (n=3). In conclusion, ClC-2 mice are blind due to a primary pigment epithelial phenotype. A similar barrier defect might cause infertility.

Institute of Cytology, Tikhoretsky 4, St. Petersburg 194064, Russia and Department of Physiology I, University of Tübingen, Gmelinstr. 5, 72072 Tübingen, Germany

Zentrum für Molekulare Neurobiologie, Falkenried 94, 20251 Hamburg P 01-13

P 01-11 BARTTIN, A NOVEL β-SUBUNIT OF CLC-K Cl- CHANNELS, IS CRUCIAL FOR RENAL Cl--REABSORPTION AND INNER EAR K+-SECRETION V. Stein * , R. Estévez* , T. Boettger *, R. Birkenhäger +, E. Otto +, F. Hildebrandt+, T.J. Jentsch* Positional cloning of the gene underlying Bartter syndrome type 4 identified barttin. The recently identified gene encoding the integral membrane protein barttin is mutated in a form of Bartter syndrome that is associated with congenital deafness and renal failure. Renal salt loss in Bartter’s syndrome is caused by impaired transepithelial transport in the loop of Henle. NaCl is taken up apically by the combined activity of NKCC2 Na-K-2Cl cotransporters and ROMK K+ channels. Cl - exits from the cell through basolateral ClC-Kb channels. When barttin was co-expressed with ClC-Ka in Xenopus oocytes, large Clcurrents were observed. No currents were seen with either barttin, ClCKa, or ClC-Kb alone. In oocytes, barttin/ClC-Kb co-expression gave small but detectable currents. Similar effects were seen in transfected tsA201 cells. The effect appeared specific for ClC-K since currents of ClC-1, ClC-2, or ClC-5 were not changed by barttin. Barttin enhanced the surface expression of ClC-Ka, but not of the muscle channel ClC1. We show that barttin acts as an essential β-subunit for ClC-Ka and ClC-Kb with which it co-localises in basolateral membranes of renal tubules and of K+-secreting epithelia of the inner ear. We describe the first known β-subunit for CLC Cl- channels and reveals that CLC-K/ barttin heteromers are essential for renal salt reabsorption and inner ear K+ recycling. * Zentrum für Molekulare Neurobiologie, Universität Hamburg, Falkenried 94, 20246 Hamburg, Germany. + Universitäts-Kinderklinik, Universität Freiburg, Mathildenstrasse 1, 79106 Freiburg, Germany

DEPOLARIZATION INDUCED Cl - CHANNELS IN BROWN ADIPOCYTES. V. Sabanov, B. Cannon, J. Nedergaard In response to norepinephrine, brown adipocytes can increase their metabolic rate over 40-fold in 1-2 minutes. The dramatic activation of metabolism is accompanied by a complex electrical response in the plasma membrane. The first component of it – a rapid 10-30 sec depolarization is due to Cl - efflux (Lee & Pappone, 1995) and the following 2-5 min repolarization is due to K+ efflux (Lucero & Pappone, 1990). Brown adipocytes can therefore represent an excellent and unique natural model of acute metabolically challenged osmoregulation. This aspect of brown fat cell physiology remains almost unexplored. We studied single-channel Cl - currents in cultured mouse brown adipocytes using traditional patch-clamp techniques in the inside-out configuration. Cl - currents could not be activated in the cell-attached mode. In excised patches, channel activity was usually induced by strong depolarization, showed the intermediate conductivity (30-50 pS) and characteristic kinetics of outwardly rectifying Cl - channels (ORCC), although brown adipocyte Cl - channels by themselves are only slightly, if at all, rectifying. When tested for calcium sensitivity, two distinct types of channel currents were identified. The activity of channels in the first group was totally inhibited by Ca-free solution, whereas channels of the second type displayed some ”flickering activation” i.e. their open probability increased but the mean time of the openings decreased. It turned out, surprisingly, that both inhibitory and activating effects were mostly determined by chelators (BAPTA or EGTA) but not by Ca, because similar effects could be induced also by the chelator solutions containing 1 mM free Ca 2+. Both BAPTA and EGTA were used in concentrations 10 µM - 2 mM and seemed to be the principal ligands. Ca could only modulate their effects to some extent. Although this does not clarify the biological role of Cl -

S 226 transmembrane currents in brown adipocytes, it nevertheless gives new information on the biochemical and biophysical properties of ORCC as an electrophysiological phenomenon.

highly sensitive to both toxins (Lqh-2 0.6 nM; Lqh-3 5.7 nM), it is expected that native Na v1.6 channels are sensitive to both toxins at nanomolar concentrations.

The Wenner-Gren Institute, Stockholm University, The Arrhenius Laboratories F3, SE-106 91 Stockholm, Sweden

A. Hansel, Klinikum FSU Jena, Molekulare und zelluläre Biophysik, Drackendorfer Str. 1, 07747 Jena

P 02-1

P 02-3

COOPERATIVE EFFECT ON FAST INACTIVATION OF S4-S5 LOOPS OF DOMAINS III AND IV OF THE Na + CHANNEL (hNa v1.4) M.O. Popa, A.K. Alekov, S. Maljevic, N. Mitrovic, F. LehmannHorn, H. Lerche The cytoplasmic S4-S5 loops of domain III and IV of voltage-gated Na + channels are important for fast channel inactivation. Using the human skeletal muscle Na + channel hNa v1.4, expression in tsA201 cells and the whole cell patch-clamp technique, we investigated cysteine mutations in III/S4-S5 (N1151C, A1152C, I1160C, N1162C) and IV/S4-S5 (F1473C, L1482C). In order to look for possible cooperative effects on channel gating of mutations located at corresponding sites in III/S4-S5 and IV/S4-S5, two double mutations (N1151C/ F1473C, I1160C/L1482C) were also electrophysiologically characterized. All mutations could be modified by intracellularly applied MTSES changing both activation and fast inactivation of the channel. Steady-state activation was shifted towards more hyperpolarized potentials by –5 to –20mV and inactivation curves in the depolarizing direction by up to +20mV compared to the wild type. In the presence of MTSES inactivation was slowed up to 4-fold and a persistent Na+ current of 1-10% of peak current occurred for all single mutations. One mutant, I1160C, showed a voltage-dependent accessibility. Its rate of modification by MTSES at –140mV was much smaller than at –20mV. The double mutant N1151C/F1473C showed a 20-fold slowing of inactivation and 5% of persistent current after modification by MTSES. I1160C/L1482C even more drastically disrupted fast inactivation with a persistent current of 27% increasing to 67% after MTSES application. Thermodynamic cycle analysis using inactivation parameters revealed an additive effect for N1151C/F1473C and a significant cooperative effect for I1160C/L1482C (coupling energy of 1.5 kcal/ mol). We conclude that both S4-S5 loops play important roles for channel activation and fast inactivation and that the two domains act interdependently concerning fast inactivation.

SINGLE-CHANNEL PROPERTIES OF THE ∆KPQ MUTANT OF THE HUMAN CARDIAC Na+ CHANNEL GENERATING LONG QT-SYNDROME S. Dugarmaa, T. Zimmer, K. Benndorf One form of the long QT-syndrome (LQT3) has been attributed to a deletion of three amino acids (KPQ) in the putative inactivation gate of the human heart Na+ channel (hH1). The functional consequence of this ∆KPQ mutation is an inactivation defect by developing persistent bursts. Although the incidence of these bursts is very low it is sufficient to account for the observed prolongation of the QT-interval. We expressed wild-type (wt) and ∆KPQ mutant channels in CHO cells and measured ionic currents in cell-attached patches containing one and only one channel. The kinetics of the channel activity were evaluated. In ∆KPQ channels the inactivation time constant, determined from averaged currents, was faster at voltages between –60 and –40 mV and less voltage dependent between –60 and 0 mV than in wt channels. From –60 to 0 mV, the mean open time in ∆KPQ channels increased from 0.15 to 0.33 ms, whereas wt channels produced the typically bell-shaped voltage dependence with a maximum value of 0.30 ms at –40 mV. Burst-like channel activity, identified by an open probability >0.5 in test pulses of 8 ms duration, was significantly higher in ∆KPQ channels than in wt channels only at –40 mV and more positive voltages, whereas it was similarly low at –50 and –60 mV. It is concluded that the ∆KPQ mutation affects multiple relevant parameters of inactivation kinetics.

Depts. of Applied Physiology and Neurology, University of Ulm, D89069 Ulm, Germany P 02-2 MOLECULAR DETERMINANTS FOR THE BINDING OF SCORPION TOXINS LQH-2 AND LQH-3 TO MAMMALIAN NEURONAL SODIUM CHANNELS A. Hansel, S.-Q. Lu, E. Leipold, S.H. Heinemann Scorpion α-toxins impair fast inactivation of voltage-gated sodium channels. The main determinant for their binding is a motif (sequence: -L-[A, S]-[D, E]-[I, L]-I-[Q, E]-[K, T]-Y-F-) located in the loop connecting the membrane-spanning segments S3 and S4 in domain IV of the channels. The specific aim of this study was to characterize molecular determinants of receptor site-3 providing specificity for the binding of the two toxins from the Israeli yellow scorpion Leiurus quinquestriatus hebraeus, Lqh-2 and Lqh-3. As a reference channel we used the rat skeletal muscle Nav1.4 channel (µI). In this background, the linker between S3 and S4 in domain IV was replaced by the respective receptor site-3 motifs of Nav1.2 (brain II), Nav1.6 (Na channel 6), and hNav1.7 (hPN1). For examining the contribution of individual amino acids within these receptor site-3 sequences, single-site mutations were included. The channels were expressed in HEK293 cells and assayed with the whole-cell patch-clamp method. For rNav1.4, the background channel used, concentrations of 1.4 nM Lqh-2 and 7.2 nM Lqh-3 lead to half-maximal toxin effects. For rNav1.2 the corresponding values were 1.8 nM for Lqh-2 and 340 nM for Lqh-3, for hNa v1.7 52 nM for Lqh-2 and 13 nM for Lqh-3. Measurements using the mutants with exchanged receptor sites as well as with single amino-acid exchanges showed that the binding of Lqh-3 is largely determined by the D/E residue in the receptor site-3 motif, whereas the sensitivity to Lqh-2 strongly depends on the K/T residue. As the mutant channel with the Na v1.6 receptor site-3 motif was

Institute of Physiology II, Friedrich Schiller University Jena, Teichgraben 8, 07740 Jena, Germany P 02-4 REGULATION OF NA+ CHANNEL TRANSCRIPTION DURING MOUSE HEART DEVELOPMENT V. Haufe, T. Zimmer, B. Günther§, K. Benndorf Cardiac cells are known to express a variety of Na+ channel α subunit transcripts. The aim of the present study was to investigate the expression level of the voltage-gated Na+ channels mH1 (SCN5A) and mH2 (SCN4A), and splice variants mH1-2 and mH1-3 during mouse ontogeny from the embryonic to the adult stage by means of RT-PCR. At all developmental stages, mH1 was the major Na+ channel isoform. Transcription appeared to be upregulated about 2.5-fold during development, similarly as shown for the Na+ channel β1 subunit. The relative amount in the adult mouse heart was 54%. The isoform of skeletal muscle cells, mH2, was not expressed in the embryonic and postnatal heart, but transcripts appeared at day 38 and later. In the adult mouse heart, the relative amount was 16%. The expression level of both mH1 splice variants (mH1-2 and mH1-3) was not regulated during development. Their relative amount in the adult mouse heart was 16% for mH1-2 and 5% for mH1-3. Interestingly, splicing of mH1 was found to be tissue-dependent. We observed mH1 and mH1-2 transcripts also in adult brain, testis, spleen and lung at variable mH1/ mH1-2 mRNA ratios (1:2 in brain, 1:8 in testis, 1:1 in lung and spleen). The ongoing research is aimed to get more insight into the functional role of the cardiac Na+ channel isoforms in these tissues. Institute of Physiology II, Friedrich Schiller University Jena, Teichgraben 8, 07740 Jena and §Institute of Laboratory Animals, Friedrich Schiller-University Jena, Beutenbergstrasse 11, 07745 Jena, Germany P 02-5 Na+ AND K+ DEPENDENT GATING OF CYCLIC NUCLEOTIDEGATED (CNG) CHANNELS OF ROD PHOTORECEPTORS J. Kusch, K. Benndorf CNG channels have been reported to be equally permeable for Na+ and K+ ions. It is generally assumed that the gating of these channels is

S 227 independent of these ions. We investigated the influence of intra- and extracellular Na+ and K+ ions on currents generated by homomultimeric channels of the α1 subunit of bovine rod photoreceptors. The currents were studied in excised patches of Xenopus oocytes expressing these channels and the voltage was set to 100 mV. The channels were activated by either the low concentration of 10 µM or the saturating concentration of 700 µM cGMP. For Na + and K+, all possible four combinations were considered (inside/outside): K+/K+; Na+/K+, K+/Na+, Na+/Na+. At 700 µM cGMP, the ion combinations K+/K+ and K+/Na + generated openings of ~2 pA amplitude whereas the ion combinations Na+/Na+ and Na +/K+ generated openings of ~3.6 pA amplitude. The open probability Po was similar (0.81-0.97) at all ion combinations. At 10 µM cGMP, however, the single channel current level was variable between 0.5 and 4 pA and the distribution of these events depended markedly on the ion composition. While in K+/K+ most of the channel openings had an amplitude near 2 pA and Po was 1.2x10-2, in Na+/Na+ all levels between 0.5 and 4 pA contributed significantly to the channels activity and Po was only 1.5x10-3. The other two ion combinations produced values in-between. In open time histograms, generated from all open levels, a second slow component was observed only in the presence of K + ions in the intracellular solution. It is concluded that intracellular K+ ions stabilize an open state in CNG channels by promoting an additional open state with longer life time. Institut für Physiologie II, Friedrich-Schiller-Universität Jena, D-07743 Jena P 02-6 MOLECULAR REGIONS OF THE β1 SUBUNIT INVOLVED IN THE MODULATION OF THE HUMAN HEART SODIUM CHANNEL (hH1) EXPRESSED IN XENOPUS OOCYTES T. Zimmer, C. Bollensdorff, C. Biskup, K. Benndorf The β1 subunit, but not the β2 subunit of voltage-gated sodium channels accelerates recovery from inactivation and increases the amplitude of human heart sodium channel (hH1) currents in Xenopus oocytes. We used β 1/β 2 subunit chimeras to identify molecular regions of the β 1 subunit involved in the modulation of hH1 and rat brain IIA channels. As a result, all chimeras containing the β1 extracellular domain modulated IIA channels similarly as the wild-type β1 subunit. However, the presence of the extracellular domain of the β1 subunit alone was not sufficient to respectively modulate hH1 channels. Instead, the putative membrane anchor plus either the extracellular or the intracellular domain of the β 1 subunit were required to modify current amplitude and recovery from inactivation of hH1 channels. When fused to the extracellular and intracellular domain of the β 2 subunit, the β 1 membrane anchor alone did not produce β 1-like effects on hH1 currents. We conclude that IIA channels interact only with the extracellular domain of the β1 subunit, whereas additional molecular regions of the β 1 subunit are required to modulate hH1.

triguing because NAPA is more hydrophobic than PA and in general this property implies a more effective Na+ channel block. Therefore, a reasonably alternative explanation for the potency of PA relies upon the electronic donating effect exerted by the amine group linked to the aromatic ring. This effect, expressed by the Hammet’s constant σ = − 0.66, leads to the increase of the electronic density of the aromatic nucleus and consequently to the improvement of its reactivity toward one site of a putative receptor for antiarrhytmic drugs in the Na+ channel: the Y1771 in the α subunit (Ragsdale et al., PNAS (93): 9270-5, 1996). NAPA lacks this electronic effect: σ = 0. Ιn addition, the acetyl group linked to the amino group, in the case of NAPA, could cause steric hindrance in the interaction of the aromatic nucleus of this molecule with the mentioned receptor site. *FAPESP(Brazil) fellow; #CAPES(Brazil)/DAAD(Germany) fellow Institut für Physiologie II, Friedrich-Schiller-Universität Jena, 07740, Jena P 02-8 EFFECTS OF THE ANTICONVULSANT DRUG SULTHIAME ON VOLTAGE-OPERATED SODIUM CHANNELS OF HIPPOCAMPAL NEURONS M. Madeja, E.-J. Speckmann The anticonvulsant properties of the sulfonamide derivative sulthiame (Ospolot®) are known for several decades. Concerning the anticonvulsant mechanisms of action, the inhibitory effect of sulthiame on the enzyme carbonic anhydrase has often been taken into consideration. Although the underlying hypothesis is experimentally well-established and convincing, evidence is growing that sulthiame exerts additional actions contributing to the anticonvulsant effects. In this study the effect of sulthiame on voltage-operated sodium channels was investigated in acutely isolated CA1 neurons from the guinea pig hippocampus. The whole-cell patch-clamp technique was applied, thus allowing to clamp both the intracellular and extracellular pH values largely. The experiments revealed the following results: (i) Sulthiame in a concentration of 10 µg/ml reduced the inactivating sodium currents without affecting potassium currents. (ii) The blocking effect of the sodium currents was not dependent on voltage. (iii) At therapeutic concentration of 1 to 10 µg/ml, sodium currents were reduced by 13 to 25 % of control. (iv) Reductions of this size (induced by the specific sodium channel blocker tetrodotoxin or by sulthiame itself) impaired repetitive generation of action potentials and thus, reduced the maximum discharge frequency by 20 to 40 %. In summary, the anticonvulsant drug sulthiame exerts blocking effects on sodium channels which can be assumed to contribute to the anticonvulsant efficacy of the substance and to be different from the anticonvulsant effects induced by blockade of carbonic anhydrase. Institut für Physiologie, Robert-Koch-Str. 27a, D-48149 Münster, Germany

Institute of Physiology II, Friedrich Schiller University Jena, Teichgraben 8, 07740 Jena, Germany P 02-9 P 02-7 DIFFERENTIAL APPARENT AFFINITY OF PROCAINAMIDE AND N-ACETYLPROCAINAMIDE FOR CARDIAC NA+ CHANNELS W. Sigler*, A. C. Oliveira#, K. Benndorf Procainamide (PA) is a class I antiarrhythmic, blocking cardiac Na+ channels. In vivo PA gives rise to N-acetylprocainamide (NAPA) which is reported not to affect Na+ channels. This study verifies whether this really holds by directly recording Na+ currents instead of the action potential Vmax so far used to compare these compounds. PA and NAPA are structurally related molecules, differing only in an acetyl group linked to the amine group of the benzene ring, in the case of NAPA. The patch clamp technique in the whole cell configuration was used in isolated cardiac myocytes of the mouse. Na + currents were generated employing pulses from – 120 mV to – 30 mV and recorded in the absence and presence of several concentrations of each drug, covering a range broader than ever studied. The IC50 for the block of Na + currents were: PA: 0.8 mM and NAPA: 24.6 mM. The NAPA concentration-response curve suggests parallel shift to the right with respect to PA’s suggesting that both drugs block Na+ channels via similar mechanisms. Thus, NAPA has been shown to block cardiac Na+ channels, although about 30 times less potently than PA. This is in-

ACID-SENSING ION CHANNELS (ASICs) ARE MODULATED BY EXTRACELLULAR DIVALENT CATIONS E. Babini, M. Paukert, H. -S. Geisler, S. Gründer The acid-sensing ion channel (ASIC) subunits are members of the amiloride sensitive Na+ channel/degenerin family of ion channels. They are expressed in the central nervous system and in sensory neurons and are activated by extracellular protons. We investigated the interaction of the ASIC subunits ASIC1a and ASIC1b with the divalent cations Ca2+, Mg2+, and Ba2+. ASIC mediated currents were measured with the two-electrode voltage clamp technique, using Xenopus laevis oocytes. Both ASIC1a and ASIC1b were inhibited by Ca2+ when the channels were activated with pH 6.0 and different concentrations of Ca 2+ During the activation. The IC50 was 0.98 mM for ASIC1a and 1.12 mM for ASIC1b; Ba2+ and Mg2+ showed a similar effect, which we investigated only for ASIC1a (IC50= 0.64 mM and 4.93 mM, respectively). Probably calcium inhibits the channel by blocking the open pore or by allosteric modulation of the channel activity. Strikingly, we have found also a potentiation by divalent cations for ASIC1a. When channels were activated with pH 6.0 and fixed concentrations of Ca2+, Mg2+, and Ba2+, After they had been exposed to different concentrations of these cations, current amplitude showed a significant dependence on the concentration of divalent cations. This potentiation can

S 228 be explained by a modulation of the properties of the channel, like the affinity for protons or recovery from inactivation. We are currently investigating the basis for inhibition as well as potentiation of ASIC1 by divalent cations. Since the extracellular Ca2+ concentration can significantly decrease from the normal value of 1.0-1.2 mM, for example in central nervous system during high nervous activity, the potentiation of ASIC-mediated currents by divalent cations may be a means to adapt ASIC activity to the metabolic state of the neuronal tissue. Dept. of Otolaryngology, University of Tuebingen, Roentgenweg 11, 72076 Tuebingen, Germany

nA on day 1 and 352 ± 68 nA on day 2 after AON injection, while GNa decreases from 13.9 ± 2.2 µS (day 0) to 11.5 ± 3.0 µS (day 1) and 5.5 ± 1.5 µS (day 2). For bAON the data (16 oocytes) are: INa from 4393 ± 695 nA (day 0) to 421 ± 236 nA (day 1); GNa from 69.3 ± 11.4 µS to 6.9 ± 3.4 µS; V m from 10.4 ± 1.2 mV to –11.3 ± 8.1 mV. Although more data are required, we believe that the variability in the time course of the inhibition is due to the existence of highly variable intracellular stores of preformed ENaC molecules. Laboratory of Physiology, KU Leuven, Campus Gasthuisberg, B-3000 Leuven, Belgium P 02-12

P 02-10 KIDNEY-SPECIFIC INACTIVATION OF αENAC SUBUNIT I. Rubera, A.-M. Mérillat, B.C. Rossier, E. Hummler The amiloride-sensitive epithelial sodium channel, ENaC, is a heteromultimeric protein made up of three homologous subunits (α, β, γ). ENaC constitutes the limiting step for sodium absorption in epithelial cells that line the distal renal tubule, distal. Gene inactivation of Scnn1a (αENaC subunit) gene demonstrated a crucial role of the αENaC subunit in lung liquid clearance (Hummler et al., Nature Genetics, 1996). Mice deficient for this subunit died within 40h after birth due to a failure to clear their lung at birth. In order to study the importance of Scnn1a expression in the adult kidney, we generated mice with a conditional knock out for Scnn1a to control gene targeting in a tissueand/or time-dependent manner. Our targeting vector contained homologous sequences of the endogenous Scnn1a locus, the vital coding exon (here exon 1) flanked by loxP sites, and a neomycin selection marker (neo), followed by a third loxP site (Scnn1aneo allele). The neomycin selection marker was selectively removed in embryonic stem cells either by transient transfection using Cre recombinase plasmid, or by breeding the mice to a germline deleter strain (EIIa-Cre). Kidneyspecific gene targeting is achieved using mice homozygous for the Scnn1a lox allele and a mouse strain specifically expressing the Cre recombinase in the kidney. Two lines have been used: constitutively expressing Cre recombinase #line 1 in almost the whole kidney, i.e in proximal and distal parts of the nephron, and #line 2, preferentially in the collecting duct. This study will allow us to determine if the inactivation of Scnn1a leads to a complete abolishment of ENaC function in the kidney demonstrating whether αENaC is a limiting subunit for ENaC activity in kidney. Using different Cre lines expressing Cre recombinase in whole versus defined segments of the nephron we can correlate ENaC activity with the expected salt-loosing phenotype (pseudoaldosteronism type 1 (PHA-1)). Institute of Pharmacology and Toxicology, Lausanne, Swizerland P 02-11 SPECIFIC EFFECTS OF ANTISENSE OLIGONUCLEOTIDES ON ENAC EXPRESSION IN XENOPUS LAEVIS OOCYTES A. Segal, W. Van Driessche, W.-M. Weber Antisense oligonucleotides (AONs) are short sequences of DNA designed to inhibit the expression of a protein by specifically binding to its mRNA. We designed two such AONs with sequences complementary to the mRNA of both rat and human α ENaC subunit. The Vienna RNA software package (http://www.tbi.univie.ac.at/~ivo/RNA/) was used to compute predicted secondary structures of rat (r) and human (h) α ENaC mRNA at different temperatures ranging from 12 to 36°C. We then searched for sequences of 16 bases that are common to the two species and are also mostly single-stranded. The two AONs we termed ”green AON” (gAON) and ”black AON” (bAON), correspond to hENaC at positions 290 and 1670, respectively. The actual molecules were synthesized as 2OMeRNA phosphorothioates. Coinjection of either AON with αβγ-rENaC in oocytes successfully inhibits channel expression, as demonstrated by the negative membrane potential (Vm) and the lack of amiloride sensitivity of the clamp current (I m) and conductance (Gm). Sequential microinjection of ENaC mRNA followed by either AON led to an exponential-like decrease in amiloride-sensitive current (INa) and conductance (GNa), with highly variable time constants ranging from 2 hours to 2 days, accompanied by a gradual shift of the membrane potential from positive to negative values. Pooled data from 10 oocytes recorded in 10 mM external Na+ shows for gAON a decrease in INa from 955 ± 128 nA before AON injection to 814 ± 176

LOCAL MECHANICAL COMPRESSION MODULATES MULTIPLE MEMBRANE CURRENT COMPONENTS IN MOUSE VENTRICULAR MYOCYTES I. Kiseleva, A. Kamkin, G. Isenberg We have reported that axial stretch activates whole-cell currents through non-selective cation channels (ISAC). Here, we asked whether ISAC can be similarly activated by vertical compression of the cell. Brick-like myocytes were rolled to attach to the glass bottom in two different positions: a) edgewise, staying on the narrow side, or b) broadwise. Vertical compression was applied. Cells were superfused by solution composed of (mM): 150 NaCl, 5.4 KCl, 1.8 CaCl2, 1.2 MgCl2, 20 glucose, 5 HEPES/NaOH (pH 7.4, 37 oC). Pipettes were filled with a solution of (mM): 140 KCl, 5 Na2ATP, 5.5 MgCl2, 0.01 EGTA, 10 HEPES/KOH (pH 7.2). Clamp pulses started from -45 mV. From the late currents the mechanosensitive difference current (Ims) was calculated. In edgewise attached cells, Ims (nA) was inward (-0.12±0.04 at 2 µm, -0.28± 0.05 at 4 µm; -0.32±0.02 at 6 µm and -0.70±0.1 at 8 µm compression), was Gd3+ sensitive and decayed rapidly upon release of compression. Ims showed two reversal potentials (Erev). Positive to -30 mV, Ims followed outward rectification with an Erev of -20 mV, as expected from a non-selective cation current ISAC. Negative to -45 mV, Ims rectified inwardly with an Erev ≈ -90 mV. This second component was blocked by replacing extracellular (e.c.) K+ by Cs+ ions, hence, we attribute it to a reduced K+ conductance. Compression reduced I Ca-L; this effect was prevented by cell dialysis of 5 mM BAPTA suggesting dependence on elevated [Ca2+]c. In broadwise attached cells, a 4 µm compression induced positive I ms that hyperpolarized the membrane (no effect on ICa-L). Positive to -45 mV, Ims showed outward rectification with an Erev of -20 mV. Negative to -45 mV rectified inwardly (maximum Ims at -70 mV) and reversed at -100 mV. Block of SACs by 5 µM Gd3+ enhanced Ims (0.27±0.01 nA). After block of K-conductance by Cs+, compression induced an Ims that was nearly voltage-independent suggesting that compression can activate a current due to electrogenic Na pumping. Our results suggest that compression modulates not only I SAC but also K+- and pump-currents. Which component is activated or deactivated depends on the orientation of the compression in regard to the long or the short diameter of the cell. Institute of Physiology, Martin-Luther-University, D-06097 Halle, Germany P 02-13 ABUNDANCE OF EPITHELIAL SODIUM CHANNEL (ENaC) INDUCTION IN ULCERATIVE COLITIS S. Amasheh, C. Barmeyer, J. Mankertz, J.D. Schulzke, M. Fromm Background & aims: Patients with ulcerative colitis suffer from diarrhea due to barrier disturbance as well as malabsorptive mechanisms. In this study, active electrogenic sodium absorption of inflamed human sigmoid colon was investigated by measuring function and expression of the epithelial sodium channel ENaC, employing electrophysiological and molecular analyses. Methods: Epithelial preparations of inflamed and control human sigma were mounted in Ussing chambers and incubated with aldosterone (3 nM) at 37° C. After 8 h, electrogenic Na+ transport (J Na) was determined by the drop of short circuit current (ISC) caused by amiloride (100 µM). Subsequently, tissues were removed and α-, β- and γ-ENaC subunit mRNAs were detected by Northern blotting. Reporter gene assays were performed for the human β-ENaC subunit promoter in the presence or absence of tumor necrosis factor α and interferon γ. Results: In colons of controls, a strong increase of JNa was induced by aldosterone (6.9 ± 1.4 µmol⋅h–1⋅cm–2, n = 8), whereas in inflamed tis-

S 229 sues, JNa was diminished (0.2 ± 0.6 µmol⋅h–1⋅cm–2, n = 7). In non-inflamed tissues a strong increase of β- and γ-ENaC mRNA was detected by Northern blotting. In inflamed tissues, aldosterone-induced transcription of β- and γ-ENaC mRNA was abolished almost completely. β-ENaC subunit promoter-mediated gene expression was strongly reduced by tumor necrosis factor α and interferon γ. Conclusion: Reduced colonic Na+ absorption contributes to diarrhea in ulcerative colitis by impaired expression of the epithelial sodium channel (ENaC) on transcriptional level of the β- and γ-subunits. In case of the β-ENaC subunit, this has been shown in reporter gene assays to be due to genomic expression regulation. Dept. of Clinical Physiology and Dept. of Gastroenterology, Benjamin Franklin University Hospital, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany P 03-1 THE P2X7-RECEPTOR CLONED FROM XENOPUS LAEVIS SUFFICIENT TO PRODUCE LARGE PORE-FORMATION IN XENOPUS OOCYTES M. Paukert, S. Hidayat, S. Gründer P2X7-receptors belong to the family of ATP-activated non-selective cation channels (P2X). They are predominantly expressed in immune cells. Repeated or prolonged activation of P2X7-receptors leads to the formation of a large pore permeable to large cations and dye molecules like NMDG+ and YO-PRO-1. Pore formation is followed by cell death, for example in macrophages. P2X7-receptors have been cloned from man, rat and mouse. Here we present cloning of this subtype from Xenopus laevis and its functional expression in Xenopus oocytes. The full-length cDNA for the P2X7-receptor was cloned from Xenopus laevis (xP2X7) using an expressed sequence tag (EST). The encoded protein has a length of 553 amino acids and its sequence shows all features characteristic to P2X7. It shares about 50% homology with the known P2X7-receptors from mammals. The tissue distribution was investigated using RT-PCR. xP2X7-RNA could be detected in all twelve tissues investigated. The presence in many tissues and especially the high expression in spleen suggest xP2X7 also in amphibians to be expressed in immune cells and to play a role similar to that in mammals. We expressed xP2X7 in Xenopus oocytes and analyzed it using two-electrode-voltage-clamp. xP2X7 showed a unique combination of properties compared to the other known P2X7-receptors. Like with the human receptor, when expressed in Xenopus oocytes, there was no large run-down in current amplitude during repeated activation. With rP2X7 the Xenopus receptor shares the property not to be blocked by the compound KN-62 at concentrations (3µM) at which hP2X7 is completely blocked. Characteristic for P2X7-receptors, xP2X7 was blocked by extracellular protons and divalent cations. Further on, it showed a low affinity to ATP (EC 50 ~ 2.6mM) and an approximately tenfold higher affinity to the agonist BzATP. The most striking difference between xP2X7 and the other P2X7-receptors is that it readily formed large pores when expressed in Xenopus oocytes. The evidence for this is threefold: i) a slowly increasing conductance between ATP applications could be blocked by hexamethylene-amiloride (HMA), ii) prolonged ATP application induced a shift in the reversal potential when extracellular Na+ was replaced by NMDG + and iii) YO-PRO-1 uptake into oocytes could be elicited by xP2X7 activation. These properties of the xP2X7-receptor might help to reveal structural domains crucial for large pore formation. Martin Paukert, Department of Otolaryngology of the University of Tubingen - Research Group of Sensory Physiology, Röntgenweg 11, D-72076 Tubingen, Germany

P 03-2 THE rP2X5 RECEPTOR IS PHARMACOLOGICALLY SIMILAR TO THE rP2X1 RECEPTOR, BUT OPERATIONALLY DIFFERENT S.S. Wildman, S.G. Brown, M. Rahman, G. Burnstock, R.J. Unwin, B.F. King The P2X5 receptor is an ATP-gated cation channel, first cloned from rat coeliac ganglia (Collo et al. 1996). P2X 5 receptor transcripts are primarily distributed in smooth muscle cells, where they are found to

coexist with P2X1 receptors as homomers and heteromultimers. We have extensively investigated the pharmacological profiles of recombinant rat P2X5 and P2X1 receptors expressed in Xenopus oocytes and studied under voltage-clamp conditions to offer a way to isolate these 2 receptor subtypes in native tissues where receptor kinetics and confounding effects from co-localisation are often difficult to assess. Superfused ATP (10nM-10µM) evoked inward currents in a concentration-dependant manner at both rP2X 5 and rP2X1 receptors. ATP-evoked currents at the rP2X5 receptor were slowly desensitising, whereas those of the rP2X 1 the currents were rapidly desensitising with dissimilar receptor kinetics. The selectivity of the receptors was tested for a broad range of nucleotides. For the rP2X5 receptor the activity order for agonists was ATP=2meSATP=ATPγS>αβmeATP =BzATP=Ap 4 A=Ap 5 A>CTP=GTP>βγmeATP=Ap 3 A=Ap 6 A>UTP (Italics-denotes a partial agonist). For the rP2X1 receptor the activity order for agonists was Ap 4 A>ATP=2meSATP>ATPγS=Ap 6 A> αβmeATP=BzATP=Ap5A>βγmeATP=CTP>UTP =Ap3A. A series of P2 receptor antagonists were tested against ATP-evoked currents (EC70 concentration). For rP2X 5 receptors the activity order of antagonists was PPADS>TNP-ATP=suramin>RB2. For rP2X 1 receptors, TNPATP>> PPADS>suramin>RB2. Both Zn2+ and H+ had modulatory effects at rP2X 5 and rP2X 1 receptors; however only H + produced distinct modulatory differences. At the rP2X5 receptor, extracellular H+ displaced the ATP C/R curve progressively rightward with a reduction in maximal ATP activity. At the rP2X1, H+ also displaced the ATP C/R curve progressively rightward but with no reduction in maximal activity. In summary, this in-depth pharmacological and kinetic comparison of rP2X5 and rP2X1 receptors demonstrates distinct operational, but close pharmacological, phenotypes. From these data it is evident that caution is necessary when assigning P2X1-like receptor responses in native tissues (e.g., smooth muscle) that is based solely on limited pharmacological profiling.Collo, G., North, R.A., Kawashima, E. et al., (1996) J. Neurosci., 16, 2495-2507. Department of Physiology and Autonomic Neuroscience Institute, Royal Free Hospital, University College London, Rowland Hill Street, London, NW3 2PF, UK

P 03-3 PURINERGIC RECEPTOR EXPRESSION IN HUMAN INSULINSECRETING CELLS IN HEALTH AND DISEASE A.-M. Gonzalez, R.M. Shepherd, A. Lee, K. Hussain, P.E. Clayton, A. Aynsley-Green, K.J. Lindley, M.J. Dunne ATP-evoked signalling events are known to promote insulin release of insulin from pancreatic ß-cells via Ca 2+-dependent signalling mechanisms. This is mediated by purinergic receptors and there is evidence for the involvement of both P2X and P2Y receptors in rodent ß-cells and insulin-secreting cell lines. In this study we have used human insulin-secreting cells to examine the expression of purinoceptors in control and disease tissue. Intact islets and isolated ßcells were obtained from control human donors (n=3 with permission) and from patients with either Hyperinsulinism in Infancy (n=3, HI) or adenoma (n=3,AD) following surgery (with permission). Experiments were also undertaken with the human insulin-secreting cell line, NES2Y. We used Fura-2 loaded cells to examine ATP-evoked cytosolic Ca2+ signals ([Ca2+]i) by microfluorimetry and cell imaging, and studied the expression of P2Y and P2X mRNA subtypes by RT-PCR and immunofluorescence. We report here that whilst ATP (100µM) consistently elevated [Ca 2+]i in control, HI, AD and NES2Y ß-cells, RT-PCR results shown that the expression of purinoceptor subtypes was variable across the different tissues studied. In control islets, we found expression for P2X4 and P2Y1 subtypes but not P2Y2 or P2Y4; in HI islets and AD ß-cells, both P2X4 and P2Y1 were expressed, yet the adenoma tissue also expressed P2Y2 mRNA. NES2Y cells ß-cells are a proliferating cell line derived from a patient with HI. In these cells, we found expression of P2Y2 and P2Y4, but not P2Y1 mRNA subtypes. NES2Y ß-cells were also found to expressed P2X4 mRNA, yet these cells failed to respond to the P2X4 receptor agonist ADPβS(100µM, n=21) and no protein was found to be localised to the cell surface (by immunofluorescence using a specific P2X4 antibody). These studies reveal that whilst control and human islets of Langerhans obtained from patients with hyperinsulinism respond to ATP through changes in [Ca2+]i signals, purinergic receptor expression is altered.

S 230 Dept Biomedical Science, Sheffield University, Sheffield; Institute of Child Health, University College London, London; Manchester Children’s Hospital, Manchester University, Manchester P 03-4 THE GLU496 TO ALA POLYMORPHISM OF THE HUMAN P2X7 RECEPTOR, LEADING TO LOSS OF FUNCTION IN HUMAN B LYMPHOCYTES, RETAINS THE NATIVE BIOPHYSICAL PROPERTIES OF THE ION CHANNEL W. Boldt1, M. Klapperstück1, G. Schmalzing2, F. Markwardt1 The Glu496 to Ala mutation of the human P2X7 receptor was reported to evoke a loss of function in human B lymphocytes as demonstrated by lacking ethidium + uptake, Ba2+ influx and induction of apoptosis. The wildtype and mutant G496A hP2X 7 receptors as well as hP2X7 receptors with the additional mutation G441R previously described by us, were expressed in Xenopus oocytes and electrophysiologically characterized. Their biophysical properties were compared. Amplitudes of ATP-induced whole cell currents characteristic for functional expression as well as kinetic properties including ATP concentration dependency were not affected by these mutations irrespective of the expressed functional receptor density. After 2 min of application of 1 mM ATP in Na +-free, choline +-containing extracellular solution hP2X7-dependent currents reversed at about -40 mV for all three receptor subtypes studied. This indicates an equal low permeability for the organic cation choline + in contrast to the formation of large pores as shown by the uptake of fluorescent dyes like YO-PRO-1 in several hP2X7- expressing cell lines. We presume that the property of hP2X7 of opening ion channels for small cations, which is similar to other P2X receptor subtypes, is not affected by the G496A polymorphism and independent of the other P2X7-dependent functions which are compromized by this mutation. 1 Julius-Bernstein-Institut für Physiologie, Martin-Luther-Universität Halle-Wittenberg, Magdeburger Str. 6, 06097 Halle (Saale), Germany; 2 RWTH Aachen, Abteilung für Molekulare Pharmakologie, Wendlingweg 2,D-52074 Aachen, Germany

P 03-6 CHANGES IN PHi AND [CA2+]i CAUSED BY GLUTAMATERGIC STIMULATION OF LEECH RETZIUS NEURONS FOLLOW DIFFERENT TIME COURSES P. Hochstrate, P.W. Dierkes, W.-R. Schlue The activation of ionotropic glutamate receptors in leech Retzius neurons evokes a multitude of physiological effects, such as an increase in [Na+]i, [Cl-]i, and [Ca2+]i, a drop in pHi and [K+]i, as well as a membrane depolarization and an increase in cell volume. All these changes appear to be primarily caused by the influx of Na + through the glutamate receptors, which depolarizes the cell membrane thereby enhancing the efflux of K+, mediated partly by the glutamate receptors themselves, and the influx of Cl-. These ion fluxes eventually lead to an increase in the intracellular osmolarity and in turn to cell swelling. Furthermore, the membrane depolarization causes the activation of voltage-dependent Ca 2+ channels and hence Ca 2+ influx and [Ca 2+]i increase. Various observations indicate that the increase in [Na+]i is the crucial event underlying the intracellular acidification, while proton influx or Ca2+-induced proton release from intracellular binding sites appears to play a minor role. We measured simultaneously pHi and [Ca2+]i in Retzius neurons by monitoring the fluorescence of the pH indicator BCECF and of the Ca2+ indicator Fura-2, ratiometrically in each case by alternate excitation with two different wavelengths. The data show that the changes in pHi and [Ca2+]i induced by glutamatergic stimulation follow markedly different time courses. In particular, no drop in pHi was detectable before [Ca2+]i had reached its maximum, rather pHi was slightly shifted to the alkaline direction. Furthermore, pHi reached its minimum during the recovery phase of [Ca2+]i. The results confirm the view that the intracellular acidification of leech Retzius neurons upon glutamatergic stimulation is not due to the displacement of protons from intracellular Ca 2+ binding sites. Furthermore, it can be excluded that the acidification is caused by proton influx, which implies that neither the glutamate receptors nor the voltage-dependent Ca 2+ channels mediate a significant proton current. Supported by the Deutsche Forschungsgemeinschaft (Schl 169/ 12). Institut für Neurobiologie, Heinrich-Heine-Universität, 40225 Düsseldorf

P 03-5 RNA EDITING (R/G SITE) AND FLIP/FLOP SPLICING OF THE AMPA RECEPTOR SUBUNIT GLUR2 IN HIPPOCAMPAL AND NEOCORTICAL TISSUE OF EPILEPSY PATIENTS W. Vollmar, J. Gloger, E. Berger, G. Kortenbruck, R. Köhling, E.-J. Speckmann, U. Mußhoff Changes in the expression of glutamate receptors are important in patho-physiological aspects of epilepsy, either in the process of epileptogenesis or in mediating seizure-induced brain injury and dysfunction. The present study analyzes the expression of (i) RNA editing at the R/G site, which controls the time course of recovery from desensitization, and (ii) flip/flop exons of the GluR2 subunit in nervous tissue of epilepsy patients. Specimens of hippocampus and neocortex from patients with refractory temporal lobe epilepsy (TLE; n=17) were obtained from neurosurgical operations and analyzed by the RT-PCR technique. Human autoptic tissue served as control (n=8). The region across the R/G sequence and the flip/flop sequence was amplified by PCR using GluR2-specific primers. The amount of RNA editing at the R/G site was analyzed by restriction analysis of the PCR products with the enzyme MseI, which identifies the sequence of the edited GluR2(flip) RNA. The relative portion of flip and flop sequences was determined by restriction analysis of the PCR products with the enzyme HgaI, which identifies the flipspecific sequence. The investigation revealed the following: (i) flip/ flop: No significant alterations were found in the relative abundance of the spliced variants containing the flip exon (hippocampus: autopsy 60%, epilepsy 53%; neocortex: autopsy 50%, epilepsy 51%); (ii) R/G editing: The relative amount of edited RNA is significantly increased in hippocampal tissue (autopsy: 56%, epilepsy: 69%; P=0.012). No significant alterations were found in neocortical tissue (autopsy: 61%, epilepsy: 70%). The data indicate that in hippocampal tissues of patients with TLE the time course of recovery from desensitization of AMPA receptors could be altered. Supported by IZKF Münster G7 Institut für Physiologie, Robert-Koch-Str. 27a, 48149 Münster

P 03-7 IONIC MECHANISM OF CELL SWELLING INDUCED BY ACTIVATION OF GLUTAMATE RECEPTORS IN LEECH RETZIUS NEURONES P.W. Dierkes, S. Neumann, G. Klees, P. Hochstrate, W.-R. Schlue Glutamate, the major excitatory neurotransmitter in the central nervous system of vertebrates, exerts its effects primarily by activating ionotropic receptors, i.e., glutamate-gated unselective cation channels. Under ischemic conditions, the persistent activation of the glutamate receptors has neurotoxic effects, which are at least partly due to the sustained membrane depolarization leading to the uptake of electrolytes and hence to cell swelling and eventually osmotic lysis. To characterize the ionic mechanisms underlying glutamate-induced cell swelling in more detail, we used multi-barrelled ion-selective microelectrodes for monitoring simultaneously the cell volume, the membrane potential as well as the intracellular concentrations of K+, Na+, and Cl- in leech Retzius neurones. In these cells activation of the ionotropic glutamate receptors by the glutamatergic agonists AMPA and kainate induced a marked increase in [Na+]i and [Cl-]i and a decrease in [K+]i. Furthermore, both agonists induced swelling of the cells by up to 20 %. The results show that about 75 % of the decrease in [K+]i was attributable to water uptake by the cells and only 25 % was related to K+ efflux. In contrast, the increases in [Na+]i and [Cl-]i were exclusively due to the influx of Na+ and Cl-. Correction for the dilution of the cytosol during cell swelling showed that glutamate receptor activation induced an electroneutral uptake of NaCl and hence an increase in the cytosolic osmolarity. The change in the osmotic gradient between intra- and extracellular medium seems to be responsible for the induction of cell swelling. In summary, our results indicate that glutamatergic stimulation induces a partial exchange of cytosolic K+ for extracellular Na+ as well as a marked electroneutral uptake of NaCl, resulting in an increase in the cytosolic osmolarity. Supported by the Ministerium für Schule und Weiterbildung, Wissenschaft und Forschung des Landes Nordrhein-Westfalen

S 231 Institut für Neurobiologie, Heinrich-Heine-Universität Düsseldorf, Universitätsstr. 1, 40225 Düsseldorf, e-mail: [email protected]

method for activation of VRL-2 and provide a reliable tool to study functional effects of VRL-2 stimulation. Neurology CEDD, GlaxoSmithKline, Harlow, Essex, CM19 5AW, UK

P 03-8 ANTIBIOTICS BLOCK MUSCLE NICOTINIC ACETYLCHOLINE RECEPTORS F. Schlesinger, K. Krampfl, R. Dengler, J. Bufler Myasthenia gravis (MG) is the most common primary disorder of neuromuscular transmission. In MG specific antibodies are directed against muscle nicotinic acetylcholine receptor (nAChR) channels. It is well known that the clinical symptoms of myasthenia gravis patients can worsen under therapy with several antibiotics. The molecular basis of this phenomenon remains elusive. In the present study we addressed the question, if there is an interaction of Penicillin G and Gentamycin with adult-type muscle nicotinic acetylcholine receptors. We employed the patch clamp technique in combination with a device for ultrafast solution exchange. To study if there is an interaction of antibiotics with nAChR channels in the unliganded state we performed preexposureexperiments with Penicillin G or Gentamycin. Repetitive 20-msec pulses of 1 mM ACh after equilibration with antibiotics revealed a dose-dependent reduction of the peak current amplitude of nAChR channels with a maximum depression of the current amplitude by 70% by 10 mM Penicillin G or Gentamycin. The kinetics of nAChR channel currents were not altered. This effect could be explained by a competitive block of nAChR channels. To test for an additional interaction of antibiotics with the liganded, open nAChR channels, coapplication experiments were performed. 200-msec pulses with 1 mM acetylcholine and Penicillin G or Gentamycin in the concentration range from 0.1 mM to 10 mM were applied. The current decay in the presence of the antibiotics was best fitted by the sum of two exponentials. The fast component of current decay represents the transition from the open state to blocked state of the open channels as recently shown. The fast time constant decreased and the proportion of the fast component increased with increasing blocker concentrations. This observed effect of antibiotics pointes to a block mechanism, like open channel block (Bufler et al., EJN, Vol. 6 pp. 677-687, 1996 and Bufler et al., J Pysiol., 495.1, pp. 83-95, 1996). Thus, we could describe the interaction between nAChR and antibiotics as a combination of open channel and competitive block on the molecular level. Therefore muscle weakness under antibiotic therapy of myasthenia gravis patients may be explained by a direct interaction of antibiotics with nicotinic receptors, resulting in a decreased safety margin of neuromuscular transmission. Supported by the DFG, Bu 938/2-3.

P 03-10 A HYPERTONICITY-INDUCED NON-SELECTIVE CATION CONDUCTANCE WITH UNUSUAL PHARMACOLOGY IN EHRLICH LETTRE-ASCITES TUMOR CELLS (ELAC) P. Lawonn1, E.K. Hoffmann2, C. Hougaard2, F. Wehner1 In many cell types studied, the regulatory volume increase following hypertonicity-induced cell shrinkage is known to involve the conductive uptake of inorganic osmolytes as one of the key mechanisms. In Ehrlich ascites tumor cells (EATC), the presence of a shrinkageactivated membrane current potentially leading to a net gain of intracellular cations was inferred from ion uptake studies (Hoffmann, 1978). The present study aimed to characterize the possible current in the very closely related ELAC by means of the conventional patchclamp technique. In whole-cell recordings, 62% of all cells tested (n=162) responded after an exposure to a 16%-hypertonic medium with the activation of an inward current (234 ± 32 pA), which reversed to resting level upon returning to isotonic medium. This current was accompanied by an approx. five-fold increase of cell membrane conductance and a shift of the reversal potential from -40.0 mV to 0.3 ± 2.0 mV (Vhold = -40mV). The current could be reversibly blocked by the ion channel inhibitors (in order of descending potency) benzamil ≅ gadolinium > amiloride >> EIPA. Ion substitution experiments revealed that the current could be carried with almost equal efficiancy by Na+, K+ or Li+, whereas it declined completely with NMDG+ or choline+ as the substitute. In cell-attached recordings, an ion channel with a single channel conductance of approx. 14 pS and a linear I/Vrelation was found to be activated by extracellular hypertonicity (21 observations in hypertonic medium vs. 0 observations in isotonic medium, 40 recordings each). This channel type is the likely candidate underlying the macroscopic whole-cell current following cell shrinkage. The data presented show consistently that in ELAC, shrinkage activates a non-selective cation conductance with a pharmacology that differs signifi-cantly from that of other volume-activated conductances reported so far. 1 Max-Planck-Institut für molekulare Physiologie, Otto-Hahn-Str. 11, D-44227 Dortmund; 2 August Krogh Institute, University of Copenhagen, Copenhagen, DK-2100 Denmark

Department of Neurology of the Medizinische Hochschule Hannover, Carl-Neubergstr.1, 30625 Hannover, Germany P 03-11 P 03-9 IDENTIFICATION OF HUMAN VANILLOID RECEPTOR LIKE 2 AND ITS ACTIVATION BY PHORBOLS G.D. Smith , P. Hayes, R. Kelsell, M.J. Gunthorpe, J.C. Jerman, J. Egerton, D. Smart, H. Watanabe, C.D. Benham, B. Nilius, J.B. Davis By screening nucleotide databases we identified a human vanilloid receptor-like channel (hVRL-2). Encoded by 2613 nucleotides, or a predicted 871 amino acids, and mapping to chromosome 12q24.1, hVRL-2 is 45% identical to hVR1 and is the human orthologue of the published OTRPC4. Although similar in structure and channel function, the vanilloid receptor gene family (VR1, VRL-1, ECAC1 and CaT2) differ in their ligand responses and, to date, VR1 is the only member that has been shown to respond to vanilloids, although several share common activators. A feature linking the receptors is response to, or modulation by, activation of PKC by phorbol esters. Hence, we have tested VR1 and VRL-2, using FLIPR and whole cell patch clamp studies, for responses to phorbols. These studies have identified 4α-phorbol12,13 didecanoate (4αPDD) as an activator of hVRL-2. Recombinant hVR1 responded to capsaicin, RTX and PMA in a concentration dependent manner (pEC 50 = 7.6±0.15, 8.89±0.11 and 7.73±0.16, respectively) whereas 4αPDD was inactive. hVRL-2 failed to respond to the vanilloids, capsaicin and RTX, but responded to PMA and 4αPDD (pEC 50 =5.77±0.16, 6.67±0.05, respectively) in a concentration dependent manner. VRL-2 responses to 4αPDD were inhibited by ruthenium red but not capsazepine. In addition, we have shown that these phorbol esters activate a current in mouse aorta endothelial cells, which endogenously express VRL-2. These findings provide a simple

INCREASED CATION CONDUCTANCE IN PLASMODIUM FALCIPARUM-INFECTED RED BLOOD CELLS V. Tanneur, C. Duranton, F. Lang , S.M. Huber Malaria-infected human red blood cells (RBCs) possess parasite-induced transport pathways, which supply the pathogen with a wide variety of nutrients. While normal RBCs exhibit a low membrane permeability, tracer flux measurements have shown that the pathways induced by the parasite dramatically increase the transport of ions and neutral solutes such as sugars and amino acids. Recent patch-clamp experiments have identified the presence of two oxidation-induced anion conductances in Plasmodium falciparum infected RBCs (Huber et al., EMBO J, 21(2): 22-30, 2002) but an infection-induced change in cation-conductance remained elusive. The patch-clamp technique was applied to identify the biophysical properties of a putative parasite-induced cation pathway. Using whole-cell configuration with symmetrical Na-gluconate solution in pipette and in bath, control RBCs exhibited a low cation conductance (250 ± 45 pS, n=6) whereas infected human RBCs had 5 to 10-fold larger Na+ currents (2.08 ± 0.40 nS, n=38). Cation substitution experiments showed that the parasiteinduced cation conductance was non-selective with a permselectivity of Cs + > K+ > Li+ = Na+, while NMDG+ induced an almost complete inhibition of the current. Significant inhibition was also observed in the presence of Gd3+, amiloride, quinidine and NPPB. Previous studies on normal RBCs has demonstrated an oxidation-induced cation conductance, which resembles that cation conductance of infected RBCs. Taken together these data suggest that an oxidative process is probably involved in the activation of endogenous erythrocyte cation channels by Plasmodium falciparum.

S 232 Department of Physiology I, University of Tübingen, Gmelinstr. 5, 72072 Tübingen

cells display increased activity and facilitated activation of L-type channels due to altered regulation by tyrosine kinase. We conclude that in AMD the increased secretory activity of the RPE is accompanied by increased activity of neuroendocrine L-type channels.

P 04-1

1

ALTERED CALCIUM-CURRENTS IN RETINAL GANGLION CELLS FROM AGRIN -/- MICE M.J. Frech, J. Schröder, S. Kröger, K.H. Backus Agrin is a key regulator during formation of the neuromuscular junction, but is also widely expressed in the CNS, including retinal ganglion cells, where its function is unknown. We analysed voltage-dependent Ca2+ currents (VDCCs) in retinal ganglion cells from organotypic cultures derived from wildtype (+/+) mice and from mice heterozygous (+/ -) or homozygous (-/-) for a deletion of the agrin gene. Morphologically identified ganglion cells from all animals showed a typical current response pattern characterized by large TTX-sensitive, voltage-dependent Na+ currents, VDCCs and outwardly directed K+ currents. In order to isolate VDCCs in ganglion cells recordings in the whole-cellclamp configuration were carried out in the presence of 0.5 µM tetrodotoxin and 20 mM tetra-ethyl-ammonium chloride to block voltagedependent Na + and K+ currents. Ganglion cells (n=58) showed two types of VDCCs: non-inactivating VDCCs and transient VDCCs that inactivated within 100 ms. Their relative proportion was dependent on the holding potential (Vh). Transient currents were observed when the Vh was set to –80 mV, but were not present at a Vh of –40 mV. 50% (9 out of 18) of the (+/+) ganglion cells and 43% (6/14) of the (+/-) ganglion cells expressed both non-inactivating and transient VDCCs during the entire time period studied (up to 26 days in culture). The other (+/+) and (+/-) ganglion cells showed only non-inactivating VDCCs. Significantly different VDCC patterns were observed in (-/-) ganglion cells and the differences depended on the age of the culture. 57% (4/7) of the (-/-) ganglion cells that were cultured up to 21 days showed transient and non-inactivating VDCCs, similar to cells from +/ + and +/- retinae. However, after more than 21 days in culture the transient VDCCs completely disappeared in all (-/-) ganglion cells investigated (n=9). We conclude that in retinal ganglion cells agrin may contribute to the expression or properties of a subtype of VDCCs by a yet unknown mechanism. Max-Planck-Institute for Brain Research, Neuronatomical Department, Deutschordenstr. 46, D-60528 Frankfurt/M., Germany

Institut f. Klin. Physiologie, UKBF, Freie Universität Berlin, 12200 Berlin; 2Augenklinik, UKBF, Freie Universität Berlin, 12200 Berlin

P 04-3 CALCIUM CURRENTS IN RAT ENTORHINAL CORTEX LAYER II STELLATE AND LAYER III PYRAMIDAL NEURONS IN ACUTE BRAIN SLICE V. Visan, U. Heinemann, W. Müller The entorhinal cortex is an essential neocortical - hippocampal relay station that plays an important role in memory formation as well as mesial temporal lobe epilepsy. Strong increases of [Ca 2+]i during epileptiform activity may be responsible for selective damage of layer III pyramidal neurons (Gloveli et al. 1999). Using infrared video-guided whole-cell patch clamp recording in the acute entorhinal cortex – hippocampus slice, we investigated voltage activated Ca2+-currents and their potential contribution to such [Ca2+]ichanges. Layer II stellate and layer III pyramidal neurons were identified by their characteristic morphology and low and/or high voltage activated (LVA & HVA) Ca2+-currents, activated by depolarizing voltage steps from a holding potential of -80 or –50 mV, respectively. Stellate and pyramidal cells had a membrane capacity of 32 pF vs. 20 pF and Ca 2+ -currents of 1100±150 pA vs. 700±140 pA, respectively. Mediation of these inward currents by Ca2+-channels was confirmed by concentration-dependent block by NiCl 2 that was more potent on stellate cell Ca2+-currents. HVA currents elicited from –50 mV were of almost the same amplitude in stellate and pyramidal neurons, while the additional LVA currents evoked from -80 mV were proportional to membrane capacitance. A steeper I-V relationship in pyramidal neurons suggests a stronger dendritic localization of Ca-channels. We conclude that not current density, but differences in cell surfaceto-volume ratio explain differences in Ca 2+-accumulations during repetitive firing and epileptiform activity. AG Molekulare Zellphysiologie, Neurowissenschaftliches Forschungszentrum der Charité, Humboldt Universität zu Berlin, Schumannstrasse 20/21, D-10117 Berlin, Germany

P 04-2 CURRENTS THROUGH L-TYPE CHANNELS IN RPE CELLS FROM CHOROIDAL NEOVASCULAR MEMBRANES: INCREASED CURRENT DENSITY AND ALTERED REGULATION R. Rosenthal1, H. Heimann2, M. H. Foerster2, O. Strauß1 Purpose: Age-related macular degeneration is accompanied by the development of new vessels in the macula. In this process retinal pigment epithelial (RPE) cells secrete the majority of angiogenic factors. Since RPE cells express neuroendocrine L-type Ca2+ channels we studied these channels in cells from surgical excised choroidal neovascular (CNV) membranes. Methods: Fresh RPE cells were isolated by trypsin/ protease digestion from CNV membranes. RPE cells were identified by staining for classIII beta-tubulin. Membrane currents were measured in the perforated-patch configuration of the patch-clamp technique. Results: Under intra-/extracellular K +-free conditions cells showed voltage- and time-dependent inward currents in the presence of extracellular Ba2+ (10 mM) which are sensitive to nifedipine (10µM). The currents showed a density of 20 pApF-1. The steady-state activation curve is shifted towards more negative values (V1/2 = -18 mV versus V1/ = -12 mV in human cells in primary culture). Cultured RPE cells from 2 CNV membranes showed a density of 3.9 ± 0.2 pApF-1 (n = 10) which is higher than in cultured cells from healthy donors (0.96 ± 0.11 pApF1 ; n = 8). In first experiments to study L-type channel regulation in cells from CNV membranes, tyrosine kinase inhibition (lavendustin A, 10 µM) increased maximal current amplitude to 124.8 ± 4.6 % of control (n = 10). Thus, tyrosinekinases inhibit L-type channels in cells from CNV membranes whereas in cells from healthy donors tyrosine kinase activate L-type channels (application of lavendustin: 79.9 ± 5.4 % of control; n = 4). Conclusions: This is the first functional study of RPE cells from CNV membranes of AMD patients. These

P 04-4 THE IMPORTANCE OF ICRAC FOR MAINTAINING RESTING CYTOPLASMIC [Ca2+] C. Kullen, I. Schulz, E. Krause Recently we have shown that “capacitative Ca2+ entry” (Putney (1990) Cell Calcium 11, 611-624) or “Ca-release-activated-Ca-current” (ICRAC, Hoth & Penner (1992) Nature 355, 353-356) which is induced by depletion of inositol 1,4,5 trisphosphate -sensitive Ca2+ stores can be also activated independently of global Ca2+ store depletion when cytoplasmic [Ca2+] ([Ca2+]cyt) is lowered below the resting [Ca2+]cyt (Krause et al. (1999) J. Biol. Chem. 274, 36957-36962). Using the fura-2 and patch-clamp techniques we now give evidence that activation of ICRAC by low [Ca2+]cyt together with the plasma membrane (PM)-Ca2+ pump determine the [Ca 2+] cyt at rest. In the presence of the membranepermeant Ca 2+ chelator BAPTA-acetoxymethylester (10 µM) a decrease in [Ca 2+]cyt from the resting [Ca 2+] cyt of 100 nM to 55 nM, (∆[Ca2+] = 45±15 nM, n = 10) resulted in activation of Ca2+ entry and in an increase in [Ca2+]cyt back to the resting level in RBL-1 cells. Moreover blockage of ICRAC at resting conditions with La3+ (100 µM) or 2-APB (50 µM) induced a decrease in [Ca2+]cyt by 15 ± 10 nM (n = 6) and 25 ± 8 nM (n = 5) in RBL cells and HEK 293 cells, respectively within 10 min. At concentrations of 2-APB > 70 µM, however, Ca2+ was released from intracellular stores and [Ca2+]cyt increased to 230 ± 70 nM (n = 5). We conclude from our data, that the PM-Ca2+ pump together with CRAC-channels control [Ca2+]cyt at rest. If ICRAC is blocked [Ca2+]cyt decreases to a new steady state which probably is determined by the Ca2+-pool-located Ca2+ pump (SERCA). This assumption suggests that the SERCA has a higher affinity for Ca2+ than the PM-Ca2+ pump. This idea is supported by our experiments that after blocking of

S 233 ICRAC with subsequent decrease in [Ca2+]cyt (up to 60 min) the stores were not depleted. The data indicate that the resting [Ca2+]cyt is determined by a minimal Ca2+ influx through CRAC counteracted by Ca2+ extrusion through the PM-Ca2+ pump. This guarantees filling of the Ca2+ pools via the SERCA.

Department for Physiology, University of Tübingen, Gmelinstr. 5, D 72076 Tübingen

Physiologisches Institut der Universität des Saarlandes, Geb. 58, 66421 Homburg/Saar

ACTIVE PROPERTIES OF LAYER 2/3 PYRAMIDAL CELL DENDRITES IN VITRO AND IN VIVO D. J. Waters, M. Larkum, B. Sakmann, F. Helmchen Pyramidal cells in layer 2/3 of neocortex are morphologically heterogenous with somata located 150-550 µm from the pia. The integration of synaptic inputs to the apical dendritic tufts in layer 1 critically depends on whether voltage-dependent ion channels are present in the dendrites. Here, we tested for active dendritic properties in these neurons both in vitro in brain slices and in vivo in urethaneanesthetized rats. Experiments were performed in somatosensory cortex of 28-day old rats using whole-cell patch-clamp recordings and calcium imaging. Both in vitro and in vivo, action potential (AP) amplitudes declined moderately along the apical dendrite as measured with dendritic pipettes (exponential space constant 360 µm). This suggests active AP backpropagation, which was confirmed by showing that in brain slices AP waveforms were attenuated strongly in the presence of TTX. Single APs evoked large calcium transients in the apical trunk but generally failed to evoke calcium influx in layer 1 branches except in very superficial neurons. Nevertheless, distal calcium influx could be induced by multiple APs delivered at high frequencies (>100 Hz). This was the case for several neurons in vitro and in vivo and could occur even in terminal branches, indicating the presence of voltage-gated dendritic calcium channels throughout the dendritic tree. In addition, all-or-none regenerative dendritic potentials could be evoked by direct dendritic current injection. These potentials could also be generated by pairing a somatic AP with distal dendritic depolarization. In some cases such pairing resulted in a second somatic AP. We conclude that layer 2/3 pyramidal cell dendrites contain voltage-dependent conductances, which are activated similarly in vitro and in vivo under the conditions studied here. These dendritic properties may allow layer 2/3 pyramidal neurons to associate basal and distal synaptic inputs.

P 04-5 INHIBITION OF CAPACITIVE CA2+ ENTRY UPON CD95 STIMULATION IN T CELL SUBPOPULATIONS L. Heil, A. Enzensberger, A. Seeger, A. Schäfer, F. Lang, A. LeppleWienhues Stimulation of the CD95 receptor inhibits capacitive Ca 2+ entry and thereby T cell proliferation. In this study, we investigated the CD95mediated Ca2+-block in subpopulations of murine thymocytes. CD4+CD8+ double positive and CD4+CD8- or CD4-CD8+ single positive thymocytes, respectively, are known to express CD95, whereas the expression in CD4-CD8- double negative cells is discussed controversially. Thymocytes of black 6 wild type mice were prepared by ficoll gradient centrifugation. The cell suspension was separated by MACS separation. The cells were incubated with primary anti-CD4 and anti-CD8-antibodies, respectively, and with secondary magnetic beads-labeled antibodies. The separation was confirmed by flow cytometry using FITC- or PE-conjugated antibodies against CD4 and CD8. According to our results thymocytes from wild type mice consist of 72,7% CD4+CD8+, 18,7% CD4 +CD8 - or 4,4% CD4 - CD8 + and 3,9% CD4 - CD8 - cells. Each subpopulation was then treated with soluble CD95 ligand for 1h. Intracellular stores were depleted by thapsigargin and reduced capacitive Ca2+ entry was assessed by FACS analysis. CD4+CD8+ murine thymocytes showed a marked inhibition of capacitive Ca 2+ entry, whereas in CD4 +CD8- or CD4-CD8+ cells Ca2+ influx was inhibited to a smaller extent. In CD4-CD8- cells hardly any CD95 mediated effect was seen. In addition, CD95-induced DNA fragmentation was analysed. CD4+CD8+ thymocytes were more susceptible to the apoptotic activity of soluble CD95 ligand than CD4+CD8- or CD4-CD8+ cells. Our data demonstrate differences of the T cell subpopulations in the sensitivity to CD95stimulation. The onset of the block of capacitive Ca2+ entry parallels the onset of CD95-induced apoptosis. Supported by DFG Le792/3-2 and Deutsche Krebshilfe 10-1825-Be2

P 04-7

Department of Cell Physiology, Max-Planck-Institute for Medical Research, Jahnstr. 29, 69120 Heidelberg, Germany P 04-8

Department of Physiology I, University of Tübingen, Gmelinstrasse 5, D-72076 Tübingen, Germany P 04-6 BA 2+ INCREASES CD95-RECEPTOR ABUNDANCE AND INDUCES APOPTOTIC CELL DEATH OF CHANG CONJUNCTIVA CELLS M. Müller, K.S. Lang, S. Fillon Among the hallmarks of apoptosis is cell shrinkage, which requires release of cellular electrolytes. Accordingly, apoptosis has been shown to trigger loss of cellular K+ and to activate K+ channels. On the other hand, stimulation of the CD95-receptor, which induces apoptotic cell death in lymphocytes, is paralleled by inhibition of voltage gated Kv1.3 K+ channels in Jurkat lymphocytes. Moreover, activation of K+ channels is one of the prerequisites for stimulation of cell proliferation by certain growth factors. Thus, the role of K + channels in the regulation of apoptosis and cell proliferation is still illdefined. In the present study FACS analysis has been performed to determine the role of K+ channel activity in the survival of Chang conjunctiva cells. Cell volume has been determined from forward scatter, DNA abundance from Syto16, permeability of the cell membrane from PI staining and phosphatidylserine translocation from annexin binding. Application of 5 mM Ba2+ to Chang cells leads to rapid cell swelling. However, within 2-3 hours the swelling is reversed in a fraction of the cells to cell shrinkage. The loss of cell volume is paralleled by increase of Syto16, PI staining and annexin binding, thus clearly pointing to the initiation of apoptosis (from 28 ± 8 % to 91 ± 4 % of the cells, n = 4). Labelling with human anti-CD95-receptor antibodies reveals that 5 mM Ba 2+ increases the CD95 receptor abundance at the cell membrane. In conclusion, sustained (3 hours) exposure to Ba 2+ (5 mM) triggers apoptotic cell death of Chang conjunctiva cells, an effect presumably in part due to upregulation of CD95 receptors.

THE REPONSE OF NG2 GLIA AND ASTROCYTES TO SPINAL β CORD INJURY IN THE ADULT RAT: A ROLE FOR TGFβ P.S. Hubbard, M. Berry, A.M. Butt Injury to the central nervous system (CNS) results in the laying down of a glial scar that is a potential ‘no-go’ zone for regenerating axons. The glial scar contains molecules that inhibit axon growth, one of which is NG2, a chondroitin sulphate proteoglycan (CSPG) that is expressed in the CNS by a mature glial cell that has many characteristics of astrocytes but has the antigenic phenotype of oligodendrocyte progenitor cells (OPC). The NG2 CSPG is a potent inhibitor of axon regeneration but also modulates receptor ligand interactions in NG2 glia (OPC). The aim of this study was to determine the effect of TGFβ on glial scar formation. A dorsal column crush of the spinal cord (T12) was performed in male rats (180-200g) under deep anaesthesia induced by a single intraperitoneal injection of combined Hypnorm/ Hypnovel anaesthetic (0.2133 mg.kg -1 fentanyl citrate, 6.75 mg.kg -1 fluanisone, 3.375 mg.kg -1 midazolam). The timecourse of the NG2 glia (OPC) and astrocyte response to injury were compared at post lesion (PL) days 2(n=6), 7(n=6), 10(n=5), 15(n=6), 30(n=5) and 60(n=6) as well as to control (n=6), unlesioned animals. TGFβ, or saline vehicle (controls) was administered via cannulae implanted in the subarachnoid space in single, daily 3µl (1.2ng TGFβ) injections before animals were sacrificed after 7 days (TGFβ n=5, Saline n=3). All rats were humanely killed by overdose of sodium pentobarbitone and perfused via the left cardiac ventricle with 4% paraformaldehyde. Spinal cords were dissected free and fixed for 2 hours in the same fixative, prior to either (1) cryoprotection in 20% sucrose overnight followed by freezing (20µm sections), or (2) embedding in gelatin (50µm sections). Sections were single or double immunolabelled with antibodies for NG2 or glial fibrillary acidic protein (GFAP, for astrocytes) using either FITC/TRITC immunofluorescence or avidin-biotin complex (ABC) methods. In the crush lesioned spinal cords, NG2 and GFAP labelling increased, and both NG2 glia (OPC) and astrocytes

S 234 exhibited the morphology of a reactive phenotype. NG2 glia reactivity was first observed after only 2 days PL with labelling concentrated around the immediate vicinity of the lesion site, whereas astrocytic reactivity was much more widespread and wasn’t apparent until 7 days PL, but remained throughout the time course. It was this result that determined Pl day 7 to be the most appropriate time point with which to study the NG2 glia and astrocytes response to TGFβ. TGFβ induced a large increase GFAP staining throughout the lesion site as well as an increase, though less marked, in NG2 expression when compared to controls. These results suggest that TGFβ augments the extent of glial scar formation by NG2 glia and astrocytes following injury. Supported by the International Spinal Research Trust. Neural Damage and Repair Group, Centre for Neuroscience, Hodgkin Building, GKT Guy’s Campus, Kings College, London, England, SE1 1UL

the substrate. Incubation of the peptide with both enzymes, however, yielded almost 100% reduction, indirectly showing that hCBS has MSRB activity. In vivo activity of hCBS was tested by coexpression with ShakerC/B potassium channels in Xenopus oocytes. Like hMSRA, hCBS accelerated the inactivation time course, which is an indication for a reduced Met residue in the inactivation ball domain of this channel. Hybridization of hCBS DNA to tissue-specific cDNA filters showed a ubiquitous expression of hCBS; strongest expression was found in the heart, followed by skeletal muscle and liver. In comparison, hMSRA expression is highest in the detoxifying organs kidney and liver as well as in the cerebellum. S. Jung, Klinikum FSU Jena, Molekulare und zelluläre Biophysik, Drackendorfer Str. 1, 07747 Jena, Germany

P 05-1 P 04-9 KIR4.1 AND POTASSIUM REGULATION IN THE MOUSE OPTIC NERVE S. Yamaguchi1, C. Knott1, G.P. Wilkin1, A.M. Butt2 There has been a long-standing proposal that CNS glial inwardly rectifying potassium channels (Kirs) are crucial for the removal of K+ from the extracellular spaces of the brain and spinal cord following neuronal firing. Among the channels, our particular interest focuses on a glialspecific protein, Kir4.1. Our immunohisotchemical study demonstrated Kir4.1 expression in optic nerve astrocytes and oligodendrocytes, suggesting that both types of glial cells may be involved in K+ regulation. Electrophysiological measurement of the compound action potential (CAP) in isolated mouse optic nerves showed that the CAP develops between postnatal 10 days (P10) and P18 and matures by P30. Stimulation of the optic nerves at high frequencies (50 Hz) caused an activity dependant decrease in CAP. This decrease was accelerated by Ba2+ , which blocks mainly glial Kir channels, and protected by tetraethylammonium chloride, which mainly blocks axonal Kv channels. To clarify the relation between the CAP decrease and extracellular K+ concentration ([K+]0), [K+]0 was measured using potassium electrodes during electrical stimulation. The activity dependant increase of [K+]0 was decreased with age and it was correlated with developmental changes in CAP. The results are correlated with K+ regulation being not developed in immature nerves. 1

Biochemistry Department, Imperial College London, London SW7 2AZ; 2Centre for Neuroscience Research, King’s College London, London SE1 1UL

MIGRATION OF NEUTROPHIL GRANULOCYTES ON FIBRONECTIN IS POTASSIUM CHANNEL INDEPENDENT B. Hinkes, M. Römer, A. Wulf, V. Dreval, A. Schwab Migration of neutrophil granulocytes from the vascular system into the surrounding tissue is a crucial step in the formation of a peripheral focus of inflammation. Little is known about the role ion channels and transporters such as Ca2+-dependent K+ channels (hIK1) or Na+/H+ and Cl -/HCO3— exchangers play in this context. It has been shown that along with cytoskeletal remodeling the function of these transport molecules is required for migration of fibroblasts and melanoma cells. We investigated in this work, whether the influence of the Ca 2+dependent K+ channel hIK1 on migration of human neutrophils depends on the extracellular matrix. Therefore, neutrophils were plated on glass coverslips coated with the extracellular matrix protein fibronectin (0-200µg/ml) and stimulated with the chemoattractant fMLP. Migration was monitored with timelapse video-microscopy. The influence of hIK1 channels on the rate of migration was determined by applying the channel-specific inhibitor clotrimazol or activator 1-EBIO. Migration of neutrophils depends critically on the fibronectin coating. A maximum speed of 6µm/min is obtained with a fibronectin concentration of 100µg/ml. Under these conditions, cells show only few adhesion complexes and migrate in an amoeboid manner. As the adhesive strength of the matrix increases (decrease of fibronectin concentration) the adhesion area increases and neutrophils migrate slowly in a fibroblast-like way. Surprisingly, the rate of migration does not change on any fibronectin coating under pharmacological modulation of hIK1 channels by clotrimazol or 1-EBIO. This is in contrast to the behavior of neutrophils migrating on poly-L-lysin. On poly-L-lysin, cells were slowed down dose-dependently by clotrimazol. Taken together we consider these results as a piece of evidence, that matrix-specific crosstalk between adhesion and transport molecules may exist.

P 04-10 Physiologisches Institut, Röntgenring 9, D-97070 Würzburg, Germany FUNCTIONAL CHARACTERIZATION AND TISSUE DISTRIBUTION OF A HUMAN METHIONINE SULFOXIDE REDUCTASE OF TYPE B: hCBS S. Jung, A. Hansel, H. Kasperczyk, T. Hoshi, S. H. Heinemann Proteins are subject to posttranslational modification by intracellularly produced reactive oxygen species. Oxidation of methionine (Met) results in methionine sulfoxide (MetO) and it is well established that excess MetO residues in proteins are found in aged tissues or under pathophysiological conditions. Methionine oxidation is a reversible process; with the aid of the thioredoxin/thioredoxin reductase system, methionine sulfoxide reductases (MSR) reduce MetO to Met. MSRs of type A (e.g. hMSRA) only reduce the stereoisomer Met-D-O, thus yielding maximally 50% reduction of a MetO-containing substrate. In contrast, it was shown recently that the YeaA gene product from E. coli displays MSRB activity, reducing Met-L-O (Grimaud et al., J. Biol. Chem. 2001, Oct 24 [epub ahead of print]). The aim of this study was to characterize the gene product of human cbs-1 which shows significant sequence homology to YeaA (52% identity). An open-reading frame coding for CBS was isolated from a human lung library. Upon deletion of the first 22 amino-acid residues and addition of a His-6-tag, hCBS protein was expressed in E. coli and isolated. Enzymatic activity was assayed by in vitro reduction and subsequent MALDI-analysis of the MetO-containing peptide KIFM(O)K. In the presence of 15 mM DTT, hCBS displayed a turn-over rate of 1.5 min-1, about three-fold lower than hMSRA. Both enzymes only reduced maximally 50% of

P 05-2 GABA IS RELEASED FROM RAT PANCREATIC β-CELLS BY Ca2+ DEPENDENT AND cAMP-REGULATED EXOCYTOSIS A.Wendt, M. Braun, J. Broman, L. Eliasson, H. Mulder, B. Birnir and P. Rorsman. Background and aim: The inhibitory neurotransmitter, γ-aminobutyric acid, GABA is present in β-cells at high concentrations but its function is unknown. It has been proposed that GABA released from the β-cells fulfils a paracrine function and may regulate hormone release within the islets of Langerhans. Here we have investigated whether GABA is released by exocytosis initiated by Ca2+-influx through voltage-gated Ca 2+-channels and whether the release is modulated by cyclic AMP. Methods: Rat pancreatic β-cells were transfected with GABAA-receptor Cl- channels and pre-incubated in 20mM GABA. Membrane currents were recorded using the perforated patch configuration of the patchclamp technique. Exocytosis was monitored as increases in membrane capacitance (∆Cm). Intracellular distribution of GABA was determined by electron microscopy and immunocytochemistry. Results: Membrane depolarisation from –70 to 0 mV evoked an inward Ca2+-current. In β-cells transfected with GABA A-receptors, but not in untransfected cells, transient outward Cl- current spikes (reminiscent of inhibitory

S 235 postsynaptic currents in neurones) were superimposed on the Ca2+current. Such transient Cl- currents (I Cl,GABA) were not observed when the depolarisations went to +60 mV, at which the Ca2+-current reverses, or in the presence of bicuculline. The latency between the onset of depolarisation and the first ICl,GABA was >10 ms, similar to that of ∆Cm. The relationship between ICl,GABA and capacitance increased was 27 fF/ event. Since an insulin-containing secretory granule produces a capacitance increase of ~2 fF, this suggest that >10 secretory granules are released per ICl,GABA. Increasing intracellular cAMP by extracellular application of forskolin (2 µM) doubled both the total capacitance increase and the number of I Cl,GABA elicited by a train of 500 ms depolarisations. Immunocytochemistry revealed that β-cells contain high levels of GABA with a distribution different from insulin. Indeed, electron microscopy demonstrated that GABA was excluded from the large insulin-containing secretory granules. Conclusions: GABA is released by Ca2+-dependent exocytosis of vesicles distinct from the LDCV, presumably the small synaptic-like microvesicles (SLMV). As previously described for the release of the insulin-containing secretory granules, exocytosis of GABA is potentiated by cAMP suggesting that the exocytotic machinery is similarly regulated for both type of vesicles. The low frequency of ICl,GABA taken together with the small diameter of SLMV (40 nm compared to ~300 nm for the insulin-containing granules) suggest that exocytosis of these vesicles contributes only marginally to depolarisation-evoked capacitance increases, which accordingly principally reflects release of insulin-containing granules. Department of Physiological Sciences Lund University, Tornavägen 10, BMC F11, 22184 Lund Sweden

P 05-3 ELUCIDATION OF THE PHYSIOLOGICAL ROLE OF GLUTAMATE DECARBOXYLASE BY SINGLE VESICLE MEASUREMENT OF GABA EXOCYTOSIS IN INSULINPRODUCING β -CELLS M. Braun, A. Wendt, H. Mulder, B. Birnir, P. Rorsman Background: The enzyme glutamate decarboxylase (GAD) is considered an important antigen for the autoimmune reactions which lead to destruction of the pancreatic β-cells in type 1 diabetes mellitus. The physiological role of GAD and its product, γ-aminobutyric acid (GABA), in the pancreatic islet is however still unclear. Within the pancreatic islet, GAD and GABA are found only in β-cells, whereas α- and δ-cells express GABA receptors. This has led to the hypothesis that GABA serves as a paracrine signalling molecule in islets. Objective: The objective of the study was to investigate, if GABA is released from pancreatic β-cells, and if so how this release is regulated. Method: The rationale of the approach is to transfect isolated β-cells with GABAA receptors, which are ligand-gated chloride-channels. GABA released from the β-cell can then bind to receptors on the same cell, giving rise to chloride currents that can be recorded using the patch-clamp method. Two adenoviral vectors encoding for the α 1 and β 1 subunits of the GABAA receptor, respectively, were constructed using the pACCMV/ pJM17 vector system. Freshly isolated rat β-cells were co-transfected with both viruses, and electrophysiological investigation was performed after 24-36 h. To increase the current amplitudes, the cells were preincubated with 20 mM GABA. Results: Successful transfection of βcells was confirmed by immunohistochemistry using an antibody against the GABAAα1 subunit. In electrophysiological experiments, application of 1 mM GABA to the bath solution evoked an inward current in transfected β-cells with a typical amplitude of 0.5 to 2 nA, while no GABA-evoked currents could be recorded in non-transfected β-cells. In recordings performed in the standard whole-cell configuration, infusion of the cell-interior with solution containing 2 µM free Ca++ resulted in spontaneous, transient inward currents of 20-150 pA amplitude that activated in 10-40 ms and inactivated in 200-400 ms. These transient currents could be reversibly blocked by 100 µM bicuculline, an antagonist of GABAA receptors, suggesting that the currents reflect the exocytosis of GABA-containing vesicles. Release of GABA-containing vesicles could be triggered by photorelease of NP-EGTA-caged Ca++, which confirms the dependence of GABA exocytosis on intracellular Ca++. Conclusion: GABA is secreted from pancreatic β-cells by exocytosis of GABA-containing vesicles, which presumably correspond to synapticlike microvesicles (SLMV). GABA secretion is triggered by a rise in intracellular Ca++. These results provide further evidence for a role of GABA as a paracrine signalling molecule in pancreatic islets.

Department of Physiological Sciences, Lund University, Tornavägen 10, BMC F11, 22184 Lund, Sweden

P 05-4 INVOLVEMENT OF CYTOSKELETON IN CELL VOLUME MAINTENANCE UNDER ANISOTONIC CONDITIONS IN LEECH NEURONS H.J. Wüsten, P.W. Dierkes, W.-R. Schlue Because cell shrinkage or swelling might impair normal cell function, animal cells possess mechanisms to regulate cell volume. Cell volume regulation is predominantly mediated by the transport of inorganic ions, such as K+, Cl-, and Na+, or of organic osmolytes across the cell membrane, or by the intracellular generation of organic compounds (molecular crowding). In addition to these mechanisms which tend to reduce the osmotic gradient, the cytoskeleton may contribute to cell volume regulation. To investigate the putative role of the cytoskeleton, we studied the effect of cytoskeleton modulators on cell volume changes induced by anisotonic bathing solutions in identified leech neurons. We used the fluorescent dye Fura-2 as a cell volume indicator by measuring fluorescence upon excitation at the isosbestic wavelength (360 nm) and as an indicator of the intracellular concentration of Ca2+ ([Ca2+]i). Changing the bath osmolarity by increasing or reducing the NaCl concentration led to cell shrinkage or swelling, respectively. The changes in cell volume were significantly increased upon disruption of microtubules by colchicine (25 nM) or vinblastine (0.2 mM). This effect was completely prevented by prior administration of the microtubule stabilizing agent paclitaxel (30 nM), which itself had no effect on cell volume. Cytochalasin B and D (0.5 mM), which are potent inhibitors of actin polymerisation, had no effect on the cell volume changes induced by anisotonic solutions. The changes in the extracellular osmolarity and the application of the cytoskeleton modulators had no significant effect on [Ca2+]i. The results suggest that in leech neurons microtubules are involved in cell volume maintenance, whereas actin filaments appear to play a minor role. Supported by the Graduiertenkolleg 320 ”Pathologische Prozesse des Nerven-systems: Vom Gen zum Verhalten”. Institut für Neurobiologie, Heinrich-Heine-Universität Düsseldorf, Universitätsstr. 1, 40225 Düsseldorf, E-Mail: [email protected]

P 05-5 CHANGES IN MEMBRANE POTENTIAL, INTRACELLULAR ION CONCENTRATIONS AND CELL VOLUME INDUCED BY ELEVATIONS IN THE EXTRACELLULAR POTASSIUM CONCENTRATION S. Neumann, P.W. Dierkes, W.-R. Schlue In both vertebrate and invertebrate nervous systems it is well known that enhanced neuronal activity raises the extracellular K + concentration ([K+]o) to an upper ceiling level of about 10 mM under physiological conditions (Baylor & Nicholls 1969, Dietzel et al. 1989). However, in various pathophysiological situations [K+]o may increase much further, e.g. to about 50 mM in rat hippocampal slices under hypoxic conditions (Müller & Somjen 2000). In many preparations it had been shown that an elevation of [K+]o is paralleled by cell swelling (Lang et al. 1998). To investigate [K+]o-induced cell volume changes and ion movements across the cell membrane of leech Retzius neurones, we used triple-barrelled ion-selective microelectrodes for the simultaneous measurement of membrane potential (Em), intracellular concentrations of K+, Na+ and Cl- ([K+]i, [Na+]i, [Cl-]i) as well as cell volume (Neumann et al. 2001). Upon raising [K+]o, a cell swelling occurred which was paralleled by a membrane depolarization The [K+]o-induced effects on cell volume and E m recovered after readmission of the physiological solution. Without correction for water uptake during cell swelling, [K+]i and [Na+]i were only slightly affected by increasing [K +]o, whereas [Cl -]i increased markedly. After calculation under the assumption of a constant cell volume the net electrolyte uptake can be quantified. The results showed that the cells take up considerable amounts of K+ and Cl-, increasing the intracellular osmolarity and thus promoting the influx of water and swelling of the cells. Supported by the Ministerium für Schule und Weiterbildung, Wissenschaft und Forschung des Landes Nordrhein-Westfahlen.

S 236 Institut für Neurobiologie, Heinrich-Heine-Universität Düsseldorf, Universitätsstr. 1, D-40225 Düsseldorf, e-mail: [email protected] P 05-6 POTENTIOMETRIC MEASUREMENT OF CELL VOLUME CHANGES UNDER VOLTAGE-CLAMP CONDITIONS IN INVERTEBRATE NERVE CELLS P. Coulon, P.W. Dierkes, S. Neumann, W.-R. Schlue Changes in cell volume can disturb the normal function of animal cells, and therefore, most cells are endowed with volume-regulating mechanisms. Experimentally induced changes in cell volume are often paralleled by changes in the membrane potential, affecting a variety of transport processes across the cell membrane and in turn volume regulation. We have shown previously that multi-barrelled ion-selective microelectrodes are a useful tool for the simultaneous measurement of cell volume and intracellular concentrations of ions that might be relevant for volume regulation [Neumann et al. (2001) Electrochim Acta, 47: 309]. In order to investigate whether voltage-dependent transport processes are involved in cell volume regulation, we combined the potentiometric method of ion-selective microelectrodes with the voltage-clamp technique. This combination allows simultaneous recordings of cell volume, relevant intracellular ion concentrations, and ion currents across the cell membrane at a fixed membrane potential. Under voltage-clamp conditions (Eh = -50 mV), reducing extracellular osmolarity by 40% induced a reversible cell swelling of 55 + 26 % and an increase in IVC of +1.2 + 0.6 nA (n = 6). Under clamp-free conditions cell volume increased by 57 + 35 % and cells hyperpolarized by -7.8 + 6.2 mV (n=6). The time course of cell volume changes was virtually identical under both experimental conditions. The results show that fixing Em to -50 mV does not affect cell swelling induced by reducing extracellular osmolarity, suggesting that changes in Em that parallel hyposmotic-induced cell swelling are not involved in cell volume regulation. Supported by: Ministerium für Schule und Weiterbildung, Wissenschaft und Forschung des Landes Nordrhein-Westfalen. Institut für Neurobiologie, Heinrich Heine Universität Düsseldorf, Universitätsstr. 1, 40225 Düsseldorf, Germany, E-Mail: [email protected] P 05-7 REGULATORY VOLUME DECREASE (RVD) BY ARTICULAR CHONDROCYTES ISOLATED FROM NORMAL AND DEGENERATE HUMAN ARTICULAR CARTILAGE A.C. Hall, M.J. Kerrigan, P.G. Bush, D. Dunlop+ An increase in articular cartilage hydration is the first macroscopic event occurring during the process of osteoarthrosis (OA; Stockwell, 1991). It is not known how this influences cartilage cells (chondrocytes) however alterations to chondrocyte volume have deleterious effects on extracellular matrix (ECM) metabolism (Urban et al., 1993). Here, we have compared the capacity of chondrocytes isolated from normal or degenerate human cartilage to undergo RVD following acute hypoosmotic challenge. Human tibial plateau cartilage was obtained with Ethical permission following knee replacement surgery and graded according to gross morphology and surface fibrillation. Chondrocytes were isolated from normal or degenerate explants following incubation with collagenase (Type I, 1.0mg/ml) in DMEM (380mosm l-1). Cells were allowed to settle on glass coverslips with fura-2AM (5µM; 30min; 37 o C), and volume measured (Kerrigan & Hall, 2000). Reducing osmolarity (to 220mosmol l-1) caused a rapid (within 2min) increase in cell volume. There was no significant difference (p>0.05; Student’s unpaired t-test) in the extent of cell swelling, or in the % of cells capable of RVD between chondrocytes from normal or degenerate cartilage. There was also no difference (p>0.05) in the rate of RVD between chondrocytes from normal or degenerate cartilage (t½ 6.3±1.87 and 6.1±0.88mins respectively; data are means ± s.e.m. for n=8 knees; 147 cells). The hypotonic challenge resulted in a transient rise in intracellular calcium ([Ca 2+]i) with no difference (p>0.05) between cells from normal or degenerate tissue. The [Ca2+]i rise did not correlate with RVD. These data suggest that in response to a hypotonic challenge there was no difference in the extent of swelling, in the proportion of cells showing RVD, or in the rate of RVD between chondrocytes isolated from normal or degenerate human cartilage. Therefore, despite the

changes to cartilage during degeneration, the chondrocytes isolated from this tissue seem to respond to hypotonic challenge in the same way as cells from normal cartilage. Previous work on in situ chondrocytes however, showed an increase in resting chondrocyte volume with advanced cartilage degeneration (Bush et al., 2000) suggesting that as the disorder develops, control of the chondrocyte volume set-point may not be maintained. Bush, P.G., Hoemann, C.D., Adams, C., & Hall, A.C. (2000). J. Physiol. 527P, 40P; Kerrigan, M.J. & Hall, A.C. (2000). J. Physiol. 527P, 42P; Stockwell, R.A. (1991). Clin. Anat. 4, 161-191; Urban, J.P.G., Hall, A.C. & Gehl, K.A. (1993). J. Cell. Physiol. 154, 262-270. Supported by the MRC, Arthritis Research Campaign (H0621) and the Wellcome Trust. Membrane Biology Group, Dept. of Biomedical and Clinical Laboratory Science, University of Edinburgh, Edinburgh EH8 9XD, UK. +Princess Margaret Rose Orthopaedic Hospital, Frogston Road West, Edinburgh, EH10 7ED P 05-8 VOLUME CHANGES OF LEECH RETZIUS NEURONS IN SITU AND AFTER ISOLATION FROM ADJACENT CELLS G. Klees, S. Neumann, P.W. Dierkes, W.-R. Schlue The cell membranes of animal cells are highly permeable to water. Under anisosmotic conditions water is flowing into or out of the cell leading to cell swelling or shrinkage, respectively. In leeches the central nervous system is located within the ventral hemocoelomic sinus. The nervous tissue is enclosed by a capsule, composed of a superficial epithelium and a fibromuscular layer, that is permeable to ions and small proteins. Physiological situations such as feeding alter the osmolarity of the hemolymph and therefore, the cells in the nervous system should change their volume. To investigate whether the capsule affects such volume changes or their kinetics, we monitored the cell volume of Retzius neurons in the segmental ganglia of the leech Hirudo medicinalis under anisosmotic conditions. The experiments were done on cells in situ and after detaching their cell bodies. For the measurements, we used the ratiometric fluorescent dye Fura-2 as a cell volume indicator by measuring changes in fluorescence intensity at the respective isosbestic wavelenght (360nm) and as an indicator of the intracellular concentration of Ca2+ ([Ca2+]i ) or double-barrelled ion-selective microelectrodes, which measured the concentration of a volume marker as well as the membrane potential (Em). Cell swelling (+46 %) was induced by raising the NaCl concentration by 50 % and cell shrinkage (-40 %) by raising the NaCl concentration by 100%. Anisotonic conditions had only small effects on [Ca 2+]i, whereas in most cells Em depolarized in hypertonic solutions (up to 10 mV) and hyperpolarized in hypotonic solutions (up to –12 mV). The results showed that both amplitudes and kinetics of the experimentally induced changes in cell volume and Em remained unchanged after the isolation of the cell bodies, indicating that the outer capsule has no limiting effect. This work was supported by the Ministerium für Schule und Weiterbildung, Wissenschaft und Forschung des Landes NordrheinWestfalen. Institut für Neurobiologie, Heinrich-Heine-Universität Düsseldorf, Universitätsstr.1, 40225 Düsseldorf, e-mail: [email protected] P 05-9 INACTIVATION OF CFTR GENE IN KIDNEY USING TISSUESPECIFIC RECOMBINATION BY CRE/LOX SYSTEM L.Hasséine, G.Bertin, O.Lorenzi, C.Poujeol, P.Poujeol, M.Tauc CFTR is considered as a Cl- channel localized in the apical membrane of epithelial cells including kidney. This protein is known to be implicated in the control of ionic channels or transporters present in renal tubules. Even if these relations are intensively studied, it remains that the role of CFTR in the kidney homeostasis is not clearly established. Thus, to evaluate the physiological role of CFTR in the kidney, we decided to generate transgenic mice to invalidate CFTR specifically in this organ by the CRE/LoxP system. To direct the expression of CRE in the distal part of the kidney in which CFTR is expressed, we chose to clone the mouse renal specific promoter directing type 2 vasopressin receptor (VR2) expression. The transgene pVR2-CRE was microinjected into the pronuclei of B6D2F1 mice. 10 transgenic strains expressing VR2-CRE transgene were detected after PCR analysis. RTPCR analysis of CRE mRNA showed that 3 lines expressed CRE in the

S 237 kidney as well as in other tissues (intestine, lung, spleen, liver) depending on strains. To assess the efficiency of CRE mediated site specific recombination, transgenic mice were mated with flox lacZ mice. Recombination events in pups were detected by PCR and Xgal staining in tissue sections. In one mouse strain (L5) CRE was functionally expressed in kidney and spleen but not in intestine or lung. Moreover, Xgal staining of kidney sections revealed a tubular expression of CRE. A second transgenic mouse strain which express a floxed CFTR gene is under construction. This is performed by homologous recombination in ES cells using 8 kilobases of the CFTR gene containing exon 11 flanked by 2 loxP sites and two selection genes (Thymidine kinase and NeomycineR). The inactivation of CFTR in the kidney will be obtained by mating of pVR2-CRE(L5) transgenic mouse strain with the floxed CFTR mouse strain. These transgenic mice will allow for the study of the physiological role of CFTR in the kidney. CNRS-UMR 6548 Parc Valrose, Université de Nice-Sophia Antipolis, 06108 Nice Cedex, France

(10µg/ml) was incubated with 3H-Ado (0.01-60 µmol/l) over night at 4°C in Tris/Hepes pH 7.4. To identify the Ado binding sites, SAH hydrolase was incubated with [2-3H]-8-azido-Ado. After irradiation the reaction mix-ture was digested by Asp-N or trypsin and the peptides, purified by HPLC, were sequenced. Results: Reconstituted SAH hydrolase with a ratio of NAD+/NADH 0.5/0.5, exhibits two binding sites for Ado with a KD1 of 22.2±1.4 nmol/l and a KD2 7.6±0.2 µmol/l, respectively, which correspond to peptides of SAH hydrolase identified as Asp307-Val325 and Tyr379-Thr410. Binding of Ado to SAH hydro-lase with NAD+/NADH ratios of 1/0 and 0/1 revealed only one binding site with low affinity of 5.3±0.6 µmol/l and high affinity of 56.7±0.6 nmol/l, res-pectively. Only one photolabelled peptide was identified as Asn312Lys318 from the NADH form, and as Asp391-Ala396 from the NAD+ form. Conclusion: Our data show that SAH hydrolase has two different binding sites for Ado. This finding may help to elucidate the enzymatic mechanism. Depts. of 1Pharmacology, Faculty of Medicine and 2Physical Biochemistry, Eberhard-Karls Universität, 72074 Tübingen, Germany

P 05-10 P 06-2 RENAL INNER MEDULLARY CELL VOLUME REGULATION DURING ACUTE ANTIDIURESIS: AN EVIDENCE-BASED HYPOTHESIS R.O. Law During sustained antidiuresis in vivo, or exposure to hypertonic media in culture, renal inner medullary cells counter shrinkage by accumulating methylamines and polyols (osmolytes) (Burg MB et al. (1997) Annu. Rev. Physiol. 59, 437; Beck F-X et al. (1998) Pflug. Arch. 436, 814). Methylamines also protect macromolecules against the destabilizing effects of urea, which accumulates during antidiuresis and readily penetrates cells. But increases in levels of these osmolytes only occurs over many hours/days. How therefore do cells maintain their volume in acute hydropenia, e.g. following severe haemorrhage? Medullary cells contain significant levels of amino acids, which provide osmoprotection in several cell types (Law, RO (1991) Comp. Biochem. Physiol. 99A, 263) but such a role in the kidney has been doubted as levels (except for taurine) are not increased by prolonged exposure to hypertonic media in vivo (Nakanishi T et al. (1991) Am. J. Physiol. 261, F957) or in culture (Burger-Kentischer A. et al. (1999) Kidney Int. 55, 1417). But medullary amino acid levels do increase rapidly and markedly following acute hypovolaemic oliguria in rats (Law RO (1991) Pflug. Arch. 418, 442). It is here hypothesized that these provide temporary osmoprotection over the longer timespan over which other osmolytes accumulate, and that the following sequence of events occurs: (1) interstitial levels of urea increase rapidly; (2) urea penetrates cells and destablizes enzymes responsible for amino acid catabolism (e.g. see Yancey PH et al. (1982) Science 217, 1214); (3) Intracellular levels of amino acids thus increase; (4) over many hours other osmolyte levels increase, with methylamines counteracting the effect of urea; (5) amino acid levels gradually return to normal, and methylamines and polyols provide long-term volume-regulatory osmoprotection.

EXPRESSION AND REGULATION OF ENOS AND INOS IN MULTICELLULAR PROSTATE TUMOR SPHEROIDS BY REACTIVE OXYGEN SPECIES M. Wartenberg, M. Schallenberg, J. Hescheler, H. Sauer Nitric oxide generated by either endothelial nitric oxide synthase (eNOS) or inducible nitric oxide synthase (iNOS) may be involved in prostate tumorigenesis. In the present study the expression of eNOS as well as iNOS was investigated in multicellular tumor spheroids of the androgeninsensitive prostate cancer cell line DU-145. With increasing size of tumor spheroids eNOS expression was downregulated whereas an upregulation of iNOS expression was observed. During the time course of tumor spheroid growth NO generation declined as evaluated by the NO-specific fluorescent dye, diaminofluorescein-2 diacethyl (DAF-2 DA), indicating that NO generation in DU-145 tumor spheroids is mainly mediated by eNOS. Elevation of intracellular reactive oxygen species (ROS) by treatment of tumor spheroids with either buthionine sulfoximine (BSO) or hydrogen peroxide resulted in upregulation of eNOS whereas iNOS was downregulated. These effects were abolished in the presence of the free radical scavengers vitamin E and ebselen. Furthermore, eNOS was upregulated by epidermal growth factor (EGF) which elevated intracellular ROS levels. The upregulation of eNOS following treatment with hydrogen peroxide was abolished by agents interfering with receptor tyrosine kinase signaling pathways, i.e. the protein kinase C (PKC) inhibitor bisindolylmaleimide I (BIM-1), the p21 ras farnesyl protein transferase (FPT) inhibitor III, the c-Raf inhibitor ZM 336372 and PD98059 which inhibits ERK1,2 activation. It is concluded that ROS involved as second messengers in receptor tyrosine kinase signaling pathways may act as positive regulators of eNOS expression. Institut für Neurophysiologie, Universität zu Köln, Robert-Koch-Str. 39, D-50931 Köln, Germany

Department of Preclinical Sciences, University of Leicester, Leicester LEI 7RH, UK P 06-3 P 06-1 CHARACTERIZATION OF ADENOSINE BINDING SITES OF SADENOSYLHOMOCYSTEINE (SAH) HYDROLASE FROM BOVINE KIDNEY D. Kloor1, A. Lüdtke1, S. Stoeva2, H. Osswald1 S-adenosylhomocysteine (SAH) hydrolase catalyzes the reversible hydrolysis of SAH to adenosine (Ado) and homocysteine and regulates Sadenosylme-thionine dependent transmethylation reactions. One mol of SAH hydrolase contains 4 mol tightly bound NAD+. Ado leads to a gradually reduction of the tightly bound NAD+ to NADH in the SAH hydrolase. The kinetics of Ado binding to SAH hydrolase suggested two or three binding sites. To decide between these possibilities we incubated in the present study modified SAH hydrolase with defined NAD+/ NADH ratios with [2-3H]-8-azido-Ado. Methods: SAH hydrolase was purified to homogeneity from bovine kidney. To prepare the enzyme with defined NAD+/NADH ratios, bound NAD was removed by dialysis and the apo-enzyme was reconstituted with NAD+/ NADH ratios between 1/0 and 0/1. For saturation binding experiments the enzyme

HAEMOGLOBIN EXPRESSION IN THE BRAINS OF ERYTHROPOIETIN-OVEREXPRESSING TRANSGENIC MICE M.H. Maurer, A. Schneider1, M. Gassmann2, W. Kuschinsky OBJECTIVE: Erythropoietin has been found in the central nervous system. A transgenic mouse model overexpressing human erythropoietin enabled us to study brain mRNA and protein expression in response to erythropoietin exposure. By gene chip analysis, we had earlier found haemoglobin-alpha mRNA being up-regulated in these brains. This study questions: (1) is haemoglobin mRNA expressed in the brain itself, (2) if so, where is it located, and (3) is haemoglobin protein expressed in the brain? SUBJECT AND METHODS: (1) 6 transgenic mice overexpressing human erythropoietin and 6 wild-type littermates were perfused with ice-cold saline to eliminate blood haemoglobin. Brain total RNA was extracted, and mRNA expression for brain haemoglobin-alpha was measured by RT-PCR in the LightCycler. (2) The expression of brain haemoglobin-alpha mRNA was localized by in situ hybridization using digoxigenin labelled oligonucleotides from the brains of 3 transgenic and 3 wild-type mice. (3) Total brain proteins were extracted and solubilized from the brains of 6 transgenic and 6 wild-

S 238 type mice. Brain haemoglobin protein expression was compared by two-dimensional gel electrophoresis. RESULTS: (1) Whole brain haemoglobin mRNA levels increased by a factor of 3.8 ± 0.2. (2) Brain haemoglobin mRNA is mainly expressed in the hippocampal formation, cerebellar cortex, and thalamic nuclei. (3) Whole brain haemoglobin protein expression increased by a factor of 3.0 ± 0.33. CONCLUSIONS : The novel finding that erythropoietin induces haemoglobinalpha expression on the mRNA and protein level in the brain indicates a function of haemoglobin in the brain. Brain haemoglobin may serve as oxygen reservoir, oxygen sensor, and/or enzymatic nitric oxide detoxifier. Dept. of Physiology and Pathophysiology, University of Heidelberg, Im Neuenheimer Feld 326, D-69120 Heidelberg, 1Axaron Bioscience AG, Heidelberg, Germany, and 2Dept. of Physiology and Veterinary Physiology, University of Zürich, Switzerland

peroxide and catalase in cerebrospinal fluid (CSF) of afebrile (n = 12) or febrile rabbits (n = 10) disclosed a positive linear correlation with those of simultaneously drawn plasma samples (y = 0.98 x ± 0. 64 µM H2O2, r2 = 0,75; y = 0.98 x ± 0.22 U catalase, r2 = 0.85). CSF endotoxin values were 7.4 ± 2.2 pg/ml 1 h after systemic injection of 1 µg/kg LPS, and 26.1 ± 5.2 pg/ml 3 h after LPS, the latter value significantly higher as the simultaneously drawn plasma value of 17.5 ± 4.1 pg/ml endotoxin. Perfusion of brains with NBT 1 h after the injection of either saline or LPS revealed that formazan was scarcely visible only within the subarachnoid space of afebrile rabbits, while in fever formazan-positively stained ramified structures, identified as microglia, were visible particularly over cortical areas, and in circumventricular organs. Methylene blue distinctly inhibited the LPS-induced formazan production in the subarachnoid space. We conclude that fever is the reaction of the brain to oxidative stress and that temperature homeostasis depends to a great extent on the activity of those nonneuronal structures, particularly the astro- and microglia, which provide the redox homeostasis of the brain.

P 06-4 FUNCTIONAL PROTON CHANNELS ARE REQUIRED FOR RESPIRATORY BURST GENERATION IN HUMAN B LYMPHOCYTES A. Gratopp, T. Schilling, C. Eder Voltage-gated proton currents were investigated in human lymphocytes using the whole-cell configuration of the patch-clamp technique. In response to membrane depolarization, activation of slowly activating, outwardly rectifying currents was detected in isolated CD 19+ B lymphocytes as well as in CD4+ and CD8+ T lymphocytes. These currents were highly selective for protons. Following increases in the extracellular pH, augmentation of proton current amplitude was observed due to shifts of activation threshold and current-voltage curve of the currents to more hyperpolarizing potentials. Time-dependent activation and deactivation of proton currents could be fitted by a single exponential. Proton currents in lymphocytes were inhibited by extracellularly applied divalent cations, including Zn2+ and Cd2+. B lymphocytes expressed proton currents of substantial amplitudes. In contrast, T lymphocytes exhibited proton currents of much smaller amplitudes than B lymphocytes. Statistical analyses revealed that the current density of proton currents in T lymphocytes was 25 times smaller than the proton current density of B lymphocytes. Following activation with the phorbol ester PMA, B lymphocytes were capable of generating reactive oxygen radicals. In contrast, superoxide anion release was not detected in T lymphocytes in response to PMA stimulation. Inhibition of proton channel activity by extracellular Zn2+ or Cd2+ caused significant reduction in superoxide anion release of human B lymphocytes, suggesting that functional proton channels are required for the respiratory burst generation in human B lymphocytes. During the respiratory burst, proton currents help maintain both membrane potential and intracellular pH. Supported by SFB 507. Institute of Physiology, Humboldt University, Tucholskystr. 2, D 10117 Berlin, Germany P 06-5 INHIBITION OF OXYGEN RADICAL FORMATION BY METHYLENE BLUE, α-LIPOIC ACID, OR ASPIRIN, PREVENTS BACTERIAL LIPOPOLYSACCHARIDE (LPS)-INDUCED FEVER W. Riedel Treatment with methylene blue has been shown to suppress lipid peroxidation and fever following systemic injection of bacterial endotoxin (LPS) in rats and rabbits (Riedel and Maulik, Mol Cell Biochem 196, 1999, 125-132). In the present study, we investigated whether aspirin and α-lipoic acid might similarly exert their antipyretic effects by diminishing oxidative stress. Incubation of human blood with nitroblue tetrazolium (NBT) showed that LPS increased the number of formazan-positive neutrophils from 10 ± 2% to 52 ± 3%. Adding LPS to blood preincubated with either methylene blue, aspirin or αlipoic acid caused the number of formazan-positive neutrophils to decrease to 0.9 ± 0.3%, 0.8 ± 0.3%, or 2.0 ± 0.4%, respectively. LPSinduced fever in rabbits is characterized by a rapid increase of plasma hydroperoxide levels and simultaneous decrease of catalase. Pretreatment with methylene blue, aspirin and α-lipoic acid significantly diminished the increase of plasma hydroperoxides following LPS and prevented the decrease of catalase, and fever. None of the three drugs diminished the elevation of plasma TNFα following LPS. Hydrogen

Max-Planck-Institute, W.G. Kerckhoff- Institute, D-61231 Bad Nauheim P 06-6 CARDIOPROTECTION OF IMMATURE HEART: THE EFFICACY OF NORMOTHERMIC CARDIOPLEGIA WITH OR WITHOUT REPERFUSION WITH TERMINAL CARDIOPLEGIA M.-S. Suleiman, H. Imura Objectives: This study was designed to investigate the cardioprotective action of normothermic cardioplegic arrest with or without terminal cardioplegia in immature rabbit hearts. Methods: Immature rabbits (710 days) were humanely killed and hearts perfused in a working mode with modified Krebs-Henseleit (K-H) buffer and maintained in a thermostatically controlled chamber at 390C. Hearts were exposed to 60 min normothermic ischemia with or without cardioplegia (K-H buffer with final KCl and MgCl2 at 20 and 16mM, respectively) and reperfused with K-H buffer for 40 minutes. In a third group, the cardioplegic arrest was followed by 2 min reperfusion with cardioplegia prior to re-introducing normal buffer. Cardiac output and release of lactate dehydrogenase (myocardial function and injury) were measured before ischemia and upon reperfusion. Results: Functional recovery (% of initial cardiac output, mean±SE, n=6 in each group) following 60 min global ischaemia was 9±6%. Ischaemic arrest with cardioplegia significantly (p<0.05, unpaired t-test) improved recovery (65±4%). Recovery was further augmented by 2 min terminal cardioplegia to 87±4% (p<0.05). Improvement in functional recovery was paralleled by a significant (p<0.05) reduction in lactate dehydrogenase release measured over 40 min reperfusion (0.5± 0.04, 0.32 ±0.03 and 0.23± 0.02 IU/g dry weight for control, cardioplegia and cardioplegia + terminal cardioplegia, respectively). Conclusions: This work shows that in the immature heart, normothermic cardioplegia confers significant protection against ischaemia and reperfusion injury. This protection can be further augmented by a terminal short period of reperfusion with cardioplegia. This additional protection is likely to be due to i) removal of accumulated waste products, ii) resuming supply of substrates and iii) decrease in energy demand by maintaining cardiac arrest. Bristol Heart Institute, Bristol University, Bristol BS2 8HW, UK P 06-7 EFFECTS OF ISCHEMIA REPERFUSION ON CONTRACTILE FUNCTION IN PROCAINE TREATED RATS F. Revnic, B. Voiculescu, S. Botea, C.R.R. Evnic, V. Voiculescu Heart muscle is largely dependent on uninterrupted blood flow which guarantees delivery of substrates and washout of harmful products of metabolism. The death of cardiac cells during ischemia and reperfusion is partially mediated by apoptosis. The aim of this study was to investigate the effect of ischemia reperfusion (IR) upon cardiac physiological parameters (CF)coronary flow, (HR) heart rate as well as LVPD (left ventricle pressure developed) as well as on apoptosis in adult Wistar rats feed on cholesterol diet and treated with Procaine (20mg/kg body weight) for 8 weeks. We used an ischemia (45 minutes) followed by 3 hours reperfusion model of isolated rat heart in Langerdoff retrograde perfusion. TACS Apoptotic DNA Laddering kit has been

S 239 used to assay heart cells for apoptosis. Our data have pointed out that following 45 minutes ischemia, myocardial excitation is decreased and action potential shortened upon opening K+ ATP channels. Inorganic phospahate, lactate and H+ accumulate and affect myofilament Ca 2+ sensitivity and maximum force development.The simultaneous activation of inward currents favours K+ loss and overload creating conditions that generate arrythmias. Reperfusion injury manifested clinically by reperfusion arrythmias and postischemic contractile dysfunction. There is a significant decrease in coronary flow (13 ml after 20 minutes) stabilization to 3 ml at the end of 3 hours reperfusion. Cardiac frequency is maintained relatively constant with minor fluctuations as well as LVPD in Procaine treated rats versus controls.

has an optical positioning system to align the center of stretch in the experimental chamber with the optical path of the microscope, thereby preventing lateral shift of the cells under study. Under observation, cells can also be perfused while mechanical stimulation occurs, enabling the additional application of fluid shear stress. The device enables for the first time quantitative and qualitative control of single cell stretch during microscopic investigation with high temporal and spatial resolution. Increasing the field of applications, the device can easily be combined with fluorescence and confocal microscopy. Supported by the FWF, grant P12974-MED and P14263-MED. Department of Medical Physics, University of Innsbruck, Müllerstraße 44, A-6020 Innsbruck

N.I.G.G.”ANA ASLAN”,9 Caldarusani Str.Sector1,78178, Bucharest, Romania P 07-1 P 06-8 ISOLATION OF ENDOTHELIAL CELLS FROM SMALL TISSUE SAMPLES L. Da Silva-Azevedo1, O. Baum2, H. Habazettl1,3, R. Gossrau2, A. Pries1,3 Introduction: Identification as well as quantification of endothelialspecific angiogenesis factors, especially at the protein level, still remains a technically demanding problem. Due to the small proportion of endothelial cells (ECs) in tissues like skeletal muscle (SkM), it is mandatory to isolate pure ECs in sufficient amounts in a relatively short period of time. Methods: Tibialis anterior and extensor digitorum longus were dissected from C57B16 mice and digested with collagenase II for 30 min at 37°C. The obtained digest was filtered through two cell strainers (l00µm and 40µm) and incubated with BS-I lectin-coated magnetic beads for 30 min at 4°C. Bound ECs were captured with a magnet. The purity of the EC fraction (e.g. its contamination with SkM fibers) was monitored by Western blot analysis using anti-eNOS and anti-αactin antibodies, whereas the enrichment and recovery of ECs were quantified by determination of the capillary-specific alkaline phosphatase activity. Results: Collagenase digestion resulted in the release of 10-15% of ECs of the microvasculature present in the dissected muscles. ECs bound strongly and specific to the BS-I-coated beads (recovery was above 95% of the released cells). ECs from the microvasculature were 10-20 fold enriched in the EC fraction, as judged from the increased alkaline phosphatase activity. Contamination of ECs with SkM fibers was negligible, as determined from the analysis of the anti-α-actin blots. 40-80% of ECs could be detached from the beads by competing with the BS-I ligand, methyl-α-D-galactopyranoside. Cell vitality was above 90%, according to trypan blue staining. This procedure was successfully adapted to heart and brain. Conclusions: We have developed a fast method for the isolation of ECs from mouse SkM based on the use of lectin-coated magnetic beads. The obtained ECs are at least free from SkM fibers. This method may be used for the isolation of ECs from small tissue samples, and is suitable for analysis of angiogenesis factors by proteomics and RT-PCR. 1

2

Departments of Physiology and Anatomy , Universitatsklinikum Benjamin Franklin, FU-Berlin, Arnimallee 22, 14195 Berlin, Germany. 3 Department of Anesthesiology, Deutsches Herzzentrum, Berlin P 06-9 COMPUTER-CONTROLLED, EQUIAXIAL STRAIN SYSTEM FOR REAL-TIME MICROSCOPIC STUDIES ON LIVING CELLS M. Frick, W. Singer, T. Haller, N. Mair, M. Ritsch-Marte, P. Dietl Strain of cells and membranes is the basis of various physiological and pathophysiological responses to mechanical stimuli. Therefore, we designed an improved stretch device enabling better investigation of living cells during mechanical distension. The stretch device is designed to permit continuous observation of single cells, while inducing equiaxial strain of variable strength and frequency to the entire substrate of the cells. To this end, cells are grown on an elastic, optically clear, inert silicone membrane. The major advance is the computer-controlled application of mechanical forces on the elastic membrane. The possibility of designing different stretch protocols offers a broad range of applications, fitting to the cell type of interest. The mechanically generated forces on the substrate are transduced to the requested strain on the cells under study. During stretch the elastic membrane is held in plane to prevent out-of-focus problems. Additionally, the stretch device

INVESTIGATION OF NEURAL CELLS USING PULSE CONTROLLED SCANNING ION CONDUCTANCE MICROSCOPY S. Mann, P. Happel, G. Hoffmann, W. Schuhmann, I.D. Dietzel Scanning ion conductance microscopy is a recently developed method to screen the surface of poorly conducting materials in electrolyte solutions (Hansma et al., Science 234, 641- 643, 1989). It has successfully been applied to investigate surfaces of cardiac and epithelial cells in culture (Korchev et al., Biophys. J., 71, 653-658, 1997). The principle of measurement of scanning ion conductance microscopy consists of measuring the resistance increase that occurs if the tip of an electrolyte-filled glass capillary closely approaches the surface of an insulator. At the point of a preset change of resistance the advancement of the electrode is halted and the coordinates digitized. The electrode is approached to the surface at adjacent positions and by registrating the corresponding coordinates an image of a surface of equal resistance changes is created, correlating with the topographical surface of the cell. We developed a variant method which uses discontinuous test voltage or current pulses to control the distance between membrane and electrode tip, thus eliminating DC electrode drifts. Furthermore, we introduced a ”backstep pattern” of electrode positioning, in which the electrode is lifted up after each point of measurement. Apart from promising more stable recordings the electrode positioning in the backstepped pulse mode allows larger lateral steps and thus ”preview” scans of the contours of whole neurons in culture. Due to the larger time constants of smaller electrode tips the maximal time needed to scan a cell increases with decreasing electrode tips. To decrease scanning time our computer program controlling the positions of the electrode tip automatically adjusts test-pulse duration to the time constant of the electrode. Using this approach, we obtained images of cultured hippocampal neurons and oligodendrocytes exceeding light microscopic resolution. In addition we could monitor volume changes in cell processes following changes in the osmolarity of the bath solution. Lehrstuhl für Molekulare Neurobiochemie, Ruhr-Universität Bochum, NC7- 170, Universitätsstraße 150, D-44780 Bochum P 07-2 MELANOMA CELL INVASION IS RELATED TO MICROTUBULE DEPENDENT SECRETION OF MATRIX-METALLOPROTEINASES (MMPs) E.-M. Schnäker, R. Ossig, T. Ludwig, M. Wilhelmi, H. Oberleithner, T. Schmeinck, S.W. Schneider Cell invasion is a prerequisite for tumor cell spreading. In previous experiments we have shown that secretion of MMPs is important for invasion of human melanoma cells into a MDCK-C7 cell monolayer. By using a highly sensitive gelatinase-based detection assay (detection limit: 3 ng) we could quantify the activity of secreted MMPs. Immunofluorescence technique was used to investigate the mechanisms of MMP exocytosis. Vesicular structures were identified in the cytoplasm of melanoma cells by staining with MMP2-specific anti-serum. Colocalization studies of MMP2 and the microtubule system (stained by α-tubulin specific anti-serum) showed MMP2 containing vesicles attached to microtubules. In fact, after destruction of the microtubule function by taxol we found an impaired secretion of MMP in melanoma cells ( reduction: 18 ± 3.2 %; n = 6). To investigate whether taxol-induced reduction of exocytosis is related to a decreased invasiveness of the melanoma cells into the MDCK-C7 monolayer we used

S 240 our recently developed transepithelial electrical resistance (TEER) breakdown assay. The addition of cancer cells to the apical surface of a highresistant C7-monolayer led within 24 h to a significant TEER breakdown from 8.7 ± 0.25 kΩ to 3.6 ± 0.39 kΩ (n=8). However, adding the same number of taxol-pretreated cells (12h; 5.9 nM) to the C7-monolayer we found that TEER breakdown was delayed within 24 h of co-culture (from 7.9 ±0.28 kΩ to 5.1±0.50 kΩ). We conclude that intracellular transport of MMP-containing vesicles depends on the microtubule system. Disruption or inhibition of this process diminishes MMP secretion and attenuates melanoma cell invasion. Department of Physiology, Robert-Koch-Str. 27a, 48149 Münster, Germany

P 07-3 ORGANIC ANION-DEPENDENT Mg 2+ EXTRUSION FROM LEECH NEURONES D. Günzel, K. Hintz, S. Durry, W.-R. Schlue Mg 2+ extrusion from Mg 2+-loaded neurones of the leech, Hirudo medicinalis, is mediated mainly by Na +/Mg2+ antiport. However, in a number of leech neurones Mg2+ is still extruded in the nominal absence of extracellular Na +, indicating the existence of an additional, Na +independent Mg2+ extrusion mechanism. In the study presented here it was attempted to further characterise this mechanism. The rate of Na+-independent Mg2+ extrusion from leech neurones was measured with ion-selective microelectrodes and found to be independent of extracellular Ca2+, K+, NO3-, HCO3-, SO42-, PO43- and of the intra- and extracellular pH. Na+-independent Mg2+ extrusion was not inhibited by DIDS and furosemide and not influenced by depolarizing the cell membrane from –40 mV to –20 mV in voltage-clamp experiments. Na+independent Mg2+ extrusion was, however, inhibited by the application of 2 mM probenecid, a blocker of organic anion transporters, indicating that Mg 2+ might be co-transported with organic anions. Of all tested organic anions (malate, citrate, lactate, α-ketoglutarate, and 4amino-hippuronic acid) only the application of high, but physiological, concentrations of malate (30 mM; Hoeger et al. 1989, J Exp Biol 147: 43-51) had a significant inhibitory effect on the Na+-independent Mg 2+ extrusion. In addition, iontophoretic injection of malate and fura-2, but not of Cl-, α-ketoglutarate, glutamate or succinate was able to stimulate Na +-independent Mg2+ extrusion from those neurones that initially did not show any Mg 2+ extrusion in Na+-free solutions. Furthermore, fura-2 extrusion from these could be inhibited by 2 mM probenecid. It was slowed in solutions containing high (30 mM) Mg2+ concentrations and accelerated in nominally Mg2+-free solutions. It is therefore proposed that Mg 2+ may be co-transported with organic anions in leech neurones, the physiologically transported anion probably being malate. Supported by the DFG (Gu447/6-1) and the German National Merit Foundation. Institut für Neurobiologie, Heinrich-Heine-Universität, Universitätsstr. 1, Geb. 26.02, D-40225 Düsseldorf, email: [email protected]

P 07-4 EXPRESSION OF FACILITATIVE GLUCOSE TRANSPORTERS 1, 3, 5, 9, 10 AND 11 IN HUMAN ARTICULAR CARTILAGE A. Mobasheri, S. Richardson, G. Neama, S. Bell, S.D. Carter Introduction: Members of the facilitative glucose transport/solute carrier 2A family (GLUT/SLC2A) mediate the transport of sugars into mammalian cells. Thus far, over a dozen GLUT family members have been identified on the basis of sequence similarity (Joost and Thorens, 2001). Monosaccharides are essential for chondrocyte metabolism and extracellular matrix synthesis and their regular provision and transport will have significant consequences for the growth and functional integrity of articular cartilage. Objective of study: The aim of this study was to determine the expression of GLUT/SLC2A family members in fully developed human articular cartilage. Methods: A human chondrocyte cDNA library was constructed from mature, disease-free femoral and knee cartilage as previously described (Neama et al., 2001). Human-specific oligonucleotide primers were designed to amplify PCR products corresponding to human SGLT1 (sodium dependent glucose transporter) and GLUTs 1, 2, 3, 4, 5, 6, 8, 9, 10 and

11. PCR products were cloned, sequenced and compared to known sequences in Genbank. Results: PCR, cloning and sequencing studies confirmed that GLUT5, GLUT9, GLUT10 and GLUT11 are expressed in human cartilage in addition to the previously identified GLUT1 and GLUT3. Expression of SGLT1, GLUT2, GLUT4, GLUT6 and GLUT8 was undetectable in mature chondrocytes. Conclusions: This study demonstrates, for the first time, that up to five distinct glucose transporters (GLUTs 1, 3, 9, 10 and 11) and a glucose/fructose transporter (GLUT5) are expressed in mature human articular cartilage. Our studies partially confirm the recent findings of Ohara et al., (2001) in rat growth plate chondrocytes where GLUTs 1-5 have been identified. Expression of the ”neuronal” GLUT3 isoform in chondrocytes is physiologically significant since this isoform has a low Km for glucose and operates efficiently at low substrate concentrations (Gould et al., 1991). This is in accordance with the low glucose levels in the extracellular matrix of avascular cartilage. The presence of GLUT5 in articular cartilage indicates that fructose is also used as an energy source and structural sugar in the cartilage matrix. This multitude of GLUT isoform diversity suggests a highly specialized handling of monosaccharides by several GLUT/SLC2A family members in chondrocytes and reflects a possible cartilage specific adaptation where glucose is a major metabolic fuel in short supply under physiological conditions. References: Gould GW, Thomas HM, Jess TJ & Bell GI. (1991) Biochemistry 30, 5139-5145. Joost H-G & Thorens B. (2001) Mol. Membr. Biol. 18, 247-256. Neama G, Bell S, Richardson S, Carter S & Mobasheri A. (2001) J. Physiol. 535P, 42P-43P. Ohara H, Tamayama T, Maemura K, Kanbara K, Hayasaki H, Abe M & Watanabe M. (2001) Acta Histochem. 103, 365-378. CTRG, Faculty of Veterinary Science, University of Liverpool, Liverpool L69 7ZJ, UK P 07-5 IN CACO-2 CELLS, MOST OF THE ”APICAL” SGLT1 RESIDES IN INTRACELLULAR, MICROTUBULI-ASSOCIATED VESICLES H. Kipp, S. Khoursandi, R.K.H. Kinne The plasma membrane abundance of many solute transporters is regulated by endo/exocytosis mechanisms between the plasma membrane and intracellular pools. We investigated the distribution of the endogenous sodium/D-glucose cotransporter (SGLT1) in polarized caco-2 cells, a model for enterocytes, in order to identify possible intracellular pools of SGLT1 which may be part of a physiological regulatory mechanism. A cellular organelle fraction was separated by free flow electrophoresis. As detected by an ELISA assay the major amount of SGLT1 resides in early endosomal fractions and only a minor amount in apical plasma membranes. The distribution-ratio between endosomes and apical membrane was approximately 2:1. Further immunochemical investigation of SGLT1 distribution by epifluorescence and confocal microscopy revealed that SGLT1 is located in small vesicles associated with microtubuli. Furthermore, the halflife of SGLT1 in caco-2 cells was determined to be 2.5 days by metabolic labeling followed by immunoprecipitation. Since SGLT1 has a relatively long residence time, it is unlikely that intracellular SGLT1 populations are part of the synthesis/degradation pathway. We, therefore, propose that intracellular compartments containing SGLT1 are part of an endo-/exocytosis process which regulates SGLT1 abundance at the apical cell surface in response to altered physiological demands for Dglucose reabsorption. Max-Planck-Institut für molekulare Physiologie, Abteilung Epithelphysiologie, Otto-Hahn-Str.11, 44227 Dortmund, Germany

P 07-6 DUAL EFFECT OF DIDS ON TAURINE TRANSPORT IN THE EHRLICH MOUSE ASCITES TUMOUR CELL: INHIBITION OF EFFLUX IN HYPOTONIC MEDIUM, AND STIMULATION OF UPTAKE IN ISOTONIC MEDIUM A.K. Jensen, L.O. Simonsen To explore the effects of DIDS on taurine transport, we have monitored the uptake of [ 3H]taurine in Ehrlich cells and the efflux from preloaded cells, in isotonic and hypotonic media (relative tonicity 0.60 by removal of mannitol), with and without the presence of DIDS (100

S 241 µM). The results (i) confirm the previous reports in the literature that DIDS in high concentrations inhibits taurine efflux in hypotonic medium (see Lambert, I.H. (1998). Taurine 3, pp. 269-276), but also (ii) unexpectedly demonstrate a stimulation by DIDS of taurine uptake in isotonic medium. In isotonic medium the efflux is very low, and thus the effect of DIDS must be caused by stimulation of the uptake via the β-system. In hypotonic medium, on the other hand, the uptake is down-regulated and the stimulation of the uptake by DIDS is very weak, and thus the inhibition by DIDS of the efflux must be caused by inhibition of the taurine efflux channel. This dual effect of DIDS is puzzling, and the underlying mechanisms are unclear. The effects could be caused by direct binding of extracellular DIDS to the β-system in isotonic medium, and to the taurine efflux channel in hypotonic medium. Alternatively, the effects of DIDS could perhaps reflect an effect somehow on a common regulatory step controlling these opposing taurine transport systems in a coordinated way. It is unclear, however, how extracellular binding of DIDS could affect such a putative intracellular control point. August Krogh Institute, University of Copenhagen, DK-2100 Copenhagen, Denmark (Email: [email protected]) P 07-7 THE EFFECTS OF NUTRITIONAL STATUS ON L-ARGININE TRANSPORT IN RED BLOOD CELLS FROM HAEMODIALYSIS PATIENTS P.F. Reis 1, L.F.A. Silva1, M.M. Yaqoob2 , T.M.C. Brunini’, N.B. Roberts3, J.C. Ellory 4, G.E. Mann5, R. Soares de Moura 1, A.C. Mendes Ribeiro1,2 Malnutrition in chronic renal failure (CRF) is a powerful predictor of morbidity and mortality despite adequate dialysis treatment1. Amongst the earliest indications of nutritional deficiency are low concentrations of plasma amino acids, including L-arginine, the precursor for nitric oxide (NO) synthesis in vascular cells [Lawson et al, 2001]. CRF is a condition associated with an increased systemic production of nitric oxide [Aiello et al, 1999, Mendes Ribeiro et al, 2001]. We have previously demonstrated that CRF patients have reduced plasma levels of L-arginine associated with increased rates of L-arginine transport, in blood cells [Mendes Ribeiro et al, 2001]. In the present study we have correlated L-arginine transport in red blood cells (RBCs) and plasma amino acid concentrations with nutritional status in CRF patients on haemodialysis. A total of 22 CRF patients and 18 control subjects were recruited with clinical consent. There were significant differences in plasma concentrations of L-arginine, but not L-arginine analogues, in malnourished and well-nourished CRF patients. Plasma L-arginine concentration was significantly decreased in both groups of haemodialysis patients, 66 ±10 µM in malnourished compared with 99 ±10 µM in well-nourished patients vs 137 ±8 µM in controls (p<0.05, Mann-Whitney test). Plasma levels of L-arginine analogues, L-NMMA and ADMA, were reduced to control values following haemodialysis of malnourished and well-nourished patients. In RBCs, L-arginine transport (µmoles. 1 cells-1, h-1) via system y+L was increased in malnourished (422 ±58) and well-nourished patients (469 ±73) compared to normal subjects (275 ±27), suggesting a similar widespread activation of system y + in the uraemic RBCs in both conditions. Transport via system y+L was 87 ±11 µmoles. 1 cells-1. h-1 in controls compared to 70 ±9 in malnourished and 110 ±10 in well-nourished CRF patients. The difference between malnourished and well-nourished CRF patients was significant (P<0.05). The present results demonstrate an increased transport of L-arginine via system y + in RBCs under conditions of reduced plasma L-arginine levels in both malnourished and well-nourished CRF patients, which may represent an adaptation of the Larginine-NO signaling pathway to sustain the enhanced synthesis of NO in chronic renal failure. Lawson JA, Lazarus R, Kelly JJ. J. Renal Nutr. 2001; 11:16-22. Aiello S, Noris M, Remuzzi G. Min. Electrolyte Metab. 1999, 25:384-90. Mendes Ribeiro AC, Brunini TMC, Ellory JC and Mann GE. Cardiovasc. Res. 2001; 49:697-712. 1

Laboratorio de Transporte de Membrana, Departamento de Farmacologia e Psicobiologia, Institute de Biologia, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil 20551-030, 2Renal Research Laboratory, St Bartholomew Hospital, London, ECIA 7BE, 3 Clinical Chemistry, Royal Liverpool Hospital, Liverpool L7 8XP, UK. 4University Laboratory of Physiology, Oxford OX1 3PT, UK, 5 Centre for Cardiovascular Biology & Medicine, GKT School of Biomedical Sciences, King’s College London, London SE1 1UL, UK

P 07-8 L-ARGININE TRANSPORT VIA SYSTEM y+L IS REDUCED IN RED BLOOD CELLS IN HYPERTENSIVE PATIENTS M.B. Moss1, L.E. Novaes Malagris1, T.M.C. Brunini1, N.B. Roberts2, J.C. Ellory3, G.E. Mann4, M.M. Yaqoob5, R. Soares de Moura1, A.C. Mendes Ribeiro1,3 Experimental and clinical studies suggest that systemic arterial hypertension (SAH) is associated with a disturbance in the L-argininenitric oxide pathway [Chowdhary et al, 2001]. NO-related endothelial dysfunction present in hypertension contributes to the increase in peripheral resistance and vascular complications of the disease. Infusions of L-arginine restores endothelium-dependent relaxation in some animal models of hypertension [Boulanger et al, 1999]. We have previously demonstrated that L-arginine transport into erythrocytes is increased in uraemia and chronic heart failure, with transport correlating with no production [Mendes Ribeiro et al, 2001]. In the present study, we determined the effects of hypertension on Larginine transport via systems y+ and y+L in human red blood cells (RBCs) and on plasma amino acid concentrations in 11 patients with SAH and 18 controls. The Vmax for L-arginine transport in via system y+L was diminished in hypertensive patients (56 ± 9 µmoles. l cells-1. h-1, means ± S.E., n=11) compared to controls (91 + 11 µmoles. l-1 cells. h-1, n=18, p<0.05, Mann-Whitney test). The Vmax and Km for Larginine influx via system y+ were not different in hypertensive patients (241 ± 26 µmol. l. cells-1. h-1 and 90 ± 17 µM, respectively) compared to controls (275 ± 27 µmoles. l -1. h -1 and 85 ± 24 µM, respectively). Plasma levels of L-arginine were increased from 137 ± 8 µM to 175 ± 19 µM in hypertensive patients (p<0.05). The plasma concentrations of ADMA were similar: 7 ± 0.7 µM in patients with SAH and 6.9 ± 0.4 µM in controls. Our present findings provide the first evidence that L-arginine influx into erythrocytes was decreased via system y+L in hypertensive patients and associated with increased plasma concentration of L-arginine. It is possible that diminished supply of L-arginine contributes to apparent decrease in NO bioactivity in hypertension. Further studies are necessary to establish a definitive role for L-arginine transport in the pathophysiology of systemic arterial hypertension. Chowdhary S., Townend J.N. J. Hum. Hypertens. 2001; 15: 219-27. Boulanger C. M. J. Moll. Cell. Cardiol. 1999; 31: 39-49. Mendes Ribeiro AC, Brunini TMC, Ellory JC and Mann GE. Cardiovasc. Res. 2001; 49: 697-712. 1

Departamento de Farmacologia e Psicobiologia, Institute de Biologie; Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil 20551-030, 2Clinical Chemistry, Royal Liverpool Hospital, Liverpool L7 8XP, UK, 3University Laboratory of Physiology, Oxford OXl 3PT, UK, 4Centre for Cardiovascular Biology & Medicine, GKT School of Biomedical Sciences, King’s College London, London SEl lUL, UK, 5Renal Research Laboratory, St Bartholomew’s Hospital, London EClA7BE, UK P 07-9 SPERMS, PROTEOMICS AND SCIENTIFIC CONCEPTS R. Mrowka, A. Patzak, H. Herzel A Dutchman, Antonie van Leeuwenhoek (1632-1723), was one of the first scientists who discovered that the male seminal fluid contains moving sperms. One theory about the human generative process claimed, that all parts of an organism preexist in the sperm cell. His drawing of a microscopic picture of a sperm (1694) was clearly influenced by this concept. Most of the current research is driven by concepts, questions, and working hypothesis. Advances in technology have enabled us to have a fresh look at data acquired with traditional single experiments and to compare them with genome-wide data. The differences can be tremendous, as we show here, in the field of proteomics. We have compared data sets of protein-protein interactions in S. cerevisiae which were detected by an identical underlying technical method, the yeast-two-hybrid system. We found that the individually identified protein-protein interactions are considerably different from those identified by two genome-wide scans. Interacting proteins in the pooled database from single publications are much more closely related to each other with respect to transcription profiles when compared to genome-wide data. This difference may have been introduced by two factors: by a selection process in individual publications and by false positives in the whole genome scans. If we assume, that the differences are due to false positives in the whole genome data, the scans would contain 47%, 44% and 91% of false

S 242 positives for the Uetz-data, Ito-core and Ito-full data, respectively. If, however, the true fraction of false positives is considerably lower than estimated here, the data from hypothesis driven experiments must have been subjected to a serious selection process. Johannes-Müller-Institut fur Physiologie, Charité, Humboldt-Universität zu Berlin, Tucholsky Str.2 10117 Berlin P 07-10 APICAL SORTING OF THE HUMAN NONGASTRIC H +,K +ATPASE: SORTING SIGNALS AND POSSIBLE TARGETING MECHANISM. D. Lemke, M. Lerner, H. Bertram, H. Oberleithner, J. Reinhardt The polarization of ion-channels, transporters and ion-pumps in either the apical or basolateral plasma membrane of epithelial cells is mediated by specific protein sorting signals which serve to specify their appropriate distribution. To understand the sorting signals and mechanisms which polarize the human non-gastric H+,K+-ATPase to the apical plasma membrane domain we take advantage of its strong structural homology to the basolaterally polarized Na+,K+-ATPase. We constructed a variety of chimeric proteins between the catalytic αsubunits of the apical H +,K+-ATPase and the highly homologous but basolaterally sorted Na +,K+-ATPase. The chimeric ion-pumps were stably transfected in Madin-Darby-Canine-Kidney cells (MDCK) and their respective spatial distribution was analysed by confocal laser scanning microscopy and surface biotinylation experiments. We could identify a short amino acid stretch between transmembrane domain 3 (TM3) and TM4 which serves to direct the H+,K+-ATPase to the apical plasma membrane of transfected MDCK cells. In particular two charged aminoacids in immediate proximity to the predicted membrane spanning region of TM3 and TM4 seem to be important for proper ion-pump polarization. Furthermore we could show that apical plasma membrane expression of the H+,K+-ATPase is dependent on an intact microtubule (MT) network. Vesicular H +,K+-ATPase can be colocalized with the MT network and destabilization of MT leads to lateral plasma membrane accumulation of the ion-pump. Therefore we conclude that apical transport of the human H+,K+-ATPase is directed along microtubules and that the specific apical targeting signal is localized to a few amino acids between the third and fourth transmembrane domain. University of Münster, Institute of Physiology, Robert-Koch Str. 27a, 48149 Münster, Germany P 08-1 HIGH FREQUENCY AIRWAY OCCLUSION CAUSES AUGMENTATION OF THE VENTILATORY RESPONSE TO INSPIRATORY RESISTIVE LOADING D. P. Sumners, S. Jackson, D. L. Turner Vibration causes a self-sustained increase in motor output in limb muscles (i.e. force or EMG), during but also after the stimulus in man (e.g. Gorassini et al., 1998). This study aimed to ”vibrate” the inspiratory muscle system via high frequency occlusions of airflow (HFO) and test for similar augmentation of neuromuscular ventilatory motor output by measuring mouth pressure (Pm, cm H2O). With local ethical approval and informed consent, breath-by-breath measurements were made in a background of inspiratory resistance (IRL = 50 cm H2O / L · s-1). Each subject breathed against IRL for 6 – 10 breaths (PRE), following which they breathed against IRL with additional high frequency occlusions (~35 Hz) for 2 breaths (VIB1). HFO was then turned off for 10 breaths (POST1). This sequence was repeated immediately afterward (VIB2 & POST2).

P m (cmH 2O)

0 -4 -8 -12 -16 Pre

Vib1 Post1 Vib2 Post2

Figure 1. Inspiratory mouth pressure before (PRE), during (VIB1 & VIB2) and after (POST1 & POST2) 2 breaths of IRL+HFO. Data are subject means ± SEM (n = 7). Filled bars indicate significantly different from PRE values (p<0.0125; Bonferroni corrected Paired T-Test). There was a significant augmentation of Pm during IRL+HFO (VIB1 & VIB2) and after HFO (POST1 and POST2) when compared to PRE (Fig. 1). The mechanism of the observed augmentation of neuromuscular drive could be due to ”traditional” respiratory modulation. Alternatively, a novel form of self-sustained firing of respiratory motoneurons may be responsible for the present finding. Funded by the Wellcome Trust (05284). Gorassini, M, A., Bennett, D. J. & Yang, J. F. (1998). Neuro. Lett. 247, 13-16. Exercise Neuroscience Research Group, School of Applied Science, South Bank University, London, SE1 0AA, UK.

P 08-2 CLASSICALLY CONDITIONED RESPONSES IN BREATHING AND IN POSTURAL REFLEXES FOLLOWING EXTERNAL MOTOR-RELATED PERTURBATIONS Th. Kaulich, W. Föhre, N. Koutsouleris, D. Timmann, F.P. Kolb Frequently, breathing is coordinated with other motor-related processes. Unexpected perturbations of the body equilibrium elicit automatic postural responses. Auditory signals preceding the motor perturbation presented in a classical conditioning paradigm change these postural responses characteristically. The aim of the current study was to establish whether in a group of young, healthy volunteers the coordination between breathing and postural responses is stable enough to result in conditioned responses in the breathing signal. Subjects stood on a platform equipped with straingauges to record vertical ground forces. Tilting this platform according to an unexpected, computercontrolled ramp-shaped movement was used as unconditioned stimulus (US) that elicited a postural reflex. The conditioned stimulus (CS) was a 1000-Hz tone, applied via head phones preceding the US and coterminating with it. The subject’s breathing cycles were recorded by a nasal thermistor. An initial block of 50 US-alone trials was followed by 80 paired trials and a final block of 10 US-alone trials, all given at intertrial intervals of 15-35 ms. The activity of the main muscle groups of both legs (tibialis anterior, gastrocnemius, rectus femoris, biceps femoris muscle) was recorded, amplified, filtered and rectified. The center of vertical pressure was displayed on a screen and subjects were asked to maintain it within a predefined range. Different strategies were used for postural responses depending on whether a preceding tone was applied as a CS. Although breathing seemed to be the lower-order motor process, during classical conditioning the respiratory cycle was changed most effectively such that the CS evoked a short expiration followed by long and deep inspiration during the US. Our data suggest that a short-term plastic process is required to coordinate autonomous functions to external perturbation. Supported by W. Sander Foundation (94.090.1-3). Institute of Physiology, University of Munich; Pettenkoferstr. 12, 80336 München P 08-3 CARDIOVASCULAR AND RESPIRATORY RESPONSES OF THE ANAESTHETISED RABBIT TO A NON-PEPTIDIC DELTA OPIOID RECEPTOR AGONIST T.A.Sears, D.Banks Morphine analgesia is due mainly to µ-opioid receptor activation which unfortunately also causes respiratory depression and tolerance. A new, non-peptidic delta receptor ligand SB235863 has good anti-hyperalgesic activity without such side effects when given i.p. or p.o. to mice (Pozzi et al., 1998). Here we investigate its respiratory and cardiovascular effects in spontaneously breathing anaesthetised rabbits (Na Pentobarbitone). Arterial BP, diaphragmatic EMG, airflow, ECG (chest lead) and end tidal CO2 were recorded via a CED 1401 interface and analysed with Spike 2 software (CED Cambridge, U.K.). Over a dose range of 0.7-12.0 mg/kg, SB235863 i.v. (ear vein) evoked a graded, brief latency (< 7.0s) fall in BP and heart rate and a striking increase in respiratory rate all lasting several minutes with the largest doses. The fall in BP persisted after vagotomy or propanolol indicating that it did not depend on a), vagal reflexes; b), increased cardiac vagal

S 243 tone; c), decreased cardiac sympathetic tone. The fall in heart rate from a typically high control rate of 300-340/min. to 260-280/min. remained after vagotomy or atropine and rarely fell below the neutral rate of 260-280 (combined blockade), indicating reduced cardiac sympathetic tone. The increase in respiratory rate, typically from 3060/min. caused hyperventilation as shown by increased peak inspiratory and expiratory airflows and a fall in end-tidal CO2. It is concluded that in contrast to the respiratory depression associated with µ-receptor dependent analgesia such respiratory stimulation by a delta receptor agonist, which also has anti-hyperalgesic activity, holds promise for a fresh approach to pain relief. O. Pozzi, O. Angelici, G. Petrone, P.Petrillo, C. Manzoni, D. Parolaro, G.D. Clarke, L. Visentin, C. Farina, M.A. Scheideler and G. Dondio. (1998) Soc. Neurosci. Abstr., Vol. 24, Part 1, 352.16 T.A.Sears, KCL School of Biomed. Sci., St. Thomas’ Campus, London, SE1 7EH, UK

P 08-4 INTERACTION BETWEEN BREATHING AND MOTOR-RELATED PERTURBATIONS N. Koutsouleris, D. Ebert, D. Timmann, F.P. Kolb Breathing, as an autonomous, periodic ongoing movement, is activated by oscillatory neuronal centers and driven by the intrinsic needs of the organism. Breathing however, is not independent but may be coordinated with other motor programs or adapted to external perturbations with coordination and adaptation assumed to be based on plastic processes. This study investigated these interactions between breathing rhythms and different types of externally applied motor programs in a group of young, healthy volunteers. The subject´s breathing cycle was recorded by a nasal thermistor, initially during rest for reference. For the coordination task subjects stood on a dynamic platform, that tilted continuously to a sinusoidal function. For the adaptation task the platform tilted, unexpectedly, according to a ramp-shaped movement, introduced randomly at inter-trial intervals of 15-35 s, but triggered at distinct times within the respiratory cycle. Physiological reactions were recorded biomechanically by measuring the vertical ground forces separately for each leg and electrophysiologically by EMG measurements of different leg muscles. The incidence and the degree of the coordination between these signals were quantified by phase-interval analysis. Expected and unexpected perturbations produced characteristic alterations of the physiological reactions. Frequency-dependent patterns and the degree of coordination between breathing rhythm, muscular activity and ground forces provide evidence that breathing is the lower-order motor-related process. The experiments with the adaptation task showed that time-dependent triggering had a different impact on temporal and amplitude parameters of the breathing cycle, thus supporting the existence of the hypothesis of a highly vulnerable period within the respiration cycle. Our data suggest that a specific short-term plastic process is required to coordinate autonomous functions to external perturbations. Supported by W. Sander Foundation (94.090.1-3). Institute of Physiology, University of Munich; Pettenkoferstr. 12, 80336 München

P 08-5 COUPLING OF RESPIRATION WITH HEART RATE VARIABILITY DURING NAIVE AND AUTOSUGGESTIVELY INDUCED RELAXATION („AUTOGENEOUS TRAINING“, MODO SCHULTZ, AT) B. Cotuk*, V. Perlitz, A. Sen, N. Kahn, S. Haberstock, H. SchmidSchönbein°, E.R. Petzold INTRODUCTION: The activity of the autonomic nervous system (ANS) in human adults is commonly attributed to three distinct frequency bands, 1. slow (0.01-0.05 Hz), 2. medium (0.05-0.25 Hz) and 3. fast band (0.25-0.35 Hz). Activity of the latter is conventionally termed respiratory sinus arrhythmia (RSA), considered to be equivalent with dominance of parasympathetic nervous system outflow during rest [1, 2]. In light of recent findings sustained on self-organised ca. 0.15 Hz activity during AT (see accompanying abstract, PERLITZ et

al), this scheme appears to be too rigid since autosuggestively induced relaxation frequencies of respiration can easily decline to frequencies well below 0.25 Hz. METHODS + RESULTS: During AT, non-invasive recordings of ECG, PPG (skin blood flow of the thermic vasoparalysed glabella region), ADMS (thoracical respiration movements) in 67% of recordings shows respiration frequencies below 0.2 Hz, invariably a tight 1:1 -coupling of respiration with cyclically varying heart rate variability (HRV). Following appropriate filtering, the use of Time-Frequency-Distribution techniques allowed the unequivocal identification of 1:1-coupling between patterns of respiration and patterns of HRV regardless of instantaneous frequency. DISCUSSION: In light of the presented findings, the exclusive association of RSA with frequencies between 0.25 and 0.35 Hz appears no longer appropriate. A modification of the present nomenclature should rather be based on strict coupling between respiration and periodically varying cardiac dynamics at a 1:1 ratio, both activities reflecting basic activity of the lower brainstem (see accompanying abstract PERLITZ et al.). 1. Mück-Weymann. M: Autonome Funktionskreise in Psychosomatik und Psychiatrie: Nicht-invasK/es Blomonitoring in der Psychopharmakotherapie. Lage Verlag Hans Jacobs; 2000. 2. Moser M, Lehofer M: Phase- and Frequency Coordination of Cardiac and Respiratory Function. Moser M, Lehofer M et al., Biological Rhythm Research 1995 Vol. 26. No. 1, pp 100-111. Klinik f. Psychosomatik u. Psychotherapeutische Medizin, Universitätsklinikum der RWTH Aachen; * Sporthochschule der Marmara-Universitat Istanbul/Türkei; °Institut für Physiologie des Universitätsklinikums der RWTH Aachen

P 08-6 NON-LINEAR DIFFERENTIATION OF GLIDING COUPLING BETWEEN RESPIRATORY AND HEMO- AND NEURODYNAMICS DETECTED IN CUTANEOUS MICROCIRCULATION DURING AUTOSUGGESTIVE PSYCHOMOTOR RELAXATION V. Perlitz, *B. Cotuk, S. Haberstock, A. Sen, °H. Schmid-Schönbein, E.R. Petzold, +R. Grebe INTRODUCTION: The trophotropic reaction (W.R. HESS) operational at rest is mediated by a disinhibition of the parasympathetic nervous system with close coupling between respiration and cardiac activity. As reported earlier, psychomotor relaxation induced by autogenic training (hypnoid, AT) was shown to be invariably accompanied by the emergence of another rhythmic activity, namely ca. 0.15 Hzrhythmic band in the vasoparalysed forehead skin in the glabella region. Another striking feature of the hypnoid state is that of freely adapted respiration frequencie clearly below those attributed to respiratory sinus arrhythmia classifications emerge with fixed ranges of frequencies ascribed to autonomic nervous system functions [2]. The present study strongly suggests that the ca. 0.15 Hz rhythm band is reflecting an expression of a phase transition between the ergotropic and trophotropic state, a feature making it necessary to abandon rigid ranges of autonomic nervous system frequencies in favour of 1:1coupling between patterns of respiration and periodic patterns of cardiac dynamics (HRV). METHODS + RESULTS: Systemic peripheral blood pressure (Finapres), skin perfusion of the glabella forehead (PPG) and ADMS (thoracical respiration movements) were recorded repeatedly during six weeks in 3 female volunteers. The 0.15 Hz rhythm band was prevalent in all PPG recordings (raw data), displayed with wavelet analysis in time-frequency-distribution portrays (TFD’s). The 0.15 Hz-rhythm band also appeared in patterns of heart rate variability (HRV), extracted from PPG- and Finapres time series (HRV in 93.3% during naive relaxation, 100% during AT). The 0.15 Hz-rhythm band was not induced by respiration, but respiration can consensualizedissociate with this rhythmic activity in this frequency band. DISCUSSION: Data presented here support the hypothesis that the 0.15 Hzrhythm band is likely to be a function of the common lower brain stem system, since detection in all cardio-vascular-respiratory recordings is possible. The biophysical mechanisms responsible for this phenomenon is highly suggestive for stochastic resonation/synchronisation. LITTERATURE: 1. Perlitz V, Schmid-Schönbein H, Schulte A, Dolgner J. Petzold E, Kruse W: Effektivität des Autogenen Trainings. Therapiewoche 1995; 26: 1536-44. 2. Moser M, Lehofer M: Phaseand Frequency Coordination of Cardiac and Respiratory Function: Moser M. Lehofer M et al., Biological Rhythm Research 1995 Vol. 26. No. 1, pp 100-111.

S 244 Klinik f. Psychosomatik u. Psychotherepeutische Medizin, Universitätsklinikum der RWTH Aachen; *Sporthochschule der Marmara-Universitat Istanbul/Turks’; °Institut für Physiologie der Universitätsklinikum der RWTH Aachen, +University de Picardie/ Amierra/Frankreich

P 08-7 PARADOXICAL REACTIONS OF NEURONS WITHIN THE ROSTRAL VENTROMEDIAL MEDULLA TO MORPHINE ADMINISTRATION J. Ellrich The rostral ventromedial medulla (RVM) plays a key role in the endogenous pain control system. RVM neurons are classified according to their behavior to noxious tail heat in rats: off-cell activity decreases, on-cell activity increases, and neutral cell activity remains unchanged. After morphine administration heat-evoked activation of on-cells and inhibition of off-cells is strongly reduced. Therefore, it is generally assumed that RVM neurons are involved in morphine analgesia. There is some evidence for divergent effects of morphine on RVM cells contradicting the classic concept. In 21 male wistar rats RVM cells were extracellularly recorded under barbiturate anesthesia. Teflon-coated stainless steel wires were inserted into the left anterior digastric muscle to record the jaw opening reflex (JOR) evoked by noxious mechanical and heat stimulation of the skin. According to their responses to noxious tail heat neurons were classified as on- (n=6), off- (n=6), and neutral cells (n=9). Cell properties were investigated under baseline conditions, after intravenous administration of morphine hydrochloride (5 mg/kg) and naloxone hydrochloride (1 mg/kg). Whereas 16 RVM cells responded to morphine in a typical way, 4 off-cells and one neutral cell showed a paradoxical response pattern to noxious stimulation after morphine administration. In these 5 cells the antinociceptive effect of morphine was proved by the naloxone-reversible strong suppression of the JOR. While noxious tail heat suppressed off-cell activity (n=4) and had no effect on the neutral cell under baseline conditions, the same stimulus evoked a reproducible activation of these cells after morphine administration. This effect was reversible by naloxone administration. The response pattern of 5 out of 21 RVM off- and neutral cells was converted into an on-cell behavior. These results raise the question whether a simple classification of neurons by their reaction to noxious tail heat is sufficient to predict their role in endogenous pain control. Department of Experimental and Clinical Pharmacology and Toxicology, University of Erlangen-Nuremberg, Emil-Fischer-Center, Fahrstr. 17, D-91054 Erlangen; [email protected]

P 08-8 EFFECT OF THERAPY ON BAROREFLEX SENSITIVITY IN PATIENTS WITH DIABETES MELLITUS AND ESSENTIAL HYPERTENSION H. Svacinová, J. Siegelová, B. Fišer, J. Dušek, J. Jancík, L. Svoboda,1J. Olšovský, 1V. Zácková Objective of study: The evaluation of the effect of pharmacological and non-pharmacological therapy on baroreflex sensitivity of the heart rate (BRS) in patients with diabetes mellitus type 2 (DM) and essential hypertension (EH). Methods: In two groups of diabetics type 2 - with hypertension controlled by ACE inhibitors or Ca-channel blockers (DMH, n = 7, age 56±4 years) and without hypertension (DMN, n = 6, age 60±8 years) BRS was evaluated before (1) and after (2) 12 weeks walking training program. Further, the BRS was measured in patients with essential hypertension without treatment (EH, n = 11, age 60±6 years), in normotensives (N, n = 11, age 58±6 years), in patients with EH and monotherapy of ACE inhibitors or Ca-channel blockers (EHT1, n = 12, age 59±6 years) and in hypertensives with combined therapy of trandolapril and dilthiazem (EHT2, n =18, age 48±5 years). BRS was determined by spectral analysis of spontaneous fluctuations of systolic blood pressure (SBP) and cardiac intervals (CI) (Finapres, 5 min record, metronome controlled breathing at frequency 0,33Hz). Statistical analysis of data: Wilcoxon paired test and ANOVA; p<0.05 was considered statisticaly significant. Results as mean ± SD are given in the following table:

BRS (ms/mmHg) SBP (mmHg) DBP (mmHg) CI (ms)

DMN1 DMN2

DMH1

DMH2 EH

N

EHT1

3.1 ± 1.1 119 ± 17 71 ± 10 775 ± 114

5.1 ± 1.8 140 ± 26 70 ± 14 946 ± 146

7.2 ± 2.3 135 ± 18 68 ± 11 947 ± 1 70

7.8 ± 3.8 121 ± 10 75 ± 8 789 ± 115

4.9 ± 8.2 ± 2.7 3.4 129 ± 9 128 ± 8 85 ± 7 82 ± 9

4.7 ± 1.2 122 ± 13 71 ± 13 788 ± 69

4.7 ± 1.8 156 ± 12 98 ± 7 759 ± 159

750 ± 93

EHT2

789 ± 60

BRS was significantly increased after 12 weeks of walking training program in diabetics without hypertension as well as in diabetics with hypertension controlled by monotherapy (p<0.05). SBP and DBP was significantly increased in EH. The value of BRS in normotensives (N) corresponds with the values of BRS in the diabetics with hypertension after training (DMH2) as well as in patients with essential hypertension with combined pharmacotherapy (EHT2). Support grant cez j037/ 98:141100004. Department of Functional Diagnostics and Rehabilitation, 1IInd Department of Internal Medicine, Faculty of Medicine, Masaryk University Brno, St. Anna Faculty Hospital Brno, Czech Republic

P 08-9 BAROREFLEX SENSITIVITY AND HEART RATE VARIABILITY IN PATIENTS WITH CHRONIC CORONARY ARTERY DISEASE: EFFECT OF EXERCISE TRAINING J. Siegelová, J. Jancík, H. Svacinová, P. Dobšák, Z. Placheta, B. Fišer, J. Dušek, R. Panovský, J. Meluzín, J. Toman, +E. Savin, +J.P. Martineaud Objective of study: The present study was designed to evaluate whether a specific program of exercise training at anaerobic threshold level may be sufficient to improve BRS and HRV of patients with chronic coronary disease (CCD). Methods: We examined 16 patients with chronic coronary artery disease (cardiography, coronarography, NYHA I-II, stable therapy, body weight of 86±6 kg, age of 63±8 years). The symptom-limited spiroergometry was provided before and after the training (maximal attained work load Wmax, peak VO2/ kg, MET). They provided 8-week long intensive aerobic training (at anaerobic threshold level, 20 min three times a week). BRS was determined by spectral analysis of pulse interval (PI), systolic (SBP) blood pressure (5 minutes lasted beat to beat non-invasive monitoring of blood pressure – Finapres Ohmeda, metronome -controlled breathing 0.3 Hz) before and after the training as well as HRV (Varia Puls T3) . Results: The results of PI (ms), SBP (mmHg), DBP (mmHg), BRS (ms.mmHg-1), Wmax (W), peak VO 2 (ml.min -1.kg-1), MET are given in the table (mean±SD) before (B) and after the training. The results of HRV (total power TP ms2, LF 0.04-0.14Hz, HF 0.15-04 Hz, LF/ HF) in table (mean±SD) before (B) and after the training (* p<0.05, Wilcoxon):

B A

B A

PI 852 ± 163 926 ± 126 * TP 299 ± 207 838 ± 1612*

SBP 123 ± 19 130 ± 16 LF 113 ± 94 455 ± 1224

DBP 66 ± 12 67 ± 9 HF 194 ± 177 390 ± 383*

BRS 2.9 ± 0.9 3.9 ± 1.1* LF/HF 0,85 ± 0,92 1,04 ± 1,01

Wmax VO2 MET 142 ± 31 20 ± 4 6.2 ± 1 21 ± 3 6.4 ± 156 ± 1 32*

Conclusions: The significant differences in PI, BRS and Wmax were found (p<0.05, Wilcoxon). It is concluded that 8 weeks lasted exercise training at anaerobic threshold level increased pulse interval, baroreflex sensitivity, heart rate variability (TP,HF) and maximal attained work load in patients with coronary artery disease. Supported by IGA MZ CR 4705-3, CEZJ037/98:14110000 Medical Faculty, Masaryk University, Brno, CZ. +Hopital Lariboisiere, Paris, France

S 245 P 08-10 SLOWER CONTRACTILE RESPONSE TO SYMPATHETIC TRANSMITTERS THAN TO KCL-INDUCED DEPOLARIZATION IN MESENTERIC PREARTERIOLES K. Wilfert, P. Martinka, H.M. Stauss Low frequency oscillations of blood pressure (Mayer waves), are mediated by periodic sympathetic inputs to the vasculature. Previously we found, that the time course of α1-mediated vasoconstriction in the isolated aorta is compatible with the frequency of Mayer waves. Here, we demonstrate, that the time course of norepinephrine (NE)- and ATP-mediated vasoconstriction in mesenteric resistance vessels is also compatible with the frequency of Mayer waves. In 10 rats, the fourth branch of the mesenteric artery (diameter 100150 µm) was mounted on a Mulvany myograph, bathed in KrebsHenseleit buffer that was temperature controlled and bubbled with carbogen. Isometric contractions were elicited by application of 100 mMol KCl, 1 µMol of the specific α1-adrenergic agonist phenylephrine (PE), 1 µMol of the physiological transmitter NE, and 30 mMol of the sympathetic co-transmitter ATP. Force development was similar for PE (3.3±0.6 mN), NE (3.0±0.6 mN), and ATP (3.5±0.7 mN). The response to cell membrane depolarization with KCl was smaller (2.5±0.5 mN, p<0.05) compared to the other vasoconstrictors. Speed of contraction elicited by KCl was greater (1.02±0.19 mN/s, p<0.05) than for the other vasoconstrictors. The speed of contraction in response to NE (0.84±0.16 mN/s) was higher than that to ATP (0.62±0.11 mN/s, p<0.05) and PE (0.66±0.13 mN/s, p<0.05). The vascular response to NE and PE depends on the formation of IP3 elicited by α1-adrenergic receptor stimulation. In contrast, the response to KCl is independent of the formation of an intracellular second messenger. Thus, vascular rings respond faster to KCl than to the two catecholamines. The response to ATP depends on fast ionotropic P2Xand slow metabotropic P2Y-purinergic receptors. The former is a cation channel that allows extracellular Ca++ to enter the cell, while the latter rely on the formation of the second messenger IP 3. The slow time course of vasoconstriction in response to ATP suggests, that the P2Y receptor dominates in small mesenteric arteries. Johannes-Müller-Institut für Physiologie, Humboldt-Universität (Charité), Tucholskystraße 2, 10117 Berlin

addition, the transduction process may be independent of extracellular calcium. University of Tuebingen, Dept. of General Surgery, ZMF, Waldhoernlestr. 22, 72072 Tuebingen, Germany P 08-12 LATERALIZED CHANGES IN RODENT BLOOD T CELL SUBSETS AFTER LEFT (L) AND RIGHT (R) SIDED STIMULATION OF TEMPORO-PARIETO- OCCIPITAL (TPO) CEREBRAL CORTEX: ROLE OF THE THYMUS AND SYMPATHETIC NERVOUS SYSTEM Y.A. Moshel, H. G. Durkin, V.E. Amassian. A model system was developed to compare L and R TPO cortical modulation of blood lymphocyte subsets in either normal, thymectomized or in sham thymectomized rats, or in rats whose thoracic spinal cord was transected at the Tl or T7 level (Wistar, M, 11 w/o). The rats were chronically implanted with an electrode assembly over L or R TPO cortex and with an indwelling R atrial catheter. Lymphocyte distributions were determined by flow cytometry before and after nocturnal stimulation electrically of either L or R TPO cortex. After L-sided stimulation in normal or spinal cord transected at T7 or sham thymectomized rats, CD5+CD4+ T cell numbers significantly increased, in normal rats to 148% of control (p<0.01). The CD5+CD4+ T cells expressing Thy1 had the largest increase (to 270%); however, levels of L-selectin (CD62L) were decreased, without changes in other adhesion/migration markers. With R-sided stimulation CD5+CD4+T cell subsets decreased in numbers to 63% (p<0.01). Lateralized changes in B and NK cell numbers were not observed after cortical stimulation. Plasma corticosterone levels increased similarly after L or R-sided stimulation (114 ± 28; 88 ± 36 ng/ml) thus not accounting for the lateralized changes observed. The T cell increases after L TPO stimulation were absent in thymectomized rats or after Tl spinal cord transection suggesting mediation by sympathetic innervation (above T7) of the thymus with emigration of cells, probably from thymus cortex. Lateralized changes in CD5+CD4+ T cells with TPO cortical stimulation were not found in the Lewis strain. SUNY Downstate Medical Center, Departments of Physiology, Pathology and Neurology, 450 Clarkson Avenue, Brooklyn, NY 11203, USA.

P 08-11 EXTRINSIC AFFERENT RESPONSES TO LUMINAL STIMULI IN MOUSE JEJUNUM IN VITRO M.H. Mueller, M.E. Kreis, W.A. Kunze, D. Grundy There is evidence that there may be ”cross-talk” between intrinsic and extrinsic sensory neurons (Jiang W et al. Gastroenterol. 120 Suppl. 1; A178 (2001)). Sensory neurons respond to a variety of mucosally applied chemicals as well as local pressure (Bertrand PP et al. Am J Physiol. 273: G422-G435; 1997). Extrinsic primary afferents may respond to the same stimuli and if so, this could be due to transmission from enteric afferents to extrinsic afferents. To investigate this we examined the effects of intraluminal acetate (pH 7), acid (pH 4 - 5) and high threshold distension on extrinsic afferent discharge in the jejunum using a novel mouse in vitro preparation in normal Kreb´s buffer and in Ca2+ free buffer to block synaptic transmission. A 2 cmlong segment of jejunum with mesenteric attachment was excised and placed in a perfusion chamber superfused in vitro with pre-gased Kreb´s buffer (pH 7, at 7 ml/min, chamber temperature 34 °C). Multiunit recordings were made from a paravascular nerve bundle within the mesenteric arcade in a separate recording chamber. The lumen was perfused separately with pH neutral buffered normal saline. Data were presented as mean and peak discharge frequency over baseline and expressed as mean (imps-1) ± SEM. A separate preparation was used for each experiment. Neutral acetate (mean 6 ± 2, peak 13 ± 3, n =8) and acid (mean 5 ± 1, peak 12 ± 4, n = 5) elicited a weak but reproducible increase in multiunit firing rates. Distension (60 mmHg) elicited a marked increase in both mean and peak discharge rates (mean 63 ± 5, peak 93 ± 7, n = 12). In calcium free Kreb´s buffer containing 10 mM Mg2+ or 2.5 mM Mg2+ the responses evoked by distension (mean 64 ± 10, peak 92 ± 9, n = 6), neutral acetate (mean 9 ± 3, peak 11 ± 3, n = 3) and acid (mean 8 ± 5, peak 10 ± 3, n = 12) were unchanged (All P > 0.01; unpaired t-test). In conclusion extrinsic afferents are able to record the presence of intraluminal acid, acetate and distension and this does not require transmission from the enteric nervous system. In

P 08-13 NITRIC OXIDE AND ANGIOTENSIN IN THE CONTROL OF THERMO-REGULATION DURING HYPOHYDRATION H. Schwimmer, R. Gerstberger, M. Horowitz Perturbation in body fluid homeostasis interferes with the ability to cope with thermal stress. Interaction of the osmo- and thermoregulatory control circuits at the level of the hypothalamic preoptic area (POA) was investigated by in vivo experiments conducted in male Sabra rats. Animals were reared under conditions of normothermia (Con; 24°C) or heat-acclimation for 30 days (AC; 34°C), combined with eu- or hypohydration (10% loss of b.w.). Subsequently body temperature (Tc) and tail blood flow were monitored during heat stress. Heat endurance and temperature thresholds for the activation of heat defense responses (VTsh for vasodilation, STsh for evaporative cooling) and relative changes in plasma volume were determined. Neurons of POA structures express neuronal nitric oxide synthase (nNOS) and angiotensin II (Angll)-specific ATI receptors. To assess the involvement of preoptic NO and Angll in the integrational processes, animals received an intracerebroventricular (5 µl bolus) aministration of the nNOS-specific antagonist 7-nitroindazole (NI, 100 nmoles), Angll (100 pmoles) or both. Results are summarized as follows: Euhydation Endurance: VTsh STsh Hypohydration Endurance: VTsh STsh

Con:

Con:

NI -

AngII -

NI ↓ ↓ -

AngII ↑ -

NII+AngII AC: ↓ ↓ NII+AngII AC: ↓ -

NI ↓ NI ↑ -

AngII ↓ ↓ AngII -

NII+AngII ↓ ↓ NII+AngII ↑ -

S 246 It is demonstrated that the modulatory role of Angll is more pronounced under conditions of euhydration, whereas NO acts primarily during hypohydration. Preliminary detection of hypothalamic ATI receptors (Western blotting) demonstrated upregulation during hypohydration in the Con NI and Angll treated rats, further supporting the involvement of NO and Angll in neuromodulation of thermoregulation upon hypohydration.

and other cellular stress proteins which interact with PASKIN, consistent with its role as a putative mammalian PAS sensor protein. From the Institute of Physiology, Medical University of Lübeck, D23538 Lübeck, Germany

P 09-3 Physiology, The Hebrew University POB 12272, Jerusalem 91120, Israel

P 09-1 A FENTON REACTION CLOSE TO THE ENDOPLASMATIC RETICULUM MODULATES THE EXPRESSION OF HYPOXIA REGULATED GENES C.Wotzlaw, Q. Liu, U. Berchner-Pfannschmidt, H.Acker, T.Kietzmann It was proposed that hydoxyl radicals (OH•) generated during H2O2 degradation in a perinuclear iron-dependend Fenton reaction are involved in the modulation of gene expression by O2 (Kietzmann et al. NIPS 2001). Thus, it was the aim of this study to localize the compartment in which the Fenton reaction takes place and to proof whether the scavenging of OH• can modulate the hypoxia-inducible factor 1 (HIF-1)-dependent gene expression. The non fluorescent dihydrorhodamine 123 (DHR123) which is irreversibly oxidized to fluorescent rhodamine 123 (RH123) while scavenging OH• was used together with gene constructs allowing expression of enhanced cyan fluorescent protein (eCFP) in mitochondria (MT), in the endoplasmatic reticulum (ER) or in the Golgi apparatus (GA) to colocalize the Fenton reaction. Threedimensional multiphoton (excitation 850nm) confocal laser scanning microscopy (2P-CLSM) revealed that in HepG2 cells within the first 10 minutes of DHR123 staining distinct hot spots of RH123 fluorescence (emission 620nm) could be observed in perinuclear ER pockets as visualised by eCFP fluorescence (emission 470nm). The OH• scavenger DHR123 (3-6 µM/24h) enhanced the levels of HIF-1α, the HIF-1α transactivation potential and activated the expression of the HIF-1 target genes heme oxygenase 1 (HO-1) and plasminogen activator inhibitor (PAI-1) in HepG2 cells. The activation of HO-1 and PAI-1 was not observed after application of 2´,7´dichlorodihydrofluorescin (DCFH,3-6µM,24h) also measuring reactive oxygen species (ROS) nor rhodamine 123 (3-6µM/24h). We assume the localized Fenton reaction housed in perinuclear ER as suited to recover the ferrous state of the prolyl hydroxylase during degrading HIF-1α under normoxia (Epstein et al. Cell 2001). Max-Planck-Institut für molekulare Physiologie, Otto-HahnStr.11,44227 Dortmund/FRG P 09-2 INTERACTIONS BETWEEN A NOVEL PAS-SER/THR KINASE (PASKIN), RELATED TO BACTERIAL OXYGEN SENSORS, AND STRESS PROTEINS P. Stengel, B. Stier, D. M. Katschinski, R. H. Wenger The PAS domain is a versatile protein fold resembling a left-handed glove which is found in many archaeal, bacterial and plant proteins. PAS domains are capable of sensing environmental changes in light intensity, voltage, oxygen concentration and redox potentials. The oxygen sensor FixL from Rhizobium species contains a heme-bearing PAS domain and a histidine kinase domain that couples sensing to signalling. Under anoxic conditions, FixL phosphorylates the transcription factor FixJ which activates expression of the genes involved in nitrogen fixation. We previously identified a novel mammalian PAS protein (PASKIN) containing a domain architecture resembling to FixL. PASKIN is encoded by an evolutionarily conserved single-copy gene which is ubiquitously expressed. The human PASKIN and mouse Paskin genes show a conserved intron-exon structure and share their promoter regions with another ubiquitously expressed gene that encodes a regulator of protein phosphatase-1. The 144 kDa PASKIN is a cytosolic protein containing a PAS region homologous to the FixL PAS domain and a serine/threonine kinase domain which might be involved in signalling. To gain insights into the biological functions of PASKIN, we used the yeast two-hybrid system to clone novel proteins interacting with distinct domains of PASKIN. We identified heat shock

AN UNUSUAL CYTOCHROME A592 WITH LOW PO2 AFFINITY CORRELATING WITH AFFERENT DISCHARGE IN RAT CAROTID BODY C. Huckstorf, T. Streller, H. Mewes, G. Heinemann, C. Pfeiffer, H. Acker The primary sensitive element triggering hypoxia induced increase in afferent discharge of the carotid body is believed to be a cytochrome with apparent low PO2 affinity. Recently, a cytochrome a592 with unusual sensitivity to PO2 already reduced at moderate hypoxia has been described (Huckstorf et al., 2001). In order to specify its origin and identity, we analyzed light absorption difference spectra and afferent discharge signal of isolated rat CBs in vitro under hypoxia and/or in the presence of the cytochrome c oxidase blockers CN- and CO. Both the spectral weight of cytochrome a 592 and the tonic component of the afferent signal decreased with increasing hypoxia and CN-, starting at high PO2 (100 mm Hg) and low CN- (10 µM). Under normoxia and high CO (PCO > 500 mm Hg) an exclusive tonic afferent signal but no a592 were detectable. The spectral weight of a592 and the phasic-tonic relationship of afferent signal induced by hypoxia could be significantly altered by CN- or CO given simultaneously. In the presence of CO, both the spectral weight of a592 and the tonic weight of afferent signal were lower than under hypoxia alone. In the presence of high CN- (100 µM), the spectral weight of a592 was lower than under hypoxia alone but higher than with CN- alone. However, in the presence of CNhypoxia could no longer evoke significant afferent responses. These results indicate that cytochrome a 592 is a redox component of the respiratory chain within the cytochrome c oxidase, but does not participate in the main electron transfer from Cu A via cytochrome a3CuB to O2. Its redox state seems to be influenced directly by hypoxia, but rather indirectly by the cytochrome a3 ligands CO and CN-, due to a shift within electron transfer rates. We hypothesize that cytochrome a592, by means of its redox state changes, is involved in the regulation of the afferent CB signal. Institute of Physiology, University of Rostock, Gertrudenstr. 9, D18057 Rostock, Germany

P 09-4 USE OF HSV-1 VP22 PROTEIN AS A TOOL FOR GENE & PROTEIN THERAPY FOR CYSTIC FIBROSIS E. Chatzidaki, J.W. Wiseman, W.H. Colledge Cystic fibrosis (CF) is a lethal inherited disease caused by mutations in the cystic fibrosis conductance regulator gene (Cftr), which encodes an epithelial Cl- channel. The most common mutation found in CF is a 3 bp deletion that causes loss of phenylalanine at position 508 (∆F508) resulting in loss of function of the CFTR protein product. The major cause of lethality is lung tissue destruction due to chronic infections. To improve gene delivery methods used to restore the function of CFTR in CF, we are exploiting the unusual properties of the Herpes Simplex Virus type 1 (HSV-1) protein VP22. VP22 is a major structural component of the tegument region of HSV-1 that can spread from initially transfected cells to surrounding untransfected cells. Reporter genes eGFP and LacZ, as well as wild type Cftr or ∆F508-cftr genes were fused to the N- or C-terminal of full length or a truncated form of VP22 (C-terminal 34 residues). These constructs were used to transfect COS-1 cells. VP22-GFP fusions spread from the transfected cells, where VP22 localized mainly in the cytoplasm, to the nucleus of surrounding cells. VP22-LacZ and VP22-CFTR fusions did not undergo intercellular trafficking. This may be due to the large size of the fused protein products or retention in the plasma membrane for CFTR protein. In normal individuals only about 25% of the produced CFTR protein localizes to the plasma membrane. We are currently testing whether VP22 can be used to increase trafficking of CFTR protein to the cell membrane. In addition, we have been testing the ability of VP22 to deliver proteins by exogenous delivery. Cell free extracts of the various VP22 constructs were made by transfecting

S 247 293T cells. VP22-GFP fusions, supplied in the growth medium, were taken up by the cells localizing in their nucleus. Currently, we are also testing whether VP22-CFTR fusion products can also be taken up by the cells and if so, where they localize. In this way, we can evaluate the potential use of purified VP22-CFTR fusion proteins as protein therapy for cystic fibrosis. University of Cambridge, Department of Physiology, Downing Str, Cambridge, CB2 3EG, UK

P 09-5 A NEW METHOD FOR NON-INVASIVE INTRACELLULAR OXYGEN MEASUREMENTS USING CONFOCAL FCS WITH SINGLE FLUORESCENT MOLECULE SENSITIVITY N. Opitz , B. Oeke , P. Schwille (*) Despite progress in identifying cellular and systemic responses to hypoxia, the precise mechanisms by which oxygen concentrations are sensed by mammalian cells still remain undetermined. One approach to elucidate these mechanisms might be the quantitative determination of cytosolic oxygen concentrations on the basis of oxygen sensitive fluorescent indicators. However, O 2-sensitive fluorescent indicators only change intensity upon varying oxygen concentrations and thus, do not allow absolute measurements. Fluorescence Correlation Spectroscopy (FCS) allows the determination of triplet fractions of suitably excited fluorescent molecules. As already described by Widengren et al. (J Phys Chem 99, 13368-79, 1995) the triplet transition rates are highly O2 sensitive and may therefore serve for quantitative determinations of molecular oxygen concentrations in biological fluids and cells. We report here about a particular useful indicator (Rhodamine Green) for FCS-based oxygen measurements via O2-dependent triplet population using a special calibration chamber enabling the equilibration of thin membrane-covered indicator layers with different O2/N2 gas mixtures. Oxygen calibration curves (i.e. triplet transition rates, triplet relaxation times and count rates per molecule as function of the oxygen content of the indicator film) can even be monitored on the basis of single fluorescent molecules traversing the open, optically confined volume element, demonstrating the unprecedented sensitivity of this technique and, simultaneously, the possibility to avoid (or at least to minimize) toxic or cancerogenic influences of the applied indicators, since few nanomolar concentrations may be used. Finally, first intracellular calibrations were performed in order to explore the potential of FCS for intrinsically calibrated absolute oxygen measurements. Max-Planck-Institute for molecular Physiology, Otto-Hahn-Str.11 , 44227 Dortmund , FRG (*) Max-Planck-Institute for biophysical Chemistry, Am Faßberg 11, 37077 Göttingen, FRG

P 09-6 ESTIMATING OXIDATIVE STRESS BY REBREATHING HYPERBARIC OXYGEN IN DIVERS G. Gronow, W. Kähler, M. Mályusz, A. Koch Pure oxygen at 170 kPa in a closed circuit oxygen rebreathing apparatus may facilitate the formation of reactive oxygen species, particulary OH.. Benzoate (BA) from food intake may trap OH. by its conversion to mono- (MHB) and dihydroxybenzoates (DHB). We measured urinary formation of MHB, DHB and thiobarbituric acid-reactive substances (TBARS) in 8 volunteers before and after 50 min seawater diving (170 kPa). TBARS rose significantly by 76% from 5.0 ±1.4 nmol to 8.8 ±1.6 nmol (all values per mmol urinary creatinine), MHB increased only moderately by 24.8% from 4.5 ± 0.8 µmol to 5.6 ± 0.8 mol, and DHB rose markedly by 87.8% from 2.5 ± 0.3 µmol to 4.7 ± 0.6 µmol. All fractional DHBs were elevated in respect to control (predive) values: 2,6-DHB by 170%, 2,3-DHB by 115%, 2,5-DHB by 85%, 3,4-DHB by 78%, 3,5-DHB by 47%, and 2,4-DHB by 34%. The observed changes in 2,6-DHB as well as in 2,3-DHB correlated significantly with changes in urinary TBA-RS despite large interindividual variations (see Fig. above). It is concluded that estimating postdiving urinary 2,6- and/or 2,3-DHB may provide a convenient measure to quantify oxidant stress by hyperbaric oxygen breathing.

Department of Physiology, Olshausenstr. 40, University of Kiel, D24098 Kiel, Germany

P 09-7 WATER CHANNEL (AQP1) AND ANION EXCHANGER (AE1) MEDIATED TRANSPORT OF CO 2 ACROSS THE HUMAN ERYTHROCYTE MEMBRANE M.E. Blank, H. Ehmke Since Nakhoul et al. (Am J Physiol 274, C543, 1998) have shown that oocyte membranes increased their CO 2 permeability after expression of AQP1, and Forster et al. (Proc Natl Acad Sci USA 95, 15815, 1998) have presented evidence on the possible involvement of red blood cell’s AE1 in CO2 permeability, the question arose whether transport through protein pores may be a general route for CO 2 exchange. Because both, AQP1 and AE1 are highly expressed in red blood cells, we studied the intracellular response of human erythrocytes to exposure of CO 2. Erythrocyte membranes (ghosts) were prepared according to Dodge et al. (Arch Biochem Biophys 100, 119, 1963). 1.5 ml aliquots of PBS buffered suspension (10 mmol/l, pH: 7.4) were transferred into cuvettes, stirred and exposed to 2.5 l/h of continuous gas flow at 37 °C. The gas mixture contained maximal 5% CO 2 . Acidification as a result of the reaction CO2+H2O⇔H++HCO3- in the presence of the fluorescent dye BCECF was taken as a measure for CO2 transport. Immediately after administration of CO2, a rapid intracellular increase in H+ concentration was observed (τ≈20 s, 5% CO2) to levels far above those in the extracellular space. Inhibition of Na +/H +exchange with 10 µmol/l EIPA showed that retarded efflux of acidic equivalents accounted for the development of an overshoot. Following the overshoot the reaction approached chemical equilibrium (τ≈1000 s). All rates of acidification except the approach of chemical equilibrium were restricted to the intracellular environment. Assays in the presence of carbonic anhydrase (CA) inhibitor ETX (10 µmol/l) revealed time constants of uncatalyzed reactions; therefore CA was solely responsible for higher rates compared to the chemical equilibrium curve. The inhibitors DIDS (which blocks AE1) and HgCl2 (which blocks APQ1), both at a final concentration of 15 µmol/l, depressed rates of acidification by 80% and 15%, respectively. Neither compound inhibited CA activity at levels below 50 µmol/l. Although it still cannot be excluded, that DIDS and/or HgCl2 may impede diffusion by acting on the lipid bilayer, our data suggest that a large fraction of CO 2 transport across erythrocyte membranes is mediated by AE1 and to a lesser extent by AQP1. University of Hamburg, University Hospital Eppendorf (UKE), Department of Vegetative Physiology, Martinistrasse 52, D-20246 Hamburg

S 248 P 09-8

P 09-10

GAP JUNCTION COUPLING AND NMDA-INDUCED OSCILLATIONS OF MEMBRANE POTENTIAL IN pH-SENSITIVE LOCUS COERULEUS NEURONES IN VITRO M. Andrzejewski, D. Ballantyne, K. Mückenhoff, P. Scheid In the en bloc isolated brainstem-spinal cord of the neonatal rat whole cell recordings from Locus coeruleus (LC) neurones reveal a subthreshold oscillation of membrane potential which, in paired recordings, is synchronized in each member of the pair. Current-induced displacement of membrane potential has no effect on the frequency of this oscillation, but lowering bath pH (7.8 to 7.2) results in a TTX-insensitive depolarization and, in the presence of Ba2+ (2 mM), a reversible increase in oscillation frequency (and so discharge frequency). These observations suggest that the pH-sensitive oscillation may be synchronized by electrical coupling. In the present experiments the spontaneous oscillation was suppressed by TTX (0.5-2.0 µM), suggesting the involvement of a voltage-gated Na+ conductance in its generation. In the continued presence of TTX bath addition of NMDA (10-40 µM, usually 10 µM) resulted in a 2-10 mV depolarization and the re-emergence of a sustained synchronous subthreshold oscillation which retained its sensitivity to a CO 2-induced acid shift in bath pH. The NMDA-induced depolarization was blocked by APV (100 µM) and the oscillation reduced in frequency or suppressed by MK801 (50 µM). Addition of the gap junction uncoupler carbenoxolone (100-300 µM) left Na+-sensitive spikes intact in these neurones, but suppressed the spontaneous oscillation. Subsequent addition of NMDA (in the presence of carbenoxolone and TTX) again elicited a depolarizing response but failed to generate oscillatory activity. These observations suggest that the capacity of depolarizing inputs to activate oscillations in LC neurones depends on intact gap junction coupling. Supported by DFG Sche 46/12-4 and by the Heinrich-und-Alma-Vogelsang Stiftung.

DYNAMIC ASSESSMENT OF FUNCTIONAL PULMONARY IMPAIRMENT AFTER LUNG RADIATION IN MINIPIGS T. Rießland, H. Dalitz, H.-J. Mädler, V. Zürich The impairment of pulmonary function after experimental lung radiation in minipigs is characterized by impeded diffusion of oxygen and restrictive signs. This was shown by us under steady-state conditions during a mild hypoxic challenge. It remained unclear, however, whether sensitivities for pathophysiological changes differ using dynamic assessment. Thus, we now assessed the dynamics by means of continuously monitored variables. The study comprised 28 minipigs radiated unilaterally (1.5-31.5 Gy) and 8 controls. In the mildly anesthetized animals breathing frequency (Bf), alveolar gas, arterial oxygen saturation (SaO2) and heart rate (HR) were monitored 3 weeks prior to and 3-37 weeks after radiation. Breathing room air was followed by 15% O2 (balanced by N2) for at least 10 min. From the momentary values, dynamic and stationary changes, means and settling times were computed. Time-dependent integrals summed up the actual differences between momentary and mean values. In highly radiated animals SaO2 markedly decreased during the hypoxic challenge, especially in weeks 9 -14. At low and medium deoxygenations, the mean Bf and the settling times of SaO2 and expired FO2 increased in parallel. Short-time integrals were negative initially for Bf and later for the alveolar-arterial-pO2-difference. During maximal deoxygenation, however, the settling times changed independently, mainly due to prolonged or, at highest doses, shortened settling of SaO2. Bf and alveolar-arterial-pO2-difference then continuously increased. In the HR, the prominent dynamic changes ended in increased long-time integrals. Generally, smaller radiation doses resulted in smaller and shorter variable alterations. Increased settling times and changed integrals, however, revealed higher sensitivity than stationary values. In conclusion, dynamic alterations may reflect functional impairments after experimental lung radiation, probably, with higher sensitivities than stationary values. They also indicate, that ventilation only partially compensates for impeded pulmonary gas exchange.

Institut fur Physiologie, Ruhr-Universität Bochum, 44780 Bochum, Germany

Institute of Physiology, Medical Faculty Carl Gustav Carus, TU Dresden, Fetscherstraße 74, 01307 Dresden, Germany P 09-9

P 10-1

CHEMOSENSITIVE MEDULLARY NEURONS AND THE CONCEPT OF LACTATE-MEDIATED ENERGY SUPPLY M.Wiemann*, M. Louvrou, D. Bingmann Lactate released by glial cells is currently discussed to be a main energy source for neurons: The shuttling of lactate is believed to involve the monocarboxylate transporters 1 and 2 (glial cells: MCT1, low affinity; neurons: MCT2, high affinity) which function as H + cotransporters. Accordingly, the neural intracellular pH (pHi) should be lowered during lactate uptake. The present study tested this concept for chemosensitive neurons of the ventrolateral medulla whose firing properties are highly pHi sensitive. Experiments were carried out on organotypic cultures (OC) of the medulla oblongata (new born rat, obex level), because superficial neurons are in tight contact with glial cells in this preparation. Membrane potentials were recorded with sharp electrodes; pHi measurements were carried out after BCECF-AM loading. In cultures being superfused with bicarbonate (24 mmol/l)-buffered artificial cerebrospinal fluid (ACSF) chemosensitive neurons were identified by their increased firing rate upon hypercapnia. Lactate or propionate (10-20 mmol/l) mimicked the hypercapnia-induced increase in bioelectric activity. Propionate was more effective than equal amounts of lactate, pyruvate, or hydroxybutyrate, pointing to a weak or missing MCT. The lactate induced fall in pHi (maximum value after 3-5 min) amounted to 0.16±0.1 and 0.18±0.05 pH units with 10 and 20mmol/l lactate, respectively. A considerable number of ventrolateral neurons (39 out of 126) exhibited no acidification upon lactate application. Double labeling studies with confocal laser scanning microscopy using antibodies against MCT1 or MCT2, and the neuronal marker MAP-2 or the astrocyte marker GFAP, revealed only few neurons expressing MCT1, while MCT2 was not found in the ventrolateral region. Similar results were obtained with the perfusion-fixed tissue of adult rats. We conclude, that - in contrast to current concepts - the acidifying uptake of lactate into ventrolateral chemosensitive neurons is not mediated by a high affinity MCT such as MCT2.

MAPPING OF NEURONAL CIRCUITRY MEDIATING THE DIVING RESPONSE IN THE WORKING HEART BRAINSTEM PREPARATION (WHBP): A C-FOS STUDY IN RAT M. Mörschel, H. Herbert, M. Dutschmann Noxious stimulation of the nasal mucosa lead to pronounced autonomic responses such as apnea and bradycardia. These responses are mediated by the trigeminal ethmoidal nerve (EN5) and represent the hallmarks of a vital defensive reflex, commonly termed the diving response. In the resent study, we investigated if physiological mapping of neuronal circuitry by using the c-fos method can be applied in the WHBP, a decerebrated and arterially perfused in situ preparation. To do so, we investigated the specific, stimulus related expression of Fosprotein. The Fos-protein is a transcription factor which is expressed by the immediate early gene c-fos in response to neuronal Ca2+ influx. To control the autonomic status of the WHBP we recorded baseline phrenic nerve activity (PNA) and ECG activity for 2.5h after reperfusion of the brainstem (control). To induce Fos-expression, we used unilateral electrical stimulation of the EN5 (10s, 20Hz, 0.11mA, 100µs) every 3 min for 45 min starting 1.5 h after re-perfusion of the brainstem. EN5 stimulation evoked consistently a stimulus locked cessation of PNA and a bradycardia. Subsequently, all preparations were transcardially perfused with 4% paraformaldehyde. Coronal sections of the brainstem (50µM) were cut for immunocytochemical detection of the Fos-protein. Cell counts (mean ± SD) revealed that, compared to controls (n=5), EN5 stimulation (n=7) evoked significant increase of Fos-expressing neurones ipsilaterally in the following brain areas: the nucleus caudalis of the spinal trigeminal tract (18±5 to 125±35, p<0.001), the rostral (24±8 to 84±36 p<0.01), intermediate (25±15 to 101±40 p<0.001) and caudal parts (22±15 to 69±31 p<0.05) of the Kölliker-Fuse nucleus. Significant increase in numbers of Fos-expressing neurones without pronounced ipsilateral dominance was observed in the commissural (12±6 to 34±13 p<0.05), medial (25±6 to 59±23 p<0.05) and rostral parts (14±6 to 39±8 p<0.01) of the nucleus of the solitary tract, within the caudal (19±10 to 50±15 p<0.01) and rostral ventro-

Institute für Phyiology, Universität Essen, Hufelandstr.55, Germany

S 249 lateral medulla (19±17 to 54±24 p<0.05), the area of the preBötzinger complex (12±6 to 38±14, p<0.05), as well as in the pontine intertrigeminal (39±5 to 62±14 p<0.05) and A5 region (12±4 to 40±11 p<0.01). We conclude that in WHBP a specific Fos-expression can be evoked by peripheral stimulation of trigeminal afferents similar to in vivo conditions. The c-fos method might therefore be a powerful tool to investigate the maturation of neuronal circuitry mediating upper airway reflexes in situ [see abstract Dutschmann & Herbert]. Supported by the SFB 430/C9 Dept. Animal Physiol., Univ. Tübingen, Morgenstelle 28, 72076 Tübingen, Germany

P 10-2 POSTNATAL MATURATION OF THE DIVING RESPONSE: A C-FOS STUDY IN NEONATAL RAT M. Dutschmann, H. Herbert Integration of regulative (e.g Hering Breuer reflex) or protective reflexes (diving response) within the respiratory network of mammals appears to be processed differently in neonatal compared to adult animals. In neonates reflexes can be evoked more easily and result in stronger autonomic responses compared to adults. The underlying maturation of neuronal pathways and mechanisms are supposed to be responsible for the occurrence of patho-physiological apnoeas during early childhood (e.g., sudden infant death). However, neurophysiological correlates of these phenomena are poorly investigated. In the present study, we investigated the maturation of the diving response, a protective reflex of the upper airways, by applying the c-fos method to the working heart brainstem preparation (WHBP) of neonatal rat (P5-P8). To do so, we evoked the diving response by repetitive electrical stimulation of the trigeminal ethmoidal nerve (EN5, 10s, 20Hz, 0.1-1mA, 100µs; every 3min for 45min) in the WHBP and monitored the autonomic responses by recording phrenic nerve activity (PNA) and heart rate. This stimulation protocol leads to the selective expression of the Fos-protein in adult rat preparations [see abstract Mörschel et al.]. In the WHBP of neonatal rat, EN5 stimulation caused stimulus locked cessation of PNA and bradycardia, comparable in strength to adult rat. In contrast to adult rat, we observed following repetitive EN5 stimulation (n=6) no or only weak decreases in the strength of apnoea and bradycardia. Histological analysis of the distribution of Fos-expressing neurones in the ponto-medullary brainstem (n=5) revealed additional differences between adult and neonatal rat. Compared to control (n=4), we detected a significant up-regulation of Fos-expressing neurones in medullary nuclei, such as the nucleus caudalis of the spinal trigeminal tract, the nucleus of the solitary tract, the rostral ventrolateral medulla. The pattern was similar to that observed in adult rat [see abstract Mörschel et al.]. However, in contrast to adult rat, neurones within pontine brain nuclei, such as the Kölliker-Fuse nucleus and the intertrigeminal region, did not express Fos protein following EN5 stimulation. This lack of Fos expression could be due to the prematurity of the KF region as indicated by the cytoarchitecture seen in thionine and fibre stained section through the neonatal pons. We speculate that cytoarchitectural reorganisation and maturational changes of cellular mechanisms during postnatal development might be the cause for differences in reflex integration between neonatal and adult mammals. Supported by the SFB 430/C9. Dept. Animal Physiol., Univ. Tübingen, Morgenstelle 28, 72076 Tübingen, Germany P 10-3 NEURONAL ACTIVITIES OF THE PONTINE RESPIRATORY GROUP OF RAT IN SITU C Kühlbach, H Herbert, M Dutschmann Neuronal activity in the pontine respiratory group (PRG) plays a key function in the mediation of inspiratory termination [see abstract Dutschmann & Kühlbach]. Although the PRG is proposed to be an essential part of the central respiratory network, cellular activity is reported to be weak, compared to medullary respiratory centres. The reason for this weak activity lies in the suppressive sensory feedback from pulmonary stretch receptors (PSR) controlling the inspiratory off-switch under in vivo conditions. In the present study, we investigated in the in situ working heart-brainstem preparation (WHBP) cel-

lular activities in the PRG. In the WHBP an eupnoea-like pattern of respiratory activity can be observed in absence of feedback from PSR. In brief, the WHBP is an arterially perfused decerebrated preparation allowing for analysis of respiratory activity. For cellular recordings, we used glass microelectrodes filled with 2M NaCl. Electrode tips were positioned on the dorsal surface of the pons at the following coordinates: 0.1-0.5 mm caudal to inferior colliculus and 2.5-3 mm lateral to midline, approximating the PRG region. To monitor ongoing respiratory activity, we simultaneously recorded phrenic nerve activity. In 15 preparations we recorded 19 inspiratory, 17 postinspiratory, 7 expiratory and 7 phase-spanning neurones in the PRG. Detailed analysis of the firing pattern of the 17 inspiratory neurones revealed 3 different types: 10 showed a decrementing and 5 an augmenting discharge pattern. The remaining 4 neurones fired with constant frequency throughout the inspiratory phase. The postinspiratory neurones could be divided into two classes: 12 showed a decrementing discharge pattern during the first three quarters of expiration (n=12), while 5 neurones fired with constant frequency restricted to the first quarter of expiration. The 7 expiratory neurones started to fire during mid expiration, however, 4 of them showed an augmenting discharge pattern while 3 were decrementing. Interestingly, we recorded different types of phase-spanning neurones in the PRG. Four were active from second stage of expiration to end of inspiration while 3 neurones fired during the inspiratory and the postinspiratory interval. We conclude that PRG neurones show distinct respiratory activity under conditions of lack of sensory feedback from PSR. However, the pattern of cellular activity observed in the PRG appears to be different from what is known from neurones located in the ventral respiratory group, as found in the WHBP of rat. (Supported by SFB 430/C9) Dept. Animal Physiol., Univ. Tübingen, Morgenstelle 28, 72076 Tübingen, Germany P 10-4 NEURONAL CONTROL OF RESPIRATORY PHASE TRANSITION AND UPPER AIRWAY PATENCY: A VITAL ROLE FOR THE PONTINE KÖLLIKER-FUSE NUCLEUS (KF)? C. Kühlbach, M. Dutschmann The dorsolateral pons including the KF was established as the pneumotaxic centre of mammals. Either lesions or NMDA-receptor blockade within the KF cause apneusis, indicative for disturbance of cellular mechanisms controlling the inspiratory off-switch. Cellulary the inspiratory off switch is irreversibly mediated by the activity of postinspiratory (PI) neurones. At the level of the upper airways, inspiratory off-switch is manifested by the PI-controlled activity of laryngeal adductor muscles, causing a transient constriction of the vocal folds. However, nor a potential role of the KF in controlling laryngeal muscles neither the role of inhibitory neurotransmission in the pneumotaxic function of the KF was investigated. The study was performed using the working heart brainstem preparation (WHBP). The WHBP is a decerebrated and arterially perfused in situ preparation allowing for analysis of respiratory function. To study the putative role of GABAergic neurotransmission in the KF, we bilaterally injected 10 mM of the GABA-receptor agonist isoguvacine (70 nl) and quantified the effects on phrenic (PNA) and recurrent laryngeal nerve (RLNA) activity. To obtain an index of glottal resistance, we recorded sub-glottal pressure (SGP). To do so, the larynx was perfused with a constant stream of air in the expiratory direction and changes in SGP were indicative of constriction or dilatation of the vocal fold. Bilateral injections of isoguvacine (n=42) into the dorsolateral pons revealed a prolongation of PNA (n=27), indicated by 111.8±6.3% increase of the time of inspiration (Ti, 0.93±0.06 to 2±0.11s p<0.001). Increased Ti was always accompanied by a significant decrease in the frequency of PNA (from 14±1.3 to 4±1.1 bursts/min, p<0.001). Simultaneous recording of RLNA (n=14) demonstrated a reduction of integrated postinspiratory activity (-85.5±29%, p<0.01). Additional recording SGP (n=14) demonstrated a complete absence of laryngeal adductor activity following bilateral isoguvacine injections (n=10). Histological analysis of the injection sites revealed a clear topography. The most effective sites (increase of T i < 250%, n=8) were centred ventrally in the intermediate KF, while less effective sites (increase of Ti < 50-150%, n=19) were located in rostral or caudal parts of the KF. Injections placed outside the KF were ineffective (n= 15). We conclude that the KF is essential for the generation of the PIphase, required for a functional modulation of upper-airway resistance

S 250 and inspiratory off switch. Furthermore, KF neurones appear to be under massive GABAergic control of a hitherto unknown source. (Supported by SFB 430/C9) Dept. Animal Physiol., Univ. Tübingen, Morgenstelle 28, 72076 Tübingen, Germany P 10-5 PATTERNS OF EVOKED ACTIVITY OF DORSAL AND VENTRAL RESPIRATORY GROUP INSPIRATORY NEURONES DURING SOMATIC AFFERENT INDUCED ENTRAINMENT OF THE RESPIRATORY RHYTHM P. Boscan, J.F.R. Paton Exercise and nociception are associated with increased respiratory rate; the latter is essential for maintaining adequate tissue oxygenation during these periods of elevated metabolic demand (Bramble & Carrier 1983, Science, 219, 251-). Further, rhythmic exercise and electrical stimulation of somatic afferents can entrain respiratory frequency (e.g. Iscoe & Polosa 1976, J.Appl.Physiol., 40, 138-). We looked at the effect of somatic afferents activation on inspiratory neurones recorded from both the dorsal and ventral respiratory groups (DRG and VRG respectively) in the rat. The brachial plexus (5-30 V, 0.5 ms, 0.31.5 Hz) or the cervical spinal cord (C3-C4) was stimulated electrically (paired pulses 20 or 10 ms apart, 0.4-2 V, 0.2 ms & 0.3-1 Hz). Stimulation of the brachial plexus entrained the frequency of PNA in a 1:1 ratio over a range of stimulation frequencies of up to 50% above the ongoing rate. The timing of the first stimulus relative to the phase of the PNA produced two patterns of response. First, if the stimulus coincided with PNA it produced a premature off-switching with subsequent stimuli entraining inspiratory bursts. Second, if the stimulus occurred during expiration it evoked an inspiratory discharge that occurred earlier than predicted from the control PNA frequency. In both cases PNA onset latency was 320±33 ms (n=13). We found two types of brachial nerve evoked cellular responses: (i) stimulation during the inter-PNA interval evoked a burst of action potentials not accompanied with a PNA discharge. This occurred most frequently in the DRG (latency: 25.2±2 ms; 18 of 25 neurones) compared to the VRG (latency: 31.6±11 ms; 5 of 20 cells tested), (ii) the remaining DRG (n=7 of 25) and VRG neurones (n=15 of 20) exhibited discharges at latencies of 374±51 ms and 282±100 ms respectively that were always associated with PNA. To determine the relative directness of these inputs, paired pulses were delivered to the spinal cord. Evoked responses in DRG neurones (n=11 of 13) followed more faithfully than VRG neurones (n=2 of 5). We conclude that there is a significant excitatory synaptic input from somatic afferents to DRG neurones. This may be relatively more direct than somatic inputs to VRG neurones. We suggest a role for DRG respiratory neurones in mediating the increased respiratory frequency associated with exercise and/or pain. BHFfiinded research. Department of Physiology, School of Medical Sciences, University of Bristol, Bristol BS8 1TD, UK P 10-6 ACTIVATION OF PERSISTENT SODIUM CHANNELS UNDERLIES RHYTHMIC „RESPIRATORY“ ACTIVITIES IN VITRO AND GASPING, BUT NOT EUPNEA, IN SITU W.M. St.- John, + I.A Rybak, *O. Pierrefiche, J.F.R. Paton Bursting or pacemaker discharge of medullary neurons underlies the rhythmic “respiratory” activity of in vitro mammalian preparations (Johnson et al. 1994, J.Neurophysiol. 72, 2598-). A similar neuronal mechanism is envisaged to generate the gasp of in vivo and in situ preparations (St.-John 1998, Prog.Neurobiol. 56, 97-). In contrast, a complex neuronal circuit involving both pontile and medullary mechanisms is considered to generate the eupneic rhythm (Richter & Spyer 2001, TINS, 24, 464-). Modeling of bursting or pacemaker discharges of medullary respiratory neurons include activation of persistent sodium channels (Rybak et al. 2001; In: Poon & Kazemi; Kluwer Academic/Plenum Publishers). When incorporated into a model circuit, activation of persistent sodium channels is not necessary for generating eupnea. Thus, we hypothesised that a blockade of persistent sodium channels should disrupt rhythmic “respiratory” activities of in vitro preparations and gasping but not eupnea of in situ preparations. In the in situ preparation, ischaemia induced gasping. A blocker of

persistent sodium channels, R056865 (50 nM; Janssen Pharmaceutica N.V.) was added to the perfusate and ischaemia-induced gasping was reduced to a single decrementing burst followed, in some preparations, by several incrementing bursts of low amplitude. In hyperoxia, the incrementing pattern of eupneic phrenic activity was un altered. In medullary slices of the neonatal rat (750 µm; 5-6 day old), rhythmic bursts of activity of the hypoglossal nerve progressively declined with augmentations in the concentration of R056865 to 16.5µM. Above this concentration of R056865, rhythmic activity was eliminated and could not be restored despite additions of NMDA. We conclude that activation of persistent Na channels of pacemaker neurons is necessary for expression of rhythmic “respiratory” activity in vitro and gasping in situ. During eupnea, this pacemaker discharge is suppressed and embedded into the ponto-medullary neuronal circuit. Following ischaemia, this latent pacemaker discharge is released to generate the gasp in situ and in vivo. Dept. Physiology, Sch.Med.Sci., University of Bristol, Bristol BS8 1TD, UK; +School of Biomedical Engineering, Drexel University, Philadelphia, USA; *Laboratoire de Neurophysiologie, 33, rue St Leu, 80039 cedex 1, Amiens, France P 10-7 PRESYNAPTIC MODULATION BY ENKEPHALINS OF EXCITATORY POSTSYNAPTIC CURRENTS IN HYPOGLOSSAL MOTONEURONES V.A. Bouryi, D.I. Lewis We have previously demonstrated that stimulation within the raphe pallidus evokes a glutamate mediated excitatory postsynaptic current (EPSC) in hypoglossal motoneurones, with this EPSC being modulated pre-synaptically by co-released 5-hydroxytryptamine (J.Physiol. 533; 84P). Since enkephalins are also co-localised within these caudal raphe neurones, the aim of this study was to determine the pre- and postsynaptic actions of enkephalins on hypoglossal motoneurones. Male Wistar rats (10-14 days) were anaesthetised with sodium pentobarbitone and 300 µm coronal slices of the medulla prepared. Whole cell voltage clamp recordings were made from identified hypoglossal motoneurones. Synaptic pathways from the raphe pallidus were activated utilising a stimulating electrode placed within this region, with pharmacological agents applied in the superfusate as required. Recordings were made from 30 hypoglossal motoneurones located within the ventral region of the nucleus. Met-enkephalin (10-20 µM) evoked a small membrane hyperpolarisation (-2.34 ± 1.7 mV, p < 0.001, n = 17) in the majority of neurones. Met-enkephalin reduced the amplitude of ESPCs evoked by electrical stimulation within the raphe pallidus (42.7 ± 23 %, P < 0.001, n = 17) whilst increasing the paired pulse ratio (1.41 ± 0.3 vs 1.16 ± 0.2, P < 0.05, n = 14); it also reduced the frequency of spontaneous EPSCs (n = 7). This inhibition of evoked EPSCs was mimicked by the µ-opioid agonist, DAMGO (1 µM, 48.3 ± 23 %, P < 0.01, n = 4). In contrast, the δ-opioid agonist, DPDPE (1-3 µM, n = 3) was less effective in inhibiting the evoked EPSCs (9.67 ± 1.7 %, P < 0.001). These data suggest that enkephalins, acting principally via µ-opioid receptors located on presynaptic terminals, inhibit the EPSCs evoked in hypoglossal motoneurones following stimulation within the raphe pallidus. These studies were supported by a grant from The Wellcome Trust. School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK P 10-8 FUNCTION OF ASTROCYTES IN RESPIRATORY NETWORK IN MOUSE BRAIN STEM SLICES D. Graß, F. Kirchhoff, D. W. Richter, S. Hülsmann The caudal medulla contains a neuronal network that generates the rhythmic activity and drives respiratory output. This function can be preserved in an acute slice preparations. Astroglial contribution to the stabilization of the respiratory activity has yet not been investigated in detail. To identify astrocytes in the rhythmic slice preparation (700 µm, postnatal day 1 to 16), we used transgenic mice in which astrocytes are fluorescently labeled by GFAP (glial fibrillary acidic protein)-promoter controlled green fluorescent protein expression. Whole-cell recordings of astrocytes in the pre-Bötzinger complex and in the hypoglossal nucleus revealed a negative membrane potential, close to the potassium equilibrium potential and a passive potassium

S 251 conductance. We recorded astrocytic responses in synchrony to rhythmic neuronal respiratory bursts activity. Under control conditions currents in synchrony to respiratory activity were hardly detectable. In contrast, during epileptiform bursts induced by blockade of glycinerigic and GABAergic inhibition burst-related currents of 50 to 100 pA amplitude could be recorded. These currents are carried by potassium ions, which are caused by an increase of extracellular potassium during neuronal activity. It remains to be clarified whether these currents are exclusively derived from astroglial somata and proximal processes or whether they represent conductance changes derived from distal processes enwrapping synaptic structures. Supported by: DFG Zentrum Physiologie und Pathophysiologie, Abt. Neuro- und Sinnesphysiologie, Georg-August-Universität, Humboldtallee 23, 37073 Göttingen, Germany P 10-9 CHRONIC INFLAMMATORY LUNG DISEASE IS ASSOCIATED WITH ENHANCED DOPAMINE RECEPTOR EXPRESSION IN AIRWAY-INNERVATING SENSORY GANGLIA IN A RAT MODEL OF BRONCHITIS C. Peiser, D. A. Groneberg, Q. T. Dinh, B. Müller, A. Fischer Background: Topical administration of dopamine or dopamine-D 2 receptor agonists has beneficial effects in chronic inflammatory lung disease. These effects are mediated through normalizing the increased activity of sensory nerves. The presence of D 2-receptors in airway projecting sensory neurons was shown previously. In the present study, a rat model of chronic bronchitis was used to examine the regulation of D2-receptor mRNA expression. Methods: Rats were exposed to nitrogen dioxide (10 ppm, 20 d) and D2-receptor mRNA levels in sensory ganglia (jugular-nodose, trigeminal, cervical dorsal root and thoracal dorsal root ganglia) were examined by quantitative real time polymerase chain reaction and compared to control tissues (n = 6 in each group). Results: Whereas for trigeminal and dorsal root ganglia the D 2 -receptor expression levels showed no difference between both animal groups, there was a significant (p < 0.05) up-regulation in the jugular-nodose ganglia with a 2.1-fold factor. Conclusion: The increase of D2-receptor mRNA in airway innervating jugular-nodose sensory neurons may represent a neurochemical basis for the effects of dopamine or dopamine-D2-receptor agonists in chronic inflammatory lung disease.

Klinik Ambrock, Klinik für Neurologie, Institut für Schlafphysiologie, Ambrocker Weg 60, D-58091 Hagen; und Abteilung für Angewandte Physiologie, Ruhr-Universität Bochum, D-44780 Bochum P 11-1 VASOPRESSIN STIMULATES MILK PRODUCTION IN GOATS K. Olsson*, C. Malmgren*, J. Häggström*, K. Hansson# and K. K. Olsson* It has generally been accepted that vasopressin at physiological doses has no effect on milk production and that high doses cause decreased mammary blood flow concomitant with no change or a decrease in milk production. However, we recently found that intravenous infusions of pressor doses of vasopressin increased milk flow and milk fat concentration in goats. It was therefore of interest to study if a doseresponse relation could be obtained during infusions of vasopressin and to compare effects of vasopressin to infusions of oxytocin. In addition, we measured blood flow during infusions of the peptides. Six goats were used in the dose-response study. The infusions were given into one of the jugular veins when the goats were 1-3 months post partum. Vasopressin (0.1, 0.3 and 1.0 ng/kg min), oxytocin (10.0 ng/ kg min), and 0.15 M NaCl were given at random order to each goat for 90 min. The goats were hand milked every 30 min. Milk flow had increased 5.3 and 5.8 times after 30 min infusion at the highest dose of vasopressin and oxytocin, respectively, and then declined. The fat concentration increased 1.6 and 1.4 times and remained elevated throughout the infusions. Milk flow increased by about half compared to the highest dose of vasopressin when the lower doses were given, but with no difference between these. Blood flow to the mammary gland was measured by pulsed wave Doppler guided by the two-dimensional image to identify the caudal branch of the external pudic artery. Four other goats were used 1-3 months post-partum in this study. Infusions of vasopressin (1.0 ng/kg min) for 30 min caused constriction of the artery, leading to reduced blood flow, whereas corresponding infusions of oxytocin (10.0 ng/kg min) caused no change. It is concluded that both vasopressin and oxytocin increased milk flow and milk fat concentration despite decreased or undisturbed mammary blood flow. The results indicate that the effects are mainly due to constriction of the myo-epithelial cells with vasopressin acting as an agonist on the oxytocin receptors.

Dept. of Pediatric Pneumology and Immunology, Charité Campus Virchow, BMFZ Forum 4, Augustenburger Platz 1, D-13353 Berlin

*Dept. of Animal Physiology and #Clinical Radiology, Swedish University of Agricultural Sciences, Box 7045, S-750 07 Uppsala, Sweden

P 10-10

P 11-2

SLEEP RELATED BREATHING DISORDERS IN PARKINSONISM: FREQUENCY, NATURE, AND THERAPEUTICAL APPROACHES D. Schäfer, W. Greulich, M.E. Schläfke Objective: In the recent literature the contribution of sleep related breathing disorders [SRBD] to the complex sleep disorder of Parkinson´s disease [PD] patients is controversially discussed. In our study we have analyzed the frequency and kind of SRBD in a group of PD patients with sleep complaints. Patients and methods: 83 patients with different types of parkinsonism have been studied polysomnographically. Respiration has been measured by means of inductance plethymography, capnography, and pulse-oximetry. Different analysing techniques have been used. Besides the evaluation of apneas and hypopneas a visual classification of the respiratory pattern and a breath-to-breath calculation of the laboured breathing index [LBI] has been carried out. Results: More than 50% of the patients showed sleep disordered breathing. Most of them had upper airway obstructions. Central apnea was observed in 5 cases, three of those suffered from a heart disease. In 25% the SRBD should be treated. The LBI calculation revealed hints for short pharyngeal obstructions in each patient. This parameter turned out to be highly sensitive to detect changes in upper airway mechanics and was therefore considered to be useful in testing dopaminergic influences on respiratory movements. Conclusion: Disturbances of the extrapyramidal motor system may lead to long lasting obstructive hypoventilation of different clinical importance. Specific measurement techniques should be used to identify these disturbances. The laboured breathing index is useful to detect dopaminergic deficits in respiratory motion.

THE EFFECT OF METABOLITES OF MDMA (‘ECSTASY’) ON OXYTOCIN RELEASE IN VITRO M.L. Forsling, J.K. Fallon, D. Shah, A.T. Kicman, A.J. Hutt, D.A. Cowan Vasopressin release is stimulated following administration of 3,4methylenedioxymethamphetamine (MDMA) in man, but the underlying mechanism is not known, nor if the response is specific for vasopressin. A study has been performed in vitro comparing the oxytocin response to a major metabolite, 4-hydroxy-3-methoxymethamphetamine (HMMA), and four other metabolites with the parent compound. Hypothalami obtained from male Wistar rats were exposed to medium alone (B1) and the medium containing MDMA or a metabolite (B2) in the dose range 10-6 to 10-10 M. Similar experiments were performed for each compound following stimulation of the hypothalami with 40mM KCl, S1 and S2 representing the two incubation periods. At the end of each experiment the viability of the tissue was confirmed by exposure to 56mM KCl. Oxytocin concentrations in the medium were determined by radioimmunoassay and t tests with Dunnett’s correction for multiple comparisons were used to compare the B2:B1 and S2:S1 ratios for compound treated hypothalami with those of controls. All compounds examined had some effect on oxytocin release, HMMA and 3,4-dihydroxyamphetamine (DHA) being more active than MDMA, while 3,4dihydroxymethamphetamine (DHMA) was least active. For HMMA the ratio B2:B1 increased from 1.02±0.04 (control values) to 1.68 for 10-8M whereas for MDMA the ratio was 1.28±0.04 (SEM, p<0.05 compared to control). Thus, MDMA does stimulate oxytocin release, possibly via metabolites of the drug.

S 252 Neuroendocrine Laboratories, New Hunt’s House, GKT School of Medicine, Guy’s Campus, London Bridge London SE1 1UL, UK P 11-3 CAPSAICIN-SENSITIVE NEURONES IN THE CONTROL OF SPLANCHNIC CIRCULATION W.W. Pawlik, R. Sendur, J. Biernat, R. Obuchowicz Objective: The splanchnic vasculature is innervated by the intrinsic fibers from enteric nervous system and extrinsic nerves. Within the extrinsic visceral vascular nerves are visceral sensory afferent nerves. Such peripheral peptide containing fibers appear to participate in the transmission of sensory impulses in visceral reflexes. This innervation plays also an important role in the modulation of splanchnic local vascular responses. These afferent neural fibers when stimulated can elicit visceral vasodilatation via a local axon reflexes releasing vasodilator peptide-neurotransmitters (mainly CGRP). Results: It has been demonstrated that afferent sensory fibers can modulate resting intestinal blood flow and oxygen uptake and intestinal vascular autoregulatory responses including reactive hyperemia, functional hyperemia and post stimulatory escape. We found that capsaicin when applied periarterially elicited a significant vasomotor response from the splanchnic arteries that appears to have two components: an initial transient dilatation that fades and a sustained constriction of lesser amplitude than observed vasodilatation. During early hyperemic response blood flow was redistributed into the mucosal-submucosal compartment of intestinal wall. Whereas during the late vasoconstrictive response blood flow was distributed into the muscular compartment. The observed changes of the blood flow distribution in the vascular compartments in the intestinal wall may be due to grater density of afferent C fibers in the mucosal-submucosal layer. Conclusion: Afferent C fibers play a significant role in the control of gastric, pancreatic and hepatic vasculature at basal conditions and during ischemia/reperfusion induced vascular damage of visceral organs. Dept. of Physiology, Jagiellonian Univ. Sch. Med., 31-531 Cracow, 16 Grzegorzecka Str., POLAND

P 11-4 CENTRAL AND PERIPHERAL HISTAMINE AND ITS N-ALPHA METHYL ANALOG IN GASTRIC SECRETION AND GASTROPROTECTION S.J. Konturek, M. Pawlik, R. Pajdo, T. Brzozowski, A. Ptak, PC. Konturek, W. Bielañski Objectives&Aim: Nα-methylhistamine (NαMH) in an unusual histamine (H) side-chain derivative that was found in Helicobacter pyloriinfected stomach and was proposed to interact with histamine H 3 receptors, but its influence on gastric secretion and acute gastric lesions induced by noxious agents remains unknown. Methods: We compared the effect of i.g. and i.c.v. administration of NαMH with that H on acid gastric secretion in rats equipped with gastric fistula (GF) and on gastric lesions induced by 100% ethanol and 3.5 h of water immersion and restraint stress (WRS). The area of gastric lesions was determined by planimetry, the gastric blood flow (GBF) was assessed by H2gas clearance method and the plasma and gastric luminal gastrin levels was measured by RIA. Results: Both, NαMH and H (0,1 -5 mg/kg i.g. dose-dependently increased gastric acid output while NαMH and H (0.05 - 5 µg/kg i.c.v.) inhibited this secretion. NαMH and H (2 mg/kg i.g. or 1 µg/kg i.c.v.) attenuated significantly the ethanol- and WRSinduced gastric lesions while elevating significantly the GBF and plasma gastrin levels. These effects of topical NαMH and H were completely abolished by antrectomy or suppression of CCKB/gastrin receptors with RPR 102681 and significantly attenuated by clobenpropit (10 mg/kg i.p.), a histamine H3-receptor antagonist and by capsaicin denervation. Vagotomy, 6-OH-dopamine (60 mg/kg s.c.) that caused chemical sympathectomy and clobenpropit reduced the protective and hyperemic effects of H and NαMH (1 µg/kg i.c.v.). Conclusions: Central H and NαMH inhibit gastric secretion and exhibit gastroprotective activity involving an activation of both vagal and symphatetic nerves and the enhancement in gastric microcirculation mediated by specific histamine H 3-receptors. Department of Physiology, Jagiellonian Univ. Sch. Med. , 31-531 Cracow, 16 Grzegorzecka Str., POLAND

P 11-5 HUMAN PROLINE RICH ACIDIC PROTEIN INTERACTS WITH TRANSTHYRETIN AND RETINOL BINDING PROTEIN D.-C. Hao, J. Zhang, S.C. Hooi Abstract We have isolated the human proline rich acidic protein(PRAP) gene by NCBI BLAST search and 5’ and 3’ RACE (EMBL/GenBank accession no. AF421885). The gene is localized on chromosome 10q23.3 and encoded a putative protein of 151 amino acids. The first 22 amino acids was deduced to be a signal peptide which is 74% identical to the signal peptide of rat PRAP and 70% identical to that of mouse PRAP. We studied the tissue distribution of this novel gene by Northern blot. The 0.8 kb transcript of human PRAP is expressed in liver, kidney, small intestine, and colon. Interestingly, PRAP was not expressed in the colon of mouse and rat. To understand further the function of human PRAP, we have used the yeast two hybrid system to screen a pretransformed human liver Matchmaker library for proteins that interact with the full length PRAP. We show that two plasma carrier proteins, transthyretin and retinal binding protein, are able to interact with PRAP in the yeast AH109. The interactions were confirmed by yeast cotransformation and in vitro and in vivo binding assays. Further characterization of the binding site was done and aa81-111 of PRAP proved to be the minimal binding region. In addition, oxysterol binding protein like protein 9, skelemin and an unknown protein with four PKD1 (repeats in polycystic kidney disease 1) regions also interact with PRAP in the yeast. These results suggest a role for PRAP in vitamin A and sterol metabolism. Department of Physiology, Faculty of Medicine, National University of Singapore, Singapore 117597 P 11-6 ROMK IS THE APICAL POTASSIUM REGULATORY CHANNEL IN THE PARIETAL CELL G.G. MacGregor1, E. Vucic*, K. Radebold*, G.E. Shull, T. Wang1, S.C. Hebert1, J.P. Geibel*1 Introduction. In order to have effective gastric acid secretion it has been postulated that the apical pole of the parietal cell would need both an active H+,K+ATPase and a K+ channel to act as a recycling pathway for potassium. Recently we have suggested that ROMK (Kir 1.1) could be this recycling channel based on immunohistochemical evidence as well as intracellular pH measurements from perfused gastric glands. ROMK has a high open probability and is highly regulated by cytosolic pH, phosphorylation, PIP and the ADP/ATP ratio. Im2 portantly we have previously shown that ROMK maintains its high open probability even with very low extracellular pH, a requirement if it is to be an effective gastric K+ recycler. The goal of the present study was to determine H +,K+ATPase activity in the complete absence of this channel protein in the ROMK knockout mouse. Methods. Gastric glands were isolated from either the ROMK knockout (-/-) mouse or the paired littermate wild-type (+/+) control. Both groups of isolated glands were loaded with the pH sensitive dye BCECF for intracellular pH measurements, and monitored using a digital imaging system. Together with BCECF loading, glands were pre-incubated either with 100 µM gastrin, histamine or carbachol, and exposed to an acid load (20mM NH4Cl) Exposure to a Na + free Ringer solution post NH4Cl loading allowed for direct observation of H+,K+ATPase activity. Results. The stomach was enlarged in (-/-) mice compared to (+/+) littermate controls. Following induction of the acid load in the continued absence of Na+ there was no pH recovery (extrusion of protons from the cell) in the (-/-) mouse. In littermate (+/+) mice, a normal rapid recovery (cell alkalinization) in the absence of Na + was observed indicative of H+,K+ATPase activity. Conclusion. We have demonstrated that with deletion of the ROMK channel gene H+,K+ATPase activity is absent. We conclude from the present study that ROMK is the associated K+ recycling channel that is critical for acid secretion in gastric glands. Dept. of Surgery*, Dept. of Cellular and Molecular Physiology1, Yale University School of Medicine, New Haven, CT 06520 USA P 11-7 STIMULATION OF DUODENAL MUCOSAL BICARBONATE SECRETION IN THE RAT BY LUMINAL MELATONIN M. Sjöblom, G. Flemström

S 253 Introduction The bicarbonate secretion by the duodenal epithelium is considered the first line of mucosal defence against hydrochloric acid discharged from the stomach. Mucosal protection is under central nervous influence and some neural stimuli increase duodenal bicarbonate secretion by release of melatonin, probably originating from intestinal enterochromaffin cells (1). Our aim was to investigate whether luminally applied melatonin stimulates the secretion. Method Rats were anaesthetised with thiobarbiturate and a 12-mm segment of proximal duodenum with an intact blood supply was cannulated in situ. Duodenal mucosal bicarbonate secretion (pH stat) and the mean arterial blood pressure were continuously recorded. Agents were administered luminally, close intra-arterially (ia), and/or intravenously (iv). Results At a concentration of 1.0 µM luminally, melatonin significantly increased (p<0.05) duodenal mucosal bicarbonate secretion from 7.29 ± 1.5 to 13.20 ± 1.5 (n=7). The MT 2 selective melatonin receptor antagonist luzindole (600 nmol·kg -1 , iv) had no effect on basal bicarbonate secretion but inhibited secretion stimulated by luminal melatonin (n=6). Close ia infusion of melatonin (20-2000 nmol⋅kg-1⋅ h-1) to the duodenum similarly increased (p<0.05) duodenal mucosal bicarbonate secretion, and luzindole prevented the response. No changes in blood pressure were observed during these experiments. Conclusions Melatonin is a potent stimulant of duodenal mucosal bicarbonate secretion when administrated from the luminal as well as from the vascular side of the epithelium. In theory, melatonin may contribute to duodenal mucosal resistance to acid injury in a circadian fashion. Sjöblom M et al. J. Clin. Invest. 2001, 108:625-633. Department of Physiology, Uppsala University, PO Box 572, SE751 23 Uppsala, Sweden

ileum (J.Physiol 475:531-537). For this reason, the ability of a contiguous loop nutrient meal to alter fluid uptake was investigated in vivo in anaesthetised (70 mg/kg i.p Sagatal) Wistar rats using recirculated perfused loops to see whether a duodenal meal had similar effects. Fluid absorption was measured by recovered volume and luminal acidification by titration. At the end of the experiment, the animals were humanely killed. 10 cm duodenal loops and 25 cm lengths of proximal jejunum were prepared. The jejunum was perfused with phosphate buffer with the duodenal loop perfused with a partially hydrolysed peptic digest of casein (2g/100mls buffer), 80 mM oleic acid in 40 mM glycocholate, 1 mM cholesterol in acidified saline (150 mmoles/1 at pH 5.0) to simulate a partially digested meal. Fluid absorption was compared in animals with and without a duodenal loop perfused with nutrient. With nutrients in the duodenal loop, jejunal fluid absorption was 47.1 ± 13.4 (8) ul/hr/cm, significantly less (P < 0.01) than the 96.6 ± 19.7 (8) µl/cm/hr from control loops. When STa (80 ng/ml) was in the jejunum, absorption was 65.6 ± 5.4(10) µl/cm/hr. Luminal acidification of 2.53 ± 0.48 (9) µg H+/cm/hr in controls resembled the 2.12 ± 0.24 (8) µg H+/cm/hr in loops with duodenal test meal perfusion. Both exceeded ( p < 0.02) the 1.12 ± 0.28 (11) µg H+/cm/hr for STa treated jejunal loops. Perfusion of the duodenal loop with nutrient reduced jejunal fluid absorption to values characteristic of exposure to STa without effect on luminal acidification rates. A new model for water absorption requires acidification to create hypotonicity at the brush border (J Appl Microbiology 90:7-26). The reduction in jejunal fluid absorption after perfusion of duodenum with nutrient solution does not seem to be acting via this mechanism. Institute of Biomedical & Life Sciences, Glasgow University, Glasgow G12 8QQ, Scotland, United Kingdom

P 11-8 P 11-10 IN VITRO AND IN SITU EXPERIMENTAL MODELS FOR X-RAY MICROANALYSIS OF INTESTINAL EPITHELIUM V. Vanthanouvong, M. Högman, G.M. Roomans Intestinal chloride (Cl) transport is disturbed in a number of diseases, e.g., cystic fibrosis. X-ray microanalysis can be used to study the distribution of Cl and other ions in intestinal epithelial cells. In this study it was attempted to set up an experimental system that retains the in vivo elemental composition of the epithelial cells. An in vitro system was set up, in which a segment of rat intestine was removed from the animal and mounted in a bath and perfused with different fluids. The chloride in the bath or in the perfusion fluid could be exchanged for gluconate or bromide to determine the direction of chloride fluxes. An in situ system was set up in which the animal was anesthetized and a segment of the intestine was perfused with different solutions. The arterial pressure was recorded during the experiment and body temperature was maintained at 37°C. Tissue for X-ray microanalysis was rapidly frozen and 16 µm thick freeze-dried cryosections were analyzed in a scanning electron microscope. Intestinal tissue that had not been exposed to any solutions was frozen and used as control. In the in vitro experiments the concentration of Na and Cl in the epithelial cells increased and that of K decreased. These changes occurred already with the first 30 min of incubation. Uptake of chloride occurred mainly from the bath, as seen in experiments where bromide was used as a chloride analogue. The concentration gradient between bath and tissue determined the extent of chloride uptake. Addition of glucose to the perfusion fluid and bath did not improve the results. In the in situ systems, preservation of the intracellular ion composition was better. Good results were obtained with perfusion with Krebs-Ringer’s buffer without glucose for 30 min. In this case, the elemental content of the cell did not change appreciably during incubation. If glucose was added, the Na concentration increased in comparison to the control, both in crypt and villus cells. It is concluded that the intestinal epithelium is a sensitive system, very prone to disturbance of its homeostasis. The in situ system can be used in studies of agonist-induced ion transport. Department of Medical Cell Biology, University of Uppsala, Box 571, 75123 Uppsala, Sweden P 11-9 THE EFFECT OF A DUODENAL MEAL ON FLUID ABSORPTION IN VIVO IN RAT PROXIMAL JEJUNUM M.L. Lucas, K.I. Mahood, D. Pande, J. McCormick The jejunum may reflexly secrete fluid in response to E.coli STa in

COLONIC ANION SECRETION IN NKCC1 KNOCKOUT MICE: CHARACTERIZATION OF AN ALTERNATIVE ANION UPTAKE PATHWAY P. Jacob, C. Weinhold, I. Blumenstein, M. Gregor, M. Flagella, G. Shull, U. Seidler Backround: Based on pharmacological experiments, we have recently demonstrated that multiple pathways are involved in cellular Cl - uptake during colonic anion secretion. Besides Na + K +2Cl cotransport, the alternative pathway predominant in crypts appears to be an exchange of intracellular HCO3- for Cl- via a basolateral DIDSsensitive anion exchanger, most likely AE2. HCO3- originated from import of HCO 3- via Na+HCO3- cotransport (NBC), and intracellular CO2 hydration via carbonic anhydrase. Aim: In order to substantiate these findings and to look for the adaptive capacity of the alternative pathway, we studied colonic anion secretion in NKCC1 knockout mice in the presence and absence of inhibitors for NBC alone, AE2 and NBC, and carboanhydrase, and compared to results to those in normal mice after acute NKCC inhibition by bumetanide. Methods: The muscle-stripped proximal colon of NKCC1 -/- and +/+ mice was placed in Ussing chambers and Isc and HCO3- secretion assessed by voltageclamp and pH-stat titration. Results: In proximal colonic mucosa of NKCC1 -/- mice, basal and db-cAMP stimulated Isc was reduced to approx. 40% of control Isc, which was the same degree of inhibition as elicited by bumetanide in +/+ mouse mucosa. 55% of the remaining Isc was inhibited by the NBC-specific inhibitor S0859 (which does not influence Cl-/HCO3- exchange activity), another 25% by azetazolamide, and the remaining 20 % were dependent on the presence of luminal Cland therefore due to apical Cl-/HCO3- exchange via DRA. Basolateral DIDS (which will inhibit NBC and AE2) caused approx. the same degree of inhibition (80%) than S0859 and azotazolamide together, suggesting that AE2 is indeed the likely basolateral Cl- uptake mechanism in NKCC1 -/- colonic mucosa. Compared to bumentanide-treated +/+ colonic mucosa, a higher percentage of Isc was inhibited by Na + HCO 3 - cotransport inhibition and a lower percentage by azetazolamide, suggesting upregulation of Na +HCO3- cotransport in these mice. In addition, the percentage of HCO 3- in the db-cAMP stimulated anion secretory response was approx 40% in NKCC1 -/mucosa compared to approx 15% in bumetanide-treated +/+ mucosa. Conclusions: Both pharmacologic inhibition and disruption of gene expression of the colonic Na +K +2Cl - cotransporter inhibits 60% of electrogenic anion secretion in murine proximal colon. Although the NKCC1 colon upregulates its HCO3- transport pathways, the remaining anion secretion is still predominantly Cl-, which is taken up via a basolateral Cl-/HCO3- exchanger in exchange for HCO3- that is supplied

S 254 via Na+HCO3- cotransport and intracellular CO 2 hydration. This pathway may be an interesting target for antidiarrheal drug development. 1st Med. Dept, University Hospital Schnarrenberg, Otfried-MüllerStr. 10, D-72076 Germany

integrin clustering. Supported by SFB 366/C3. Dept. of Physiology, Freie Universität, Arnimallee 22, D-14195 Berlin, Germany P 12-3

P 12-1 CASPASE-10 IS CRITICALLY INVOLVED IN SPONTANEOUS APOPTOSIS OF HUMAN POLYMORPHONUCLEAR NEUTROPHILS F. Göpel, P. Weinmann, B. Walzog Mature human polymorphonuclear neutrophils (PMN) undergo apoptosis spontaneously within hours to days and this is thought to contribute to the high turnover of these cells in the circulation. However, the molecular mechanisms which underlie the induction of spontaneous apoptosis of PMN in physiological settings are not completely understood. Since caspases are known to be critical for the control of the apoptotic machinery, we studied the role of these enzymes in spontaneous apoptosis and in tumor necrosis factor α (TNFα)-induced acceleration of apoptosis. Apoptosis was measured by detection of CD16 expression on the cell surface, by analysis of DNA content and nuclear morphology as well as by detection of DNA degradation. By means of RT-PCR and western blotting technique, we were able to identify caspase-10 in mature human PMN. Moreover, spontaneous apoptosis was accompanied by a specific activation of caspase-10 as detected by measuring the activity of the isolated enzyme using a fluorogenic substrate. The involvement of caspase-10 in spontaneous apoptosis was further confirmed by the finding that this process was absent in the presence of a caspase-10 inhibitor. In contrast, TNFαinduced acceleration of spontaneous apoptosis but not spontaneous apoptosis by itself was dependent on activation of caspase-8. Together, this shows that the control of spontaneous apoptosis and TNFα-induced acceleration of spontaneous apoptosis underlie different molecular mechanisms: Whereas caspase-8 seems to represent the critical checkpoint for the activation of the apoptotic machinery of human PMN in inflammation, activation of caspase-10 seems to be critical for spontaneous apoptosis which controls the life-span of PMN in physiological settings. Supported by SFB 366/C3 Dept. of Physiology, Freie Universität, Arnimallee 22, D-14195 Berlin, Germany P 12-2 A ROLE FOR SYK-KINASE IN THE CONTROL OF THE BINDING CYCLE OF THE Β2 INTEGRINS (CD11/CD18) IN HUMAN NEUTROPHILS P. Prange, T. Willeke, B. Walzog During recruitment of human polymorphonuclear neutrophils (PMN) to sites of inflammation, leukocyte adhesion molecules of the β2 integrin family (CD11/CD18) mediate firm adhesion of these cells to the endothelial cell monolayer lining the vessel wall. A fine control of the binding of the β2 integrins to their ligands is required to allow shape change and spreading as well as migration of PMN on the endothelial cell monolayer to the intercellular junction between neighbouring endothelial cells where PMN eventually transmigrate into the extravascular space. To elucidate the molecular mechanisms which allow ordered β2 integrin-mediated adhesion and de-adhesion of PMN, intracellular protein tyrosine signalling was studied subsequent to β 2 integrin-mediated ligand binding. By means of immunoprecipitation and western blotting technique, a 72 kDa protein which was tyrosine phosphorylated upon adhesion was identified as Syk-kinase. Moreover, Syk-kinase was found to be directly associated with CD18, the βsubunit of β2 integrins. Inhibition of Syk-kinase by piceatannol enhanced adhesion and spreading of PMN but diminished fMLP-induced chemotactic migration. The analysis of the molecular mechanism which underlies this enhancement of PMN adhesiveness revealed that inhibition of Syk-kinase induced the clustering of the β2 integrins on the cell surface which is known to result in increased integrin avidity. In contrast, inhibition of Syk-kinase had no effect on CD18 expression or expression of activation-specific neoepitopes on β 2 integrins. Altogether, the present study suggests that Syk-kinase may control alternation of strong β2 integrin-mediated ligand binding with integrin detachment which is necessary to allow ordered migration of PMN to sites of inflammation by controlling the avidity of the β2 integrins via

LOCOMOTORY BEHAVIOR AND MOTILITY OF PERIPHERAL HUMAN LYMPHOCYTES (CD4) UNDER THE INFLUENCE OF 50 HZ MAGNETIC FLUX DENSITIES OF 10 MT J. Reißenweber, E. David, F. Gholamrezaei, M. Pfotenhauer, A. Wojtysiak Lymphocyte migration under the influence of rather strong magnetic fields in comparison to that without field should be investigated. Extremely-low-frequency (ELF) and low-frequency (LF) electric and magnetic fields and high-frequency (HF) electromagnetic fields of our everyday life are discussed to provoke impairments of well-being in some percent of the population. In this context the term of electromagnetic hypersensitivity must not be neglected. Until now investigations using magnetic flux densities of up to 100 µT (50 Hz) in human and animal experiments normally did not result in significant biological effects. Therefore we elucidated the question whether lymphocyte migration and velocity – important parameters of the body´s immune surveillance – might be modified by magnetic flux densities in the intermediate intensity range. So magnetic 50 Hz flux densities of 10 mT were applied. This value represents the 100 fold of the German threshold value of 100 µT (50 Hz). A major objective of this preliminary study was to test if magnetic flux densities above that threshold value could cause significant effects in terms of changes in cell velocity or motility. Peripheral human lymphocytes (CD4) were examined in regard of average velocity and percentage of moving lymphocytes under and without exposure in extremely low frequency magnetic fields. CD4 cells were chosen which represent T-helper cells or T-inducer cells. Homogeneous sinusoidal flux densities of 10 mT at a frequency of 50 Hz were applied by a magnet. By time-lapse video microscopy groups of exposed and sham-exposed (control) lymphocytes were filmed during migration for 120 minutes each. For assessment of the moving percentage of 30 lymphocytes the individual pathway (track) was recorded. Finally the velocity of all lymphocytes could be determined. A temperature of 37 °C was exactly maintained throughout the experiments independently of field status. A clear, however small, difference between mean percentages of moving lymphocytes under and without field exposure could be demonstrated: the mean values with field were more elevated than those without field. However, no significant difference in cell velocities appeared between status field on/field off. Possibly lymphocytes´ locomotion in vivo through the tissue and even in vitro is influenced by adhesion phenomena between cells and cells or between cells and connective tissue or matrix. It is conceivable that adhesion forces might be weakened by the external magnetic field. Thus the average percentage of moving lymphocytes could indirectly be increased by the magnetic field. Another explanation: a quite small not measurable enhancement of temperature (heating effect) eventually caused by the magnetic coil could enhance lymphocyte metabolism. The present preliminary results point towards a visible difference between curves representing the lymphocytes´ migratory behavior with and without magnetic field influence. However, heating effects of magnetic coils cannot be excluded from the first. Institute of Physiology and Physiopathology, Witten/Herdecke University, Stockumer Straße 12, D-58453 Witten, Germany P 12-4 EFFECT OF RED BLOOD CELLS TO PLATELET RESPONSES INDUCED BY SECONDARY FLOW S. Körfer, K. Mottaghy Objective of study: The objective of our work was to evaluate the effects on platelet reactions especially receptor activation due to secondary flow appearing in blood. Secondary flow is characterised by curved streamlines involving changes in velocity and direction. This leads to eddy formation and flow separation. In vivo secondary flow occurs e.g. at aortic T-junctions, at carotid bifurcation and particularly under pathological conditions. Method used: Secondary Flow and also laminar flow were generated in a new introduced Taylor-Couette system under defined blood flow conditions with same shear rate (G=600

S 255 1/s). Different haematocrit (hct) values were adjusted by centrifugation and resuspending in native plasma. Platelet activation was measured using flow cytometry (GP Ilb-IIIa activation, CD62P expression, micropaticle and aggregate formation) and ELISA technique (PF4 release). Adhesion profile was studied using scanning electron microscopy. Results: Platelet responses at different hct values were determined and compared with measurements resulting from studies with native whole blood. With 0 % hct no difference between platelet response to laminar flow and to secondary flow can be detected. In whole blood secondary flow leads to a higher fibrinogen receptor activation and a raised aggregate formation. On the other hand laminar flow induces enhanced platelet adhesion, higher PF4-release and increased microparticle formation in whole blood. All of this parameters excluding micro-particle formation are dependent on hct value. Platelet adhesion, PF4-release, GPIIb-IIIa activation and aggregate formation increases with increasing hct whereas expression of CD62P measured in the bulk fluid decreases. Already at 10 % hct the differences in platelet response between the two flow modes are detectable. Conclusions: Platelet responses to secondary flow were evaluated subject to occurrence of red blood cells. Platelet activation is affected by red cells and not directly influenced by secondary flow as shown with 0 % hct. Department of Physiology and IZKF „Biomat“, University Hospital RWTH Aachen, Pauwelsstraße 30, 52074 Aachen, Germany P 12-5 NO THROMBOEMBOLISM IN MICE AT A HEMATOCRIT OF 0.85: THROMBOPOIESIS AND ENDOGENOUS HYPOCOAGULABILITY J. Shibata, J. Hasegawa, D. M. Katschinski, E. Wolber, W. Jelkmann, R.H. Wenger, K.F. Wagner* Objective: A high hematocrit correlates with thrombembolic events in polycythemia vera. But tg6 mice, which have an inborn high hematocrit of 0.85 from overexpression of the human Erythropoietin transgen (1), do not suffer from thrombosis or embolizations. Therefore, we investigated thrombopoiesis, primary hemostasis, plasma coagulation factors, platelet function and whole blood coagulation in tg6 mice. Materials and Methods: Male tg6 and their wild type (wt) littermates were investigated at 1, 2, 4, 6, and 8 months. Morphology of thrombopoiesis and erythropoiesis were characterized from bone marrow smears and thrombopoietin (TPO) and EPO protein and TPO mRNA levels measured. Whole blood coagulation and platelet function were assessed by thrombelastography and functional tests investigating the influence of elevated red cell concentration and platelet function were performed. Results: Tail bleeding time was not significantly different between wt and tg6 mice. Plasma-EPO was markedly elevated in tg6 mice while TPO levels were not different. Unexpectedly, blood platelet concentration was significantly reduced from 1,000,000 ± 300,000 /µL (wt) to 450,000 ± 150,000 /µL (tg6). Activated PTT was increased (29.1 ± 5.9 sec vs. 70.3 ± 20.7 sec) and the Quick-value was markedly in tg6 mice. TEG analyses revealed a significantly reduced clot strength in tg6 mice which further deteriorated with age. Conclusions: While primary hemostasis was normal, the function of the extrinsic and intrinsic system was severely impaired in tg6 mice. Reduced megacaryocyte numbers due to a displaced thrombopoiesis resulted in a reduced platelet count. Chronic right heart congestion most likely did account for a liver cell damage resulting in a reduced production of hepatic coagulation factors. Together with a mild hemolysis tg6 mice showed a marked hypocoagulability of whole blood preventing thromboembolism. Reference: (1) Wagner KF, Katschinski DM et al. Blood 2001; 97:536-42. K.F. Wagner, Physiology/Anesthesiology, MUL, 23538 Luebeck, Germany

P 12-6 THE –786C VARIANT OF THE HUMAN ENDOTHELIAL NITRIC OXIDE SYNTHASE GENE PROMOTER IS A RISK FACTOR FOR CORONARY HEART DISEASE M. Cattaruzza, W. Stodowski, A. Pelvan, 1J. Becker, M. Halle, A.B. Buchwald, M. Hecker Background: Coronary heart disease (CHD) is defined as an insufficient blood supply of the myocardium due to atherosclerotic stenosis

of coronary arteries. In addition to the known risk factors for developing atherosclerosis, gene polymorphisms may predispose to CHD. One candidate gene for such a polymorphism is the endothelial isoform of NO synthase (NOS-3) where a C/T variance, which is thought to cause impaired expression of the gene, has been found at position – 786 in the 5’-flanking region. Methods: DNA isolated from umbilical arteries and from patient blood collected during heart catheterization was genotyped for the -786C/T variance of the NOS-3 gene by restriction fragment length polymorphism analysis. Moreover, laminar shear stress-induced NOS-3 expression in umbilical vein primary cultured endothelial cells (HUVEC) was determined by Western blot and RTPCR analyses. Results: Among the 338 artery specimens analyzed, 36% exhibited a -786T/T, 55% a -786C/T and 10% a -786C/C genotype. Of the 84 patients with angiographically proven CHD (37-64 years old), -786 C/T and especially the -786C/C genotype were more frequent the with 67% and 17%, respectively. In contrast, in patients without CHD, the -786T/T genotype was most frequent with 50% as compared to 44% and 5% for the -786 C/T and -786 C/C genotype, respectively. Moreover, when the increase in NOS-3 expression in response to laminar shear stress was analyzed in HUVEC isolated from genotyped donors, this was highest for cells with -786T/T genotype (7-fold increase over 24 h as compared to the static control), intermediate in cells with -786C/T genotype (4-fold) and absent in cells with -786C/C genotype (0.4-fold). Conclusions: The observed association of the -786 C/C genotype with CHD thus appears to be causally linked to the lack of adaptation of NOS-3 expression, hence nitric oxide formation, to an increase in shear stress brought about, e.g. by vasoconstriction or an exercise-induced increase in coronary blood flow. Departments of Cardiovascular Physiology and Cardiology and Pneumology, Heart Center, University of Göttingen and 1 Clinics NeuBethlehem Göttingen, Göttingen, Germany

P 12-7 THE QUANTITATION OF BUFFERING ACTION–A FORMAL AND GENERAL APPROACH B.M. Schmitt The term ”buffering” originated in acid-base physiology, but was extended to other phenomena, e.g. Ca++, Mg++, redox, or blood pressure buffering. The quantitation of buffering action, however, has remained controversial and problematic. Units may be missing altogether (e.g., for blood pressure buffering), there may be multiple, incommensurate units for identical phenomena (e.g., for H+ buffering), and units may not allow for scientific scales of a higher type (e.g., dBase/dpH). Here, I propose a ”unified” concept of buffering that provides a single ratio scale for any type of buffering. It is the follow-through, systematization, and generalization of an approach pointed out (and discouraged) by Van Slyke (J.Biol.Chem. 52:2, 1922) and practiced by Neher & Augustine (J.Physiol. 450, 1992). Briefly, a ”buffered system” is defined as any set of two functions τ(x) and β(x) that satisfies the ”buffering condition”, i.e. τ’(x)+β’(x)=1. The ratio β’(x)/τ’(x), termed ”buffering odds” B(x), is then taken as a measure of buffering action. The advantages are: i) The concept is general, because it is purely formal and mathematical, and the ”buffering odds” B are dimensionsless, even if x is not. Like other mathematical concepts, a ”buffered system” is not a ”real” thing, but a quantitative pattern exhibited (or not) by ”real” things, and thus a tool for their description. ii) The concept allows to build a ratio scale, i.e., an equal interval scale with an absolute zero. Of all scientific scales, this is the most versatile and useful type. iii) The concept is simple, intuitive, and allows to visualize buffering processes by physical models. For instance, a fluid partitioning into two communicating vessels represents a ”buffered system” if total fluid volume is taken as x, and the volumes in the two individual vessels as τ(x) and β(x). With proper design, the cross-sectional area of one vessel is proportional to B(x) at any given fluid level. The concept provides stringent and intuitive, though unorthodox insights into “classical” buffered systems, e.g. H+ buffering by pure water or by weak acids. Herein, x denotes total H+ ions, and τ(x) and β(x) the free and bound H+ ions, respectively. The concept affords a similarly rigorous quantitative treatment of “non-classical” buffering phenomena for which units were problematic or missing (e.g., redox or blood pressure buffering, ”cognitive buffering” of stress), or that are not usually interpreted as buffering phenomena (e.g. saturation of electrolyte solutions, rectification by ion channels, or autoregulation of kidney and brain perfusion).

S 256 Anatomy & Cell Biol. I, Univ. of Würzburg, Koellikerstr.6, D-97070 Würzburg P 12-8 ALBUMIN-INDUCED PROINFLAMMATORY NF-ΚB EXPRESSION AND ACTIVATION IN HUMAN RENAL PROXIMAL TUBULAR CELLS (IHKE-1) K. Drumm, B. Bauer, B. Gassner, S. Silbernagl, M. Gekle Chronic renal diseases with enhanced glomerular protein filtration are accompanied by tubulointerstitial inflammation and progression to renal function deterioration. Albumin seems to be a pathogenetic factor per se in the progression of renal diseases, as several studies support a role of albumin in the progression of renal interstitial inflammation and fibrosis. The transcription of numerous inflammatory or fibrotic mediators as, e.g. IL-1β or TNF-α is mediated by the nuclear factor ΚB (NF- ΚB). To investigate the hypothesis, that NF-ΚB may be involved in albumin-induced renal inflammatory pathomechanisms, we exposed human renal proximal tubule cells (IHKE-1) to 50 or 500 mg bovine serum albumin (BSA)/I up to 12 hrs. After BSA-exposure, total RNA, total protein or nuclear proteins were extracted. The NF-ΚB and TNFα specific mRNA expression was detected by RT-PCR. NF-ΚB specific protein expression was analysed by Western blot. Electromobility shift assays (EMSAs) were performed to determine the NF-ΚB specific DNAbinding activity in nuclear protein extracts of BSA-exposed IHKE-1 cells. Albumin-exposure induced an increase in NF-ΚB and TNF-α specific mRNA expression NF-ΚB protein expression and activity. These effects are decreased by BIM and Herbimycin A. We suggest, that albumin-exposure induces an increase in NF-ΚB and TNF-α specific mRNA expression, NF-ΚB specific protein expression and protein activity in renal proximal tubule cells in culture, which is at least in part protein kinase C and tyrosine kinase dependent. Most probably this NF-ΚB and TNF-α activation is involved in albumin-associated inflammatory or fibrotic renal pathomechanisms in vivo. Supported by the DFG Ge 905/4-1.

P 12-10 HYPEROSMOLAR MEDIA INHIBIT INTRAERYTHROCYTIC GROWTH OF PLASMODIUM FALCIPARUM A.-C. Uhlemann, V. Brand, P. Kremsner, S.M. Huber, F. Lang The intraerythrocytic survival and growth of Plasmodium falciparum depends on the properties not only of the pathogen but as well of the host cell. Pathogen survival is expected to be modified by alterations of intracellular electrolyte composition, metabolism and transport across the cell membrane, all functions sensitive to host cell volume. To study whether the host cell volume is important for parasite development, infected human erythrocytes were grown in vitro in increasing NaCl concentrations. The replication of parasites was assessed microscopically and quantified by FACS analysis after ethidiumbromide staining. The viability of the host cell was determined by determining the rate of hemolysis. Addition of 200 and 450 mM of NaCl (for 48 h) to the isotonic culture medium killed the parasites completely (n = 3) while addition of 10, 20, 50, and 100 mM NaCl decreased parasite replication to 0.61 ± 0.03 (n = 9), 0.43 ± 0.05 (n = 3), 0.35 ± 0.02 (n = 9), and 0.17 ± 0.02 (n = 9), respectively, compared to growth in isotonic control medium (1.0 ± 0.02, corresponding to an increase in parasitemia by factor 3.7 ± 0.3 within 48 h; n = 9). Even short exposure (20 min) to medium with increase of NaCl by 100 mM followed by 48h of culture in isotonic medium induced a decline in parasite replication to 0.52 ± 0.05 (n = 3) of the control. In contrast, culturing of non-infected cells (for 48 h) in medium with increase of NaCl by 100 mM did not alter cell number or hemolysis. In conclusion, increase of ambient osmolarity by addition of NaCl to the culture medium impairs intraerythrocytic growth of P. falciparum while (up to a concentration of 100 mM) does not affect the viability of the host. Department of Physiology I, University of Tübingen, Gmelinstr. 5, 72072 Tübingen, Germany P 13-1

Dr. Karina Drumm, Physiologisches Institut der Universität Würzburg, Röntgenring 9, D-97070 Würzburg, Germany

P 12-9 CYCLIC AMP-REGULATED PROTEINS IN EMBRYONIC RED BLOOD CELLS OF THE CHICK EMBRYO S. Dragon, R. Baumann Red blood cells (RBC) of the avian embryo complete their terminal differentiation in the circulation. In the past, we could show that several key events of late erythroid differentiation, like changes in the glycolytic pathway, nucleotide degradation and the induction of carbonic anhydrase are controlled by the signaling molecule cAMP. Since transcriptional activation is absolutely required for the cAMP responses, we looked for cAMP-induced genes by differential cDNA library screening: We identified three cAMP-induced genes: The transcription factor fos, the anti-proliferative tob protein, and the differentiationpromoting gene ifr1. In addition to mRNA expression studies we detected the gene products by western blotting analysis. The proteins are expressed in definitive and primitive RBC populations with characteristic differences in the temporal induction profile: The Fos protein is hardly detected in erythroid cells without cAMP stimulation, but is induced fast and stable upon hormonal stimulation for at least 24 h. Ifr1 which is already expressed in not stimulated cells, becomes upregulated by cAMP much slower. In contrast, Tob is expressed transiently between two and twelve h of stimulation. Finally, we tested the cAMP-dependent induction of the three proteins in response to shortterm hypoxia in vivo: Hypoxic incubation of the embryo causes a rise in the plasma norepinephrine concentration thereby raising the cAMP levels of circulating RBC: While we observe no significant change in the Ifr1 level, Fos and Tob proteins are strongly up-regulated by hypoxia. Although the precise role of the three proteins in erythroid cells is still unknown we think that they are important components of terminal erythroid differentiation. Physiologisches Institut der Universität Regensburg, Universitätsstr. 31, D- 93040 Regensburg

TIME CONSTANT (ττ ) ESTIMATES OF THE CARDIAC ISOVOLUMIC PRESSURE DECAY BECOME ERRONEOUS IF PRESSURE ASYMPTOTE IS PREFIXED INSTEAD OF COESTIMATED S.F.J. Langer, H. Habazettl, W.M. Kübler, A.R. Pries Objective: The time constant τ of the exponential part of the isovolumic left ventricular pressure (LVP) decay in hearts is mostly estimated by fitting the empirical data by the three-parameter regression LVP(t)=P ∞ +(P 0 -P ∞ )exp(τ -1t). However, some authors use a twoparametric regression with preset P ∞ =0 instead of estimating the asymptote from the data. The present study tests the hypothesis that these three- and two-parameter approaches (τ 3 and τ 2, resp.) on the time constant τ yield equivalent information.Methods: A mathematical formula was established to exactly precalculate τ2 for arbitrarily given τ and P∞ to design a hypothesis-challenging protocol. Eleven guinea pig and rat hearts were repeatedly preloaded by rising left atrial inflow in an isolated working left heart apparatus; τ3 and τ2 were separately estimated from the LVP data obtained by an intraventricular subminiatur manometer. The sequences of τ3 and τ2 during each preload test were statistically compared by a modified two-sided sign-sequence test. Results: Most of the preload tests with freshly isolated hearts present with significant (p <0.01) concordance between τ3 and τ2. This pattern reversed in spontaneously depressed hearts (>3h after isolation): τ 3 (closed circles) and τ 2 (open circles) proceed in opposite direction during P∞ increases with preloading (Figure).

S 257

Institute of Physiology, Freie Universität Berlin, Arnimallee 22, D14195 Berlin

sion to static control values. As for TSP-1, the mRNA for CD36 is down-regulated by flow exposure. The time dependence as well as the reversibility of the shear stress effect show a parallel modulation pattern for CD36 and TSP-1. Compared to TSP-1, a slightly lower shear stress threshold (1 dyn/cm2) appears to be sufficient for CD36 downregulation. Conclusions: Our data point out the relevance of shear stress as modulator of TSP-1 and CD36 expression. Furthermore, the parallel down-regulation of TSP-1 and CD36 indicates that shear stress affects especially the angio-adaptive features of TSP-1. This supports the hypothesis of an involvement of TSP-1 in the control of flowdependent angio-adaptation. Supported by DFG (Za 184/1-2).

P 13-2

Dept. of Physiology, Freie Universität Berlin, Arnimallee 22, D-14195 Berlin

Conclusion: Obtaining the time constant of isovolumic pressure fall from models with preset or with estimated pressure asymptote leads to quantiatively and qualitatively different measures. This may result in directly contradictory information about „lusitropy“ if the estimated asymptote changes in opposite direction to τ. Thus estimating τ with preset zero asymptote provides erroneous results if changes in the actual asymptote occur. Supported by the Sonnenfeld Foundation, Berlin

PRESSURE-DEPENDENT ENDOTHELIAL NITRIC OXIDE PRODUCTION IN LUNG VENULAR CAPILLARIES W.M. Kuebler, H. Habazettl, H. Kuppe, A.R. Pries Introduction: Endothelial dysfunction is a prominent feature in pulmonary hypertension and characterized by an imbalanced release of endothelial vasoactive mediators. However, the effects of pressurestress on endothelial nitric oxide (NO) production are still discussed controversially. Methods: Here, we determined the endothelial concentration of NO in intact capillaries of the isolated blood-perfused rat lung by fluorescence imaging of the NO-sensitive dye DAF-FM. Since endothelial NO synthase is regulated by the cytosolic Ca2+ concentration ([Ca2+]i), we separately quantified endothelial [Ca 2+] i by in situ fura-2 ratio imaging technique (J Clin Invest 104:495, 1999). Images were obtained during baseline at left atrial pressure (PLA) of 5 cmH2O, then during 30 min of PLA elevation to 20 cmH2O, and after return to baseline PLA. Experiments were replicated after addition of gadolinium chloride (GdCl3, 10 µΜ), a Ca2+ channel blocker, to the blood. Results: During elevation of PLA, endothelial DAF-FM fluorescence increased from 60±9 fluorescence units (FU) to 125±15 FU after 10 min and 218±42 FU after 30 min (both p<.05 vs baseline). Concomitantly, endothelial [Ca2+]i increased from 105±4 nM to 118±6 nM and 172±8 nM, respectively (both p<.05 vs baseline). Return to baseline PLA rapidly (<5 min) reset DAF-FM fluorescence to 70±8 FU and [Ca2+]i to 100±2 nM. [Ca2+]i and DAF-FM fluorescence showed significant correlation (rs = 0,976, p<.001). GdCl3 did not affect baseline DAF-FM fluorescence (79±8 FU) or [Ca2+]i (100±5 nM), but completely blocked the pressure-induced responses (77±4 FU, 101±4 nM [Ca2+]i after 30 min, both p<.05 vs control). Conclusion: We propose that elevation of vascular pressure increases endothelial NO production in lung venular capillaries by Ca2+ influx through GdCl3-inhibitable channels. This effect may constitute a vasoregulatory response to pressure stress within the lung vasculature. W.M. Kuebler, Institute of Physiology, Free Univ. of Berlin, Arnimallee 22, 14195 Berlin, Germany P 13-3 DOWN-REGULATION OF THE THROMBOSPONDIN-1/CD36 SYSTEM IN SHEAR STRESS EXPOSED ENDOTHELIAL CELLS M. Bongrazio, E. C. Bergmann, A.R. Pries, A. Zakrzewicz Introduction: Shear stress-dependent modulation of genes controlling angiogenesis is supposed to play an important role in the process of vascular adaptation to altered flow conditions. The multifunctional protein thrombospondin (TSP)-l is a potent inhibitor of angiogenesis. This activity is mediated by the interaction of the TSP/typel domain with the CD36 receptor. We previously reported the shear stress-dependent induction of another protein (METH-1) containing TSP/typel repeats. In this study we investigated if also TSP-1 and/or CD36 may be modulated by shear stress in endothelial cells (EC) and thus potentially be involved in flow-dependent angio-adaptation. Methods: Human umbilical vein EC (HUVEC) or cardiac microvascular EC (HCMEC) were exposed to laminar shear stress (HUVEC: 0.6-6 dyn/cm2, 4-72 h; HCMEC: 6 dyn/cm2, 24 h) using a cone-and-plate apparatus. In additional experiments, HUVEC were exposed to shear stress (6 dyn/cm2; 24 h) and then returned to static conditions for different times (1, 4, 24 h). Gene expression was analyzed by RT-PCR, Northern, and Western blot. Results: Exposure of EC to shear stress markedly reduces TSP-1 expression, both at the mRNA and the protein level. The downregulation of TSP-1 mRNA becomes significant after 24 h (6 dyn/ cm2), is sustained for up to 72 h, and requires a threshold shear stress of 2-3 dyn/cm2. Restoration of no-flow conditions returns TSP-1 expres-

P 13-4 MATURATION OF MICROVASCULAR NETWORKS: A SIMULATION APPROACH B. Reglin1, T. W. Secomb3, A.R. Pries 1,2 Introduction: Microvascular networks, generated under physiological conditions, e.g. during embryogenesis, as well as upon topical application of VEGF, are initially characterized by a high vessel density. During subsequent maturation, vessel diameters change and a significant portion of vessels is removed by „pruning“. Here, a previously developed simulation model of structural diameter adaptation in complete microvascular networks is used to improve understanding of principles underlying this maturation process. Methods: A model is used, in which any vessel segment changes its diameter in response to hemodynamic and metabolic conditions according to an optimized set of adaptation rules. If the diameter of a given segment falls below a threshold value, this segment is removed („pruning“). Diameter adaptation is simulated for A) mature networks, with topology derived from the rat mesentery in vivo, and B) immature networks, represented by dense, regular, hexagonal („honeycomb“) grids which correspond e.g. to capillary plexus resulting from vasculogenesis. In both cases, adaptation starts with equal diameters for all vessel segments. Results: A) Simulated adaptation in mature networks results in diameter distributions very similar to those observed in vivo. No segment is removed, i.e. no additional pruning occurs. B) In immature networks, pronounced vessel pruning is observed leading to treelike arteriolar and venular structures. Compared to the initial state, oxygen supply to more distal network regions is improved and overall pressure drop is reduced. However, adaptation rules tested so far did not yield satisfactory agreement between immature networks after simulated adaptation and experimental networks. Conclusion: The maturation process of terminal vascular beds may be guided by similar rules as the diameter adaptation in mature microvascular networks, but for network maturation these rules seem to be not sufficient. Supported by DFG (FOR341/1) and by NHI Grant (HL-34555). 1 Dept. of Physiology, Freie Universität Berlin, Arnimallee 22, D 14195 Berlin; 2 Dept. of Anaesthesiology, Deutsches Herzzentrum Berlin; 3 Dept. of Physiology, University of Arizona, Tucson, USA

P 13-5 DILATIONS DURING REACTIVE HYPEREMIA IN MICE ARE MEDIATED PARTIALLY VIA CX40-DEPENDENT GAP JUNCTIONAL COUPLING, BUT NOT NO OR PROSTAGLANDINS H. Boye, U. Pohl, C. de Wit Short term cessations of blood flow lead to arteriolar dilation and increases in blood flow (reactive hyperemia, RH). Different mechanisms are involved in this dilatory response, e.g. release of metabolites, flow-induced dilation, and myogenic responses. However, none of these mechanisms is the sole mediator. Additionally, signals generated in capillaries may travel upstream via gap junctional coupling and contribute to upstream dilation, as shown for active hyperemia. Gap junctions (GJ) in vascular cells are composed of different connexins (Cx) and Cx40 is especially important in endothelial cells. To study the role of Cx40-dependent signalling in reactive hyperemia we examined dilations upon short term cessations of blood flow in wildtype (wt) and Cx40-deficient (Cx40–/–) mice. To this end, a main feeding arteriole in the cremaster microcirculation was occluded with a micropipette for 10, 30 or 60 s and changes of arteriolar diameter were

S 258 measured in a downstream vessel by intravital microscopy. This was repeated after blockade of cyclooxygenase and NO-synthase. The maximal amplitude and the duration of the dilation increased with prolonged time of occlusion (10s: 48±5%, 18±3s; 30 s: 66±4%, 21±2s; 60s: 71±6%, 35±5s; n=16) in wt mice. Neither the amplitude (47±5%, 60±7%, 65±7%) nor the duration (16±2s, 42±15s, 44±16s) were attenuated after indomethacin (3µM) and L-NA (30 µM). In Cx40 –/– mice the amplitude of the reactive hyperemic response was similar to that in wt mice (38±8%, 58±5%, 61±8%; n=12), however, the duration of the responses was shorter (11±2s, 14±1s, 22±5s; p < 0.05 vs. wt). The data show that dilations in RH are either not dependent on NO or prostaglandin synthesis in this model or that other mechanisms compensate. In contrast, the duration of the dilation in RH was attenuated in Cx40-deficient mice. This is consistent with a role of GJ coupling via Cx40 in arteriolar dilations upon short term ischemia. Physiologisches Institut, LMU, Schillerstraße 44, D-80336 München

the entire pressure range; maximal improvement of flow was obtained at a pressure level of about 20 mm Hg. During compression the arterial inflow was significantly increased (8.8 ± 2.01 ml/min per 100 ml tissue) compared to baseline (3.7 ± 0.85 ml/min per 100 ml tissue) and to the control arm (3.1 ± 0.76 ml/min per 100 ml tissue) (p < 0.001). The arterial reserve was also significantly increased (28.3 ± 7.03 ml/ min per 100 ml tissue) compared to baseline (17.7 ± 4.68 ml/min per 100 ml tissue) and to the control arm (18.1 ± 3.82 ml/min per 100 ml tissue). During the time of intervention (total time of application of the compression stockings was 60 min) there was no change in the forearm temperature and no report of discomfort by the volunteers. The results clearly show that during short term application of compression stockings arterial perfusion of the human forearm is significantly increased in healthy volunteers. 1 Institut für Physiologie und 2Institut für Textil- und Bekleidungstechnik, Technische Universität Dresden, Fetscherstr. 74, 01307 Dresden, Germany

P 13-6 P 13-8 CHANGES OF INTRACELLULAR PH IN RAT TAIL SMALL ARTERIES DURING THE MYOGENIC RESPONSE U. Krien, R. Schubert The myogenic response of small arteries is one of the major mechanisms determining blood flow autoregulation. It is characterised by a contraction after pressure increases and a dilation after pressure decreases. During agonist-induced contractile responses changes of intracellular pH (pH i) have been observed. Whether such changes accompany contractile reactions induced by pressure changes is unknown. Thus, alterations of pH i ,during pressure-induced responses of rat tail small arteries have been investigated by measuring diameter changes of isobaric vessel preparations using videomicroscopy and measuring pHi using fluorimetry. Application of increasing pressure steps produced typical myogenic responses: a transient increase in diameter followed by an active constriction. These diameter reactions were accompanied by a decrease of pHi. A pressure step from 10 to 80 mmHg decreased pHi by 0,072±0,009 units. In order to identify the mechanisms responsible for the regulation of pHi in these vessels, standard pHi changes were induced with 10 mM NH4CI during various manoeuvres eliminating pHi-regulating mechanisms. It was observed that at 10 mmHg and at 80 mmHg the pHi-recovery after NH4CI-induced pHi-changes was considerably reduced in sodium-free extracellular solution and in the presence of 3 x 10-5M EIPA, an inhibitor of the Na+-H+ exchanger. For example, compared to control the pH i -recovery rate was only 25,38±11,84 % at 80 mmHg in sodium-free solution and 65,45±4,11 % in the presence of EIPA. Further, at 10 mmHg the recovery rate was 35,31±12,47 % in sodium-free extracellular solution and 42,45±18,33% during application of EIPA. Additionally, in sodium-free solution a pressure step from 10 to 80 mmHg increased pH i by 0,347±0,041 units, i.e. reversed the pressure-induced pHi-response seen in physiological extracellular solution. Thus, the myogenic response of rat tail small arteries is accompanied by changes of pHi and sodium-dependent pH i-regulating mechanisms seem to play an important role in this process. Institute of Physiology, University Rostock, PSF 100888, 18055 Rostock

RBC-MEMBRANE TANKTREADING IN VIVO AND BLOOD DISPLACEMENT AT BIFURCATIONS : DUALISTIC EFFECTS OF STABLE DEFORMATION AND ORIENTATION ON MICROVASCULAR BLOOD FLOW H. Schmid-Schönbein, H. Heidtmann, D. Ruder, M. Baumann BACKGROUND: In applying the Aachen in vivo rheoscope displacement, deformation, orientation and cell-cell and cell wall collision of RBC under controlled shear conditions can be observed at maximum temporal and spatial resolution. Novel behavioural traits of normal human (non-nucleated and highly fluidal), rigidified human and nucleated avian RBC (anser anser) can be quantified, allowing objective monitoring of adaptive behaviour of the most abundant cellular blood component as it prevails in living microvascular networks (hydrodynamically isolated rat mesentery). METHODS and RESULTS: Guided by three work stations, three stepper motors displace the microvessels on the microscope stage in X, Y and Z-direction, allowing to keep RBC in focus while they travel along complex trajectories from arterioles (15µm>diameter>20 urn) via capillaries to venules. Using HEINZ-bodies (globin precipitates) as produced by phenylhydrazin treatment, 1,5 mMol for 30 min), the tanktreading movement can best be observed directly (1000 x magnification, interference contrast, strobe illumination) by cytosolic movement patterns as the RBC cells collide with flow arteriolar flow dividers. In addition, the obligatory combination of orientation, deformation and axial drift can be objectively monitored off-line from high-resolution video-films: the highest values of orientational order parameters and the maximum axial drift velocity is found at the immediate entrance of arteriolar mother segments into capillary daughter segments (p<0.001). CONCLUSION: Contrary to currently held microrheological concepts, immediate and rapid drift from the wall (high shear, low velocity) to the axial stream (high velocity of RBC, low shear) is the main reason for the pronounced FAHRAEUS-effect and FAHRAEUSLINDVIST-Effect in vivo. These features are not found in avian nucleated RBC and in human RBC subsequent to membrane rigidification [p<0.001]. Membrane tanktreading thus represents the major cause of rapidly fluidizing heterophase effects at bifurcations which minimise cell-cell and cell wall-collision and become more pronounced as RBC approach the terminal microvascular bifurcations.

P 13-7 MECHANICAL INTERVENTION BY MEANS OF COMPRESSION STOCKING INCREASES FOREARM PERFUSION R. P. Bochmann1, W. Seibel1, E. Haase2, H. Rödel2, A. Deussen1 Application of compression stockings to the lower extremities is a widely used therapeutic intervention to improve venous outflow but there are only a few data about the effects of such intervention on arterial blood flow. To evaluate the relation between increasing pressure induced by compression stockings (consisting of 12% Elastan and 88% Cotton) and perfusion the arterial inflow and the arterial reserve (after 5 min occlusion) of the human forearm was measured in a group of healthy young males (n=10). Flow was measured with venous occlusion plethysmography at both forearms while the compression stockings were applied to the right forearm and the left forearm remained unaffected and served as control. Flow measurements were performed after a 10 min application of 6 stockings with increasing compression pressure ranging 13 – 32 mm Hg. Arterial blood flow was improved in

Department of Physiology, RWTH Aachen, D 52174 Aachen FRG [email protected] P 13-9 MAMMALS HAVE IMPROVED THERMODYNAMIC EFFICACY OF OXYGEN TRANSFER IN PULMONARY AND PERIPHERAL MICROVESSEL: RBC-MEMBRANE TANKTREADING AND HYDRO-THERMODYNAMIC EXPENDITURE FOR HEMOGLOBIN TRANSFER H. Schmid-Schönbein, H. Heidtmann, H. Horstkott BACKGROUND: Since FAHRAEUS’ secular work (1928, 1931) on “heterophase effects” associated with axial drift of RBC (described in its pivotal kinematic details by POISEUILLE (1842) it is known that the viscosity of mammalian blood in the paracapillary bed is not significantly higher than that of cell free plasma. WHITTACKER and

S 259 WINTON (1934) corroborated these in vitro findings by measurement of relative blood viscosity in the hindlimb of dogs. The microrheological cause has now been in vivo rheoscopy has now been shown to follow from minimised cell-cell and cell wall collision due to RBC-membrane tanktreading (see accompanying abstract). Expressed in hydrothermodynamic terms, this behaviour results in the minimisation of the energetic expenditure for oxygen transfer, a selforganised physiological transport response greatly enhancing the well known benefits of enhancing transport capacity by reversible O2-binding to Fe2+ in haemoglobin. METHODS and RESULTS: Using the isolated rat intestine as a “in vivo rheometer”, the pressure-flow rate relationship for controlled suspensions (10, 20, 30% feed hematocrit) under variable pressure heads (10-100 cm H2O) of normal human RBC, human RBC with fluid cytosole but mildly stiffened membranes, heat treated spherical RBC and avian nucleated RBC were compared. The maximum oxygen transfer rates (Q’ oxygen. Mol/s) were calculated from flow rate, haemoglobin concentration and HÜFNER’s number, the entropy production rate (S’, Nm/s) follows from the product of pressure gradient and flow rate; thence hydrothermodynamic efficiency κ (Q’ oxygen / S’, Mol/Joule) can be calculated. There are small differences in the κ-values of avian and rigidified human RBC at 20 cm driving pressure, showing that the murine microvasculature offers similar resistances to the four types of cells. However, at physiological pressure gradients, the ê-value dramatically increases by more than 250%, p<0.001) exclusively in the experiments with normally fluidal RBC. CONCLUSION: Non-nucleated RBC and thereby rheologically beneficial haemoglobin transfer cells endow their marsupial and mammalian species with a dramatic decrement in the energetic cost of oxygen transfer through restricted microvessel: this is a phenotypic benefit of highest significance, which is not reducable to genomic “mutations” since the molar binding capacity of the mammalian and avian haemoglobin is about identical. This is an evolutionary benefit exclusively based on “synergetic transfer consensualisation” by way of membrane tanktreading: in dramatically reducing cell-cell and cell wall collision in the microvessels, the most essential of all physiological transports is optimised.

P 13-11 RESPONSES OF CORONARY TERMINAL ARTERIOLES TO CHANGES IN PERFUSION PRESSURE; RESULTS OF A NEW VIDEO MICROSCOPIC IN SITU APPROACH S. Goebel1, W.M. Kübler1,2, A.R. Pries1,2, H. Kuppe2, H. Habazettl1,2 Introduction: To date, investigation of coronary arterioles by intravital microscopy or after isolation of arterioles has been restricted to vessels > 30-40 µm. However, terminal arterioles, i.e. < 25 µm, may considerably contribute to microvascular blood flow distribution. We, therefore, established a new technique to study regulatory mechanisms of coronary terminal arterioles by in situ video microscopy. Methods: Isolated, Langendorff-perfused rat hearts were arrested by tetrodotoxin (TTX). The closed-loop perfusion system (filling volume 20ml) was placed on a computer-controlled microscope stage. Tilting of the microscope by 90° allowed access to the ventricular surface. The 20x objective was optically coupled to the heart by contact gel and a cover glass. A CCD-camera (Kappa C8/1FMC) generated the images, and recorded diameters were measured offline. With the onset of recirculation, after a 30 min. stabilization period, FITCdextran was added. TTX (50 µM) was used to arrest the heart and maintained oxygen demand at basal levels. Arterioles were identified by the flow direction of fluorescent beads (1 µm). Responses of terminal arterioles (< 25 µm) to changes in perfusion pressure (60, 40, 100, 120, 140 mmHg) after baseline recordings at perfusion pressure (PP) of 80 mmHg were investigated. Results: Smallest arterioles (< 10 µm) dilate more prominently at low PP, while larger arterioles show a more sustained contraction after an increase in PP. Conclusions: The newly developed technique allows investigations on regulatory mechanisms of coronary terminal arterioles. The data indicate, that consistant with transmural pressure distribution the pressure range for effective myogenic activity is lower in the smallest arterioles (< 10 µm) than in upstream larger vessels. Supported by DFG (Ha 1651/8-1) and by Medtronic GmbH.

40

diam ete r ( ∆% )

Department of Physiology, RWTH Aachen, D 52174 Aachen FRG [email protected] P 13-10 STANDING WAVES IN TRIDOMAIN VIRTUAL VENTRICULAR TISSUE O.V. Aslanidi, R.A. Gray, A.V. Holden, O.A. Mornev Standing excitation waves have been produced by periodic forcing of isolated rabbit hearts in vitro by voltage pulses applied in the bathing solution (Gray et al., 2001, Phys. Rev. Lett. 87, 168104). Such waves are steady periodic patterns of depolarization seen at the heart surface, eliminating propagation activity. The phenomenon cannot be reproduced by simple mono- or bidomain models of cardiac tissue. Extention of the bidomain model (Roth, 1991, Circ. Res. 68, 162173) for the intra- and extracellular spaces of the tissue to account for existance of the bathing solution - tridomain model - allows simulation of the standing waves in one dimension. Objectives: Our aim is to utilize the tridomain model and simulate essentially two-dimensional phenomena, such as transition from re-entry and fibrillation to standing waves upon increase of the forcing strength, as seen in experiments. Methods: We use the Luo-Rudy family of virtual ventricular tissues (Aslanidi et al., 2001, J. Physiol., in press), modified to account for the tridomain structure of the tissue. Spiral waves are first initiated in 3x3 cm 2 square uniform isotropic tridomain virtual tissue by the phase distribution method, then after 200 ms of simulation external forcing is applied to the bathing solution locally along two opposite boundaries of the tissue. Results: For low forcing strength (<12 V) spiral wave rotation is stable both before and after the stimuli application. However, increasing the forcing strength leads to formation of standing waves in the tissue which eliminate spiral waves. Simulations show that mechanism of such „defibrillation“ is based on formation of a distributed electric current source in the bathing solution forced by external voltage. Conclusions: 1. Simulations with the tridomain virtual ventricular tissue reproduce standing waves observed in experiments. 2. There is an external forcing strength threshold between re-entry and standing waves in the heart. Acknowledgements: This research was funded by the MRC and EPSRC. School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK

30 20 10 0

-1 0 -2 0 40

60

80

100

< 10 µ m ± S E M > 10 µ m ± S E M

12 0

1 40 PP

1 Dept. of Physiology, Freie Universität Berlin, Arnimallee 22, 14195 Berlin. 2 Dept. of Anaesthesiology, Deutsches Herzzentrum Berlin

P 13-12 PINACIDIL AND ADENOSINE-INDUCED DILATION OF HUMAN AND RAT PRESSURIZED CORONARY RESISTANCE ARTERIES; EFFECTS OF GLIBENCLAMIDE F.M. Lynch, A.S. Izzard, C. Austin, A.M. Heagerty Autoregulation and metabolic dilation are major determinants of tone in the coronary circulation. Pressure-dependent myogenic tone may be important for autoregulation and KATP channels may be involved in metabolic dilation. Human coronary resistance arteries were dissected from atrial appendages obtained from consenting patients undergoing cardiac surgery. Wistar rats were killed by cervical dislocation and septal coronary arteries were dissected. Arteries were pressurised to 60mmHg and checked for leaks. The inner diameter and wall thickness was continually monitored using a video dimension analyzer. Once myogenic tone stabilized the myogenic reactivity of vessels was determined by reducing pressure to 20mmHg and then increasing it in 20mmHg increments to 100mmHg. Vessels were returned to 60mmHg before addition of pinacidil, a K ATP channel opener (5*10 -6M) or adenosine, a metabolic dilator (10 -6 – 10-4M). Glibenclamide (10 -6, 5*10 -6 M) was subsequently added. Mean lumen diameters (µm) (± SEM) are shown in the table below.

S 260 Group Human Pinacidil (n=5) Adenosine (10-5M) (n=4) Rat pinacidil (n=7) Adenosine (10-4M) (n=8)

Pre-drug (µm) 90.3 (15.6)

With pinacidil/ adenosine (µm) 100.6 (16.6)*

With Glibenclamide (µm) 89.9 (15.7)+

85.8 (2.9)

114.8 (6.6)**

114.4 (7.5)

141.2 (10.7) 164.2 (14.5)

166.1 (14.1)** 193.8 (13.3)**

142.3 (10.5)+ 192.7 (14.3)

*, ** = significantly different from pre-intervention value; P < 0.05, 0.01 respectively, unpaired t test. + = significantly different from intervention; P < 0.05, unpaired t test. This study implies that KATP channels are not necessary for adenosine-mediated relaxation of coronary resistance arteries in the rat or human. Funded by The British Heart Foundation. The authors would like to acknowledge the staff of the Cardiology Dept. MRI for donation of the atrial appendages.

(TNF) has a permissive role in VEGF-induced vascular permeability, we hypothesised that inhibition of TNF action is capable of preventing hypoxia- and ischemia-induced brain oedema formation. Vascular permeability was measured in mice exposed to various degrees of hypoxia or to normoxia using a well established hypoxic-chamber model. Subsequently, a neutralising anti mouse TNF antibody was injected intraperitoneally prior to the hypoxic or normoxic exposure and changes in oedema formation were analysed. Furthermore, the protective effect of a low molecular weight inhibitor (SB 203580) that interferes with the SAPK-2/p38 MAPK signalling pathway mediating the permissive effect of TNF on VEGF-induced permeability was studied. Our results demonstrate that inhibition of TNF activity completely blocks the hypoxia-induced increase in vascular permeability and identify the p38 MAPK as an essential signalling molecule of oedema formation. In conclusion, our data show that TNF has an important permissive role in VEGF-induced vascular permeability and provide the basis for new therapeutic concepts for treatment of oedema formation. 1

Dept. of Medicine, University of Manchester, Manchester Royal Infirmary, Oxford Rd, M13 9WL, UK

Institute of Physiology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland, and 2Department of Molecular Cell Biology, Max-Planck-Institute for Physiological and Clinical Research, Parkstrasse 1, D-61231 Bad Nauheim, Germany

P 13-13 P 14-2 AUTOMATIC SEGMENTATION AND FRACTAL ANALYSIS OF THE MICROCIRCULATION OF THE LIVING CAM C. Arlt, M. Baumann, H. Schmid-Schönbein Objective of study: The chorioallantois membrane (CAM) of the avian chick is an in vivo assay to characterise the angiogenesis of tumours and angiognetic factors. We use this assay to characterise biomaterials. We developed an image analysis to segment the microvascular network and evaluate it objectively. The fractal anlaysis is a new method used to quantify the biomaterial induced angiogenesis. Methods used: Fertilised eggs are used for the CAM assay. They are opened at day 6 and a biomaterial is applied to the membrane. Six to nine days later we examine the capillary network of the membrane. The in vivo rheoscopy is used to acquire images and film sequences. For analysis of fractal geometry the box-, information- and correlation-dimension were determined. Results: The films of the in vivo rheoscopy show the flow of the red blood cells (RBC). Due to the movement of the embryo the films must be stabilised. After this the movements of the RBC are used to identify the capillary network. The result is a binary image of the microvascular network.The binary image shows the capillaries where RBCs flow through. We compare three different fractal dimensions of these binary images. We calculate the mean and the standard deviation of these dimensions. It is shown that the correlation-dimension has the lowest standard deviation due to their definition. Conclusions: The developed image processing is a new method to examine the angiogenesis of the CAM. This noninvasive method extracts the change of the CAM vessels. Especially the segmentation of the capillaries offer a sensitive method to determine changes of the microvascular network. The determination of the correlation-dimension is a useful measurement to quantify the differences in the angiogenetic behaviour. Christian Arlt, Institute of Physiology, RWTH Aachen, Pauwelsstraße 30, 52074 Aachen, Germany P 14-1 PREVENTION OF HIGH ALTITUDE CEREBRAL OEDEMA IN MICE BY INHIBITION OF TUMOUR NECROSIS FACTOR SIGNALLING H.H. Marti1, G. Springmann2, H.J. Schoch2, M. Clauss2 Oedema formation is a major life-threatening complication of various diseases of the central nervous system (CNS) such as head injury, tumours, stroke and high altitude illness. We have recently shown that vascular endothelial growth factor (VEGF) is responsible for the hypoxia-induced vascular leakage and oedema formation in the brain that occurs during high altitude exposure. However, VEGF is also an important angiogenic and neuroprotective factor. Thus, blocking VEGF action to prevent brain oedema formation might lead to unwanted side effects in the CNS. Our aim was to identify strategies for inhibition of VEGF-induced oedema formation without affecting the angiogenic and neuroprotective properties of VEGF. Since tumour necrosis factor α

VASOMOTION IN INTRAPARENCHYMAL ARTERIOLES IN BRAIN SLICES L.A. Brown, B.J. Key*, T.A. Lovick Increases in neuronal activity are associated with a local increase in cerebral blood flow. The mechanisms which underly this process remain intact in brain slices (Lovick TA et al 1999, Neuroscience 92, 47-60). Thus slices afford a convenient preparation to study the reactivity of individual intraparenchymal arterioles in situ. Recently we observed that intraparenchymal arterioles displayed rhythmic contractions (vasomotion) that were inhibited during periods of increased synaptic activity (Brown LA et al 2001, J Physiol 536P, 151P). In that study a thromboxane A2 agonist was added to the bathing fluid in order to increase vascular tone. Under conditions in vivo, intraluminal flow and pressure would be expected to be the major determinants of resting vascular tone. We have therefore perfused individual arterioles in situ in brain slices to ascertain whether vasomotion occurs in the absence of exogenous constrictor agents. Coronal slices of parietal cortex 400µm thick, were prepared from 150-200g male Wistar rats anaesthetised with urethane (1.5g 100g-1 body weight i.p.). Slices were maintained at 33oC in a chamber perfused with artficial cerebrospinal fluid (ACSF) and images of vessels captured as described previously (Lovick et al, 1999). Penetrating cortical arterioles 15-20µm inside diameter, inside/external diameter =0.7 were located which lay 3070µm below the surface of the tissue. The vessel wall was punctured with a glass micropipette 2.5-3µm outside diameter filled with ACSF containing 1% bovine serum albumin and perfused in the anterograde direction at 0.2-2.0µl min-1. Prior to cannulation vessels were quiescent. However, once perfusion started, rhythmic contractions (8-20min-1) developed in the smooth muscle wall of each vessel (n=5). The results suggest that vasomotion is an intrinsic property of intraparenchymal arterioles. Inhibition of vasomotion could contribute to local hyperaemia during increased neuronal activity. This work was supported by the British Heart Foundation Departments of Physiology and Pharmacology*, University of Birmingham, Birmingham B15 2TT, U.K. P 14-3 REGULATION OF CAPILLARY DIAMETER IN ACUTELY ISOLATED CEREBELLAR SLICES C. Auger, D. Attwell Functional brain imaging measures changes in blood flow. Since blood flow regulation in the brain is assumed to be tightly coupled to changes in neuronal activity, fMRI signals are interpreted as reflecting the local level of neuronal activity. The mechanisms linking neuronal activity to the microvasculature, however, have not been worked out in detail at the cellular level. In general, modulation of blood flow is thought to take place at the level of pre-capillary resistance arterioles, and the possible regulation of local blood flow at the level of the

S 261 capillaries is still controversial. We investigated the effect of norepinephrine (NE) and glutamate on unperfused capillaries in acutely isolated cerebellar slices. Capillary morphology and diameter were imaged using DIC-IR optics. While part of the capillary network appeared collapsed or damaged in the slices, some capillaries remained open and clearly visible. For the latter, healthy looking capillaries, local application of NE (100µM) decreased the diameter, occasionally leading to complete closure. Glutamate (100µM), on the other hand, when applied during contraction with NE, relaxed capillaries towards their uncontracted diameter, often completely. These results indicate that local release of neurotransmitters could play a significant role in the regulation of blood flow at the level of the capillary network.

SEM). This suggests greater pooling of blood during standing. MABP at heart level was unchanged, but at brain level it fell by 20±2 on sitting and by 18±2 when standing. MCAFV fell by 2.4±1.5 cm s-1 when the subjects sat up and by 5.3±1.2 when they stood (~4 and ~8% respectively). petCO2 fell by 1.5±0.3 mmHg and 3.6±0.2 on sitting and standing. Conductance of the cerebral vasculature increased by 36.3±5.4% on sitting up but by only 19.5±3.8% when standing. These results demonstrate that marked autoregulation of MCAFV, and by inference cerebral blood flow, results from postural change. The larger reduction of MCAFV when standing could represent reflex cerebral vasoconstriction consequent on the greater orthostatic stress. Equally it could be due to the larger reduction in arterial pCO2

Physiology Department, University College London, Gower Street, London WC1E6BT, UK

Department of Human Physiology and Aerospace Medicine, GKT School of Biomedical Sciences, King’s College London (Guy’s Campus), LONDON SE1 1UL, United Kingdom

P 14-4 P 14-6 CORRELATION BETWEEN DENSITIES OF MONOCARBOXYLATE AND GLUCOSE TRANSPORTERS IN RAT BRAIN M.H. Maurer, M. Canis, W. Kuschinsky, R. Duelli Under normal conditions, glucose is the major source of energy in the brain. As alternative substrates, monocarboxylates (lactate, pyruvate and beta-hydroxybutyrate) can also be used under certain circumstances like starvation or high fat diet. In mammalian cells, the transport of monocarboxylates across plasma membranes is facilitated by a family of monocarboxylate transporters (MCTs). Previous studies have shown the expression of two isoforms in the rat brain, MCT1 and MCT2. MCT1 is mainly expressed in brain endothelial cells and astrocytes whereas MCT2 is located in the endfeet of perivascular astrocytes. Little is known about the local MCT1 and MCT2 distribution in rat brain. Therefore, autoradiographic methods were developed for the determination of local MCT1 and MCT2 densities. 10 µm coronal cryosections were incubated with MCT1 or MCT2 antibody. Local densities of MCT1 and MCT2 were detected with [35S]-coupled IgG. The mean densities of MCT1 and MCT2 were determined and set as 100% for each brain section. The local densities of MCT1 and MCT2 were compared to the local Glut1 density. MCT1 as well as MCT2 densities showed an uneven distribution throughout the brain. 16 brain structures were investigated. MCT1 densities ranged from 59% - 133% of mean values and MCT2 densities from 60% - 140%. Low values were found in the white matter structures, the highest values in the auditory cortex (MCT1) and the dentate gyrus (MCT2). A close correlation could be shown between MCT1 density and Glut1 density (r=0.82), as well as between MCT2 density and Glut1 density (r=0.78). The uneven distribution of MCT1 and MCT2 shows a heterogeneous capacity of the brain vessels to transport monocarboxylates. Brain structures with a high glucose transport and glucose utilization can take up more monocarboxylates than structures with a low glucose transport and glucose utilization. Supported by the Deutsche Forschungsgemeinschaft (WK). Dept. of Physiology and Pathophysiology, University of Heidelberg, Im Neuenheimer Feld 326, 69120 Heidelberg, Germany

REGULATION OF TONE OF PORCINE MEDIAL CEREBRAL ARTERY IS STRONGLY DEPENDENT ON ENDOTHELIAL NOPRODUCTION G. Judio, W.Reimann, H.Heinle The function of cerebral arterial vessels is still less described than that of cardial or peripheral arteries. Therefore we investigated by in vitro contrac- tion studies the effects of depolarisation, serotonine (5HT), or phenyleph- rine (Phe) on excised segments of porcine medial cerebral artery and deter- mined the impact of inhibition of endothelial NO synthesis as well as of inhibition by the calcium antagonist verapamil, the serotonin antagonist naftidrofuryl and the alpha antagonist prazosin. Endothelial function was tested by application of acetylcholin (Ach) and measuring relaxation of contraction force, nitro-arginine was used to inhibit endothelial function. Depolarisation was achieved by adding a KCl enriched solution (30 mM), all other agonists and antagonists were applied in concentrations between 0.1 and 10 µM (nitroarginine up to 100 µM). The results show that the agonists induced phasic (5HT>Phe) or tonic (KCl) contractions. Direct addition of Ach under the stimulation conditions had only marginal relaxing effects. However, addition of nitro-arginine increased all evoked contraction forces 2- to 4-fold. Application of the different antagonists showed that inhibition of calcium channels had blocking effects on all three stimulations (KCl>Phe>5HT). Prazosin had a strong inhibitory effect on Phe, a weak on KCl, but a stimulatory on 5HT. Naftidrofuryl was without significant effect, even when applied with 5HT. Whereas the findings from the studies with the different antagonists suggest that various co-activations of receptors and ion channels might be possible in smooth muscle and endothelium of this cerebral artery, the most striking result shows the involvement of endothelial NO production which is obvi- ously strongly activated by all agonists used here. The lack of Ach response under normal stimulation seems to be the result of maximal activation of endothelial NO production already achieved by depolarisation, 5HT or Phe. We could not find similar properties in many other arteries obtained from pig heart, rabbit or rat. Institute of Physiology, Gmelinstr. 5, 72076 Tübingen, Germany

P 14-5 P 14-7 CEREBROVASCULAR RESPONSES TO SITTING AND STANDING F.J. Imms, R. Mehta When a person sits up or stands from a lying position, blood pools below the level of the heart. Although venous return and cardiac output are reduced, mean arterial blood pressure [MABP] at the level of the heart is unchanged in young, healthy subjects. However MABP at the level of the brain will be reduced by ~20 mmHg because the cerebral vessels are vertically above the heart. Previous studies have shown that during both standing and passive head-up tilt there are small reductions of cerebral blood flow [CBF] which are less than might be predicted from the reduction of cerebral MABP. The small fall in alveolar pCO2 which occurs on tilting or standing could contribute to the fall in CBF. We investigated in ten healthy young males the effects of sitting from a lying position and then of standing on heart rate [HR], MABP, end tidal pCO2 [petCO2] and flow velocity in the right middle cerebral artery [MCAFV]. The results are given for the steady state responses from one to five minutes after postural change. HR increased by 7±2 beats min -1 on sitting and increased further to 23±3 on standing (mean ±

MECHANISMS OF ACTION OF NITRIC OXIDE IN THE BRAIN STEM: ROLE OF LOCAL MEDULLARY BLOOD FLOW S. Passingham, M. Dewhurst, S. Westbrook, J. Zanzinger The exact mechanisms by which nitric oxide (NO) mediates its neuromodulatory effects within the central control of cardiovascular functions are still unclear. Both excitatory and inhibitory actions of NO in different regions of the brainstem have been reported that could be caused by direct actions of NO on neurones and/or by NO-mediated changes in local cerebral blood flow. Microinjection studies suggest that NO directly modulates neuronal activity via cGMP-dependent pathways. To test whether, in addition, changes in local medullary blood flow by NO-dependent mechanisms indirectly alter sympathetic activity, we studied local and systemic potentiation of NO-dependent vasodilation with UK-343,664, a specific inhibitor of phosphodiesterase V (PDE5i) (>1000-fold selectivity over PDE1, which is the primary PDEi in neurones) in chloralose anaesthetised dogs (n=6). As a measure of sympathetic excitability, somato-sympathetic reflexes were evoked by electrical stimulation of the left sciatic nerve (8-15V, 1ms, 20Hz,

S 262 for 30 sec). Stimulation at 2x-threshold caused pronounced increases in systemic blood pressure (BP) and heart rate (HR). Local intracerebroventricular administration of UK-343,664 (3-30 mcg in 0.5 ml) to the ventral surface of the lower brain stem did not change baseline BP or HR nor the responses to sciatic nerve stimulation. In contrast, previous studies have shown that inhibition of NO-synthases in the brain stem potentiate, whereas NO-donors significantly attenuate somato-sympathetic reflexes, respectively. The data demonstrate that inhibition of PDE5 has no effect on the somato-sympathetic reflex. Given the demonstrable potentiation of NO mediated vasodilation by PDE5 inhibitors in vivo, we conclude that the sympathoinhibitory effects of NO on sympathetic outflow from the brainstem may be independent of changes in local blood flow that could be caused by NOrelease within the brain. Pfizer Global Research & Development, Ramsgate Road, Sandwich, Kent, CT13 9NJ, United Kingdom P 14-8 LOSS OF CLAUDIN-3 BUT NOT OF CLAUDIN-5 FROM CEREBRAL ENDOTHELIAL TIGHT JUNCTIONS DIRECTLY CORRELATES WITH BLOOD-BRAIN BARRIER LEAKYNESS IN VIVO AND IN VITRO 1 H. Wolburg, S. Hamm, K. Wolburg-Buchholz, B. Dehouck, M.-PDehouck, S. Liebner, M. Furuse, R. Cecchelli, W. Risau, B. Engelhardt In the central nervous system (CNS) endothelial cells are connected by complex tight junctions (TJs) forming the blood-brain barrier (BBB). During inflammation in the CNS, vascular permeability increases forming brain edema. The molecular changes at BBB TJs involved in the development of BBB leakyness have not been defined. Here we report the specific loss of the TJ-protein claudin-3 from cerebral vessels during CNS inflammation in experimental autoimmune encephalomyelitis (EAE), whereas the localization of other TJ proteins such as claudin-5 or occludin and ZO-1 remained unchanged. A direct correlation of the reduction of claudin-3 from BBB TJs with an increase in paracellular permeability is supported by our in vitro findings. Brain endothelial cells in co-culture with astrocytes formed a tight permeability barrier for 3H-inulin and 14C-sucrose, removal of astrocytes resulted in breakdown of this barrier as determined by increased diffusion of both tracers across the endothelial monolayer and opening of the TJs to HRP. Opening of the endothelial TJs was accompanied by the specific loss of junctional claudin-3 immunostaining, whereas junctional localization of claudin-5, occludin, ZO-1, ZO-2, ß-catenin and p120cas was conserved. Our data demonstrate a direct correlation of loss of claudin-3 from endothelial TJs with their opening and suggest that claudin-3 but not the endothelial cell specific claudin-5 is involved in regulating the tightness of BBB TJs and thus of BBB leakyness in vivo. 1

Inst. Pathol. Univ. Tübingen, Liebermeisterstr. 8, D-72076 Tübingen

P 14-9 ELEVATED INTRAVASCULAR VOLUME DOES NOT CONTRIBUTE TO THE DEVELOPMENT OF RENAL POST-TRANSPLANTATION HYPERTENSION O. Grisk, M. Heukäufer, R. Rettig Previous studies have shown that the development of renal posttransplantation hypertension is associated with elevated renal sodium retention but not with activation of major neurohumoral pressor systems. Therefore we investigated the potential role of intravascular volume expansion for the development of renal post-transplantation hypertension. F1-hybrids bred from Wistar-Kyoto and spontaneously hypertensive rats (SHR) were unilaterally nephrectomized and transplanted either with a kidney from SHR (group 1, n = 11) or with a kidney from syngeneic donors (group 2, n = 12). Animals were instrumented with telemetric arterial pressure recording devices as well as arterial and venous catheters. Six days after transplantation, the remaining native kidney was removed. Blood volume was estimated from measurements of plasma volume (dye dilution technique) and hematocrit. Prior to completion of bilateral nephrectomy, mean arterial pressure (MAP) was similar in both groups. Two days after bilateral nephrectomy, MAP was 125 ± 3 mmHg in group 1 vs. 103 ± 3 mmHg in group 2 (p < 0.01). At this time blood volume was 7.01 ± 0.32 ml/ 100g bw in group 1 and 6.48 ± 0.20 ml/100g bw in group 2, respectively (n.s.). Central venous pressure was 0.9 ± 0.4 cmH2O in group 1

vs. 0.8 ± 0.3 cmH2O in group 2 (n.s.). Six days after bilateral nephrectomy MAP was 137 ± 5 mmHg in group 1 vs. 102 ± 2 mmHg in group 2 (p < 0.01). Blood volumes were 6.56 ± 0.36 ml/100g bw in group 1 vs. 7.28 ± 0.32 ml/100g bw in group 2, respectively (n.s.). Hematocrit did not differ between groups. In conclusion elevated intravascular volume does not contribute to the development of renal post-transplantation hypertension. Dept. of Physiology, Greifswalder Str. 11c, 17495 Karlsburg, Germany P 14-10 FREQUENCY DOMAIN OF PLASMA RENIN ACTIVITY IN THE FREELY MOVING RAT B. Nafz*, M.F. Beier, F. Schweda, P.B. Persson In the recent past, it is becoming generally appreciated that fluctuations in blood pressure can have major pathophysiological importance in the development of hypertension. Nonetheless, little is known regarding the influence of blood pressure oscillations (BPO) on the renal renin - angiotensin -system, a crucial element of blood pressure regulation. Basically, stepwise reductions in renal perfusion pressure (RPP) induce significant increases in renal renin release and plasma renin activity (PRA) at RPP < 95mmHg. The phenomenon has been well characterised for fixed levels of RPP, but the influence of physiological variations in RPP remain unknown. To clarify this question RPP was reduced to 80mmHg in freely moving rats (N=10, CON). In addition, BPO (±10mmHg) with one of the following frequencies were superimposed on this reduced RPP: 2.5, 5, 10, 20, 40, 80, 160 and 320mHz. Frequency and length of each experimental phase (about 27min) were chosen in such a way that the mean RPP was identical among all protocols. To investigate to which extend renal nerve signals are involved in buffering the influence of the induced RPP oscillations on PRA, the same protocols were repeated in renal denervated animals (N=7, DEN). Under resting conditions PRA was similar in both groups (CON: 4.6±0.9; DEN: 5.0±1.2ng AI/ml/h). Superimposing BPO induced a frequency dependent decrease in PRA (CON, 2.5mHz: 10.4±1.4; 320mHz: 5.7±1.0ng AI/ml/h). This decrease showed a different pattern with a “peak” around 40mHz and was more pronounced in renal denervated animals (2.5mHz: 14.6±2.1; 320mHz: 7.2±1.3ng AI/ml/h). In conclusion our data provide evidence, that BPO induce frequency dependent changes in PRA. This influence of immediate changes in blood pressure on PRA can be buffered by the activity of renal nerves in the freely moving rat. Medizinische Fakultät der Humboldt Universität (Charité), Johannes Müller - Institut für Physiologie, Tucholskystraße 2, 10117 Berlin, email: [email protected] P 14-11 EFFECT OF THE ACE GENOTYPE ON ANGIOTENSIN II AND RENIN LEVELS IN PLASMA AND KIDNEY TISSUE B. Klanke, M. Porst, R. Veelken, K.F. Hilgers Background: It is unknown whether genetic variants of the angiotensin (Ang) converting enzyme (ACE) gene affect plasma and tissue levels of Ang II formation. We investigated plasma and kidney tissue Ang II content in mice with different numbers of Ace gene copies. Methods: We studied mice with 1 or 4 copies of the ACE gene (Ace) in strains originally described by Krege, Smithies and coworkers. Blood for measurement of plasma renin activity (PRA) and Ang II was drawn from carotid artery catheters in conscious mice. Immediately after the mice were sacrificed, peptides were extracted from kidney tissue and prepurified by CIS columns. The extracts were then subjected to high performance liquid chromatography (HPLC), and the Ang II content in HPLC fractions was determined by radioimmunoassay. PRA was measured by Ang I radioimmunoassay after incubation. Kidney renin content was estimated by counting the percentage of glomeruli with a juxtaglomerular apparatus staining positive for renin. Results: Plasma renin was not significantly affected by the number of Ace gene copies. Plasma Ang II tended to be higher in Ace-4 mice (17.5±7.4 fmol/ml, n=7) than in Ace-1 mice (5.8±1.5 fmol/ml, n=9, p=0.07 by 2-tailed t-test). The recovery of Ang II from extraction and HPLC was 80-90%. Kidney tissue content of Ang II was higher in Ace-4 mice (424±58 fmol/g, n=8) than in Ace-1 mice (277±34 fmol/g, n=6, p=0.05). In contrast, kidney renin staining was decreased in Ace-4 mice (24±3 % of glomeruli, n=7) compared with Ace-1 mice (34±3 %,

S 263 n=6, p<0.03). Conclusion: A higher number of Ace gene copies causes elevated kidney tissue Ang II levels, despite a downregulation of kidney renin content. The Ace genotype has a much more pronounced effect on tissue Ang II levels than on the plasma renin-Ang system. Dr. B Klanke, IV Medizinische Klinik, Universität Erlangen-Nürnberg, Universitätsstraße 17, 91054 Erlangen P 14-12 GENETIC AND PHARMACOLOGICAL DISSECTION OF PATHWAYS INVOLVED IN THE ANGIOTENSIN II-MEDIATED DEPRESSION OF BAROREFLEX FUNCTION IN THE NUCLEUS TRACTUS SOLITARII J.F.R. Paton, L.-F. Wong1, J. Polson1, D. Murphy2, S. Kasparov Arterial hypertension is associated with increases in angiotensin II (ANG II) activity (Phillips et al. 1997, Nature 270, 445-). One area in which this may occur is the nucleus of the solitary tract (NTS). Located in the dorso-medial medulla, the NTS is the termination site of baroreceptor afferents and is essential for mediating the baroreflex. Recently we showed that the baroreflex depressant action of ANG II in the NTS was mediated by activation of endothelial nitric oxide synthase (eNOS) and enhanced GABA release (Paton et al. 2001, J.Physiol., 531.2, 445-). Here, we have determined the intracellular pathway(s) in the NTS by which ANG II activates eNOS. Where dominant negative proteins were employed, the NTS was transfected and five to seven days allowed for expression. ANGII was microinjected (500 fmol) bilaterally into the NTS and baroreflex gain assessed. We used both conventional pharmacological agents and a range of adenoviruses to induce expression of dominant negative proteins to block intracellular pathways selectively. The control baroreflex was 1.95+0.2 bpm/mmHg. This was reduced by ANGII delivered into the NTS to 0.8+0.1 bpm/ mmHg (n=6; P<0.01). In contrast, the following all prevented the depressant effects of ANGII in NTS on the baroreflex gain: (i) NTS expression of a dominant negative form of the α-subunit of the Gq protein (n=7); (ii) blockade of phopholipase C (PLC) with U73122 (15 µM, systemically administered; n=7); (iii) blockade of 1,4,5-inositol triphosphate receptors (IP3) with xestospongin C (15 pmol; n=6); (iv) antagonism of calmodulin (W-7; 5-25 pmol; n=6). The latter two agents were microinjected into the NTS prior to ANGII. In contrast, NTS microinjection of ANGII after multiple site disruption of the protein kinase B pathway, which can lead to eNOS activation, always attenuated the gain (P<0.01). Our data pertain that AII acting in the NTS depresses the baroreflex via a Gq protein-mediated activation of PLC and that via IP3 receptors causes release of calcium from intracellular stores and calcium-calmodulin formation. Dept. Physiology, University of Bristol, University Walk, Bristol BS8 1TD, UK; 1University Research Centre for Neuroendocrinology, University of Bristol, Bristol Royal Infirmary, Marlborough Street, Bristol BS2 8HW, UK P 15-1 CALCIUM DEPENDENCE OF SYNAPTIC VESICLE FUSION ESTIMATED FROM THE TIME-COURSE OF CAPACITANCE CHANGES AT THE CALYX OF HELD M. Wölfel, R. Schneggenburger Based on measurements of excitatory postsynaptic currents (EPSCs) and Ca2+ uncaging, the intracellular Ca2+ ([Ca]i) dependence of vesicle fusion at the calyx of Held synapse has recently been reported (Bollmann et al. 2000, Science; R.S. & Neher 2000, Nature). Here, we have used Ca 2+ uncaging in combination with membrane capacitance (C m) to measure the Ca 2+ dependence of the rate of depletion of a readily releasable pool (RRP) of vesicles in presynaptic terminals. Whole-cell patch-clamp recordings were made from calyces of Held in brainstem slices from P8 – P10 day rats, with pipette solutions containing Ca2+loaded DM-nitrophen (DMN; 1.5 mM) and fura-2FF (0.1 mM), a low affinity Ca2+ indicator. Cm was measured with the sine+dc technique (Pulse, HEKA-elektronik), using a 70 mV peak-to-peak sine wave at 2 kHz. Light flashes with varying intensities elevated [Ca]i to values of 2 – 40 µM. Below 5 µM [Ca]i, only small increases in Cm were observed (∆Cm 28 ± 3fF; n = 6). In a range of 5 – 40 µM [Ca]i, ∆Cm was 348 ± 200 fF (n = 54) and did not significantly depend on [Ca]i (slope of 5 fF/µM), indicating that the RRP was largely depleted at [Ca]i > 5 µM. This was confirmed in experiments in which strong presynaptic depolarizations

(0 mV for 32 ms), which likely deplete the RRP, were paired with Ca2+ uncaging stimulation. The time-course of pool depletion after Ca2+ uncaging was analyzed by fitting the rise of C m with exponential functions. In about one-half of the flash responses (n = 23/45), monoexponential fits gave satisfactory results. In a range of 10-15 µM [Ca]i, the average time course of pool depletion was found to be 2.2 ± 1.1 ms, confirming the relatively high Ca2+ sensitivity of vesicle fusion determined previously at the calyx of Held synapse. Max-Planck-Institut für biophysikalische Chemie, Abt. Membranbiophysik 37077 Göttingen, Germany P 15-2 POSTFUSIONAL REGULATION OF QUANTAL CURRENT SHAPE C. Pawlu, M. Heckmann Whether release from synaptic vesicles happens in an all-or-none manner or is subject to postfusional regulation, is a matter of debate. During a study of the glutamatergic currents at neuromuscular junctions of Drosophila, we noticed that the compound EPSCs decayed faster than the quantal currents in this preparation (in recordings with a macropatch electrode at 20°C). The decay of quantal currents was often protracted, contradicting the assumption of a rapidly declining glutamate concentration after a quantal release. To quantify this protraction we measured the area of the quanta (Q) and the area of idealized quanta with a pure monoexponential current decay (F) and calculated the hump ratio h = Q/F. The mean h was 1.48 ± 0.03 SE. The hump ratio was reduced in quantal currents from receptor mutants with faster desensitisation. Computer simulations of quantal currents indicate that the hump is due to prolonged glutamate release. Postfusional regulation of transmitter release appears to be calcium dependent in chromaffin cells (Elhamdani et al., 2001, Neuron, 31:819). We studied the calcium sensitivity of our quantal current shapes and found that the protraction of the current decay was highly calcium sensitive, indicating a shorter glutamate release at high calcium. In conclusion it appears that glutamate is released over a period of several milliseconds through a calcium sensitive fusion pore. Supported by DFG SFB 391 Physiologisches Institut I, Albert-Ludwigs-Universität, Hermann-Herder-Str. 7, D-79104 Freiburg, Germany P 15-3 ROLE OF Na + /Ca 2+ EXCHANGE IN Ca 2+ EXTRUSION MECHANISM IN AXON TERMINALS OF NEUROHYPOPHYSIS W.K. Ho, M.H. Kimm, S.H. Lee, Y.E. Earm Mammalian neurohypophysis (NHP) represents a model system for studying physiology of peptidergic nerve terminals. Ca 2+ is a key molecule which couples action potential with exocytosis. Although it is generally accepted that Na +/Ca2+ exchanger and Ca 2+-ATPase are two primary Ca2+ clearance mechanisms in axon terminals, the factors that control Ca2+ decay in axon terminal is not fully understood. In order to estimate role of Na+/Ca2+ exchange in Ca2+ extrusion in axon terminals, we investigated Ca2+ decay rates in Ca2+ transient evoked by depolarizing pulses, using whole-cell patch clamp method and fura2 microfluorometry. Ca2+ decay phases were well fitted with single or double exponential curves. Average Ca2+ decay rate was 1.39 ± 0.32/ sec (n=6) in control condition that NaCl and K-gluconate solution containing 50 µM fura-2 were used as bathing and pipette solution, respectively. Decrease in [Na]o from 143 mM to 20 mM by isomolar replacement with Li+ or N-methyl-D-glucamin (NMG+) slowed down the decay rate to 0.50 ± 0.08 / sec (n=4). The extent of slow-down in external low Na+ solution was not dependent on peak Ca2+ range we examined. When internal K+ was replaced with NMG+, the decay rate became significantly slower (0.43 ± 0.13 / sec; n=11). Under this condition, replacement of external Na+ with Li+ decreased the decay rate further (0.36 ± 0.10; n=10). From the results, we concluded that 1) Na+/Ca2+ exchanger plays a significant role in extrusion of Ca2+ in axon terminals of neurohypophysis, 2) intracellular K+ is involved in the Na +/Ca2+ exchange mechanism. Department of Physiology and National Research Laboratory for Cellular Signalling, Seoul National University College of Medicine, Seoul, 110-799, South Korea

S 264 P 15-4 DELAY BETWEEN FUSION PORE OPENING AND PEPTIDE RELEASE FROM LARGE DENSE-CORE VESICLES S. Barg, C.S. Olofsson, P. Rorsman Peptidergic neurotransmission is slow compared to that mediated by classical neurotransmitters. We have studied exocytotic membrane fusion and cargo release by simultaneous capacitance measurements and confocal imaging of single EGFP-labeled secretory vesicles in neuroendocrine cells. Depletion of the readily releasable pool (RRP) correlated with exocytosis of 10-20% of the docked vesicles. Some remaining vesicles became releasable after recovery of RRP. Expansion of the fusion pore, seen as an increase in luminal pH (and EGFPfluorescence), occurred after ≈0.3 s and peptide release was delayed by another 1-10 s. We conclude that 1) RRP-refilling involves chemical modification of vesicles already in place; 2) release of large neuropeptides via the fusion pore is negligible and only proceeds after complete fusion; and 3) sluggish peptidergic transmission reflects the time course of vesicle emptying. A

B

10 m s

C

0.2 s

D

1s

Schematic explaining exocytosis and hormone release. (A-B) Exocytosis of a docked vesicle involves the rapid (<20 ms) opening of the fusion pore. The fusion pore is wide enough to enable detection of increased cell capacitance but not for release of vesicle contents. (C) Following the expansion of the fusion pore, protons equilibrate with the extracellular medium resulting in an increase in EGFP fluorescence. The expanded fusion pore has a mean lifetime of ~1.6 s. (D) Marked dilation of the fusion pore or even complete fusion is required for release of bulky peptides. Department of Molecular and Cellular Physiology, Institute for Physiological Sciences, Lund University, Sweden P 15-5 RAB3 PARALOGS ARE ESSENTIAL REGULATORS OF SYNAPTIC TRANSMISSION IN MICE O.M. Schlüter, F. Schmitz, T. Jo, M. Geppert, R. Jahn, T.C. Südhof, C. Rosenmund Rab3A, B, C, and D constitute a family of small GTP-binding proteins that are implicated in regulated exocytosis. Using quantitative immunoblotting we show that Rab3A, B, and C are differentially expressed in the brain, whereas Rab3D is the most abundant isoform of exocrine glands and present only in minor amounts in the brain. Although Rab3A is the most abundant Rab protein in brain, the Rab3A knockout mice developed a relative mild phenotype, that is obvious only in a subset of synapses. To test whether the remaining three Rab3 paralogs share a redundant function, we generated knockout mice lacking all four Rab3 paralogs. These mice died soon after birth, whereas triple knockout animals were viable. In a detailed electro-physiological characterization of single neurons in microdot island cultures, the quadruple knockout animals demonstrated an increased paired pulse fascillitation. Synaptic amplitudes, synaptic pool sizes, and the basal release probability however were unchanged, indicating a role for Rab3 proteins in encoding synaptic transmission during fast trains of action potentials. Max-Planck-Institut fur biophysikalische Chemie, Am Faßberg 11, 37077 Göttingen und Max-Planck-Institut fur experimentelle Medizin, Hermann-Rein-Straße 3, 37075 Göttingen P 15-6 β PHORBOL ESTER -INDUCED AUGMENTATION OF NEUROTRANSMITTER RELEASE FROM HIPPOCAMPAL NEURONS IS MEDIATED BY MUNC-13s AND NOT BY PKCs J.S. Rhee*, A. Betz, S. Pyott, K. Reim, F. Varoqueaux, I. Augustin, D.

Hesse, T.C. Südhof, M. Takahashi, C. Rosenmund, N. Brose The presynaptic proteins Munc13-1 and Munc13-2 are essential for formation of fusion competent synaptic vesicles. They contain a diacylglycerol (DAG)/β phorbol ester (βPE) binding C1-domain and are therefore a potential target of the DAG second messenger pathway acting in parallel with PKCs. To understand the roles of PKCs and Munc13-1/2 in the βPE induced enhancement of neurotransmitter release, we employed genetically modified mice that express a DAG/ βPE binding deficient Munc13-1H567K (m/m) variant instead of the wild type (WT) protein. This mutation was perinatally lethal. Using wholecell patch clamp recordings from neonatal primary hippocampal autaptic cultures, we observed in WT neurons a twofold enhancement of evoked EPSC amplitudes by application of 1µM β phorboldibutyrate (PDBU). This enhancement was strongly reduced in m/m neurons and completely blocked when the contribution of wild type Munc13-2 in mutant cells was abolished. PKC dependent phosphorylation in the presynaptic terminal was unchanged and Munc13-1 was neither a PKC substrate nor impaired in its expression. Analyses of vesicular dynamics in m/m neurons revealed that a pool of vesicles with low basal release probability, which shows in WT neurons strong enhancement of release probability when stimulated at high frequency or when stimulated by βPE, was eliminated by the mutation. As a result, m/m neurons showed unusually strong depression of synaptic amplitudes when evoked release was initiated at high frequency. These data demonstrate that that Munc13s are the main presynaptic DAG/β phorbol ester receptors in hippocampal neurons and that modulation of Munc13-1 activity by second messengers via the DAG/β phorbol ester binding C1 domain is essential for use dependent alterations of synaptic efficacy and, ultimately, for survival. MPI für biophysikalische Chemie, Abt. Membranbiophysik, Am Faßberg 11, 37077 Göttingen, Germany

P 15-7 NSF IS PRESENT IN RAT PANCREATIC B-CELLS AND IMPORTANT FOR SUSTAINED EXOCYTOSIS L. Eliasson, J. Schriever-Abeln, X. Ma Background and Aims: Insulin is characteristically released in a biphasic pattern where the first rapid phase is thought to be attributable to primed granules immediately ready to be released upon an increase in intracellular Ca2+, and the second phase to the ATP-dependent mobilisation of granules from a reserve pool. The aim of this study was to investigate the role of NSF (NEM-sensitive factor) in exocytosis of insulin-containing granules using a monoclonal antibody against the protein (Mab 2E5). Material and Methods: For the investigations of the exocytotic process capacitance measurements were combined with the standard whole-cell configuration of the patch-clamp technique. This technique allows an intracellular application of the antibody. The identification of the protein was obtained using immunocytochemistry and Western Blot analysis preceded by subcellular fractionation. Results: An ~80 kD band corresponding to the size of NSF was mainly found in the cytocolic fraction using Western Blot analysis. Immunostaining confirmed the presence of NSF in the cytosol. Further, NSF was shown to co-localize with insulin, mainly close to the plasma membrane. For the temporal measurements a train of ten 500 ms depolarising pulses from -70 to 0 mV were applied on single rat B-cells. Under control conditions the first depolarisation evoked an increase in membrane capacitance of 202+21 fF (n=6) corresponding to the release of 100 granules using a conversion factor of 2 fF/granule. The exocytotic response was significantly (P<0.02) reduced by 33% to 134±13 fF (n=7) in the presence of the antibody. A more pronounced decrease was observed during the latter pulses where an 80% reduction in the increase in membrane capacitance was observed in the presence of Mab 2E5 (11±5 fF vs 57+17 fF under control conditions; P<0.03). The exocytotic response during the latter pulses evoked by a second train was almost abolished in the presence of the antibody. Conclusions: The data suggest that NSF is involved in Ca2+-dependent exocytosis in rat pancreatic B-cells. The protein might either influence an early step of the granule recruitment or take part in the rearrangement of the exocytotic complex after fusion of already primed granules. Department of Molecular and Cellular Physiology, Lund University, BMC Fll, Tornavägen 10, S-221 84 Lund, Sweden. Email: Lena. [email protected]

S 265 P 15-8

P 15-10

THE DEPENDENCE OF EXOCYTOSIS RATE IN PANCREATIC βCELLS ON THE CA2+ SENSING LATENCY TIMES, CA2+ CHANNEL KINETICS AND CLUSTERING J. Galvanovskis, B. Soderberg The release of insulin from secretory granules requires a transient increase of intracellular Ca2+ concentration in the vicinity of the fusion site. The characteristic feature of pancreatic β-cells is that a short time delay (~10 ms) is observed in these cells between the actual increase in [Ca 2+]i and the onset of membrane fusion. It is understood that this delay or the latency time in the fusion of membranes reflects a certain structural organization of the Ca2+ sensing system of the secretory machinery. Obviously, in a situation where the Ca2+ is supplied through randomly opening channels, the existence of the latency time affects in an essential way the probability for a certain granule to fuse with the plasmalemma after a depolarizing pulse and to release its contents into the extra-cellular space. In this project we have investigated, theoretically and numerically, the extent of this influence and its dependence on the parameters of Ca2+ channels, channel clustering, the Ca2+ sensing system, and, in experimental conditions, on the length of depolarizing pulses. We used a linear probabilistic model for a random opening and closing of channels, that yields an explicit expression for the Laplace transforms of the waiting time distributions for at least one channel to be open during the latency time. This allows one in principle to calculate the probability that a vesicle will fuse with plasmalemma during the action potential. The theoretical results were used to test our numerical simulations of the exocytosis in pancreatic β-cells. Several rigid assumptions about the organization of a secretory machinery are to be made in order to legally apply this mathematical model for the analysis of exocytosis rate. The latest experimental data show that these assumptions may be indeed implicated in the structural organization of secretory proteins in pancreatic β-cells. The applicability and limitations of these ideas to other cell types, such as neurons, is discussed.

ROLE OF PROTEINASES IN TUMOR CELL INVASION DETERMINED BY ELECTRICAL RESISTANCE BREAKDOWN ASSAY T. Ludwig*, R. Ossig, E.M. Schnaker., S. Graessel, H. Oberleithner, S.W. Schneider Invasion is the most distinctive feature of malignant tumors characterized by sensitive method is available to measure cancer cell invasion in a functional cell-based assay. We applied the recently developed resistance breakdown assay measuring changes in transepithelial resistance of a tight epithelial MDCK-C7-monolayer in response to tumor cells. This electrophysiological approach provides a continuous assay system for cancer invasion that demonstrates changes prior to visible morphometric changes. By means of this assay system we addressed the question whether physical contact between tumor cell and epithelial monolayer is a prerequisite of tumor cell invasion. When human melanoma cells were seeded directly (i.e. physical contact) on top of an electrically tight epithelial cell layer (5800 ± 106 Ωcm2) electrical monolayer leakage occurred fast and efficient (18 ± 3 % reduction in 24 h). When melanoma cells were seeded close to the basolateral surface of the epithelial cell monolayer but separated by a filter membrane (i.e. no physical contact) electrical leakage occurred even more effective (42 ± 3 % reduction in 24 h). Atomic force microscopy detected discrete structural changes between cells. Electrical leakage could be effectively blocked by inhibitors of matrix metalloproteases (MMPs). Zymography and immunoblotting detected three soluble MMPs (MMP1, MMP2 and MMP9) secreted by the melanoma cells. We conclude that exocytosis of soluble proteases causes electrical monolayer leakage that is the first step for tumor cell invasion. Physical contact between tumor cells and monolayer is not sufficient to attack the intact epithelium.

The Department of Molecular and Cellular Physiology, the Institute of Physiology, Lund University; Tornavägen 10, BMC F11, 221 84 Lund, Sweden

P 15-9 DEVELOPMENT OF AN ASSAY TO MONITOR HOMOTYPIC FUSION OF ISOLATED SECRETORY VESICLES BY FLUORESCENCE MICROSCOPY H. Sammer, M. Frick, W. Singer, N. Mair, H. Moser, M. Ritsch-Marte, P. Dietl, T. Haller The investigation of the basic mechanisms underlying exocytotic vesicle fusion has progressed by the use of in vitro fusion assays. However, many studies relied on the use of synthetic lipid bilayers as a substitute for the plasma membrane, and fusion assays using native membranes from mammalian cells are sparse. Homotypic vesicle-vesicle fusion, which is reported to occur in many secretory cell types, is thought to follow similar or even identical mechanisms as in exocytotic fusion. Therefore, we used the secretory vesicles of alveolar type II cells (lamellar bodies; LB’s), the size of which (~ 2 µm) allowed a direct visual control and manipulation with laser traps. LB’s were isolated from homogenised rat lungs by ultracentrifugation in a continuous sucrose density gradient. The vesicles could be brightly stained with LysoTracker Green and Lysotracker Red, which both accumulate within these organelles due to their acidic interior. The simultaneously added amphiphilic dye FM 1-43 did not stain the lipidic LB contents (surfactant). Moreover, the fluorescence intensity of LysoTracker stained LB’s decreased to background levels by adding nigericin, which dissipates the pH-gradient across biological membranes. Thus, these protocols demonstrate that an intact limiting membrane of LB’s is preserved following isolation. Vesicle fusion was stimulated by applying either centrifugal forces, or by forces exerted with an optical stretching device. Fusion was detected as the mixing of the red and green LysoTracker dyes, resulting in an intermediate spectral emission. This assay should allow to study the fusion process at the level of individual vesicles. Supporte by ÖNB and FWF grants, No. 7413, 12974 and 14263 Department of Physiology, University of Innsbruck, Fritz-Pregl-Str. 3, A-6020 Innsbruck, Austria

Physiologisches Institut, Robert-Koch-Str. 27a, 48149 Münster, Germany P 15-11 MONITORING OF ANP SECRETION FROM SINGLE ATRIAL MYOCYTES USING DENSITOMETRY S.Y. Ryu, S.H. Lee, W.K. Ho, Y.E. Earm Atrial myocytes secrete atrial natriuretic peptide (ANP) in response to mechanical stretch, and can serve as a challenging model for studying stretch-secretion coupling. Most studies about ANP secretion have been carried out with measurement of ANP every tens of minutes from the perfusate of isolated atria using radioimmunoassay. This method has a limitation for studying cellular mechanisms of stretchsecretion coupling, due to the poor time resolution and translocation barrier between transcellular fluid and perfusate whose integrity can be affected by atrial contraction, extracellular [Ca 2+], and paracrine hormones. Here, we developed a technique for monitoring ANP secretion from a single atrial myocyte with relatively high time resolution, using neutral red and video-CCD camera. After incubation of atrial myocytes with 0.005% neutral red for 30 min at 36 oC, atrial specific granules were well observed under the conventional inverted microscope. They were often clustered around nucleus, and stained orange to red color according to their densities. In order to quantify the amount of secretion, we monitored the change of gray values of the cell, while it was illuminated with monochromatic light (550 nm). We set the region of interest (ROI) surrounding the atrial specific granules and calculated ODmean using the following equation: ODmean = log [(intensity – black level)/(incident light – black level)]. ODmean is considered to be proportional to intracellular concentration of neutral red, and the change of ODmean can be regarded as the total amount of secretion from the cell. Under the control condition without stimulus, ODmean spontaneously decreased with linear time course. Average slope of the basal decrease for 10 min was 19.7 ± 1.4 % (n = 14). When we applied direct mechanical stretch to the atrial myocyte with two micropipettes in longitudinal direction by 20 % of resting cell length, the ODmean decrease was significantly accelerated (48.7 ± 7.4 % / 10 min ; n = 3). Hypotonic swelling (200 mOsm) also accelerated the OD mean decrease significantly (47.2 ± 2.4 % / 10 min; n=7). In conclusion, this method allows the monitoring of ANP secretion with relatively high time resolution while mechanical stress is applied. Furthermore, patch clamp technique or intracellular perfusion technique can be combined for studying cellular mechanisms of stretch-secretion coupling.

S 266 Department of Physiology and National Research Laboratory for Cellular Signalling, Seoul National University College of Medicine, Seoul, 110799, South Korea P 16-1 HETEROMERIC AMPA RECEPTORS ASSEMBLE WITH A PREFERED SUBUNIT STOICHIOMETRY AND SPATIAL ARRANGEMENT M. Mansour*, N. Nagarajan, R.B. Nehring, J.D. Clements, C. Rosenmund AMPA-receptors are thought to be a tetrameric assembly of GluR2 with GluR1/3/4 subunits. This heteromeric assembly controls the permeability of AMPA channels. If subunits combine randomly, heterogenous assembly populations may be formed. Alternatively, specific mechanisms may exist that promote the preferential assembly of heteromeric channels incorporating GluR2 subunit. We have investigated the inter-subunit interactions that direct AMPA receptor assembly and its influence on channel gating. The GluR1 and GluR2 subunits carried markers controlling ion permeation (e.g. GluR2 Q607R) and desensitization (e.g. GluR1 L497Y). Responses to saturating Glutamate concentrations from homomeric or heteromeric GluR1/2 receptors were quantified in outside-out patch configuration excised from transfected HEK cells in conjugation with fast flow application system. The rectification properties and the effect of the desensitization blocker cyclothiazide were monitored as a function of subunit composition and relative expression level. The results were described with binomial or preferential subunit assembly models. Homomeric receptors assembled according to a binomial distribution while heteromeric receptors preferentially formed with a stoichiometry of two GluR1 and two GluR2 subunits positioned on opposite sides of the channel pore. This structure will predominate if GluR1 binds to GluR2 more rapidly during receptor assembly than other subunit combinations. The selective assembly of AMPA receptors represents a case study in spontaneous self-assembly of protein nanomachines. Its practical outcome is a more homogenous receptor population in vivo. MPI für biophysikalische Chemie, Abt. Membranbiophysik, Am Faßberg 11, 37077 Göttingen P 16-2 LIGHT-DEPENDENT REGULATION OF NMDA RECEPTORS IN RAT RETINAL GANGLION CELLS IS NOT MEDIATED BY INCREASING OF NR2A SUBUNIT EXPRESSION G. Albach, S. Schmid, Th. Wheeler-Schilling, S. Fauser, E. Guenther Purpose: Changes in N-methyl-D-aspartate receptor (NMDAR) properties are reported to regulate periods of plasticity in different cortical areas. In order to learn whether similar changes occur in the retina, structure-function relationships of NMDAR were investigated in rat retinal ganglion cells (RGCs) during retinal development. Methods: Perforated-patch clamp recordings of NMDAR-currents were performed on identified RGCs in a slice preparation of the pigmented rat retina, followed by single-cell RT-PCR and semi-quantitative dot blots to analyze the NMDAR subunit distribution in individual RGCs. Drugs were applied by a fast superfusion system via a pipette located in a distance of about 100 µm from the RGC recorded. Results: Before eye opening, all RGCs investigated between postnatal day (P) 9-12 had large NMDAR currents with mean amplitudes of -110 pA (±53 pA SD). When animals were kept in a 12 h light-dark cycle, NMDAR current amplitudes decreased after eye opening to a mean value of -13 pA (±7 pA SD) at P30 and could be elicited in only 65% of the RGCs. In contrast, dark rearing animals, up to P30, prevented this decrease of NMDAR currents. In dark-reared animals, NMDAR currents could be evoked in 93% of RGCs with mean amplitudes of -81 pA (±25 pA SD). Subsequent light exposure for as little as 5 days induced a down regulation of NMDAR current amplitude to a mean of -19 pA (±7 pA SD). Single-cell RT PCR analysis revealed distinct developmental alterations in the composition of NMDAR subunits in RGCs; however, none of them were induced by the onset of light. At P30, there is no statistical difference in quantitative and qualitative aspects of NR2A expression, that could explain the different NMDAR current amplitudes under different light treatments. Conclusions: Our data indicate that the light-induced decrease of NMDAR currents in rat RGCs is not caused by alterations in the molecular structure of NMDARs. This is in contrast to cortical neurons where alterations in NMDAR proper-

ties during development seem to be related to an increased substitution of subunit NR2A for NR2B in the NMDA channel complex. Since the down-regulation of NMDAR in RGCs occurs after eye opening, when the retinal network is reported to be mature, we hypothetise that NMDAR function in RGCs is related to periods of retinal plasticity and is regulated by a post-translational mechanisms acting on the NMDAR protein rather than a change in NMDAR subunit expression and composition. Support: German Research Council, SFB 430 G. Albach, Dept. Exp. Ophthalmology, Univ. Eye Hospital, Roentgenweg 11, D-72076 Tuebingen, Germany P 16-3 ANION PERMEATION IN EAAT2 GLUTAMATETRANSPORTERS A. Biela, Ch. Fahlke Glutamate, the predominant excitatory neurotransmitter in brain, is removed from the synaptic cleft by specialized transporters located in the plasma membrane of nerve and glial cells. Five EAAT isoforms have been cloned so far, EAAT1 to 5, and all exhibit in addition to the carrier-mediated glutamate transport a novel thermodynamically uncoupled anion conductance. To get insights into basic properties of this anion-selective conduction pathway, we studied hEAAT2, heterologously expressed in tsA201 cells using whole-cell patch clamp recordings under various ionic conditions. In the absence of external substrate, currents are very small in symmetrical Cl-. External substitution of I-, NO3- and SCN- causes an increase of the current amplitudes with a SCN > NO3 > I > Cl selectivity. More permeant anions do not only carry a larger anion influx at positive potentials, but also cause an increased efflux of Cl- at negative potentials. Upon addition of external glutamate, the current amplitudes at positive potentials increased in a variable extent in external I-, NO3- and SCN-, but not in external Cl. External substrate does not simply increase the percentage of time the transporter spend in its anion conducting mode, but interferes with a functional property of the anion conduction pathway, the selectivity between anions. External glutamate furthermore increased the current amplitude at negative potentials. In external Cl-, this glutamateinduced current entirely consists of the carrier-mediated uptake current. For external I-, the current increase is smaller than this current amplitude, it is roughly the same for NO3-, and much larger than the glutamate-uptake current for SCN- indicating that external anions also affect the carrier mode of EAAT2. Our results are in agreement with the notion of two distinct, but intimately interacting functional domains, an anion-selective pore and a glutamate-carrier. Activation of one functional domain is not necessary for the function of the other, but modifies its properties. Institut für Physiologie, RWTH Aachen, Pauwelsstr. 30, 52074 Aachen P 16-4 INHIBITION OF NEURONAL AND GLIAL GLUTAMINE TRANSPORTERS ATA1 AND SN1 BY AMMONIUM CHLORIDE (NH 4 Cl) C.A. Wagner, A. Bröer, S. Bröer Glutamate is the major excitatory neurotransmitter. After release into the synaptic cleft it is taken up by neighbouring astrocytes and converted into glutamine. Glutamine is then released from the astrocytes, is accumulated by neurons and after conversion back into glutamate stored in presynaptic vesicles. We have recently identified several transporters involved in the release and uptake of glutamine, namely ASCT2, ATA1, SN1, and 4F2/y + LAT2. During hepatic encephalopathy, serum ammonia levels increase to 1 - 2 mM due to the loss of hepatic clearance and cause a variety of neurological symptoms. Here we show that glutamine uptake by the glial SN1 and neuronal ATA1 glutamine transporters expressed in Xenopus oocytes is specifically inhibited by low concentrations of NH4Cl with Ki values of 1.1 mM and 0.5 mM, respectively. The glutamate transporter GLAST and the glutamine transporters ASCT2 and 4F2/y +LAT2 transporters, however, were not affected. Inhibition was not due to depolarisation of oocytes as demonstrated by current and voltageclamp experiments. Reduction of transport was at least in part due to inhibition of Na+ -binding to the carriers which is an essential part of the transport mechanism. The efflux of glutamine via the glial transporter SN1 was not inhibited by NH4Cl suggesting that neuronal glutamine uptake may be inhibited rather than release from astrocytes.

S 267 Thus, these data suggest an involvement of glutamine recycling transporters in the brain in the pathogenesis of hepatic encephalopathy.

Department of Neurophysiology, Heinrich-Heine-Universität, POB 101007, D-40001 Düsseldorf, Germany

Division of Biochemistry and Molecular Biology, Faculty of Science, Australian National University, Canberra, ACT 0200, Australia P 16-7

P 16-5 INTRAOPERATIVE MEASUREMENT OF EXTRACELLULAR AMINO ACIDS IN THE EPILEPTIC HUMAN NEOCORTEX USING MICRODIALYSIS (EPILEPSY SURGERY) H. Straub*, J. Wölfer, D. Kuhlmann, H.W. Pannek, F. Behne, B. Rambeck, U.H. Jürgens, E.-J. Speckmann There is much experimental support for a link between activation of the excitatory amino acid transmitter system and the generation and propagation of seizure activity. Epileptic discharges may result from a mismatch between excitatory and inhibitory amino acid systems. In order to investigate biochemical and electrophysiological relationships in the human epileptic focus, microdialysis was used together with electrocorticography focussing excitatory amino acids. Extracellular fluid was obtained from patients (n=4) during surgical resection of epileptic foci using microdialysis. The microdialysis probe (cut off 20.000 kD; flow 1 µl/min) was placed into the cortex which had to be removed and fixed by a pressure foot. An equilibration period of 35 min was allowed to obtain basal amino acid levels. Sampling was performed in 10 min fractions. Samples were analyzed by HPLC with fluorescence detection. Parallel to the micodialysis the electrocorticogram (ECoG) was recorded. In the dialysate of three patients showing low frequency asynchronous spikes in the ECoG the following basal levels of amino acid concentrations (in µM) were observed: glutamate: 2.86 –8.80, aspartate: 0.98 –2.42, glutamine: 122 –232, alanine: 8.09 –22.85, taurine: 7.07 –20.18. In the dialysate of one patient exhibiting the EEG pattern of a generalized tonic clonic seizure during microdialysis altered levels of amino acids were detected. GLU displayed the largest increase ranging from 2.5 to 8 times the basal levels. Aspartate increased slightly to three times the basal level. Glycine, alanine and taurine displayed the same basal concentration range. The in vivo microdialysis provides the advantage to relate intraoperative results to those from postoperative functional and morphological (immunohistochemistry, receptorautoradiography) analyses of identical tissue. Institut für Physiologie, Robert-Koch-Str. 27a, 48149 Münster, Germany P 16-6 GABA A RECEPTOR ASSEMBLY AND FUNCTION IN HYPOTHALAMIC NEURONS O.A. Sergeeva, K.S. Eriksson, I.N. Sharonova, V.S. Vorobjev, H.L. Haas The structure-function relationship of GABAA receptors is studied in native hypothalamic neurons of the rat. Whole-cell recordings and single-cell RT-PCR from acutely isolated rat tuberomammillary (TM) neurons were used to characterize GABA -evoked currents and to correlate them with the expression pattern of 12 GABA A receptor subunits. Occurrence and combination of subunits (α1, α2, α5, β1-3, γ1, γ2 and ε) was highly variable. Three distinct groups were apparent according to GABA-sensitivity and γ-subunit expression: 1. neurons containing γ1- and γ2-subunit-mRNA yielded an EC50= 6 µM, 2. neurons lacking γ-subunit-mRNA an EC50= 19 µM, 3. neurons coexpressing γ2with various α- and β-subunits an EC50= 45 µM. These three groups showed no difference in cell soma size or expression of glutamic acid decarboxylase (GAD). Immunocyto-chemistry demonstrated the γ2 subunit in the vast majority of TM neurons. A pharmacological analysis suggested that the γ1- and γ2-subunit-proteins are present in the1st and 2nd group, but only the γ2-subunit in the 3rd group. TM neurons from the 1st and 2nd groups differed in their zinc sensitivity. Zolpidem prolonged the decay time of spontaneous IPSCs and GABA-responses more in the 1 st than the 2nd group. Neurons with different expression levels of GABA A receptor γ-subunits displayed a difference in cooperativity for GABA- and zolpidem-binding which we explain by the presence of one versus two γ-subunits (modulatory sites) in one receptor. Thus, we describe here for the first time native GABA A receptor function in relation to its stoichiometry.

PARS COMPACTA OF THE SUBSTANTIA NIGRA: IDENTIFICATION OF GABAA RECEPTORS ON DOPAMINERGIC NEURONS IN THE HUMAN MESENCEPHALON S. Petri, K. Krampfl, R. Dengler, J. Bufler Pathomechanisms of neurodegenerative diseases affecting the basal ganglia are poorly understood, but there are many hints that ligandgated receptor channels are involved in the pathogenesis. As the degeneration of the dopaminergic neurons of the substantia nigra (SN) is crucial for the pathogenesis of parkinson‘s disease (PD), synaptic transmission in this brain region is of major neurologic interest. GABA is the most widespread inhibitory neurotransmitter. GABAA-receptors maintain fast postsynaptic inhibition at excitable neuronal membranes. The functional receptor channels are made up of five subunits, each contributing specific structural, physiological and pharmacological properties to the receptor protein. We employed a combination of in situ hybridization histochemistry (ISH) and immunohistochemistry (IHC) on sections of human postmortem mesencephalon in order to investigate the expression of GABAA-receptor subunit mRNAs in SNcells immunopositive for tyrosine hydroxylase (TH), the pivotal enzyme of dopamine-synthesis. In colocalization experiments the mRNA of the GABA A-receptor subunit α1 was detected in cells strongly immunopositive for TH. ISH-experiments showed the GABAA receptor subunits α1, β2 and γ2 being coexpressed in human SN. Thus we could demonstrate that dopaminergic neurons in human SN receive inhibitory input via GABA A receptors of the α1β2γ2-subtype. This may represent a negative feedback loop between the striatum and the SN, which is impaired in PD by depletion of dopaminergic cells in SN. Dr. S. Petri, Dept. of Neurology, Medizinische Hochschule Hannover, Carl-Neuberg-Str. 1, 30625 Hannover

P 16-8 CHANGES IN THE SYNAPTIC TRANSMISSION IN THE SPINAL VENTRAL HORN OF THE MOUSE MUTANT SPASTIC J. v.Wegerer, K. Becker, H. U. Zeilhofer, C.-M. Becker, D. Swandulla Synaptic transmission was examined in two strains of the glycine receptor mutant mouse spastic displaying a severe (C57BL/6J) and a mild phenotype (F1(C57BL/6J x C3H/HeJ) x C57BL/6J). Electrically evoked postsynaptic whole-cell currents were recorded from αmotoneurons in the ventral horn of lumbar spinal cord slices. Amplitudes of glycinergic inhibitory postsynaptic currents (IPSCs) were reduced in both mutants. The degree of inhibition correlated with the severity of the neurological phenotype. The impairment was not only limited to glycinergic neurotransmission, but also reduced significantly the GABAA receptor-mediated IPSCs and the glutamatergic EPSCs in both mutants (holding potential –80 mV). The NMDA receptor-mediated EPSC was increased in the mutants with the severe phenotype. The ratio of each transmitter component to the total PSC was unchanged in the mutants with the mild phenotype, while the relative glycinergic contribution was reduced in those mutants with the severe phenotype. Although the primary cause for the disturbance in synaptic transmission in spastic mice is the postsynaptic glycine receptor defect, an analysis of variance suggested in addition presynaptic changes for the glycinergic transmission in the mutants with the severe phenotype, and for the GABAA receptor-mediated and glutamatergic neurotransmission in both mutants. The decay phase of the postsynaptic currents was fitted with a bi-exponential function. In both mutants the fast decay time constants τ1 were slightly reduced for the glycinergic IPSC and significantly longer for the GABAA receptor-mediated IPSC. In the mutant with the severe phenotype the glutamatergic EPSC was significantly prolonged, while it was not changed in the mild phenotype. The NMDA receptor-mediated EPSC kinetics were not altered in the mutants. Our findings suggest that the mutation spastic does not only affect glycine receptors, but influences spinal neurotransmission in a complex manner. Physiologisches Institut II, Universität Bonn, Wilhelmstr. 31, 53111 Bonn, Germany

S 268 P 16-9

P 16-11

TWO-PHOTON CHLORIDE IMAGING IN SINGLE CELLS OF BRAIN SLICES N. Marandi, O. Garaschuk, A. Konnerth Two-photon laser scanning microscopy has been used successfully for imaging activity-dependent changes of calcium and sodium levels in single cells. Here we introduce a simple approach of two-photon chloride imaging in individual neurons in rat brain slices. This approach involves the use of the membrane permeant chloride indicator dye N(6-methoxyquinolyl) acetoethyl ester (MQAE). Besides a high chloride affinity (in situ Kd = 31 mM), MQAE is insensitive to bicarbonate and to changes of pH in the physiological range. We find that incubation with MQAE produces a robust loading of cells within the upper layer of slice preparations obtained from various brain regions, including hippocampus, cortex and cerebellum. In contrast to conventional fluorescence measurements, two-photon imaging using MQAE does not suffer from dye bleaching and phototoxic damage. By using pyramidal neurons in cortical and hippocampal slices, we demonstrate that the new approach allows prolonged recordings of time-resolved changes in intracellular chloride concentration with a good signal-tonoise ratio. Thus, application of GABA on neurons stained with MQAE evoked marked fluorescence changes (10-20%) that were well resolvable in sub-cellular compartments, e. g. cell bodies and dendrites. The GABAA receptor antagonists picrotoxin (20µM) and bicuculine (20µM) blocked these responses. Our results indicate, that under conditions of two-photon excitation, MQAE reports intracellular chloride dynamics with significantly better signal-to-noise ratio than earlier approaches. Thus, MQAE-based two-photon microscopy promises to be a valuable method for many applications requiring chloride imaging in single cells.

FUNCTIONAL POSTSYNAPTIC NICOTINIC ACETYLCHOLINE RECEPTORS ON PYRAMIDAL NEURONS IN THE RAT NEOCORTEX G. Zolles*, B. Sutor Nicotinic acetylcholine receptors (nAChR) have been implicated in a variety of brain functions, including neuronal development, attention, learning and memory formation. Until recently nAChRs were thought to serve primarily as modulators in the brain regulating neurotransmitter release from nerve terminals via preterminal and presynaptic receptors, respectively. While new evidence has revealed functional postsynaptic nicotinic receptors and fast acetylcholine(ACh)-mediated synaptic transmission onto interneurons in the neocortex, there is still controversy regarding a nAChR-mediated postsynaptic action of ACh on neocortical pyramidal neurons. Here, we report the discovery of functional postsynaptic nicotinic receptors on pyramidal neurons in the rat neocortex in vitro. The whole-cell mode of the patch-clamptechnique was used to record currents evoked by pressure-application of the non-selective agonist ACh on pyramidal neurons in layer V of the frontal and somatosensory cortex in brain slices. The experiments were performed in the presence of TTX, glutamate receptor and GABA receptor antagonists and atropine (1 µM). Under these conditions, ACh evoked inward currents (holding potential: -70 mV) displaying a biphasic decay. Methyllycaconitine (MLA), an antagonist selective for nAChRs composed of α7 subunits, inhibited 90-100% of the AChevoked current. In some cases a residual slowly rising and decaying current was observed, which could be blocked by additional application of dihydro-β-erythroidine (DHβE), a non-α7 nAChR antagonist. These results suggest a heterogeneous nAChR population expressed on pyramidal neurons of the rat neocortex. Since most of the ACh-induced inward current was suppressed by MLA, we conclude that the nicotinic currents were predominantly mediated by nAChR composed of a7 subunits. Due to its high calcium permeability, this nAChR subtype may play an important role in calcium dependent processes including long-term potentiation. (supported by the FHG)

Institut für Physiologie, LMU München, Biedersteiner Str. 29, 80802 München P 16-10 APPLICATION OF SEMICONDUCTOR LASER LIGHT VIA GLASS FIBRES FOR MEASURING INTRACELLULAR Ca 2+ TRANSIENTS St. Dörschel, J. Eichler, G. Wallukat, R. Mrowka, A. Patzak, E. Storch The measurement of intracellular Ca2+ concentrations using fluorescent dyes in three dimensional biological objects demands an excitation light that is spatially positioned. This has been put into effect in the following study by the application of a glass fibre in which a laser light with the desired wavelength is coupled. Fura 2 was used as an indicator of calcium fluorescence whose excitation wave lengths are 340 and 380 nm. The change in the intensity of the emission light of about of 510 nm correlates with the intracellular changes in the Ca2+concentration. Fura 2 was monochromatically stimulated with ultraviolet light of 390 nm produced by a seminconducting laser with 5 mW of optical power (NLHV500A, NICHIA VIOLET LASER DIODE). The laser light was coupled into the glass fibre (core diameter:10 µm, covering diameter:125 µm, and numeric aperture: 0.05) via a lens system. The fibre can be positioned in all three dimensions toward the biological object with the help of a micromanipulator. The set up of the experiment was established on an inverse microscope (Axiovert 35, Carl Zeiss). A photometer (R928, Hamamatsu) served to measure the intensity of the fluorescence. The intensity of the fluorescence excited with the laser light (390 nm) by means of a glass fibre was compared to the fluorescence classically excited with Xenon light through the light way of a microscope with an optical filter system (380 nm). The measurement of standard Ca 2+ solutions with concentrations between 0 to 70 µM using Fura 2 salt resulted for both methods in qualitatively equal courses of fluorescent intensity dependent on the Ca 2+ concentration. In further experiments, neonatal spontaneously active cells as well as stimulated cardiomyocytes were incubated with Fura 2 ester. The change of Ca2+ dependent fluorescence of the cells was measured and showed similar patterns while using both excitation methods. The study demonstrates that semiconductor lasers in combination with the light wave conductors can be used for the measurement of intracellular Ca2+ transients. The application for studying three dimensional objects is useful, since the excitation light can be positioned within the space. Johannes-Müller-Institute of Physiology, Charité, Humboldt University, Tucholskystr. 2, 10117 Berlin

Physiologisches Institut der LMU München, Pettenkoferstr. 12, 80336 München, Germany

P 16-12 MUSCARINIC ACETYLCHOLINE RECEPTORS ENHANCE THE TRANSMISSION AT GABAERGIC SYNAPSES IN THE INFERIOR COLLICULUS OF THE RAT M. Yigit, K.H. Backus The inferior colliculus (IC), is a major relay station in the auditory pathway of mammals. In the IC GABAA receptors (GABAARs) play an important role in the processing of auditory information. Muscarinic acetylcholine receptors (mAChRs) are also expressed in IC neurons, but their contribution to the processing of auditory information is still unknown. In order to investigate whether their activation could influence synaptic transmission we recorded spontaneous inhibitory postsynaptic currents (sIPSCs) in neurons of acutely isolated IC slices of rats (P4-P14) using the patch-clamp technique. The experiments were performed in the presence of kynurenate (1 mM) and strychnine (0.5 µM). The sIPSCs could be blocked with bicuculline (10 µM) and reversed close to the Cl--equilibrium potential (n=8) indicating that they were mediated by GABAARs. These GABAergic sIPSCs occurred at a frequency of 0.86 ± 0.26 Hz (mean ± SEM; n=28 cells) with a peak amplitude of –45.9 ± 3.3 pA. Muscarine increased the IPSC frequency in a dose-dependent manner. 10 µM muscarine significantly increased the mean sIPSC frequency to 4,9 ± 1.0 Hz (p < 0.001), but did not significantly change the peak amplitude -50.7 ± 3.9 pA (n=28; p=0.352). When tetrodotoxin (0.5 µM; n=7) was present, muscarine did not affect the mean sIPSCs frequency and amplitude suggesting that it mediates its effect by increasing the presynaptic spike activity. In the presence of 4-DAMP (1 µM), an antagonist prefering the M3 subtype of mAChRs, muscarine failed to increase the sIPSCs frequency (n=5). The M1 antagonists telenzepine (50 nM; n=6) and pirenzepine (n=5), and the M4 prefering antagonist himbacine (n=4) partially inhibited the muscarine-induced increase of the sIPSC frequency. The M2 antagonist methoctramine (n=5) showed only a very weak inhibition of the muscarine-evoked increase of sIPSC frequency and amplitude. We conclude that the GABAARs in the IC can be modulated by activation of the M3 subtype of mAChRs.

S 269 Institute of Physiology II, Cellular Neurophysiology, University of Frankfurt/M., Theodor-Stern-Kai 7, D-60590 Frankfurt/M., Germany P 17-1 INVESTIGATION OF NEUROTRANSMISSION IN VAS DEFERENS FROM α2A/D-ADRENOCEPTOR KNOCKOUT MICE L. Cleary, C. Vandeputte, J.R. Docherty We have investigated pre- and postjunctional responsiveness in vas deferens from wild-type and α2A/D-adrenoceptor knockout mice. The isometric contraction to 10Hz stimulation for 4 sec was significantly larger in vas deferens from knockout as compared with wild-type, but was reduced to a significantly greater degree in wildtype (to 29.2±4.6% of control, n=4) than knockout (to 68.5±9.5% of control, n=4) by the α1-adrenoceptor antagonist prazosin. The maximum potentiation of 10 Hz stimulation-evoked contractions by yohimbine was to 206.2±38.0% of control in wild-type but to 135.8±13.6% of control in knockout. The α2A/D-adrenoceptor selective antagonist BRL 44408 significantly increased the 10Hz stimulation-evoked contraction in wild-type but not knockout, and the reverse was true for the α 2c adrenoceptor selective antagonist spiroxatrine. The α2B-adrenoceptor antagonist imiloxan had no effect on the evoked contraction except at high concentrations, and only in wild-type. Following cocaine (3 µM) and BRL 44408 (1 µM) 10 Hz responses were similar in shape and maximum between wild-type and knock-out. The α2-adrenoceptor agonist xylazine virtually abolished the early component of the contraction to 10Hz stimulation in the presence of nifedipine (10 µM) in vas from knockout mice in a way consistent with a change of receptor subtype but without clear evidence for a reduced receptor number. In summary, although non-α2A/D- replace α2D-adrenoceptors in this knockout, antagonist data suggest a major loss of prejunctional α 2 adrenoceptors, but this is not supported by the agonist data. Department of Physiology, Royal College of Surgeons in Ireland, 123 St. Stephen’s Green, Dublin 2, Ireland

P 17-2 ALPHA1-ADRENORECEPTOR ACTIVATION FACILITATES INHIBITORY TRANSMITTER RELEASE DURING THE NORADRENERGIC MODULATION OF SWIMMING IN XENOPUS LAEVIS TADPOLES H. Fischer, S.D. Merrywest, K.T. Sillar In hatchling Xenopus tadpoles, the neuromodulator noradrenaline (NA) slows down the locomotor rhythm by affecting glycinergic inhibitory pathways within the spinal cord (McDearmid et al., 1997, J Physiol 503:111). The modulatory effects of NA on swimming are mediated via different adrenoreceptors (Fischer et al., 2001, Eur J Neurosci 13: 977). This study focuses upon a possible interaction between the α1 receptors and the particular spinal inhibitory pathways (glycine and GABA) during the noradrenergic modulation of the locomotor rhythm. Tadpoles immobilised in α-bungarotoxin can generate a self-sustained ”fictive” swim pattern. Ventral root activity during swimming was monitored extracellularly by suction electrodes positioned over the intermyotomal clefts and intracellular recordings from spinal motoneurons were made using KCl filled electrodes. The increased cycle period after activating α1-receptors by bath-applying the specific agonist phenylephrine, (PHEE, 150µM) was reversed by blocking glycine receptors (strychnine, 2µM, N=10 animals) but not by blocking GABAΑ receptors (by bicuculline, 50µM, N=10). Pre-application of strychnine (N=9), but not of bicuculline (N=9), prevented the effects of PHEE. Bath-application of PHEE (150µM, in presence of 0.2µM TTX and 50µM bicuculline) increased the frequency, but not the amplitude of spontaneous glycinergic IPSPs observed in intracellular recordings from motoneurons (N=5). In contrast, PHEE (150µM, under 0.2µM TTX and 2µM strychnine) did not affect the frequency or amplitude of spontaneous GABAergic IPSPs (N=5). This suggests that NA modulates swimming by interacting with presynaptic α1-receptors which in turn facilitate the release of the inhibitory transmitter glycine onto spinal motoneurons. Supported by The Wellcome Trust School of Biology, Department of Biomedical Sciences, Bute Medical Buildings, University of St Andrews, Fife, KY16 9TS, Scotland (UK)

P 17-3 POSTSYNAPTIC MODULATION OF SYNAPTIC TRANSMISSION VIA 5HAT-RECEPTOR SUBTYPES IN THE HIPPOCAMPUS OF MOUSE M. Heine, U. Bickmeyer, D.W. Richter Serotonin, an important transmitters in hippocampus formation acts through 5HT 1A , 5HT 7 and the 5HT 4 -receptors on CA1-pyramidal neurones. All these receptor subtypes are G-protein-coupled, the 5HT1Areceptor being negatively and the 5HT 7 and 5HT 4-receptors being positively coupled to an adenylyl cyclase isoform (Barnes, Sharp 1999). This raises the question why the same transmitter activates antagonistic intracellular pathways. For investigation of this aspect, we used the patch-clamp technique in the whole cell mode and extracellular bipolar stimulation of the Schaffer collateral pathway. Measuring I-V relationship clearly showed that the 5HT1A-receptor decreased and the 5-HT 4 and 5HT7-receptors increased the hyperpolarisation-activated current (Ih). BIMU8, 5HT4-receptor agonist, additionally increased a non-specific cationic conductance that is not blocked by CsCl (13mM), BaCl2 (2 mM) AP4 (2 mM) or TEA (10 mM). Stimulation of the Schaffer collateral pathway at frequencies between 10 and 30 Hz and local application of 50 µM Serotonin at the local dendritic tree induced a reduction of EPSP-amplitudes and an increase of their decay time constant (τ). However, after the end of drug application a longlasting amplification of the excitability of the CA1 neurons and the increase of τ remained. 50 µM 8-OH-DPAT mimicked this effect, but BIMU8 did not show a clear effect when applied to remote dendrites. BIMU8 (50 µM) applied locally on the soma of CA1 neurons clearly enhanced the EPSP amplitude in a frequency dependent manner. Application of 100 µM ZD7288, a specific blocker of the Ih, increased τ of single EPSPs. This enhanced the summation of EPSPs at frequencies between 10 and 30 Hz as already shown by Magee 1998. We conclude that the 5HT 1A-receptor induced decrease of I h is at least partly responsible for the long lasting increase in the excitability of the CA1neurons that is supported by an additional modulation of an nonspecific cationic conductance activated by the 5HT 4 -receptor. Imunhistochemical staining of the hippocampal region with a 5HT 7receptor specific antibody showed that the receptor subtype is preferentially expressed in the soma and proximal dendrites. This result supports our view that the 5HT 7-receptor adds to the effect of the 5HT 4 -receptor. References: Barnes, N.M., Sharp, T. (1999). Neuropharmacology 38, 1038; Magee, J.C. (1998). J. Neurosci. 18, 7613. Supported by the GRK: Organisation and Dynamic of neuronal networks and the SFB 406 Abt. Neuro- u. Sinnesphysiology, Zentrum Physiology, Humboltallee 23, 37073 Goettingen P 17-4 MONOAMINE OXIDASE INHIBITION IS ASSOCIATED WITH NEURONAL ACIDIFICATION D. Büsselberg, M. Wiemann, D. Bingmann, U. Bonnet Recently, we have demonstrated that moclobemide and clorgyline, both MAO-A inhibitors with a benzamide-structure, reduced intracellular pH (pHi) of CA3-neurones. Here we used further MAO-A and also MAO-B inhibitors to test the hypothesis that MAO-inhibitors more generally induce pHi-changes. Intracellular recordings were obtained from CA3 neurones (guinea pig, hippocampal slice). Microelectrodes were filled with 2M potassium-methylsulfate and slices were superfused with ACSF (4,5 ml/min, 32°C) equilibrated with 5% CO 2 in O 2 . Bicuculline (20 µM) or 4-aminopyridine (50 µM) were added to ACSF to induce epileptiform activity mostly consisting of periodic burst activity. To analyze changes in pHi, optical measurements were carried out on BCECF-AM loaded superficial neurones intermittently excited with 440 and 490 nm (sampling rate 0.05 Hz). For pHi-measurements, moclobemide (300-500µM), clorgyline (10µM), harman (200µM), selegiline (5-50µM), sulpiride (500µM) or idazoxan (500µM) were added to ACSF. The MAO-A inhibitors harman (n=5, 200µM) and norharman (n=3, 100µM) reversibly reduced pHi by up to 0.3 pHunits. Harman impaired pHi-regulation and/or acid-extrusion systems as was concluded from prolonged periods of pHi-recovery subsequent to a ammonium prepulses (n=3). Also, the MAO-B inhibitor selegiline (n=7, 5µM) reversibly reduced pHi. In contrast, sulpiride (n=4, 200µM), a benzamide without action on MAO, failed to affect pHi. As MAOenzymes are closely located to imidazoline I2 receptors of the outer mitochondrial membrane we furthermore tested the I2 receptor

S 270 modulator idazoxan (250µM, n=5), which, however, failed to change pHi too. We suggest that MAO-A and MAO-B inhibition is generally followed by a slight intracellular acidification of CA3-neurones. Taking into account that a modest (in contrast to an excessive!) pHi-decrease is sufficient to reduce the excitability of CA3-neurones, MAO inhibitors might act neuroprotective via a limited excitotoxicity and energy expenditure.

preceded by a rise in [Ca2+]i, but an elevation of aKe was not found. This indicates that most of the effects of 100 µM ATP in an intact peripheral nerve trunk do not involve activation of P2X receptors. In summary, our data indicate that changes in extracellular K+ activity can be used to detect activation of P2X receptors and to distinguish them from other purinergic receptors. Supported by the Deutsche Forschungsgemeinschaft

Institut für Physiologie, Universität Essen, D-45122 Essen

Department of Physiology, University of Munich, Pettenkoferstr. 12, 80336 Munich, Germany

P 17-5 P 17-7 FUNCTIONAL P2X7 RECEPTORS IN RAT RETINA S. Pfeiffer, B. Innocenti, T. Wheeler-Schilling, E. Guenther Purpose: The P2X7 receptor belongs to a family of cation channels which are activated by ATP. Different from other P2 receptors, prolonged activation of P2X7 can result in an additional opening of a membrane pore that is permeable to molecules up to 900 Da and has been shown to induce cell death in the immune system. We have previously reported that P2X7 receptors are expressed in the rat retina and located in the inner nuclear and ganglion cell layer (GCL). The goal of the present study was to investigate whether activation of P2X 7 in the inner retina induces similar events as in the immune system. Methods: The experiments were performed on intact retinas isolated from Brown Norway rats using fluorescence imaging. The specific agonist BzATP was used to activate P2X7 receptors. In parallel, the fluorescence dye YO-PRO-1 was applied, which as a large molecule (MW 375 Da), can enter into cells only by passing through large pores in the plasma membrane and subsequently bind to DNA or RNA. Results: After prolonged application (10 min) of BzATP together with YO-PRO-1 many cell nuclei were labeled. The fluorescence signal was detected in the GCL and in the nerve fiber layer, suggesting that different cell types were responding to P2X 7 stimulation. We succeeded in the identification of at least one cell type involved in P2X 7 stimulation. Retinal ganglion cell (RGC) were identified by fluorescent beads injected into the Superior Colliculus of living animals and clearly showed an uptake of YO-PRO-1 under stimulation of P2X7 with BzATP. However, RGC were only a small percentage of the retinal cell population that is responsive to P2X7 stimulation. Further investigations are necessary in order to clarify the identity of these other cell types. Conclusions: We clearly have shown that prolonged activation of P2X7 receptors can result in a permeabilization of retinal ganglion cells similar to what has been observed in the immune system. Whether this activation indeed induces RGC death has to be shown. However, it is attractive to speculate that P2X 7 receptor activation can be involved in the elimination of RGC during retinal development and may induce RGC death during retinal pathologies when ATP is released in excess. Supported by: EU IHP-Programme. Dept. of Experimental Ophthalmology, University Eye Hospital, Röntgenweg 11, 72076 Tübingen, Germany P 17-6 PURINERGIC RECEPTORS IN PERIPHERAL NERVE TRUNK CAN BE DISTINGUISHED BY CHANGES IN EXTRACELLULAR K+ ACTIVITY D.J. Tracey, P. Lang, O. Windmueller, P. Grafe P2X receptors are non-selective cation channels and contribute to the transduction of sensory, including nociceptive, stimuli. In a complex neural tissue, such as the peripheral nerve trunk, identification of P2X activation by the natural agonist ATP is difficult because ATP also acts on other nucleotide receptors such as P2Y and adenosine A2 receptors present on various cellular elements. Activation of all these purinergic receptors increases axonal excitability. In the present study, we tested whether receptors for ATP can be distinguished by effects on the extracellular K+ activity. Experiments were performed on segments of isolated rat vagus nerves. The experimental set-up enabled simultaneous recording of axonal excitability, intracellular Ca 2+ ([Ca2+]i) and extracellular K+ activity (aKe). Agonists at P2X receptors (α,β-meATP 20 µM; Ap5A 50 µM) produced a transient rise in aK e with a peak amplitude of about 250 µM above bath K + concentration (3 mM). Agonists at metabotropic ATP receptors (ADP, UTP, 2-MeSADP 50 – 100 µM) did not change aKe although a clear increase in [Ca2+]i was observed. Adenosine, too, did not influence aKe. The increase in axonal excitability seen during bath application of ATP (100 µM) was

ADENOSINE, ACTING AT A1 RECEPTORS, SUPPRESSES SEIZURE-LIKE EVENTS IN SLICES FROM RAT ENTORHINAL CORTEX E. Avsar, R.M. Empson The entorhinal cortex is a highly seizure-prone region of the limbic system. Here we demonstrate that adenosine, a neuromodulator that is released during seizures, acts at A1 receptors to suppress seizure-like events within the entorhinal cortex. Horizontal combined hippocampal entorhinal cortex slices (500 µm) were prepared and wedges of the entorhinal cortex (2-3 mm wide) were transferred to a twocompartment chamber. This was continuously perfused (2-2.5 ml/ min) with a solution containing in mM, NaCl 135, KCl 3, NaH2PO4 1.25, MgCl2 1, CaCl2 2, glucose 10 & NaHCO3 26, pre-equilibrated with 95%O2/5%CO2 at room temperature. Removal of Mg2+ lead to the appearance of repetitive depolarising seizure-like events that were displayed on a chart recorder. All drugs were applied via the perfusate and mean changes in the frequency and duration of the seizure-like events were measured before and after drug application. NECA a metabolism-resistant form of adenosine, (applied at concentrations ranging from 0.1 µM to 30 µM), suppressed the seizure-like events, reducing both their frequency and duration in a concentration dependent manner. 30 µM NECA reduced the frequency and duration of the events by 62 ± 8 % and 34 ± 6 % (n=9) respectively. This was also the case for the adenosine A1 receptor agonist, CPA, (concentration range 0.1 µM to 30 µM) where 10 µM reduced the frequency by 63 ± 9 % of the control value. Overall CPA had little effect upon the duration of the seizure-like events although in 5/11 slices we observed an increase in the duration of the events. Consistent with these findings, application of the adenosine A1 receptor antagonist, DPCPX, increased the frequency of the seizure-like events in a rather dramatic manner, exerting its effects at concentrations as low as 3 nM (concentration range tested 0.3 nM to 1 µM, n=18). At 100 nM, DPCPX increased the frequency of the seizure-like events by 540 ± 82 %. Moreover, this was accompanied by a conversion of the pattern of the epileptiform activity from the large, discrete seizure-like events into a fast and often continuous pattern of epileptiform activity. Our result strongly suggests that in the entorhinal cortex, adenosine suppresses seizurelike activity in the low Mg2+ model of epilepsy through an action at adenosine A1 receptors. Our results with the A1 receptor antagonist also suggest that endogenous adenosine is released during the seizurelike events and that this influences their shape and pattern. We acknowledge the support of the Epilepsy Research Foundation, UK. School of Biological Sciences, Royal Holloway University of London, EGHAM, Surrey, TW20 0EX P 17-8 BDNF-EVOKED CALCIUM TRANSIENTS IN CULTURED ASTROCYTES B. Pichler*, C.R. Rose, K.W. Kafitz, A. Konnerth Neurotrophins regulate the survival, differentiation and maintenance of function of different neuronal populations and play an essential role in the modulation of activity-dependent neural plasticity. Furthermore, it was recently shown that neurotrophins excite central neurons as rapidly as the neurotransmitter glutamate. Growing evidence suggests that neurotrophins may exert important effects not only on neurons, but also on glial cells. Here, we analysed the effect of neurotrophin application on the intracellular calcium concentration in cultured astrocytes. Primary cell cultures of rat astrocytes were prepared from the hippocampi of newborn rats. Calcium measurements were performed on cells from 9-15 days in vitro using fluorescence ratio imaging with the Ca 2+ -sensitive dye Fura2-AM.

S 271 Neurotrophins (NTs) were applied focally to single cells for 10-150 ms by a picospritzer coupled to standard micropipettes. Application of brain-derived neurotrophic factor (BDNF, ≤50 ng/ml) reliably induced Ca2+ transients in cultured astrocytes. The BDNF-evoked Ca2+ signals persisted in Ca2+-free saline and were blocked following store depletion by the SERCA-inhibitors CPA or Thapsigargin. This strongly suggests that they are mediated by calcium release from internal stores. Neurotrophin-4/5 (NT-4/5, ≤50 ng/ml) evoked similar calcium transients as BDNF, whereas both Nerve Growth Factor (NGF) and Neurotrophin-3 (NT-3) were ineffective even at a threefold higher concentration. This agonist profile indicates that the BDNF-induced calcium signals are mediated by TrkB-receptors. Our results demonstrate that BDNF evokes calcium transients in cultured astrocytes. Furthermore, our results suggest that these calcium signals are mediated by release of calcium from intracellular stores following activation of TrkB-receptors. Thus, the mechanisms of BDNF-evoked calcium transients in glial cells differ significantly from those observed in neurons (see also Abstract by Rose et al.). Physiologisches Institut, Ludwig-Maximilians-Universität München, Biedersteiner Str. 29, 80802 München

vitro and in vivo. Human and rat VR1 receptors were stably expressed in CHO or HEK cells, respectively. Heat or ligand induced currents were measured in the voltage clamp patch clamp configuration at -60 mV. Although I-RTX differs by only one iodine from the parent agonist resiniferatoxin, no current was induced when tested at 1 µM between -80 and +80 mV. On rat VR1, I-RTX (IC 50 0.69 nM) more potently blocked capsaicin induced currents than capsazepine (IC 50 562 nM), and was 10-fold more potent on rat than on human VR1. Proton induced currents in rat VR1 were reduced by capsazepine to 35% of the control value but fully blocked in human VR1. I-RTX also abolished proton and heat activated currents in human and rat VR1. Thus, I-RTX was more effective than capsazepine in vitro. However, only capsazepine reduced capsaicin-induced paw flinching behaviour in rats in that I-RTX (0.001-1 nmol/paw) was ineffective. Thus IRTX, whilst having limited utility for in vivo studies, is a valuable tool to study the functional properties of cloned and native VR1 receptors in vitro. Merck Sharp & Dohme Research Laboratories, Neuroscience Research Centre, Terlings Park, Eastwick Road, Harlow, Essex, CM20 2QR, UK P 17-11

P 17-9 BDNF REGULATION OF NMDA RECEPTOR FUNCTION IN THE RAT RETINA T. Ladewig, S. Fellner, E. Guenther* OBJECTIVE: Growth factors have been implicated in the regulation of neuronal survival and differentiation during development. However in recent years the idea of neurotrophic factors modulating synaptic transmission has gained increasing interest (Sherwood & Lo, 1999). In order to assess the effect of brain-derived neurotrophic factor (BDNF) on glutamate receptors of the NMDA type, reported to be involved in synaptic transmission and plasticity, we have performed patch-clamp recordings on retinal gangion cells in wildtype and BDNF knockout mice. METHODS: BDNF -/- mutants were obtained by cross breeding heterozygous mice from a well characterized line of BDNF null-mutant mice (Korte et al., 1995). The animals were genotyped by performing PCR on genomic DNA. Retinas from wildtype, BDNF +/- and -/- mice were examined at postnatal days P3 to P12. Electrophysiological responses to the externally applied glutamate agonist NMDA (in concentrations from 500 µM up to 2 mM) of ganglion cells in a retinal slice preparation were recorded using the whole-cell patch-clamp technique, while AMPA and GABA receptor activity were blocked by the selective antagonists CNQX and bicuculline, respectively. RESULTS: We found that peak amplitudes of NMDA-evoked currents in -/- mutant retinal ganglion cells at postnatal days P3 - P12 in vitro were significantly decreased compared to that of wildtype mice at the same age. Heterozygous mutant mice showed an intermediate reduction of NMDA responses. A restoration of the decreased NMDA receptor activity by a 1 to 3 hour incubation of -/- mice retinal slices in BDNF (100 ng/ml) was not observed. CONCLUSIONS: BDNF seems to be responsible for the expression of functional NMDA channels in retinal ganglion cells. Our results suggest that BDNF has long-term (days to weeks) but not short-term (minutes to hours) effects on excitatory synaptic transmission in retinal ganglion cells. The mechanism of BDNF regulation of NMDA receptor activity remains to be determined. Supported by Grimmke-Foundation and German Research Council (SFB 430). * Department of Experimental Ophthalmology, University Eye Hospital, Roentgenweg 11, 72076 Tuebingen, Germany P 17-10 CHARACTERIZATION OF THE POTENT VR1 ANTAGONIST IODO-RESINIFERATOXIN IN VITRO AND IN VIVO W. Jarolimek*, K.G. Sutton, S. P. Tleague, G. J. Hollingworth, J. Webb, S. Boyce, J. Kerby, A. B. Jones, G.R. Seabrook Vanilloid receptors are sensory neurone specific cation channels involved in the sensation of pain. Selective and high affinity VR1 receptor antagonists will be valuable to help further characterize the physiological role(s) of VR1 receptors. We have synthesized one of the most potent VR1 receptor antagonists described to date, iodoresiniferatoxin (I-RTX; Wahl et al., 2001 Mol. Pharmacol. 59, 9), and evaluated its properties on heterologously expressed VR1 receptors in

DIFFERENTIAL DISTRIBUTION AND TRANSCRIPTIONAL CONTROL OF nNOS VARIANTS IN THE GASTROINTESTINAL TRACT D. Saur, B. Seidler, H.D. Allescher Neuronal nitric oxide synthase (nNOS) is the predominant isoform of NOS in the enteric nervous system. Several different human nNOS mRNA transcripts containing either different first exons derived from alternative promoters or exon deletions/insertions generated by alternative splicing have been described recently. They encode for full length nNOSα or the N-terminally truncated nNOSβ and nNOSγ proteins. nNOSα, but not nNOSβ/γ contains a N-terminal PDZ/GLGF motif for interaction with regulatory proteins. Aims: To study the expression and transcriptional regulation of nNOS variants in the gastrointestinal tract. Methods: Expression of nNOS variants was analyzed by quantitative ”real time” RT-PCR, Western blotting and immunohistochemistry. nNOSα-, β- and γ-cDNAs were expressed by transfection of COS7 and HeLa cells and protein dimerization was analyzed by low temperature SDS-PAGE. nNOS promoter gene constructs of the 5´flanking region of exon 1c, 1f , 1g and the triple promoter complex of exon 1a/1b/1c were transfected in to HeLa and TGW-nu-I cells. Luciferase activities were measured after incubation of the transfected cells with an active or an inactive phorbol ester. Results: We observed a site specific expression of nNOS mRNA variants in distinct functional regions of the human gastrointestinal tract with nNOS exon 1c, 1f and 1g being the predominant forms. Immunohistochemistry showed a differential distribution of C- and Nterminal nNOS immunoreactivity in distinct layers of the duodenum. Expression of nNOS proteins revealed bands of ~160 kDa for nNOSαcDNA, 135 and 125 kDa for nNOSβ-cDNA and 125 kDa for nNOSγcDNA in the Western blot. Expressed nNOSα and the 135 kDa protein generated by nNOSβ-cDNA formed SDS resistant dimer, whereas the 125 kDa protein of nNOSβ-cDNA and expressed nNOSγ was not SDS resistant. The promoter activity of exon 1f was stimulated by PMA (+46%, n=3, p<0,001), whereas the triple promoter complex of exon 1a/b/c was downregulated (-21%, n=3, p<0,001) in HeLa cells. In TGW-nu-I cells promoter activities of exon 1a/b/c (+139%, n=3, p=0,01), exon 1c (+171%, n=3, p<0,01) and exon 1f (+61%, n=3, p=0,001) were stimulated. Conclusions: The observed differential distribution of nNOS mRNA and protein variants argue for a complex control of nNOS expression by usage of distinct promoters, cell and site specific splicing mechanisms and translational initiation. Dieter Saur, Technische Universität München, 2. Medizinische Klinik, Ismaningerstr. 22, 81675 München, Germany P 17-12 SENSITIVITY OF SYNAPTIC PLASTICITY TO NITRIC OXIDE SYNTHASE INHIBITORS GREATLY VARIES BETWEEN 3 DIFFERENT RAT STRAINS C. Hölscher Previous research has produced great inconsistencies in results of studies of the dependence of synaptic plastic processes such as long-term

S 272 potentiation of synaptic transmission (LTP) to nitric oxide synthase (NOS) inhibitors. The present work demonstrates that the efficacy of inhibitors of NO synthase to block LTP varies greatly between rat strains. Male Sprague-Dawley, Wistar, and Long-Evans rats (220-300 gr) were implanted with electrodes in the hippocampus area CA1 under urethane anaesthesia. High frequency stimulation was applied (10 trains of 10 stimuli at 200Hz) to Schaffer-collaterals, and the field excitatory postsynaptic potentials (EPSPs) were measured in area CA1. Injection of the NO synthase inhibitor 7-nitro-indazole (30mg/kg i.p.) greatly impaired the development of LTP in Wistar rats (118± 10% of slope of baseline EPSP) but not in controls that received vehicle injections (148± 9 % of slope of baseline, n=6 per group, p<0.01, Wilcoxon test). In contrast, Sprague-Dawley rats showed virtually no reduction of LTP after administration of 7-NI (141± 8 % of slope of baseline EPSP) compared to controls (138± 7 % of slope of baseline, n=6 per group). Furthermore, Long-Evans rats did not exhibit any impairment of LTP after 7-NI application: 144± 11 % of slope of baseline EPSP, controls that received vehicle injections (138± 8 % of slope of baseline, n=5 per group) Since NOS is an enzyme that is highly regulated in activity and in expression it is feasible that the actual rate of baseline expression/activity varies greatly between strains. The results presented here can explain to some degree why results of studies of the effects of nitric oxide on synaptic plasticity that have been published previously are highly inconsistent. Dept cog Neurosci, University of Tübingen, Auf der Morgenstelle 28, 72074 Tübingen, Germany P18-1 A MICROSTRUCTURED GLASS CHIP FOR WHOLE-CELL RECORDINGS FROM MAMMALIAN CELL LINES N. Fertig, R.H. Blick, J.C. Behrends Electrophysiology (i.e. patch clamping) remains the gold standard for pharmacological testing of putative ion channel active drugs (ICADs). However, patch clamping has low throughput, and requires highly trained personnel. We present a new ion channel recording device in which a microstructured chip electrode provides the aperture over which current is recorded. Advanced microstructuring techniques are applied to produce patch clamp apertures in planar glass or quartz substrates, which replace the patch pipette in the electrophysiology recording setup. On-chip whole-cell voltage clamp recordings from mammalian cell lines (N1E-115 neuroblastoma, CHO transfected with BK channels) are taken, recording K+, Ca2+, and Na+ currents. The low access resistance and low capacitance of the chip device enable good voltage clamping. The quality of the data obtained with the patch clamp chip is comparable to standard patch clamp recordings. As such a planar electrode design can be scaled up to form arrays with many apertures, it holds potential as a platform for automated, rapid and parallel screening of ion channel active drugs. In addition, the planar geometry of the chip electrode make this approach very attractive for optical/spectroscopical experiments and the application of scanning probe techniques with concomitant electrical recordings. Supported by the DFG, SFB 486, and the BMBF. Center for Nanoscience der LMU & Physiologisches Institut der Universität München, Pettenkoferstr. 12, 80336 München

oscillations (Braun et al., Int J Bifurc & Chaos 8: 881-889, 1998). With addition of noise it successfully mimicks the experimentally observed impulse patterns and their stimulus dependent modifications including the occurrence of UPOs. Moreover, simulation runs without noise clearly exhibited a broad range of deterministic chaos at tonicto-bursting bifurcations (W Braun et al., Phys Rev E 62: 6352-6360, 2000; Feudel et al., Chaos 10: 231-239). These chaotic dynamics can be seen as the result of ”critical” interactions between two oscillatory subsystems, i.e. the subthreshold oscillations and the spike-generating processes (Braun et al. Neurocomputing 32: 51-59, 2000). However, in ”noisy” simulations the unstable regime seems to be considerably extended towards the deterministically regular spiking range. Indeed, noise effects are extremely blown up in tonically firing modes close to the bifurcations. Also, ”stochastic synchronisation”, as an indication of stable limit cycles, was only obtained far away from the bifurcations. These data indicate the occurrence of unstable dynamics which are already detectable in the tonically firing mode in anticipation of burst discharges. Supported by the Volkswagen Stiftung. Institute of Physiology, Deutschhausstr. 2, D-35037 Marburg, Germany P 18-3 INTRINSIC MEMBRANE POTENTIAL OSCILLATIONS AND RESONANCE BEHAVIOUR OF RATS NEOCORTICAL NEURONS IN VITRO C. Biber*, B. Sutor The subthreshold integrative properties of central nervous system neurons is strongly influenced by voltage- and time-dependent intrinsic membrane conductances. It is known that such conductances may lead to spontaneous membrane potential oscillations, particularly near spike threshold, and to a frequency-dependent amplification of intracellulary injected alternating current stimuli revealing a phenomenon called membrane resonance. To investigate the ionic conductances underlying membrane potential oscillations and resonance behaviour, we performed intracellular recordings from rat neocortical neurons in vitro. Recordings were taken at different levels of membrane potentials before and during the intracellular injection of wobbled alternating current stimuli. Near spike threshold we observed spontaneous voltage oscillations with preferential frequencies between 2 and 15 Hz. Fourier Analysis of the voltage responses to injected a.c. stimuli revealed a membrane resonance with a maximum between 1 and 5 Hz. Application of a cocktail consisting of glutamate receptor and GABA receptor antagonists had no effect on spontaneous oscillations and resonance behaviour excluding a synaptic contribution to these phenomena. Spontaneous oscillations were suppressed by TTX, but not by a variety of potassium channel blockers indicating that these oscillations are generated by the persistent sodium current INaP. TTX reduced the magnitude, but did not prevent the resonance response of the neurons. Application of the I h channel blockers Cs + or ZD7288 led to the dissappearence of the membrane resonance at every membrane potential. In contrast, the IIRK channel antagonist Ba2+ (l00µM) increased the resonance magnitude resulting in a clearly detectable membrane resonance at resting membrane potential. GABAB - and Al- receptor activated GIRK channels did not contribute to the membrane resonance. These results indicate that the resonance behaviour of rat neocortical neurons is determined by an interplay between the activation of I h and I IRK channels. (Supported by the DFG, SFB 462, TPA4)

P 18-2 INDICATIONS OF ”CHAOTIC” DYNAMICS AT TONIC-TOBURSTING TRANSITIONS IN NEURONAL RECORDINGS AND HH-TYPE COMPUTER SIMULATIONS H.A. Braun, M.T. Huber, K. Voigt Tonic-to-bursting transitions in neuronal recordings often exhibit transient occurrences of rather irregular impulse patterns. Recently developed methods for the detection of unstable periodic orbits (UPOs) revealed that the irregular firing is not only caused by stochastic components (”noise”, high-dimensional dynamics) but also seems to reflect ”deterministic chaos” (low-dimensional dynamics) (e.g. Braun et al. J Comp Neurosci 7: 17-32, 1999). For a better understanding of the underlying dynamics we have analyzed ”noisy” and ”deterministic” computer simulations of a Hodgkin-Huxley (HH) type model in comparison to experimental data from peripheral sensory receptors and hypothalamic neurons (rat brain slices). The model consists of a minimal set of ionic conductances for spike-generation and subthreshold

Physiologisches Institut LMU München, Pettenkoferstr. 12, 80336 München P 18-4 CORRECTING CONDUCTANCE MEASUREMENTS IN NONSPACE-CLAMPED STRUCTURES A.T. Schaefer1, M. Helmstaedter1, B.Sakmann1, A. Korngreen1,2 To understand functions of a single neuron, such as propagation and generation of synaptic or action potentials, it is imperative to obtain a detailed description of the kinetics and distribution of the underlying ionic conductances. Incomplete space clamp in voltage-clamp experiments distorts the recorded currents, rendering accurate analysis impossible. Here we present a simple numerical algorithm that corrects these distortions. The method is based on a stepwise approximation of the conductance density at the site of a local voltage clamp; this is

S 273 achieved by estimating membrane conductances in a simulation yielding simulated clamp currents, which are then fitted to the recorded spaceclamp-distorted ones. The algorithm was tested both on artificial channels with varying activation parameters and on a set of realistic potassium-channel models taken from the literature that were simulated in a model of either an unbranched cable or in a reconstruction of a neocortical pyramidal neuron. We show that for all channels tested, the method enables accurate retrieval of local somatic and dendritic channel densities and kinetics at the site of the voltage clamp, rather than resulting in an average of these parameters across the structure. In addition, conductance gradients could be calculated correctly requiring only a small number of recordings. Neither the addition of noise nor variation of passive parameters significantly reduced the performance of the correction algorithm. The generality and robustness of the algorithm make it a useful tool for voltage-clamp analysis of voltagegated currents in any morphological structure. 1

Max-Planck-Institut für Medizinische Forschung, Abt. Zellphysiologie, Jahnstraße 29, D-69120 Heidelberg; 2Faculty of life sciences, Bar-Ilan University, Ramat-Gan 52900, Israel

which was not explicable by variable protein expression. Both parameters had faster time constants than expected. Activation time constants were identical, whereas recovery from desensitization showed a tendency towards slower time constants with increasing content of GluR1 cDNA. Coexpression of GluR2IRN and the kainate receptor subunit GluR6 produced biexponential time-courses of recovery from desensitization suggesting the formation of separate homomeric kainate and AMPAR channels. Conclusion: The kinetic of an AMPA receptor is composed of the AMPA subunit kinetics with a predominance of faster deactivation and desensitization. This could reflect a slight subunit interaction during the gating-process of the channels, which has been proposed using slow agonist application on GluR3/GluR6-chimeras with concanavalin and cyclothiazid added to the solution (Ayalon et al., Neuron 6/2001). Supported by the Deutsche Forschungsgemeinschaft. Dr. J. Grosskreutz, Dept. of Neurology, Medizinische Hochschule Hannover, Carl-Neuberg-Str. 1, 30625 Hannover, Germany P 19-1

P 18-5 RETRIEVAL OF COMPLETE HODGKIN-HUXLEY CHANNEL KINETICS IN NON-SPACE-CLAMPED STRUCTURES A.T. Schaefer1, M. Helmstaedter1, A. Korngreen1,2 Classical quantitative analysis of channel densities and kinetics with voltage-clamp protocols requires voltage control over the entire measured structure, so-called space clamp3. We have implemented recently a computer algorithm in NEURON that cirumvents this principal experimental requirement in realistic morphologies, allowing correct retrieval of potassium-channel densities and activation kinetics. Here, we present subsequent algorithms permitting correction of other voltage protocols, which then aid to retrieve correctly channel inactivation states, kinetics, and distributions for other ion species. The algorithm is tested for various potassium and calcium conductances. In order to account for more complex spatial and temporal voltage profiles that occur in typical voltage protocols used for analysis of inactivation states, we implemented a stepwise correction approach that exploits previously achieved correction results. This yields the following multi-step algorithm for retrieval of complete Hodgkin-Huxley kinetics: Correct for 1. kinetics at depolarized voltages 2. steady-state inactivation 3. inactivation kinetics, recovery from inactivation 4. activation kinetics (tail currents). The algorithm successfully corrects both synthetical and realistic neocortical potassium conductances without any prior knowledge of channel parameters, yielding errors in channel parameters of approximately 5-10 %. Thus, the devised correction approach may be a useful tool for voltage-clamp analysis of complete channel kinetics of various ion species in non-space-clamped morphologies. 3Hodgkin, A.L. and A.F. Huxley 1952, A quantitative description of membrane current and its application to conductions and excitation in nerve, J Physiol 117:500-544. 1

Max-Planck-Institut für Medizinische Forschung, Abt. Zellphysiologie, Jahnstraße 29, D-69120 Heidelberg, Germany; 2Faculty of life sciences, Bar-Ilan University, Ramat-Gan 52900, Israel P 18-6 KINETICS OF HETEROMERIC GLUR1/GLUR2 AMPA RECEPTORS: EFFECTS OF FRACTIONED COASSEMBLY J. Grosskreutz, A. Zoerner, K. Krampfl, R. Dengler, J. Bufler Recent evidence suggests that AMPA-type glutamate receptors (AMPAR) are composed of four subunits contributing individually and independently to the desensitization and ion permeation of the functional receptor (Mansour et al, Neuron 12/2001). Objectives: To determine the dependence of activation, deactivation, desensitization and recovery from desensitization on subunit composition as induced by fractional cDNA tranfection. Methods: Recombinant AMPAR subunits (GluR) transiently expressed in HEK 293 cells were studied with Patch-Clamp-techniques using ultrafast agonist application. GluR2 flip R-edited (Glu2IR) subunits were coexpressed with the GluR1 flip (GluR1I) isoform at cDNA ratios of 6:1, 3:1, 1:1, 1:3 and 1:6, or with the kainate GluR6 subunit at a ratio of 10:1. Results: In GluR1I/ GluR2IRN channels, time constants of deactivation and desensitization displayed a non-linear dependence on the amount of cDNA transfected

GLUCOCORTICOIDS DIFFERENTIALLY MODULATE BK CHANNEL SPLICE VARIANTS L. Tian, M.S.L. Hammond, H. Florance, F.A. Antoni, M.J. Shipston Adrenal glucocorticoid hormones exert rapid (<3h) effects on cellular excitability in neuroendocrine cells and neurons by regulation of ion channels through poorly defined genomic mechanisms. Largeconductance calcium- and voltage-activated potassium (BK) channels are important regulators of cellular excitability and are targets for regulation by glucocorticoid hormones. In this study we demonstrate that the protein kinase A (PKA) dependent regulation of BK channel splice variants is differentially modulated by glucocorticoids. BK channel splice variants (STREX and ZERO) were expressed in HEK293 cells and channel regulation was examined using patch clamp electrophysiology. Application of 1 mM cAMP to the intracellular face of isolated inside-out patches reduced activity of STREX BK (125 pS) channels but increased activity of ZERO BK channels under the same recording conditions. These effects were through activation of PKA closely associated with the channels as inhibition of STREX, and activation of ZERO, channels was not observed in the presence of the specific PKA inhibitor peptide, PKI(5-24). When the cells were pretreated with 1 µM of the synthetic glucocorticoid, dexamethasone (DEX, 2 h), PKA-dependent inhibition of STREX BK channels was significantly blocked. This action was through a mRNA and protein synthesis dependent mechanism. In contrast, PKA-dependent activation of ZERO channels remained unchanged upon DEX pretreatment, suggesting that the STREX insert is required for glucocortcoid regulation. We also demonstrate that the action of glucocortcoid was dependent on protein phosphatase 2A-like activity intimately associated with STREX channels as DEX had no effect when ATP-γS was used as phosphate donor, or when 10 nM okadaic acid, but not the specific protein phosphatase 1 inhibitor peptide PPI2, was used to inhibit phosphatase activity in isolated patches. Taken together these data suggest a common genomic mechanism for glucocorticoid regulation of cellular excitability through regulation of protein phosphatase activity associated with STREX BK channels. This work is supported by grants from The Wellcome Trust (046787) and the Biotechnology and Biological Sciences Research Council (15/ C11774). LT: Membrane Biology Group and MRC ‘Membrane and adapter protein’ COOP, University of Edinburgh Medical School, Hugh Robson Building, Edinburgh EH8 9XD P 19-2 CLONING, ALTERNATIVE SPLICING AND ELECTROPHYSIOLOGICAL CHARACTERIZATION OF A PERIPLANETA AMERICANA LARGE CONDUCTANCE CA 2+-ACTIVATED POTASSIUM CHANNEL (PSLO) C. Derst, S. Messutat, C. Walther, D. Wicher A 5949 bp cDNA was isolated from Periplaneta americana encoding a 1124 amino acid homolog of the Drosophila large conductance calcium-activated (dSlo) potassium channel. The isolated channel protein named pSlo was 56 % identical to the Drosophila ortholog and showed all typical motifs of slo channels, including one pore region,

S 274 one S0 transmembrane domain, six further transmembrane regions and a large C-terminal region with several putative Ca2+ binding sites. RTPCR analysis of isolated ganglia and muscle cells of Periplaneta showed a differential pattern of alternative splicing of five exons in the Cterminal part of the protein. The splice sites are highly conserved since they are identical to those in Drosophila. Heterologous expression of pSlo channels for electrophysiological characterization was demonstrated in HEK293 cells by means of whole cell recordings using a pipette solution with 100 micromolar free [Ca2+] and 140 mM [K+] and a bath solution with 1 mM [Ca2+] and 5 mM [K+]. Cells lacking green fluorescent protein - which was used as a transfection marker exhibited only a small (~200 pA) outward current component at V=+60 mV on application of a ramp command (-70 to +60 mV; rate: 0.27 mV/ ms) which did not considerably grow within 3 min after breaking into the cell. However, fluorescent cells exhibited an outward component which was 800 to 2500 pA (at +60 mV) a few seconds after breaking into the cell and then grew within 1 min by 25 - 200 %, reflecting the rise of free intracellular [Ca2+].

(Oliver et. al.: Neuron 2000, 26(3):595-601) we developed a strategy for identification of additional proteins that are directly or indirectly associated with the intracellular domains of these channels. The Cterminal cytoplasmic domains of rat SK2 and SK3 were expressed in E. coli as His-tagged fusion proteins, purified by Ni-NTA agarose chromatography and preparative SDS polyacrylamide gel electrophoresis, and covalently coupled to CNBr-activated sepharose. Control proteins were coupled equivalently. Immobilized proteins were allowed to refold, and the beads were then incubated with cytosolic rat brain extracts. After washing extensively, the bound proteins were eluted, separated by two-dimensional gel electrophoresis and visualized by silver staining. Spots containing proteins significantly more abundant in SK eluates than in controls were excised and identified by mass spectrometry and database searches. Selected candidates are further characterized in biochemical binding assays and functionally studied in coexpression experiments. University of Freiburg, Physiology II, Hermann-Herder-Straße 7, D79104 Freiburg, Germany

Institut für Physiologie, Deutschhausstr. 2, 35037 Marburg; Sächsische Akademie der Wissenschaften, Erbertstr. 1, 07743 Jena, Germany P 19-5 P 19-3 ENDOCYTOSIS OF CALCIUM-SENSITIVE POTASSIUM CHANNELS (hIK1) IN MIGRATING CELLS M. Römer, B. Trost, B. Hinkes, W. Kessler, B. Gassner, A. Wulf, B. Heppelmann, A. Schwab Cell migration is important for immune defense, wound healing or formation of tumor metastases. It relies on the coordinated remodeling of the cytoskeleton and activity of ion channels like Ca2+ sensitive K+ channels (IK1). Migration is accompanied by rapid endo- and exocytotic recycling of plasma membrane. Thus, the entire plasma membrane of migrating ameoba recycles within 10 min. However, it is unknown whether membrane proteins like ion channels are also recycled with the plasma membrane in migrating cells. We addressed this question for the human isoform of Ca2+ sensitive K + channels hIK1 in migrating epithelial (MDCK-F) cells. Here, we focused on endocytosis of hIK1 channels as the first step in channel recycling. Electron microscopy revealed the presence of many ”omega”-shaped structures typical for endo- or exocytotic events in the plasma membrane of MDCK-F cells. We introduced a HA-tag into an extracellular domain of hIK1 channels in order to allow their in vivo-labeling with anti-HA antibodies in immunofluorescence or ELISA experiments. Living cells transfected with this channel construct were incubated with anti-HA antibodies for 30 or 60 min at 37°C. Immunofluorescence revealed the typical channel distribution with concentration of the hIK1 protein in the ruffled membrane at the front of migrating cells. In addition, vesicle-like structures were seen in the cytoplasm. ER and/or Golgi apparatus were not stained. In contrast, vesicle-like structures were never observed when cells were incubated for up to 16 h at 4°C. Moreover, untransfected cells or cells transfected with hIK1 channels without extracellular HAtag were not stained. We quantified uptake of anti-HA antibodies bound to hIK1 channel proteins with direct-ELISA technique. Cells transfected with extracellularly labeled hIK1 channels take up anti-HA antibodies in a time-dependent way. This is not the case for untransfected wildtype cells. Taken together, our results are consistent with the endocytotic internalization of hIK1 channels in migrating cells. Physiologisches Institut, Röntgenring 9, 97070 Würzburg, Germany

PROTEIN KINASE A REGULATION OF BK CHANNEL SPLICE VARIANTS L.S. Coghill, H. Florance, F.A. Antoni, M.J. Shipston Large conductance calcium- and voltage- activated potassium (BK) channels are important determinants of endocrine, vascular and neuronal cell excitability. Despite being the product of only one gene (KCNMA1), native BK channels display diverse functional properties. Such diversity results from alternative splicing of the nuclear premRNA, and is exemplified by the different response of two major BK channel splice variants to protein kinase A (PKA). Channels expressing a unique 59 residue, cysteine-rich insert (STREX) located within the channel’s intracellular C-terminal domain are inhibited by PKA, whereas channels lacking this insert (ZERO) are activated. The aim of this study was to investigate the interaction of PKA with STREX BK channels. Hemagglutinin (HA)-tagged STREX BK channel splice variant was co-expressed with PKA catalytic subunit (PKAc) in HEK293 cells, and immunoprecipitated (IP) with anti-HA or anti-PKAc antibodies. Final IPs were analysed by SDS-PAGE, blotted, and probed for HA or PKAc. IP with anti-HA resulted in co-IP of PKAc and, reciprocally, IP with anti-PKAc resulted in co-IP of HA-tagged STREX channels suggesting that PKAc and the STREX channel exist as a complex. To investigate whether the STREX insert is a possible PKAc binding site, the 59 residue STREX insert was expressed as a GSTfusion protein (GST-STREX) in E.coli and purified by glutathione sepharose (GS4B) chromatography. PKAc did not directly interact with GST-STREX in an in vitro pull-down assay. To investigate whether PKAc can interact with GST-STREX as part of a larger complex pulldown assays were performed using rat brain lysate or HEK293 cells as sources of endogenous PKAc. However, although PKAc was not found to associate with the GST-STREX fusion protein, PKAc phosphorylated GST-STREX in an in vitro phosphorylation assay. In vitro phosphorylation of GST-STREX was completely abolished on mutation of a single serine residue (S4A) within a putative PKAconsensus site within STREX. These data suggest that PKAc associates with the STREX channel as observed by IP. However, although the STREX insert is a target for phosphorylation by PKAc, affinity pulldown assays indicate that the STREX insert is not required for interaction of PKAc with the channel. This work is supported by grants from the BBSRC to MJS. LSC is supported by a Crichton Scholarship from the University of Edinburgh Medical School.

P 19-4 PROTEOMIC APPROACH FOR IDENTIFICATION OF PROTEINS ASSOCIATED WITH SMALL CONDUCTANCE CA2+ACTIVATED POTASSIUM CHANNELS W. Bildl, U. Schulte, H. Molina, S. Weidemann, E. Reisinger, M. Mann, B. Fakler Small conductance Ca2+-activated potassium (SK) channels are responsible for the afterhyperpolarization (AHP) that follows the action potential in many types of central neurons and is subject to extensive modulation by second messenger pathways. SK channels are voltageindependent and gated solely by intracellular Ca2+ in the submicromolar range. This high affinity for Ca2+ is mediated by calmodulin constitutively associated with the SK α-subunit. Because of their important physiological role and their known colocalization with Ca 2+ sources

Membrane Biology Group, Section of Biomedical Sciences, University of Edinburgh, Hugh Robson Building, George Square, Edinburgh, Scotland, EH8 9XD P 19-6 CHARACTERIZATION OF THE CYTOPLASMIC N-TAIL OF THE SMALL Ca2+ ACTIVATED POTASSIUM CHANNEL, hSK3 I. Spindler, 2E. Wanker, 3K.G. Chandy, F. Lehmann-Horn, S. Grissmer, H. Jäger Small Ca 2+ -activated K + channels (SK) generate the slow afterhyperpolarization (AHP) that follows an action potential. To understand the physiological role of the hSK3 in this scenario it is

S 275

Department of Applied Physiology, University of Ulm, Albert-Einstein-Allee 11, D-89081 Ulm, Germany, 2MPI for molecular Genetics, Ihnestr. 73, D-14195 Berlin, Germany 3Department of Physiology and Biophysics, University of California, Irvine, CA 92717, USA

The voltage-gated K +-channel Kvl.3 plays an important role in T lymphocyte activation and hence the use of Kvl.3-blocking substances in humans could be clinically useful. It has been shown previously that phenylalkylamines (PAA) like verapamil potently block mouse Kvl.3 channels and it was suggested that the binding of verapamil takes place in the inner vestibule of the channel (Rauer & Grissmer, 1999. Brit. J. Pharm. 127:1065). Our goal was to obtain further information about the structural properties of the verapamil binding site in the human Kvl.3 channel by a mutant cycle analysis approach. For that purpose we wanted to combine different channel mutants with various verapamil derivatives. Based upon computer modelling we expected possible candidate amino acids responsible for PAA-channel interaction to be in the hydrophobic S6 region of the channel. The whole-cell configuration of the patch-clamp technique was used to examine the current properties obtained by injection of in vitro transcribed mRNA in rat basophilic leukemia (RBL) cells. Measurements were done with an external solution containing (in mM): NaCl 160; KCl 4.5; MgCl 2 1; CaCl2 2; HEPES 5; NaOH ad pH 7.4 and with patch-pipettes containing (in mM): KF 155; MgCl2 2; HEPES 10; EGTA 10; KOH ad pH 7.2. We successfully expressed the human Kvl.3 channel protein (a generous gift from O. Pongs, ZMN Hamburg) in RBL cells and found that application of verapamil accelerated the current decay during depolarizations and resulted in an accumulation of block. These observations are qualitatively similar to the effects known for the mouse Kvl.3 channel. It has also been shown that the mouse Kvl.3 mutant H404T lacks intrinsic inactivation which interferes with measurements of blocking affinities (Röbe & Grissmer, 2000. Brit. J. Pharm. 131:1275). Therefore we want to introduce the analogous mutation in the human channel, serving as a basis for further mutations in the S6 region. Supported by grants from the DFG (Gr 848/8-2) and the 4SC AG, Martinsried, Munich.

P 19-7

Dept. of Applied Physiology, University of Ulm, 89081 Ulm, Germany

important to characterize the modulating factors and to identify the localization. hSK3 channels are widely expressed, in liver, heart, brain, skeletal and smooth muscle. Recently, it was shown that SK channels bind calmodulin (CAM) at the conserved C-terminal segment conferring the Ca2+ sensitivity to the channel (Xia et al., Nature 395:503-507, 1998). We used the N-terminal region of the hSKCa3 channel to look for putative interacting proteins because this segment contains prolinerich motifs, which are known to interact with SH3 domain proteins. We used a LexA based yeast two-hybrid system to find interacting proteins. Therefore we cloned a fragment coding for the first 296 amino acids of hSK3 into a LexA vector as DNA binding domain fusionprotein and tested with different proteins as activation domain fusions. First, different SH3 domain containing candidates were tested: Abl, Lck, Fyn, Grb2, and PI3K P85, however an interaction could not be detected between these SH3 domain containing proteins and the N-tail of hSK3 in yeast. Second, we tested one of the huntingtin (HD) binding proteins, SH3GL3 (Sittler et al., Mol Cell 2:421-436, 1998), in this system for interaction because of a structural similarity of the Nterminal segment of hSK3 and HD. Our data indicate that SH3GL3 does interact with the N-terminal tail of hSK3 in yeast. Interestingly SH3GL3 is expressed in brain and testis and is concentrated in nerve terminals. Further the hSK3 is found predominatly in the presynaptic membrane of mouse hippocampal neurons (Obermair et al., Biophys. J. 80:234a, 2001) and therefore the SH3GL3/hSK3 interaction might be possible under physiological conditions. Supported by a grant from the DFG (Gr848/8-2) and the BMBF (IZKF B7).

ALTERNATIVE SPLICING OF THE HUMAN SK3-GENE IN VARIOUS MUSCLE CELL LINES E. Frei, O. Wittekindt, F. Lehmann-Horn, S. Grissmer, H. Jäger The SK3-gene is expressed in various tissues e.g. heart, liver, brain, smooth and skeletal muscle. Especially in skeletal muscle of patients with myotonic dystrophy there is an elevated expression of the SK3gene (Kimura et al., 2000 Neurosci Lett, 295(3):93-6). Alternative splicing is a process, which can generate multiple transcripts encoding proteins with subtle or opposing functional differences that can have profound biological consequences. The human SK3-gene shows four different transcripts (SK3-ex1, SK3-ex4, SK3-1a, SK3-1a/4) (Wittekindt et al., 2001 Pflügers Archiv 441:R212). Splice variants containing exon 4 will increase the length of each P-loop of the outer pore by 15 amino acids. The other alternative splice-site results in a replacement of exon 1 by exon 1a and codes for a channel with a shorter N-terminal segment. The splice variants SK3-ex1 (exon 1 without exon 4) and SK3-ex4 (exon 1 with exon 4) form functional Ca2+ activated and apamin-sensitive K + channels, when expressed in oocytes. However, they do have slightly different Ca2+ sensitivities in excised inside-out oocyte patches. The SK3-1a splice variant shows no functional expression and is mainly localized in the cytoplasm (Wittekindt et al., 2001 ibid.). To test if there is an age-dependent or disease-dependent difference in the expression pattern of these transcripts, RT-PCR was made with a variety of muscle cell lines. We investigated the mouse muscle cell line C2C12 and human muscle cell lines generated from fetal (F) and adult (A) muscle as well as muscle from patients with myotonic dystrophy (MD). The splice variant SK3-ex4 was found in F and MD muscle cell lines. The splice variant SK3-1a/ex4 (exon 1a with exon 4) could be detected predominantly in cells of the MD-muscle. Therefore one can expect that some of the immunological reactive SK3 protein is non functional in the muscle of MD patients. Supported by grants from the DFG (Gr848/8-2) and the BMBF (IZKF B7) Dept. of Applied Physiology, University of Ulm, Germany P 20-1 CHARACTERIZATION OF THE PHENYLALKYLAMINE BINDING SITE IN THE HUMAN VOLTAGE-GATED POTASSIUM CHANNEL Kvl.3 T. Dreker, F. Lehmann-Hom, S. Grissmer

P 20-2 ROLE OF N-TERMINAL DOMAINS IN Kv4 CHANNEL INACTIVATION GATING M. Gebauer, D. Isbrandt, K. Sauter, O. Pongs, R. Bähring Kv4 channels in coassembly with Kv channel interacting proteins (KChIPs) represent the molecular substrate of both the transient outward current (Ito) in cardiac myocytes and the neuronal somatodendritic A-type current. Inactivation of Kv4 channels differs from the one of Shaker channels. Kv4 channels can inactivate from both open- and closed-states. Kv4 channel inactivation is well described by a gating model that assumes allosteric coupling of inactivation to partially activated closed-states. However, neither the Kv4 channel inactivation mechanism per se nor the effects of KChIPs on Kv4 channel gating, like the slowing A-type current decay, are well understood on a structural basis. We expressed Kv4.2 deletion and chimeric replacement mutants in HEK 293 cells and studied various gating paramters using the patchclamp technique. To elucidate the role of N-terminal domains in Kv4 inactivation gating we first replaced the whole cytoplasmatic Nterminus of Kv2.1 by that of Kv4.2 (180 amino acids long), asking which Kv4 inactivation properties are transfered to the chimeric channel Kv2.1(4.2NT). We observed a fast component of inactivation (t = 80 ms), and the coexpression of KChIP led to a dramatic slowing of inactivation. These results show that the attachment of the cytoplasmic Kv4.2 N-terminus onto Kv2.1 channels transfers the ability to rapidly inactivate from the open-state. Furthermore, like in Kv4.2 wild-type channels, this reaction is strongly inhibited in the chimeric channels by the coexpression of KChIPs. KChIPs interact with the proximal N-terminus of Kv4.2 channels. We studied the gating characteristics of a 40 amino acid N-terminal deletion mutant (Kv4.2D40), which leaves the tetramerization-domain intact. The results were compared to the ones obtained for Kv4.2 wild-type channels coexpressed with KChIP. Most interestingly, both N-terminal deletion and KChIP association with the proximal N-terminus resulted in a similar slowing of A-type current decay. By contrast, closed-state inactivation remained unaffected by the deletion but was enhanced by KChIP. We conclude that the proximal Kv4.2 N-terminus is directly involved in inactivation from the open-state, but not in inactivation from closed-states. Institut für Neurale Signalverarbeitung, ZMNH, Martinistr. 52, 20246

S 276 Hamburg, Germany P 20-3 KCHIP INTERACTION WITH A CONSERVED RETENTION SIGNAL CONTAINING N-TERMINAL DOMAIN OF Kv4 CHANNELS B. Callsen, D. Isbrandt, K. Sauter, J. Dannenberg, O. Pongs, R. Bähring Kv channel interacting proteins (KChIPs) associate with Kv4 channels, which leads to modulation of these A-type potassium channels. Our goal was to identify and characterize the KChIP interaction domain on Kv4 α-subunits. For this purpose we studied KChIP2.2, a splice variant which we cloned from human ventricle. By heterologous coexpression in CHO and HEK 293 cells we showed, using electrophysiological and immunocytochemical methods, that KChIP2.2 modulates Kv4.2 inactivation gating and functional cell-surface expression. KChIP2.2 slows the onset of inactivation, accelerates the recovery from inactivation, and shifts the Kv4.2 steady-state inactivation curve to more positive potentials. KChIP2.2 stimulates the Kv4.2 current densities, which correlates with a redistribution of immunoreactivity from perinuclear areas to the plasma membrane. Increased Kv4.2 cellsurface expression and current densities were also obtained in the absence of KChIP2.2, when the highly conserved proximal Kv4.2 N-terminus (40 amino acids) was deleted. The same domain is required for the association of KChIP2.2 with Kv4.2 α-subunits, and therefore the Nterminal deletion mutant Kv4.2∆40, shows none of the typically observed effects when coexpressed with KChIP2.2. We hypothesized that the Kv4.2 proximal N-terminus bears a retention signal, which can be masked by the association of KChIPs. The results we obtained with a Kv4.2∆40C40 construct, in which we transfered the N-terminal 40 amino acids to the C-terminal end, support this idea. Like wild-type channels, Kv4.2∆40C40 shows low functional expression, which can be increased by coexpression with KChIP2.2. These results also suggest that the proximal Kv4.2 N-terminus alone can serve as a KChIP binding site. The KChIP binding characteristics in the Kv4.2∆40C40 mutant may be different from wild-type since the effects of KChIP2.2 on the functional cell-surface expression and the inactivation behaviour of the mutant channel are less pronounced.

P 20-5 PRESENILIN-1 MUTATIONS ALTER K+ CURRENT EXPRESSION IN THE HUMAN NEUROBLASTOMA CELL LINE, SH-SY5Y L.D. Plant, J.P. Boyle, N.M. Thomas, N.J. Hipkins, N.M. Hooper, Z. Henderson, C. Peers, R.F. Cowburn, H.A. Pearson Mutations in presenilin 1 (PS1) are the major cause of autosomal dominant Alzheimer’s disease (AD). Previous reports suggest that transient overexpression of PS1 in a non-neuronal cell line (HEK293) can increase the density of K + channel currents (Malin et al, 1998). Furthermore, this increase is lost if the PS1 bears a mutation known to give rise to AD. Since this cell line ordinarily expresses only low levels of K+ channel current, we studied this effect of PS1 in a neuronal cell line that normally expresses a substantial K+ channel current. Voltagegated K+ currents in the human cell lines SH-SY5Y and HEK 293 were measured using whole-cell patch-clamp as described previously (Ramsden et al, 2001). Cells stably transfected to overexpress either PS1 or a deletion mutation (∆E9) of PS1 were grown and prepared as described in Shukla et al (2001). Values given are means±s.e.m. and statistical significance was assessed using Student’s unpaired t-test. Mean K+ channel current density at a test potential of +50mV in untransfected SHSY5Y cells was 76±10pA/pF (n=39) and was not significantly different from that of SH-SY5Y cells overexpressing PS1 83±9pA/pF (n=33). In contrast, cells that had been transfected with PS1 bearing the ∆E9 mutation had a significantly smaller K+ channel current density 43±6pA/ pF (n=43, p<0.01). These changes were channel specific since no change in the sodium current could be observed in the same cells. This finding is consistent with the notion that PS1 can modulate the activity of ion channels in central neurones and implicates a compromise in electrical signalling as an underlying factor in the pathogenesis of familial Alzheimer’s disease. References: Malin SA, Guo WX, Jafari G et al., Neurobiol Dis. 4, 398-409 (1998); Ramsden M, Plant LD, Webster NJ. et al J Neurochem. 79, 699-712 (2001); Shukla C, Plant LD, Finch LH et al J Cell Path 5, 241-249 (2001) School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK P 20-6

Institut für Neurale Signalverarbeitung, ZMNH, Martinistr. 52, 20246 Hamburg, Germany

P 20-4 STIMULATION OF KCNE1 BY THE SERUM AND GLUCOCORTICOID DEPENDENT KINASE SGK1 H. Embark, B. Brehm, S. Wolf, C. Boehmer, F. Lang The serum and glucocorticoid dependent kinase sgk1 is expressed in all human tissues tested including the heart. Besides serum and glucocorticoids, mineralocorticoids, angiotensin II and TGFß1 upregulate transcription of sgk1. The kinase is activated by oxidative stress and IGF1. The present study has been performed to possibly disclose a functional significance of altered sgk1 activity in the heart. Specifically, the influence of sgk1 on minK (KCNE1) has been tested by coexpression studies in Xenopus oocytes and subsequent determination of the currents (ImK) induced by a voltage step from -80 mV to -10 mV in dual electrode voltage clamp experiments. ImK approached 0.8 ± 0.2 µA (n = 31), 1.4 ± 0.1 µA (n = 31), 1.3 ± 0.1 µA (n = 30), 1.4 ± 0.1 µA (n = 26) and 1.1 ± 0.1 µA (n = 29) in oocytes injected with minK alone, with minK + sgk1, with minK + sgk2, with minK + sgk3 and with minK + PKB, respectively. Thus, coexpression of wild type sgk1 moderately upregulates minK-channel activity, a property shared by the sgk isoforms but less so by PKB. To test, whether the upregulation of I mK was secondary to Na + /K + -ATPase activation, experiments were performed in oocytes pretreated with ouabain. In the presence of ouabain, ImK was enhanced by a factor of 5.1 ± 0.6 by coinjection of the constitutively active S422Dsgk1, while coinjection of the inactive mutant K127Nsgk1 was without significant effect on ImK. In conclusion, sgk1 and the other isoforms are capable to stimulate the activity of minK+. Thus, sgk1 may modify the electrical properties of myocytes by altered regulation of Isk and thus at least in theory participate in the regulation of cardiac function. Departments of Physiology, and Nephrology, University of Tübingen, Germany

MODAL GATING OF KV1.1 CHANNELS S. Kalina, A. Scholle, R. Koopmann, T. Zimmer, K. Benndorf The patch-clamp technique was employed to investigate currents of the rat voltage-dependent potassium channel K V1.1 in transiently transfected mouse L cells. Measurements of whole-cell currents revealed variable time courses of activation among the cells. In order to identify the origin for this variability single-channel currents were measured in the cell-attached configuration. The patches were depolarized from a holding potential of -80 mV to a test potential of +40 mV at a pulsing frequency of 9.5 Hz. Analysis of the these recordings revealed the existence of two unitary current levels: Openings with a large conductance were dominant with respect to openings with a small conductance. The latency after which half of the channels were opened (first latency) was shorter (0.7 ms) in the small openings compared to that in the large openings (1.0 ms). Large and small openings were therefore considered to be caused by different gating modes, L and S mode, respectively. The modes developed characteristic mean open times (τo,L= 12.9 ms, τo,S =5.6 ms), whereas the mean closed times were indistinguishable (τc,L = 1.1 ms, τc,S = 1.2 ms). The channels gated either in the pure modes or the open channels switched between both modes. The existence of the S mode might explain some of the differences in the activation speed observed in the whole-cell currents. Institute of Physiology II, Friedrich Schiller University Jena, Teichgraben 8, D-07740 Jena, Germany P 20-7 INTERACTION OF THE S4 SEGMENT WITH THE CONSERVED GLUTAMATE (E418) OF THE SHAKER POTASSIUM CHANNEL PORE DOMAIN D.J.S. Elliott, E.J. Neale, M. Hunter, A. Sivaprasadarao Voltage-gated potassium channels are integral membrane proteins comprising a central pore domain (S5-H5-S6) and a surrounding voltage-

S 277 sensing domain (S1-S4 segments). Upon depolarization, the voltage sensor detects the change in membrane potential and transmits the signal to the pore domain, leading to the opening of activation gate(s); this results in the flux of K + ions through the pore. The positively charged S4 transmembrane segment plays a central role in voltage sensing. During depolarization, it moves out of the membrane bilayer, exposing its positive (gating) charges to the extracellular phase. In this study, we have asked whether during exposure, S4 approaches the highly conserved, negatively charged glutamate (E418 in Shaker) present just outside the outer membrane border of the pore domain (extracellular end of S5). To test this, we have mutated E418 to a cysteine and combined this with a second mutation of residues (358-362) in the extracellular end of S4. These mutations were made in the inactivation removed Shaker channel. The resulting double mutants were expressed in Xenopus oocytes and the effect of the extracellular application of 100 µM Cd2+ examined using two-electrode voltage-clamp. The rationale was that if the engineered S4 cysteines move into the close proximity of 418C, Cd2+ would form a metal bridge and thereby produce a measurable change in the functional properties of the channel. Of the five double mutants tested, L358C-E418C, I360C-E418C and L361CE418C showed a significant decrease (79 ± 8 %, 34 ± 5 % and 70 ± 9 %, respectively) in the conductance of the channel, when Cd2+ was applied. Such effects were not seen with any of the corresponding single mutants. Cd 2+ showed an additional effect on L361C-E418C by shifting the voltage- dependence of channel activation to the left by –42 ± 3.8 mV. The effect of Cd2+ on A359C-E418C was similar to that on the A359C single mutant. R362C-E418C did not produce functional channels. We next examined the voltage-dependence of the Cd2+ effect. Cd2+ bound L358C-E418C at –80 mV, where the channel was in its closed state, while binding to I360C-E418C and L361C-E418C occurred at more positive potentials, where the channel begins to open. These data suggest that L358 of S4 is physically close to E418 of S5 in the closed state of the channel, and that residues at positions 360 and 361 move into the proximity of E418 during membrane depolarisation, when S4 moves out of the membrane bilayer. Furthermore, since residues 358 and 360 are present on the opposite faces of the S4 helix, the data suggest that S4 rotates during its outward movement. The work was supported by the Wellcome Trust School of Biomedical Sciences, Worsley Building, University of Leeds, LS2 9JT

mediate the inhibition of I Ks after selective stimulation of the β 3 adrenoceptor in guinea pig ventricular myocytes. These findings support the idea of a counterregulatory mechanism to the prominent β1/2-AR activation of IKs via proteinkinases A and C in high sympathetic tone, as can be observed in heart failure. Department of Cardiology, University of Tübingen, Otfried-Müllerstrasse 10, 72076 Tübingen, Germany

P 20-9 AMINO-TERMINAL DETERMINANTS OF U-TYPE INACTIVATION OF VOLTAGE-GATED K+ CHANNELS H.T. Kurata, G.S. Soon, D. Fedida The T1 domain is a cytosolic N-terminal domain present in all Kv channels, and is highly conserved within Kv channel subfamilies. A truncated form of Kv1.5 (Kv1.5∆Ν209) expressed in myocardium, which comprises deletion of more than 80% of the T1 domain, exhibits a U-shaped inactivation-voltage relationship characteristic of U-type inactivation. This observation prompted us to investigate the Nterminus as a regulatory site for slow inactivation of Kv channels. We examined the macroscopic inactivation properties of several N-terminal deletion mutants of Kv1.5 expressed in HEK 293 cells, demonstrating that deletion of residues up to the T1 boundary (Kv1.5∆N19, Kv1.5∆N91, and Kv1.5∆N119) did not alter Kv1.5 inactivation. However, deletion mutants which disrupted T1 structure exhibited inactivation phenotypes resembling Kv1.5∆N209. Chimeric constructs of the transmembrane domains of Kv1.5 and the N-termini of either Kv1.1 or Kv1.3 preserved the inactivation kinetics observed in fulllength Kv1.5, again suggesting that the Kv1 T1 domain influences slow inactivation. Furthermore, disruption of residues near the intersubunit T1 interface, by mutation of residues E131 and T132 to alanines or residue I126 to aspartate, consistently resulted in channels exhibiting features of U-type inactivation. Fusion of the N-terminus of Kv2.1 to the transmembrane segments of Kv1.5 imparted a Ushaped inactivation-voltage relationship to Kv1.5. In contrast, fusion of the N-terminus of Kv1.5 to the transmembrane segments of Kv2.1 decelerated Kv2.1 inactivation and abolished the U-shaped voltagedependence of inactivation normally observed in Kv2.1. These data suggest that intersubunit T1 domain interactions influence U-type inactivation in Kv1 channels, and suggest a generalized influence of the T1 domain on U-type inactivation between Kv channel subfamilies.

P 20-8 TYROSINE KINASE-DEPENDENT INHIBITION OF THE SLOW DELAYED RECTIFIER K+ CURRENT IKS BY BETA3-ADRENERGIC ACTIVATION IN GUINEA PIG VENTRICULAR MYOCYTES A.C. Schneck, R.F. Bosch, V. Kühlkamp Recently, we provided evidence for a functional coupling of the β3adrenoceptor (β3-AR) to the delayed rectifier IKs channel in guinea pig cardiomyocytes. In this study, we evaluated the signaling cascade responsible for the observed reduction of IKs current amplitudes induced by β3-AR stimulation. Whole-cell patch-clamp techniques were applied to assess the effects of various protein kinase inhibitors on IKs regulation in isolated left ventricular myocytes at 36 °C. After incubation (>2h) of the myocytes with the unselective protein kinase A/C/G inhibitor staurosporine (100 nM), there was no significant change in the inhibitory effect of the β3-agonist BRL 37344 (1µM) on IKs step and tail current amplitudes. However, the responses on β 3-AR activation were fully abolished by pretreatment with the potent membrane permeable protein tyrosine kinase inhibitor genistein (50 µM). Lavendustin A (10µM) and tyrphostin A23 (10 µM), other PTK inhibitors structurally distinct from genistein, also strongly attenuated the inhibition of IKs. Change of IKs current amplitudes in [%] of control (con); n=4-7, *p<0.05:

con Step current Tail current

100 100

BRL 37344 60.8 ± 6.3* 60.4 ± 7.1*

Stau + BRL 64.0 ± 4.6* 66.9 ± 6.3*

Gen + BRL 93.7 ± 3.4 95.3 ± 4.4

LavA + BRL 94.6 ± 4.6 103.7 ± 5.0

TyrA23 + BRL 91.0 ± 2.4 85.7 ± 2.1

In conclusion, activation of the tyrosine kinase signaling pathway can

Department of Physiology, University of British Columbia, 2146 Health Sciences Mall, Vancouver, B.C., V6T 1Z3, Canada

P 20-10 PDZ DOMAINS OF PSD-95 BINDS TO BOTH THE C- AND NTERMINI OF THE VOLTAGE-GATED POTASSIUM CHANNEL KV1.5 J. Eldstrom, D. Steele, D. Fedida PDZ containing proteins are involved in the clustering of ion channels at sites of functional significance. Based on reports of a Kv1.4 Cterminal interaction with PSD-95 (Kim et.al.1995 Nature 379:85-88) we examined whether this PDZ containing protein also interacted with the C-terminus of hKv1.5. A GST fused C-terminal peptide of Kv1.5 interacted in an in vitro binding assay with a N-terminal construct of PSD-95 that included the three PDZ domains. However, the Nterminus of Kv1.5 also interacted with the PDZ domains of PSD-95. This interaction was confirmed by yeast two-hybrid studies, and shown to be true also of the N-termini of Kv1.1, 1.2, 1.3 and less strongly 1.4. Deletion constructs within the N-terminus of Kv1.5 localized the interaction to a region within or near the T1 domain. Confocal microscopy of Kv1.5 and PSD-95 transfected HEK293 demonstrated overlapping staining at the membrane of these cells, while staining of a C-terminal truncation mutant showed co-localization throughout the cell. These results indicate potential roles for both the C- and Ntermini of Kv1.5 in cellular localization by PDZ domain containing proteins. University of British Columbia; 2146 Health Sciences Mall, Vancouver, B.C. Canada, V6T 1Z3

S 278 P 21-1

P 20-11 +

COEXPRESSION STUDIES OF ERG K CHANNEL SUBUNITS WITH KCNE β SUBUNITS IN CHO CELLS J.-H. Buhk, S. Burgarth, I. Wulfsen, J.R. Schwarz, C.K. Bauer Members of the KCNE gene family (MinK, MiRP1, MiRP2) have recently been reported to interact with different voltage-gated K + channels, including KCNQ1 and HERG1. In the present study, we coexpressed all three cloned rat erg channels with the three rat KCNE ß subunits in CHO cells. Whole cell membrane currents were recorded following microinjection of the different cDNAs. In external high K+ low Ca2+ solution no significant alterations of erg current properties were detected after coexpression of the three erg channels with any of the KCNE subunits. In „Ringer“ solution (5 mM K +; 2 mM Ca2+), coexpression of erg1 with KCNE2 induced a small, but significant shift in the voltage dependence of erg1 current activation to more negative potentials. This effect was absent in external solution with buffered Ca2+. KCNE1 and KCNE3 did not change the gating kinetics and voltage dependence of erg1 channels determined in „Ringer“ solution. Since no dramatic changes of erg current properties were induced by KCNE subunit coexpression, KCNQ1 was coinjected with the KCNE subunits as a test for their functional expression. In these experiments, KCNE coexpression induced the reported strong alterations of the KCNQ1 currents (Schroeder et al., Nature 2000, 403:196; Tinel et al., EMBO J 2000, 19:6326). The distribution of the different erg and KCNE proteins was investigated by immunostaining using CHO cells transfected with epitope-tagged cDNAs. All three erg channels were unable to abolish the clustering of KCNE proteins in intracellular compartments which was more pronounced for KCNE1 and KCNE3 than for KCNE2. Also the immunostaining of erg1 and erg2 was not obviously altered by the cotransfection of the different KCNE subunits. In contrast, KCNQ1 clearly changed the distribution of KCNE1 and KCNE3 suggesting a translocation to the plasma membrane. The distribution of KCNE2 was less affected by KCNQ1 cotransfection. Conclusion: In CHO cells, KCNE subunits do not exert dramatic effects on erg channels and the detection of putative interactions might depend on the presence of external free Ca2+. Supported by the DFG (GK255)

ERG1 CURRENT PROPERTIES DEPEND ON THE TYPE OF EXPRESSION SYSTEM J.-H. Buhk, C.K. Bauer, J.R. Schwarz The endogenous ether-à-go-go-related gene (erg) K+ current in native rat lactotrophs is characterized by a pronounced slowly deactivating component when measured in isotonic KCl solution. Similar erg currents were observed in clonal lactotroph MMQ cells (Rosati et al., Soc Neurosc Abstr, p1984, 1998), whereas in clonal somatomammotroph GH 3/B6 cells the slowly deactivating erg current component is very small. All these cell types contain at least mRNA for erg1 (Schäfer et al., J Physiol 518:401, 1999; Wimmers et al., Pflügers Arch 441:450, 2001). The present experiments were performed to further analyze the slowly deactivating erg current component. Endogenous erg currents were measured in isotonic KCl solution as E-4031-sensitive currents in GH3/B6 and MMQ cells. The potential-dependent availability curves in GH3/B 6 cells could well be fitted with a single Boltzmann function (V0.5=-77 mV; n = 10), whereas in MMQ cells two Boltzmann functions were necessary (V0.5=-67 mV; V0.5= -115 mV; n = 8). These results found in MMQ cells are very similar to those reported for the slowly deactivating erg current component in native lactotrophs. After overexpression of erg1 in GH3/B6 and MMQ cells by cDNA injection erg currents were recorded in 40 mM external K+ with pulses to 120 mV from a holding potential of -20 mV. The erg current amplitudes were larger by a factor of up to 20 as compared to the native erg currents. The time constants of the fast and slowly deactivating current components of the overexpressed currents were similar to those of the respective native erg currents. In most MMQ cells with overexpressed erg1 channels the relative amplitude of the slowly deactivating current component was about 50% in comparison to 11% in GH3/B6 cells independent of the erg current amplitude. This indicates that cell type specific properties change the characteristics of the expressed current. Abteilung für Angewandte Physiologie, Institut für Physiologie, UKE, Universität Hamburg, Martinistraße 52, D-20246 Hamburg

Abteilung für Angewandte Physiologie, Institut für Physiologie, UKE, Universität Hamburg, Martinistraße 52, D-20246 Hamburg P 20-12

P 21-2

DIFFERENTIAL EFFECTS OF EXTERNAL POTASSIUM AND CALCIUM CONCENTRATIONS ON ERG CHANNELS S.Wimmers, J.R. Schwarz, C.K. Bauer Three members of the erg K+ channel subfamily have been cloned so far from rat (Bauer et al., Receptors & Channels 1998, 6:19; Shi et al., J. Neuroscience 1997, 17:9423). Whereas in the heart only erg1 is expressed, in pituitary lactotroph cells transcripts for all three erg channel subunits have been found (Schäfer et al., J. Physiol. 1999, 518: 401). In the brain, the three erg subunits are differentially expressed with some overlapping regions (Saganich et al., J. Neuroscience 2001, 21:4609). The aim of the present study was to investigate the effects of changes in the external Ca2+ and K+ concentrations on the biophysical properties of the three erg channels. Following heterologous expression of erg1, erg 2 or erg3 in CHO cells, erg currents were measured with the patch-clamp whole cell technique using external solutions containing a free Ca2+ concentration of either 66 nM or 2 mM, and a K+ concentration of either 5 or 150 mM. In accordance with previous studies with HERG channels expressed in Xenopus oocytes (Ho et al., J. Physiol. 1998, 507:631) we found that the potential of half-maximal activation of all three erg channels was shifted to more positive values by elevating the external Ca2+ concentration. This effect was stronger for erg1 than for erg2 and erg3. While Wang et al. (J. Physiol., 1997, 502:45) did not observe an effect of the external K+ concentration on the activation of HERG, we found that the potential of halfmaximal activation of erg1 and erg3 was shifted to more negative values by an elevation of the K+ concentration. This effect could not be observed with erg2. Since erg3 channels exhibit an already more negative threshold of activation compared to erg1 and erg2 channels, a further shift of the activation curve to more negative values induced by an elevated external K+ concentration could be of importance e.g. in preventing neurons from firing under hyperkalemic conditions.

POTASSIUM CURRENTS IN NERVE TERMINALS OF THE RAT POSTERIOR PITUITARY G. Glassmeier1, M. Schweizer2, O. Pongs2, J.R. Schwarz1 Secretion of neuropeptides from nerve terminals in the posterior pituitary is controlled by the frequency of action potentials. K+ channels take part in this process because they modulate spike duration and frequency adaptation. To identify the types of K+ currents present in these specialized presynaptic structures electrophysiological experiments were performed. K+ currents were recorded with depolarizing 250 ms pulses from a holding potential of either -60 or -100 mV. At least one delayed rectifying (IKD) and three transient K+ currents (IKA1IKA3) could be distinguished: 1. IKD was recorded from a holding potential of -60 mV, where the transient K+ currents were completely inactivated. IKD was activated positive to -40 mV, the activation curve exhibited half-maximal activation at 1.0 ± 1.1 mV (mean ± SEM; n=7) and a slope factor of 11.2 ± 1.2 mV. The time constants of IKD activation decreased from 25 ± 3.6 ms at -20 mV to 2.2 ± 0.3 ms at +50 mV. 2. Subtraction of membrane currents recorded from a holding potential of -60 mV from those recorded from a holding potential of -100 mV yielded the transient K + currents. IKA1 was activated at -40 mV, the activation curve exhibited half-maximal activation at -32 ± 3 mV (mean ± S.E.M; n=5) and had a slope factor of 7.6 ± 1.5 mV. Inactivation of IKA1 was fast and voltage-dependent (τ decreased from 20.8 ± 3.2 ms at -30 mV to 3.8 ± 0.6 ms at 20 mV). 3. IKA2 activated at membrane potentials > 0mV and exhibited voltage-independent inactivation kinetics (τ ~ 70 ms). 4. IKA3 was activated near -20 mV, 50% activation occurred near +10 mV and inactivation proceeded with voltage-independent kinetics (τ ~ 20 ms). This current was blocked by BDS I, a selective blocker of Kv3.4. IKA1 and IKA2 occurred together in most nerve terminals, whereas IKA3 was found only in a subpopulation of nerve terminals.

Abteilung für Angewandte Physiologie, Institut für Physiologie, UKE, Universität Hamburg, Martinistraße 52, D-20246 Hamburg

Institut für Physiologie1 and ZMNH 2, UKE, Universität Hamburg, Martinistraße 52, D-20246 Hamburg, Germany

S 279 P 21-3 DELAYED RECTIFIER K+ CURRENTS IN ENDOTHEL CELLS ARE BLOCKED BY HEPTANOL IN THIN SLICE OF GUINEA PIG HEART C. Malik, A. Scholz Electrical cell-cell coupling between endothel cells in vessels is based on gap junction channels and as typical gap junction blocker is used 1heptanol. In a thin slice preparation from guinea left ventricle it was found that endothel cells in small vessels (20 – 80 µm) possessed voltage dependent delayed rectifier K+ currents. In a subgroup of these cells it was impossible to measure these currents due to large background leakage current probably owing to electrical cell-cell coupling. Therefore we wanted to investigate in these cells the ionic currents after uncoupling the cells with the 1-heptanol. We prepared thin slices (100 µm) from the left ventricle of guinea pigs (290–380 g) with a vibratome. Endothel cells in cut-open small vessels on the top of the slices were investigated with the patch-clamp technique. After application of control solutions containing 1 to 3 mM 1-heptanol we found a 10-20 times reduced leakage current but we could no longer measure a voltage dependent K+ current. Hence we measured in tight cells the effect of 1heptanol and found a concentration dependent blockade of the delayed rectifier K+ currents with half-maximal inhibition at around 1.5 mM (n=15 cells). This K+ current blockade resulted in a destabilized resting potential whenever a depolarisation occurred. The result suggests that those experiments where 1-heptanol was used as gap junction blocker might be revisited, taking into account that 1-heptanol potently blocked delayed rectifier K+ channels. Physiologisches Institut, Justus-Liebig-Universität, Gießen, Germany P 21-4 K + -DEPENDENCE OF k-CONOTOXIN PVIIA BINDING TO SHAKER K+-CHANNELS WITH FAST C-TYPE INACTIVATION E.D. Koch, F. Conti, B.M. Olivera, H. Terlau k-conotoxin PVIIA (k-PVIIA), a 27 amino acid peptide from the venom of Conus purpurascens, is the first member of a new family of conotoxins blocking voltage-gated K+-channels. The analysis of the block of Shaker K+-channels by k-PVIIA showed that it binds to the extracellular mouth of the ion channel pore in a potassium dependent manner. The binding of k-PVIIA to Shaker K + -channels is state dependent in that the affinities to the open and closed states are different. Furthermore k-PVIIA binding is incompatible with fast C-type inactivation of mutant Shaker K +-channels, leading to an apparent slowing of inactivation. By using the Xenopus oocyte expression system we here studied the binding of k-PVIIA to a pore mutant (M448K) of the Shaker K+-channel in low (2.5 mM) and high (115 mM) external K+ concentrations. In the absence of N-type inactivation (D6-46), this channel displays rapid C-type inactivation with a time constant of ~12 ms at +40 mV in low K + . K + is known to interfere with C-type inactivation and increasing external K + to 115 mM slows the time constant for inactivation of these channels to 50 ms. Applying kPVIIA in the presence of high external K+ leads to a further slowing of the inactivation. The total charge flowing during a depolarizing pulse is not reduced by the toxin. This is similar to what had been obtained for 2.5 mM K+. But in contrast to 2.5 mM K+ no charge increase can be observed in the presence of the toxin at high external K+. Raising the extracellular K + concentration from 2.5 to 115 mM results in a ~10 fold decrease of the IC 50 for the closed state (75 and 720 nM, respectively) which is due to a decrease in kon. In contrast to this, an increase in extracellular K + has no major effects on open channel binding. It is concluded that the K+-dependence of k-PVIIA binding to M448K channels despite the strongly modified inactivation behaviour of these channels is similar to what had been described for wild-type Shaker K +-channels.

channels. This peptide inhibits Shaker K+ channels by interacting with amino acids at the outer vestibule of the ion channel pore. The block has a 1:1 stochiometry and is different for the closed and the open state of the channel. When the toxin is bound it interacts intimately with the ion(s) sitting in the pore as shown by the dependence of the block properties on the solution composition and on voltage. Here we investigated the role of the permeant ion species on the interaction of PVIIA with the open and closed Shaker-∆6-46 channel by using the Xenopus oocytes expression system and outside-out patch clamp experiments. When K is the only permeant ion, increasing the extracellular K-concentration decreases the affinity of PVIIA binding to closed channels by decreasing the „on“ binding rate, whereas the open channel block is only affected by the intracellular Kconcentration. In contrast, extracellular Rb affects both closed- and open-channel binding, and with different modalities: The binding to the closed channel is only slightly destabilized but has faster kinetics at increasing Rb-concentrations. The block of open channels depends on the intracellular solution and is maximally destabilized when the pore is likely filled only by Rb ions. Our observations suggest that equally permeant cations may have quite different occupancy configurations within the pore permeation pathway. This implies that the functional relevant pore of the Shaker potassium channel which consists of the amino acids lining the pore and the ion(s) present differs for different permeating ions. Max-Planck-Institut für Experimentelle Medizin, Molekulare & Zelluläre Neuropharmakologie Hermann-Rein-Str. 3, D-37075 Göttingen, Germany P 21-6 DYNAMIC CLAMP ANALYSIS OF THE FUNCTIONAL IMPACT OF KV3-LIKE K+ CHANNELS ON THE ACTION POTENTIAL PHENOTYPE OF HIPPOCAMPAL INTERNEURONS C.-C. Lien, P. Jonas Previous studies indicated that the K+ current in GABAergic interneurons of stratum oriens alveus (OA) consist of three major components: a fast delayed rectifier, a slow delayed rectifier, and an inactivating A-type component. Furthermore, single-cell RT-PCR analysis suggested that the dominating fast delayed rectifier component was largely mediated by Kv3.2 channels (Lien et al., J. Physiol., in press). However, it is unknown how the specific gating properties of Kv3-like channels are related to the characteristic action potential (AP) phenotype of these neurons. To address this question, we first developed Hodgkin-Huxley type gating models of these channels. Second, we developed a dynamic clamp system that allows us to apply artificial conductances to neurons. The system is based on a digital signal processor card (Texas Instruments TMS320C), is driven by programs written in C, and runs at a frequency of ≈ 50 kHz. Finally, we blocked the native Kv3-like channels in OA interneurons by 1 mM tetraethylammonium ions (TEA) or 0.3 mM 4-aminopyridine (4-AP), and added conductances that were either identical or modified in comparison to the original Kv3-like channels using dual whole-cell recordings from the same cell (n = 7; 22°C). Application of either TEA or 4-AP led to a reduction of mean AP frequency during 1-s depolarizing pulses (from 45 ± 3 Hz to 28 ± 1 Hz) and induced a marked AP broadening within a series of spikes (50 Hz) evoked by a train of pulses (halfduration 229 ± 46 % of the control value for the 10th spike). Adding Kv3-like conductances restored the original AP pattern. Adding K + conductances with altered gating properties revealed that (1) a deactivation time constant (≈ 10 ms at -50 mV) is optimal to generate spiking with maximal frequency, (2) the lack of inactivation is favorable for repetitive spiking without adaptation, and (3) the high midpoint potential of activation (≈ -10 mV) is favorable for the repetitive spiking. Physiologisches Institut der Universität Freiburg, Hermann-HerderStr. 7, D-79104 Freiburg, Germany

Max-Planck-Institut für Experimentelle Medizin, Molekulare & Zelluläre Neuropharmakologie, 37075 Göttingen, Germany P 21-7 P 21-5 INFLUENCE OF POTASSIUM AND RUBIDIUM ON THE BINDING OF K-CONOTOXIN PVIIA TO SHAKER K+ CHANNELS A. Boccaccio, F. Conti, B.M. Olivera, H. Terlau κ-conotoxin PVIIA is the first conotoxin known to interact with K +

A NOVEL LONG-QT5 SYNDROME GENE MUTATION IN THE C-TERMINUS (V109I) IS ASSOCIATED WITH A MILD CLINICAL PHENOTYPE M. Schwarz*1'3, E. Schulze-Bahr2, W. Haverkamp2, S. Hauenschild3 , D. Isbrandt3 Mutations in the human minK gene KCNE1 have been linked to auto-

S 280 somal dominant and autosomal recessive long-QT (LQT) syndrome, a cardiac condition predisposing to ventricular arrhythmias. minK and KvLQTl, the LQT1 gene product, form a native cardiac K + channel that mediates the slowly delayed rectifier potassium current I K s- We used single-strand conformation polymorphism and sequencing techniques to identify novel KCNE1 mutations in patients with a congenital LQT syndrome of unknown genetic origin. In 150 unrelated index patients a missense mutation (VI091) was identified that significantly reduced the wild-type IKS current amplitude (by 36%) when coexpressed with KvLQTl in Xenopus oocytes. Other biophysical properties of the IKs channel were not altered. Since we observed incomplete penetrance (only one of two mutation carriers could be diagnosed by clinical criteria), and the family’s history was unremarkable for sudden cardiac death, the 109I allele most likely causes a mild phenotype. This finding may have implications for the occurrence of „acquired“ conditions for ventricular arrhythmias and thereby the potential cardiac risk for asymptomatic mutation carriers still remains to be determined.

concentrations of donor and acceptor molecules. In order to evaluate the power of this approach we measured the time resolved emission spectra of constructs, in which spectrally distinct variants of the green fluorescent protein, CFP and YFP, were linked together by a short amino acid chain. These constructs were expressed in HEK293 cells and investigated with a confocal laser scanning microscope (LSM 510, Zeiss). As pulsed excitation source a frequency doubled mode-locked Titanium:Sapphire laser was used. Fluorescence light was guided via an optical fiber to a spectrograph (250is, Chromex) and a streak camera (C5680, Hamamatsu). Alternatively, a 16 channel photomultiplier detector head (PML-16, Becker&Hickl) and a time-correlated single photon counting imaging module (SPC-730, Becker&Hickl) was used to record fluorescence data over time, wavelength, and image coordinates. Fluorescence decay analysis revealed a double exponential decay for both, CFP and YFP. The average lifetime of CFP was shortened and, interestingly, the amplitude of the fast component was negative for YFP, indicating that energy was transferred from CFP to YFP.

1

Institut fur Physiologie, Universität Münister, Robert-Koch-Str. 27a, 48149 Münster; 2Klinik für Innere Medizin C, Molekulare Kardiologie, Universität Münister, Domagkstr. 3, 48149 Münster; 3Institut für Neurale Signalverarbeitung, ZMNH, Martinistr. 52, 20251 Hamburg

Friedrich-Schiller-Universität Jena, Institut für Physiologie II, Teichgraben 8, D 07740 Jena, and 2Becker&Hickl GmbH, Nahmitzer Damm 30, D 12277 Berlin, Germany

P 21-8

P 21-10

KNOCKDOWN OF Kvβ β 1.1 DECREASES EXCITABILITY IN HIPPOCAMPAL NEURONS T. Falk, R.K. Kilani, A.J. Yool, S.J. Sherman Voltage-gated K + channels are important determinants of neuronal excitability and synaptic transmission. Some pore forming α-subunits are expressed in association with accessory β-subunits. The Kvβ1.1 subunit is known to confer fast inactivation to members of the Kv1.x family of α-subunits and contribute to the A-type current in hippocampal neurons. We transduced cultured rat hippocampal neurons with a herpes virus antisense-vector designed to knockdown the expression of Kvβ1.1. Transduced neurons were identified using fluorescence microscopy to detect expression of a Green Fluorescent Protein marker (eGFP) coupled to the Kvβ1.1-antisense sequence. Whole-cell voltage-clamp recordings were performed to evaluate the effects of expression of the Kvβ1.1-antisense sequence in hippocampal neurons as compared with untreated controls or neurons transduced with an eGFP-control vector. These data show a developmental decrease in the time constant (τ) of fast inactivation of K+ currents in untreated neurons: τ = 17.9±1.5 at 5-7 DIV (n=53); τ = 9.9±1.0 at 12-15 DIV (n=72, mean ± SEM, p<0.01). However, 48h after transduction of older (12-15 DIV) neurons with the Kvβ1.1-antisense vector, the τ of fast inactivation increased to 18.3±2.4 (mean ± SEM, n=20, p<0.01 compared to controls). This value is comparable to that of untreated young (5-7 DIV) hippocampal neurons. Electrical excitability was studied using current clamp protocols with spike trains elicited by injecting depolarizing current in increasing steps or in a continuous ramp. Older neurons (>13 DIV) showed a significant increase in the interspike interval (ISI) when treated with the Kvβ1.1-antisense viral vector. The increase in ISI was most pronounced in the range of Vm = -40 to +10 mV. Thus, virally induced knockdown of Kvβ1.1 in rat hippocampal neurons reduces the decay of the A-type K+ current and decreases electrical excitability. Supported by NINDS K08 NS 62015 (SJS) and NIH RO1 MH59747 (AJY).

FRET-BASED OPTICAL VOLTAGE IMAGING OF CULTURED HUMAN KERATINOCYTES R. Burgstahler, F. Rucker, H. Koegel, D.J. Tracey, C. Alzheimer, P. Grafe Electrophysiological techniques enable recordings of membrane potential with high sensitivity and temporal resolution. However, it is difficult to compare changes in membrane potential of different elements in a complex biological tissue or of several cells simultaneously. Recently, measurements have been described using a combination of two-component fluorescent indicators of cell membrane potential that use fluorescence resonance energy transfer (FRET) with a chargecoupled device camera (T.W. Cacciatore et al., Neuron: 23, 1999, 449f). In an attempt to develop a viable alternative for recording membrane potentials in more complex tissue systems, we combined the high spatial resolution of confocal imaging and these novel voltagesensitive dyes. As a first step, we used a violet diode laser (Ex 408 nm) connected to a confocal laser microscope for confocal ratiometric imaging of a monolayer of human keratinocyte cell line HaCaT. HaCaT cells were stained with voltage-sensitive dyes CC2-DMPE (Em 460 nm) and DiSBAC4(3) (Em 560 nm). We validated this approach by comparing the optically recorded voltage signals in HaCaT keratinocytes to electrophysiological signals determined by nystatinperforated patch whole-cell recordings. We obtained sensitivities of ∼15% change of emission ratio / 100 mV and a high temporal resolution. We demonstrate that the optical signals allow precise multi-site recordings of voltage changes evoked by the extracellular signaling molecules ATP and bradykinin. Supported by the Deutsche Forschungsgemeinschaft

1

Physiologisches Institut der Universität München, Pettenkoferstr. 12, 80336 München, Germany P 21-11

Departments of Physiology and Neurology, College of Medicine, Tucson, Arizona 85724, USA P 21-9 SPECTRALLY RESOLVED RESONANCE ENERGY TRANSFER MEASUREMENTS IN LIVING CELLS C. Biskup1, T. Zimmer1, A. Bergmann2, W. Becker2, K. Benndorf1 Resonance energy transfer (RET) is one of the few tools available to estimate distances in the nanometer range in living cells. When RET occurs, both the intensity and the lifetime of the donor fluorescence are decreased, whereas the intensity of the acceptor emission is increased. This offers two different approaches for the determination of RET efficiency. One is to detect changes in the fluorescence intensity of the donor or the acceptor emission, the other is to measure changes in the fluorescence lifetime of the donor molecule. The latter approach has the advantage to be independent of changes in the respective

GENE EXPRESSION OF RCX46 IN XENOPUS OOCYTES IS NOT AFFECTED BY 50 HZ ELECTROMAGNETIC RADIATION M. Steffens, H.-A. Kolb There is growing concern about the effects of electromagnetic radiation on biological structures. If there exists an interaction between extremely low frequency electromagnetic fields (ELF-EMF) and biological structures the corresponding mechanisms are still obscure. We applied 50 Hz powerline in double blind experiments to Xenopus oocytes expressing rodent lens connexin Cx46 (rCx46) which after expression form voltage-dependent hemichannels. In parallel to an expression for 16 h the following two exposure protocols to ELF-EMF were applied, either 16 h continuous exposure to a field intensity of 2.3 mT or an intermittent exposure (5 min On/ 10 min Off) to a field intensity of 2.3 mT or 1.0 mT. The expression of rCx46 was followed by kinetic analysis of time-dependent macroscopic currents which were evoked upon depolarisation. For the different exposure conditions and the control more than 60 different oocytes were analysed,

S 281 respectively. The results show neither significant changes of the leak current nor of the mean steady state current conductance. It is known that the most evident effects induced by magnetic waves are the mobilisation of intracellular calcium, although the cell signals when present are extremely variable. Therefore we studied the additional effect of external Ca2+ and H+ which change the membrane conductance in a concentration dependent manner. This functional relationship was not affected by exposure to ELF-EMF. In general, if a detectable cellular modification exists, it is not constant but occurs after activation of a complex cell mechanism especially for cells under stress conditions. Therefore, in addition we activated protein kinase C by addition of 200 µM l-oleoyl-2-acetyl-sn-glycerol (OAG) which is known to significantly change the macroscopic voltage-jump current-relaxation curves. It turned out that exposure of ELF-EMF did not significantly influence the recorded macroscopic currents. It is concluded that low frequency ELF-EMF is not linked to pathological changes of gene expression followed on the cellular level. Institute of Biophysics, University Hanover, Herrenhaeuser Strasse 2, D-30419 Hanover, Germany

serum-free media (MDSF). Once seeded, cells were incubated at either fetal PO2 (23mm Hg) or adult alveolar PO2 (100mm Hg) and they formed resistive monolayers within 48 h. Monolayers that had been maintained at 100mmHg in PC1 exhibited an amiloride sensitive Isc that was ~ 2.6 fold greater than that of cells maintained at 23mmHg (23mmHg: 3.4±0.2 µA.cm2, 100mmHg: 8.9±0.2 µA.cm2). FDLE cells maintained in MDSF also displayed amiloride sensitive Isc that was sensitive to PO2 (~ 1.7 fold increase) and the Isc was greater than that obtained from cells grown in PC1 (23mmHg: 6.1±0.4 µA.cm 2 , 100mmHg: 10.6±0.5 µA.cm2). Interestingly, the cells maintained in MDSF also displayed a 3-fold increase (3.0±0.85) in γ ENaC mRNA levels compared to supplemented cells. Cells grown at 100mmHg in PC1 showed a 2.91(±0.2) fold increase in γ ENaC mRNA expression compared to those grown at 23mmHg. Preliminary data has shown that there is no difference in γ ENaC mRNA expression at different PO2 when cells were grown in MDSF media. These results could indicate that a growth factor or hormone in the supplemented media is having an inhibitory effect on γ ENaC transription. 1. Ramminger S.J. et al. (2000) J. Physiol. 524, 539-547 Tayside Institute of Child Health, Ninewells Hospital and Medical School, University of Dundee, Dundee, DD1 9SY

P 21-12 IMPROVED HYBRID CLAMP: RESOLUTION OF TAIL CURRENTS FOLLOWING SINGLE ACTION POTENTIALS D. Dietrich Hybrid clamp protocols, in which a discontinuous single electrode voltage clamp (dSEVC) amplifier is switched from current to voltage clamp during the recording, are frequently used to investigate conductance changes after high frequency trains of action potentials. This technique is advantageous because it combines the physiological stimulation of the cell by action potentials with the possibility of analyzing the consecutive conductance changes quantitatively. Up to now this technique could not be employed to fast conductance changes after single action potentials because those currents are small and fast decaying. In this study an improved hybrid clamp protocol, called dynamic hybrid clamp, is developed that enables the experimentator to study fast conductance changes after single spontaneous action potentials. This is achieved by introducing two major advancements: The protocol employs real time detection of action potentials to assure precise timing of the mode switch and utilizes an external sample and hold amplifier to avoid voltage steps during the switch from current to voltage clamp. Together with the use of whole-cell patch clamp recordings and high switching frequencies (≥ 25 kHz), dSEVC could easily be started with a minimal delay (< 1.5 ms) after single action potentials of CA 1 pyramidal cells in situ and tail currents underlying fast and medium afterhyperpolarizations and afterdepolarizations ensuing single spikes are clearly resolved. The dynamic hybrid clamp should be useful for analysis of other spontaneously occurring events such as intrinsic or population bursts in a large variety of neurons. Exp. Neurophysiol., Klinik und Poliklinik für Neurochirurgie, Universitätsklinikum Bonn, Sigmund-Freud-Str. 25, 53127 Bonn, Germany P 22-1 EFFECT OF GROWTH CONDITIONS ON EXPRESSION OF EPITHELIAL SODIUM CHANNELS IN RAT FETAL DISTAL LUNG EPITHELIAL CELLS K. Richard, S.J. Ramminger, R.E.Olver, S.M.Wilson At birth, the lung switches from fluid secretion to absorption. The active pumping of Na + across the alveolar epithelium drives the absorption of lung liquid. Na+ crosses the alveolar cells through epithelial sodium channels (ENaC). ENaC is composed of three homologous subunits (α, β and γ) and channels can be formed from different combinations of these subunits. Levels of many hormones increase during late gestation and may increase transcription of ENaC subunits. There is also evidence to suggest that oxygen levels affect the numbers of ENaC present in lung. An increase in oxygen in combination with an increase in hormones may play a crucial role in stimulating lung liquid absorption. In this study we investigated how growth conditions and PO2 affect ENaC expression in rat fetal distal lung (FDLE) cells. FDLE cells were isolated as previously described (1). Cells were plated onto Transwell membranes in either media supplemented with growth factors and hormones normally found in serum (PC1) or minimally defined

P 22-2 HORMONAL REGULATION OF APICAL MEMBRANE Na + CONDUCTANCE IN RAT FETAL DISTAL LUNG EPITHELIAL CELLS S.J. Ramminger, R.E. Olver, S.M. Wilson In this study we investigated the role of glucocorticoids and other hormones as physiological regulators of amiloride-sensitive Na + transport in rat fetal distal lung (FDLE) cells. FDLE cells were isolated from day 20 fetal rats (full term = 22 days) that were removed from anaesthetised rats (3% halothane, killed before regaining consciousness) as previously described (Ramminger et al, J Physiol 524:539-547, 2000). Cells were placed onto Transwell col membranes in a defined serum-free media consisting of DMEM/Hams F12 (1:1) supplemented with 1.25mg/ml BSA, 2.0mM glutamine, 0.1mM non-essential amino acids and antibiotic. Once seeded cells were then incubated at fetal (23mmHg) or adult alveolar (100mmHg) PO 2 where they formed resistive monolayers within 48h. Isoproterenol, a drug that normally stimulates Na + transport in these cells, failed to increase amiloride-sensitive ISC (n=15) in cells maintained under these conditions. Moreover studies of basolaterally permeabilised cells showed that this drug also failed to elicit any change (n=3) in apical Na+ conductance (G Na+ ). Treatment (24h) with a combination of dexamethasone (200nM) and T3 (10nM) however, allowed isoproterenol to stimulate amiloride-sensitive ISC (n=5, 23mmHg: ∆ISC= 1.8±0.7µA.cm-2, P<0.05; 100mmHg: ∆I SC = 5.8±1.6µA.cm-2, P<0.01) and increase GNa+ (n=3, 23mmHg: ∆G Na+ = 18±8µS.cm -2 , P<0.05; 100mmHg: ∆G Na+ = 41.9±8.4µS.cm -2 , P<0.04). It is also known that β-adrenoceptor agonists can increase apical Cl - conductance (G Cl-) although the physiological role of this response is unclear. Isoproterenol (n=5) stimulated GCl- in cells maintained at both PO2 regardless of hormonal status (23mmHg: ∆GCl- = 23.7±4.0µS.cm-2, P<0.01; 100mmHg, ∆GCl= 17.2±1.1µS.cm -2, P<0.03). Therefore when rat FDLE cells are maintained in the absence of hormones or growth factors the βadrenoceptor agonist isoproterenol can only increase GCl-. Simultaneous exposure to dexamethasone and T 3 however, allows such drugs to regulate GNa+ and hence control amiloride-sensitive ISC. Dept. Child Health, University of Dundee, Scotland, DD19SY

P 22-3 STIMULATION OF XENOPUS OOCYTE NA+/K+-ATPase BY THE SERUM AND GLUCOCORTICOID DEPENDENT KINASE SGK1 I. Setiawan, G. Henke, Y. Feng, C. Böhmer, LA. Vasilets, W. Schwarz, F. Lang The serum and glucocorticoid dependent kinase sgk1 is expressed in a wide variety of tissues including renal epithelial cells. As it is upregulated by aldosterone it is considered to participate in the regulation of renal Na+ reabsorption. Indeed, co-expression of sgk1 with the renal epithelial Na+ channel ENaC augmented Na+ channel activity.

S 282 The present study has been performed to test for an effect of sgk1 on Na +/K +-ATPase. According to voltage-clamp experiments hyperpolarization and outward-directed current created in the presence of K+channel blocker Ba2+ by addition of extracellular K+ after 10 minutes K+-free conditions are both significantly larger in oocytes expressing the active kinase S422Dsgk1 (-18±2 mV, n=18 and 28±4 nA, n=26, respectively) compared to those expressing the inactive mutant K127Nsgk1 (-9±2 mV, n=17 and 9±2 nA, n=24 respectively). As both phenomena were fully inhibited by ouabain, the results reflect altered transport by the Na +/K+-ATPase. According to ion-selective microelectrodes the stimulation of pump current is not the result of altered cytosolic Na+activity ( S422Dsgk1: 2.6±0.3 mmol/l, n=13, K127Nsgk: 3.0±0.3 mmol/l, n=12) or pH (S422Dsgk1: 7.27±0.02 pH, K127Nsgk: 7.27±0.06 pH, n=5). Thus, the present results point to an additional action of sgk1, which may participate in the regulation of renal tubular Na+ transport. Moreover, sgk1 may be involved in the regulation of Na+/K+ ATPase in extrarenal tissues. Departments for Physiology, Universities of Tübingen and Halle, and Max Planck-Institute for Biophysics, Frankfurt, Germany P 22-4 EFFECT OF PURINES AND PYRIMIDINES ON ION CONDUCTANCES IN XENOPUS LUNG M. Fronius, A. Berk, M. Schnizler, W. Clauss We used excised lungs from the clawed frog Xenopus laevis to investigate ion transport across a native lung epithelium. The lungs were dissected as flat sheets and mounted in adapted Ussing-chambers. Both sides of the tissue were perfused with Na +-Ringer solution (NRS) and short circuit current (Isc) was recorded. The most prominent part of the Isc was amiloride sensitive, indicating the presence of highly selective epithelium sodium channels (ENaCs). Both, UTP and ATP, stimulated Isc in a dose dependent manner when added to the apical side. Half maximal stimulation was achieved at 59 µM for UTP and 96 µM for ATP. Stimulation with either one of these trinucleotides was abolished in the presence of 10 µM amiloride. Perfusion with 200 µM UTP increased the amiloride sensitive current (Iami) significant about 19.1 ± 3%. An additional treatment with 200 µM ATP produced a supplementary rise in Iami of about 24.4 ± 6%. Overall stimulation was 48.7 ± 10%. The same procedure was applied on the basolateral side of the tissues. The effect of the nucleotides was the same – an increase in Iami was observed. Further experiments were done using 100 µM adenosine. The nucleoside was applied to the apical side and stimulates Isc. Our data suggest two different conductances to be responsible for this effect. One part of the adenosine induced current could be blocked with amiloride, suggesting an upregualtion of ENaCs in the apical membrane. The residual part was not suppressed even in the absence of Na+. This indicates a stimulation of Cl- conductance across the apical membrane. Our experiments demonstrate the presence of different types of purinergic receptors in the epithelium of Xenopus lung and underlines the importance of purines and pyrimidines as paracrine/autocrine secreted extracellular signaling molecules. This work was supported by DFG. Institute of Animal Physiology, Justus-Liebig Universität Giessen, Wartweg 95, D-35392 Giessen

this stimulating effect of PGE2 as well as of forskolin. To prove the influence of intracellular Ca2+, we incubated Xenopus lungs in 100µM BAPTA/AM, which did not significantly affect ISC or Iami. These findings suggest that the PKA-pathway seems to be involved in the stimulatory action of PGE2 in Xenopus lung epithelia. Comparative studies, using two-electrode voltage-clamp technique at Xenopus laevis oocytes, expressing Xenopus kidney ENaC, demonstrated that perfusion with PGE2 and forskolin had no effect on Iami. Supported by Stiftung VERUM Institut für Tierphysiologie, Justus-Liebig Universiät, Wartweg 95, 35392 Gießen P 22-6 COLONIC NA + ABSORPTION IS ENHANCED IN A MOUSE MODEL FOR LIDDLE’S SYNDROME J.E. Cuffe, M. Bertog, J.P. Koch, E. Regardsoe, S. Pradervand, E. Hummler, B.C. Rossier, C. Korbmacher We used a recently established mouse model for Liddle’s syndrome (Pradervand et al.,1999, J. Am. Soc. Nephrol. 10: 2527-33), an hereditary form of arterial hypertension, to measure the amiloride-sensitive (10 µM) short-circuit current (∆ISC-AMI) in late distal colon from wild-type (+/+), heterozygote (+/L), or Liddle (L/L) mice. To modify plasma aldosterone (Paldo) animals were maintained on a regular, low, or high Na+ diet two weeks before they were sacrificed in accordance with Home Office regulations. Data are given as mean ± S.E.M. and significance was evaluated using the appropriate Student’s t-tests (*p < 0.05; **p < 0.01).

Genotype regular Na+ ∆ISC-AMI (µA/cm2) Paldo (nM) low Na+ ∆ISC-AMI (µA/cm2) Paldo (nM) high Na+ ∆ISC-AMI (µA/cm2) Paldo (nM)

+/+ n=9 64 ± 14 0.56 ± 0.14 n=9 168 ± 30 1.07 ± 0.29 n = 12 67 ± 15 0.18 ± 0.02

+/L n = 25 82 ± 14 0.17 ± 0.04** n = 20 184 ± 19 0.41 ± 0.06** n = 22 94 ± 11 0.09 ± 0.02**

L/L n = 12 100 ± 19 0.17 ± 0.07* n = 11 280 ± 32* 0.45 ± 0.12* n = 11 152 ± 36* 0.10 ± 0.02*

Under low and high Na+ diet ∆ISC-AMI was significantly larger in L/L animals compared to +/+ animals with a similar trend also under regular Na+ diet. In contrast, Paldo was significantly suppressed in L/L animals compared to +/+ in all three dietary groups. +/L animals had an intermediate phenotype. These observations are consistent with increased epithelial Na+ absorption in Liddle’s syndrome due to a gainof-function mutation of the epithelial Na+ channel (ENaC) which leads to volume expansion and arterial hypertension while the reninangiotensin-aldosterone system is suppressed. University Laboratory of Physiology, Parks Road, Oxford OX1 3PT, UK. P 22-7

P 22-5 EFFECT OF PROSTAGLANDIN E2 ON AMILORID-SENSITIVE SODIUM-CHANNELS IN LUNG EPITHELIA OF XENOPUS LAEVIS A. Berk, M. Fronius, W. Clauss, M. Schnizler We used Ussing-chamber technique to investigate ion-transport processes across Xenopus laevis lung epithelium. We focused our interests on the regulation of the amiloride-sensitive epithelial sodium channel (ENaC) by prostaglandin E2 (PGE 2), which is known to mediate inflammatory processes in many tissues. Apical application of 1µM PGE2 resulted in a 73% stimulation of the amiloride-sensitive sodium current (Iami) and in 57% stimulation of the short circuit current (ISC). Forskolin, an activator of the adenylate cyclase, increasing the cellular concentration of the second messenger cAMP, stimulated Iami in Xenopus lung in a similar manner. Exposure to forskolin and PGE2 revealed no additive effect neither on ISC nor on Iami. Preincubation with 50 µM 8-bromo-cAMP-RP-isomer, an inhibitor of the PKA-pathway blocked

HYPOXIA AFFECTS THE EXPRESSION OF NA/K PUMPS AND ENAC OF A549 CELLS S. Höschele, P. Bärtsch, H. Mairbäurl Hypoxia decreases the activity of transporters Na transporters of primary alveolar epithelial cells. This decrease has been found to be was associated with a decrease in the amount of transport proteins in the plasma membrane. We wanted to know, whether the decrease in transport capacity by hypoxia is caused by a decrease rate of the expression of these transporters. A549 cells were exposed to normoxia and hypoxia (1.5% O 2) as well as cobalt-Cl (Co, 100µM). Since the expression of ENaC was very low, cells were also treated for 48 hours with dexamethasone (DEX, 1µM). Expression was measured by real time PCR. Hypoxia and Co increase the expression of GAPD significantly, expression of 28S-ribosomal RNA was not altered, whereas the expression of Hif-1α decreased slightly. DEX-treatment increases the mRNA-levels of α1-Na/K pumps increased 2-fold, they were not affected by hypoxia and Co in absence and presence of DEX. β1-Na/K pump mRNA was increased significantly by hypoxia and Co; these

S 283 effects were not seen in presence of DEX. α-ENaC was not affected by hypoxia and Co in control cells. As expected, treatment with DEX augments β- and γ-ENaC expression considerably. In presence of DEX hypoxia activates their expression but Co does not. These results indicate that the decrease in activity and transport capacity upon exposure to hypoxia of A549 cells is not associated with changes in mRNA levels and the level of expression. The hypoxia-induced decrease in the number of copies of transport proteins must therefore be caused by internalization and/or degradation of membrane associated transporters. Medical Clinic VII, Sports Medicine, University of Heidelberg, Hospitalstr. 3/4100, D – 69115 Heidelberg, Germany P 22-8 S23E MUTATION OF THE PKC PHOSPHORYLATION SITE OF THE α-SUBUNIT OF Na,K-ATPase LEADS TO INCREASE OF SURFACE EXPRESSION OF PUMPS AND REDUCTION OF THE TRANSPORT RATE L.A. Vasilets, A. Spielmann, W. Schwarz It has been shown that acidic replacement in the PKC phosphorylation site Ser-23 mimics PKC-mediated inhibition of cation transport by the Na,K-ATPase (Vasilets et al. FEBS Lett, 495(1999), 8-12). In addition, substitution of Ser-23 by Glu increased cellular expression of the α-subunit as judged from immunoblots of oocyte homogenates with anti-α antibody. To understand whether targeting of functionally active pumps into the surface membrane is also enhanced, we performed voltage-clamp analysis of ouabain-resistant rat α1 wild-type and S23Emutant Na,K pumps expressed in Xenopus oocytes. From the voltage dependence of the transient charge movements in Na +/Na+ exchange mode, the effective charge zeffe moved by a single pump molecule has been calculated, and from the ratio between total charge and effective charge Qtot/zeffe the number of functioning pump molecules in the cell membrane has been determined. Western blot analysis showed that the total number of expressed S23E, pumps is by a factor of 2.2 higher than the number of wild-type pumps. As with Western blot a much higher number of S23E pumps compared to wild type (factor 1.9) was obtained from the ratio Qtot/zeffe. The E 1/2 value for the voltage-dependent charge distribution was 12 mV in the wild-type and 28 mV in the S23E mutant. Maximum transport activity has been determined by measurements of steady-state pump currents and of ouabain-dependent 86Rb + uptake scaled up to the number of s-active Na,K pump complexes in the surface membrane. Turnover of the reaction cycle of the wild-type Na,K-ATPase was 25 s-1 and was reduced to 11 s-1 in the S23E mutant. The results confirm that the S23E mutation functionally mimics inhibition of transport mediated by PKC-dependent phosphorylation at Ser-23, and that mutation enhances cellular expression of the pump subunit as well as targeting of the active Na,K-ATPase complexes to the surface membrane. Julius-Bernstein-Insitut für Physiologie, Martin-Luther-Universität Halle/Wittenberg, D-06112 Halle/Saale, Germany

SGK coexpression increases the Na+ current carried by pumps containing the rat alpha1 Na,K-ATPase subunits (1.95 ± 0.06 fold) in oocytes. The increase in pump current failed to occur if the kinase site of SGK was disrupted (0.90±0.16), supporting the notion that the kinase activity of SGK is involved. This finding suggest the possibility that SGK plays a key role for the aldosterone-induced increase in Na,KATPase cell-surface expression which is observed in cortical collecting duct and the cell line mpkCCD. The possible role of SGK for this effect is now being investigated by expressing SGK constructs in mpkCCD cells using a retroviral expression system. In conclusion, results obtained in the Xenopus oocyte expression system suggest the possibility that, besides a role for ENaC regulation, SGK also plays a role in the mediation of aldosterone-induced Na,K-ATPase surface expression. 1 Institute of Physiology, University of Zurich, Winterthurerstrasse 190, 8057 Zürich, Switzerland

P 22-10 INTRACELLULAR Na+ AFFECTS TRANSCRIPTIONAL REGULATION OF THE EPITHELIAL Na+ CHANNEL (ENaC) IN RAT LATE DISTAL COLON S. Amasheh, J. Kunkel, H.J. Epple, J. Mankertz, J.D. Schulzke, M. Fromm In colon and kidney, regulation of the epithelial sodium channel (ENaC) is attributed primarily to the mineralocorticoid aldosterone. In rat late distal colon, induction of ENaC activity by aldosterone can be explained on transcriptional level alone. An antagonistic mechanism has not been described so far. We studied the influence of intracellular Na+ (Nai) on ENaC induction. Methods: Mucosal preparations of rat late distal colon (LDC 2, 1-2 cm apart from the anus) were mounted in Ussing chambers. ENaC activation was induced by incubation with 3 nM aldosterone. Electrogenic Na+ transport (JNa) was determined by the drop in short circuit current induced by 100 µM amiloride. Subsequently, tissues were removed and α-, β-, and γ-ENaC subunit mRNAs were detected by Northern blotting. Results: Application of aldosterone (3 nM, 8 h) revealed a distinct induction of JNa. Incubation with 3 nM aldosterone enhanced the expression of β- and γ-ENaC, whereas α-ENaC was reduced. Elevation of Nai by incubation with 1000 U/ml nystatin reduced JNa from 19.2±1.4 to 9.7±2 µmol⋅h–1⋅cm–2 (n=8). On the transcriptional level, a reduction of β- and γ-ENaC by nystatin incubation was detected. A decrease of Nai by the combined incubation for 8 h with aldosterone and amiloride prevented the induction of JNa but induced a pronounced increase of β- and γ-ENaC mRNA, indicating a counter-regulatory effect of Nai on ENaC transcription. Conclusion: Here we report a newly identified regulative mechanism showing that intracellular Na+ concentration influences ENaC regulation on the transcriptional level in a main target tissue of aldosteroneinduced Na+ absorption, the late distal colon of the rat. Dept. of Clinical Physiology and Dept. of Gastroenterology, Benjamin Franklin University Hospital, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany

P 22-9 P 23-1 BOTH APICAL (ENaC) AND BASOLATERAL (Na,K ATPase) Na+ TRANSPORT PROTEINS ARE UPREGULATED BY SGK IN IN VITRO SYSTEMS M. Zecevic1, D. Heitzmann1, D. Pearce2, F. Verrey1 Aldosterone increases Na+ reabsorption across segment-specific cells of the aldosterone-sensitive distal nephron (ASDN). This is mediated by a transcription-dependent mechanism which leads to an increase in the number of active apical epithelial Na+ channels (ENaC) as well as to an activation of basolateral Na+ pumps (Na,K-ATPase). In vitro studies using Xenopus oocytes and cultured epithelial cells strongly suggest that the aldosterone-induced serum and glucocorticoid activated kinase (SGK) plays an important role in this action by mediating an increase in ENaC cell-surface expression. The rapid induction of SGK by aldosterone in the segment-specific ASDN cells of adrenalectomized rats and the shift of ENaC to the cell surface of the same cells (however not along the entire ASDN) are compatible with this hypothesis (Loffing J, Zecevic M et al,. Am. J. Physiol. 280: F675, 2001). We have now addressed the question whether SGK might play a role as a mediator of Na,K-ATPase activation as well using, in a first approach, the Xenopus laevis expression system. Electrophysiological experiments show that

EVIDENCE FOR H+/K+-ATPASE EXPRESSED IN HUMAN POLYMORPHONUCLEAR LEUCOCYTES AND HL60 CELLS S.Chwatal, I. Mühlhahn, A. Dörge, M. Jakab, H. Rossmann, U. Seidler, J. Schirmer, J.Fürst , M Paulmichl., M. Ritter In a previous study we have shown that migration of human polymorphonuclear leucocytes (PMNs) is dependent on cell swelling mediated by the NHE-1 isoform of the Na +/H+ exchanger. In addition we presented evidence for the existence of an H+/K+ATPase which supports chemotaxis in a cell volume dependent manner. The aim of the present study is to further elucidate the functional and molecular properties of this putative granulocyte H+/K+ATPase. Step omission of extracellular Na+ causes intracellular acidification of PMNs, which is accelerated by subsequent addition of SCH 28080, an inhibitor of gastric H+/K+-ATPase or by withdrawal of extracellular K+ ions. The acidification is paralleled by an increase in cell volume (CV), which is further enhanced by fMLP. Gastric H+/K+ATPase blockers inhibit chemotaxis and blunt the fMLP induced increase in CV. In PMNs the fMLP induced CV increase is counteracted by simultaneous exposure of the cells to hypertonicity. Under these conditions chemotaxis is inhibited. Osmotic cell shrinkage

S 284 is followed by regulatory volume increase (VRI), partially due to activation of the Na+/H+ exchanger. The absolute CV is smaller and the initial rate of VRI is significantly retarded in HL60 cells treated with SCH28080. Western blot analysis with a monoclonal anti gastric H +/K+ATPase αsubunit antibody detectes a glycosylated 35 kD protein in lysates of mouse and human gastric mucosa as well as in human PMNs and HL60 cells. An anti gastric H+/K+-ATPase α-subunit antibody detects a band of expected size in HL60 cells. RT-PCR with primers specific for H+/K+ATPase α-subunits reveals products of expected size. Sequencing of the products revealed low homology to identified H+/K+-ATPase α-subunits. The results indicate, that PMNs and HL60 cells express an H+/K+ATPase which contributes along with Na+/H+ exchanger to cell volume regulation and regulation of chemotaxis.

ured, and the virtual short-circuit-current (ISC), and the transepithelial resistance (RT) were calculated from superimposed current-pulses. The integument is a high-resistance epithelium (RT=9.3±1.5 for PW and 8.2±1.9 for ERS (kΩ cm2, n=4 for all values). VT was –6.3±6.3 for PW and 2.9±0.9 for ERS (mV), and I SC was –0.68±0.28 for PW and 0.49±0.17 for ERS (µA/cm 2). Under PW-conditions there was a pronounced inhibition of ISC by 50 µM/l of amiloride (-1.6±0,28 µA/cm2) which was not present under ERS-conditions (0.61±0.16 µA/cm 2). This indicates the up- and down-regulation of apical Na+ channels. Apical furosemide (100 µM/l) was also effective, increased ISC and thus indicated electrogenic Cl - absorption. We were also able to isolate epithelial cells and keep them in primary culture for further investigations. Supported by the DFG.

Department of Physiology, Fritz Pregl Str. 3, University of A-6020 Innsbruck, Austria

Institute of Animal Physiology, Justus-Liebig-University Giessen, Wartweg 95, D-35392 Giessen

P 23-2 P 23-4 HYPOXIA DECREASES ACTIVE NA TRANSPORT ACROSS PRIMARY RAT ALVEOLAR EPITHELIAL CELL MONOLAYERS H. Mairbäurl1, K. Mayer1, K.-J. Kim 2, Z. Borok 2, P. Bärtsch 1, E.D. Crandall2 Hypoxia has been reported to inhibit activity and expression of ion transporters of alveolar epithelial cells. This study extended those observations by investigating the mechanisms underlying inhibition of active Na transport across primary cultured adult rat alveolar epithelial cell monolayers grown on polycarbonate filters. Cell monolayers were exposed to normoxia and hypoxia (1.5% and 5% O2, 5% CO2) and resultant changes in bioelectric properties (i.e., short circuit current (ISC) and transepithelial resistance (Rt)) were measured in Ussing chambers. Results showed that ISC decreased with duration of exposure to hypoxia, the degree of inhibition after 24 hours reaching 54% at 5% O2 and 65% at 1.5% O2, while relatively little change was observed for Rt. In normoxia, amiloride inhibited about 75% of ISC. The amiloridesensitive portion of ISC decreased over time of exposure to hypoxia, whereas the magnitude of the amiloride-insensitive portion of ISC was not affected. Na pump capacity measured after permeabilization of the apical plasma membrane with amphotericin B decreased in monolayers exposed to 1.5% O2 for 24 hours, as did the capacity of amiloridesensitive Na uptake measured after imposing an apical to basolateral Na gradient and permeabilization of the basolateral membrane. These results demonstrate that exposure to hypoxia inhibits alveolar epithelial Na reabsorption by reducing the rates of both apical amiloridesensitive Na entry and basolateral Na extrusion. 1

Medical Clinic VII, Sports Medicine, University of Heidelberg, Hospitalstraße 3, Geb. 4100, 69115 Heidelberg, Germany, and 2Will Rogers Institute Pulmonary Research Center, Department of Medicine, University of Southern California, 2011 Zonal Avenue, Los Angeles, CA 90033, USA

P 23-3 ELECTROGENIC SODIUM AND CHLORIDE TRANSPORT ACROSS INVERTEBRATE EPITHELIA S. Krumm, S. Goebel, M. Schnizler, W. Clauss Many ion transport systems have anchestors found early in evolution, as for example the epithelial sodium channel ENaC archetype is already present in nematode worms. As invertebrates have ADH-like peptide hormones, but lack the specific mineralocorticoid regulation by aldosterone, it is of particular interest to use them as simplified models for elucidating the regulatory pathways and components of epithelial transport regulation. Since some years we have established such a model from annelid integument (leech). We now report about another model (earthworm) from this phylum, which may even be more useful for this purpose. Dissected segments of earthworm integument were mounted in modified Ussing chambers and perfused with earthworm-Ringer-solution (ERS) on the basolateral side, and with either artificial pond water (PW) or ERS on the apical side. In order to investigate ion transport under near in vivo physiological conditions, the epithelia were not voltage-clamped, but investigated under current-clamp conditions. The transepithelial potential (VT) was meas-

ELECTROGENIC CHARACTERISTICS OF THE TYPE IIB NA+/ PI COTRANSPORTER REVEAL A UNIQUE KINETIC BEHAVIOR I.C. Forster, G. Stange, J. Biber, H. Murer The type IIb Na +-coupled P i cotransporter (NaPi-IIb), a potential candidate for mediating Pi uptake in the mammalian small intestine, is also expressed in brain, lung, liver and testes, but not in kidney. Previous studies (Hilfiker et al., (1998) PNAS 95:14564-69;de la Horra et al., (2000) J. Biol. Chem. 265: 6248-6287) reported that NaPi-IIb has significantly higher apparent substrate affinities and reduced pH dependency compared to the type IIa isoform (NaPi-IIa), the key player in renal Pi reabsorption. These kinetic differences warranted a detailed study of NaPi-IIb by expressing it in Xenopus oocytes and assaying electrogenic behavior by means of the two electrode voltage clamp. In the steady-state (external pH 7.4), P i activation was Michaelian with a Hill coefficient close to unity over a wide voltage range. The apparent affinity for Pi (KmPi ) increased with decreasing external Na +, consistent with the ordered substrate binding model previously proposed for NaPi-IIa and Na+ activation was sigmoidal with a Hill coefficient >1.5. Surprisingly, in the voltage range –100 to 0 mV, KmPi decreased from 14 µM to 5 µM under saturating external Na + (100 mM), whereas like NaPi-IIa, the apparent Na + affinity increased (from 20 to 40 mM) at saturating Pi (1 mM). Unlike NaPiIIa, steady-state I-V curves were curvilinear with no apparent voltage independent rate limiting behavior in the range –160 mV to 0 mV. Moreover, the I-V curves obtained at pH 6.2 were little changed from those at pH 7.4 (1 mM total Pi). However, presteady-state relaxations, which are purported to reflect binding/debinding of Na+ ions and carrierintrinsic charge movements, were markedly altered by increasing [H+]: in 100 mM Na+, a change in external pH from 7.4 to 6.2, completely suppressed charge movements, whereas in the absence of external Na+, only the midpoint voltage of the empty carrier charge distribution V0.5 was altered. These findings suggest that protons chiefly interact only with the first Na+ binding step, but neither prevent binding of Pi nor interact with the final Na+ binding steps. The reduced pH dependency, suggests that NaPi-IIb is able to transport both divalent and monovalent Pi with high binding affinity in an electrogenic manner, which would thereby allow transepithelial P i transport in the acidic microenvironment of the mammalian gut. Institute of Physiology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland.

P 23-5 MOLECULAR CLONING OF A NOVEL Na/P i COTRANSPORTER FROM ZEBRAFISH KIDNEY C. Graham, P. Nalbant, B. Schölermann, H. Hentschel, R.K.H. Kinne, A. Werner Zebrafish express two isoforms of the type IIb Na/P i cotransporter. Type IIb1 has previously been cloned and functionally characterised (Nalbant et al, 1999). Here, we report the cloning of the IIb2 transcript from zebrafish kidney, its localisation and functional characterisation. Total RNA was extracted from kidney. Primers were designed against highly conserved regions of the type IIb transporter. RT-PCR yielded

S 285 a specific product. 3’ RACE and subsequently 5’ RACE resulted in a specific fragment which was cloned and sequenced. The combined clone displayed a high degree of homology with published type IIb cotransporter sequences. Specific antibodies were raised against a Cterminal epitope of both NaPi-IIb1 and IIb2 isoforms. Antisera specificity was confirmed by Western blot of zebrafish tissue lysates and labelling sections of Xenopus oocytes expressing NaPi-IIb1/IIb2, respectively. Immunohistochemical analysis revealed expression of both isoforms in renal and intestinal epithelia, as well as in bile ducts and gills. Both IIb1 and IIb2 were detected in the apical membrane of epithelial cells. The novel clone was expressed in Xenopus oocytes and transport was analysed by two-electrode voltage clamp technique. Phosphate induced currents were Na- and voltage dependent. Preliminary data suggest that the IIb2 isoform has a greater affinity for Pi (Km 30 µM) than NaPi- IIb1 (Km 250 µM). This difference in Km may indicate different physiological roles for NaPi-IIb1 and IIb2. We hypothesise that IIb1 could act as a low affinity/high capacity system for absorption of the bulk of the filtered phosphate complemented by IIb2 fine tuning phosphate reabsorption. Nalbant, P. et al, 1999. J. Physiol. 520: 7989. Department of Physiological Sciences, University of Newcastle, Newcastle, UK and Max-Planck-Institut für Molekulare Physiologie, Dortmund, Germany P 23-6 HETERODIMERIC AMINO ACID TRANSPORTERS OF CAENORHABDITIS ELEGANS 1 E. Veljkovic, P. Skelly, S. Stasiuk, 1F. Verrey , C. Shoemaker The mammalian heterodimeric amino acid transporters are composed of a type II glycoprotein heavy chain, linked by a disulfide bond to a multi-transmembrane domain light chain that belongs to the new family of glycoprotein-associated amino acid transporters (gpaATs). Subunit interaction, but not the disulfide bond per se, has been shown to be required for functional surface expression. Amino acid transporters of this family play important roles for the transport of various amino acids, in particular in proliferating cells and transporting epithelia. Sequence analysis of the C. elegans genome reveals the presence of nine light chain (cPRM1-9) and two heavy chain (cBAT1,2 ) homologues. Only three C. elegans light chains have the conserved cysteine responsible for disulfide bond, suggesting the possibility that the six others might be able to function in the absence of covalent heterodimerization. cDNAs for each of the C. elegans transporters have been isolated and their heterodimerization and transporter function are now being tested in the Xenopus oocyte expression system. SDSPAGE analysis of cPRM1, cPRM3, cBAT1 and cBAT2 immunoprecipitated from metabolically labeled oocytes shows specific bands with the expected Mr. In addition cPRM1 co-precipitates with both cBAT’s but not with the mammalian heavy chain rBAT. Functionally, cPRM1 and cPRM3 induce an uptake of neutral amino acids when co-expressed with cBAT2, but not when co-expressed with any other C. elegans or mammalian heavy chain. The localization of the different heavy and light chains is being studied by immunofluorescence in C. elegans and Xenopus oocytes. These experiments show that the light chains cPRM1 and cPRM3 require the coexpression of cBAT2 for surface expression in the oocytes. In vivo localization as well as dsRNA interference and gene targeting approach will shed light onto the functional roles played by selected transporters in the context of the whole organism. In conclusion, we have identified two pairs of polypeptide chains (cPRM1-cBAT2 and cPRM3- cBAT2) that form the first functional, non-mammalian heterodimeric amino acid transporters to be characterized. The study of all members of the C. elegans heterodimeric amino acid transporter family will provide new insight into the role of this protein family within a whole organism. 1

Institute of Physiology, University of Zürich, Winterthurerstrasse 190, 8057 Zürich Switzerland P 23-7

RESIDUES IN TWO OPPOSING LINKER REGIONS OF THE RENAL TYPE IIA NA+/PI COTRANSPORTER CODETERMINE A PUTATIVE TRANSPORT PATHWAY K. Köhler, I.C. Forster, G. Stange, J. Biber, H. Murer Intrasequence comparison of the NaPi-IIa cotransporter revealed two

regions with high similarity in the 1st intracellular (ICL-1) and the 3rd extracellular loop (ECL-3). The functional importance of both ECL3 and ICL-1 has been shown previously (Lambert et al., J. Gen. Physiol., 2001, Jun; 117(6): 533-46., Kohler et al., Amer. J. Physiol., 2001, in press) by applying the substituted cysteine accessibility method (SCAM). Modification of cysteines in both ICL-1 and ECL-3 led to full inhibition of cotransport activity. Given that there is only one Pi translocation pathway within the NaPi-IIa protein, these findings would indicate that the linker regions are functionally associated to contribute to a common transport pathway. To further elucidate the common role of both regions, we constructed a double mutant (A203C-S460C) with cysteines introduced in both linker regions and assayed the effect of modification on the transport activity under two electrode voltage clamp conditions. We investigated two electrogenic transport modes before and after exposure to MTS reagents: a Na+-dependent slippage mode in the absence of external P i (stoichiometry = 1 Na+), and a cotransport mode (assumed stoichiometry = 3 Na +: 1 HPO 42-). To elucidate sidedness and charge dependency of MTS reaction, we used permeable (MTSEA) versus non-permeable (MTSES, MTSET) reagents that carry either a positive (MTSEA, MTSET) or a negative charge (MTSES). Application of MTSEA and MTSES led to a full inhibition of cotransport of the A203C-S460C mutant and increased the slippage mode, whereas incubation with MTSET suppressed the cotransport mode only. Determination of Pi and Na+ dose responses showed that the apparent binding affinities for both substrates were unchanged after modification. Further mutagenesis at these was performed. The double mutant A203S-S460C was fully inhibited after incubation with MTSEA and MTSES, but no increase in slippage mode was observed, whereas other mutations (A203L/E/R-S460C, A203CS460E/R) led to non-functional transporters. These findings indicate that both cysteines within the double must be modified to alter the slippage mode. Both single mutants, A203C and S460C exhibit a higher slippage than the WT protein already before Cys modification. As previously reported, neither of the single mutants shows an increase in slippage after incubation with MTS reagents. However, in the double mutant, modification of both cysteines resulted in an increased slippage, which may be explained by a cooperative effect between these two sites in determining the kinetics of the slippage pathway of NaPiIIa. Institute of Physiology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland. P 23-8 EXPRESSION OF THE TYPE IIB NA/PI COTRANSPORTER IN Sf9 CELLS T. Radanovic, H. Murer., J. Biber In mammals the type IIb Na/Pi-cotransporter is expressed in various tissues such as intestine, lung and testis. In intestine, the type IIb Na/ Pi-cotransporter is involved in absorption of phosphate and its abundance is regulated by vitamin D3 and phosphate diet. The type IIb cotransporter shows 51% identity with the renal type IIa Na/Picotransporter for which a detailed model of the secondary structure has emerged based on recent structure/function studies. To make the type IIb Na/Pi-cotransporter available for the future structural studies, we have expressed this cotransporter in Sf9 cells. Sf9 cells were infected with recombinant baculovirus containing the IIb protein as a 6His fusion protein. Infected cells expressed a polypeptide of 80 kDa corresponding to the ungycosylated form of the type IIb cotransporter. Transport studies demonstrated that the type IIb protein expressed in Sf9 cells mediates the transport of the phosphate in a Na-dependent manner with similar kinetic characteristics (apparent Km’s for Pi and Na) as previously described. Moreover, Na/Pi cotransport was not dependent on external pH as was also shown for the type IIb protein expressed in Xenopus laevis oocytes. Solubilization experiments demonstrated that, in contrast to the type IIa cotransporter, the type IIb could be solubilized by a number of non-ionic detergents. Solubilized type IIb Na/Pi-cotransporter was purified by Ni-NTA chromatography. These results show that the type IIb Na/Pi-cotransporter can be functionally expressed in Sf9 cells and that expressed type IIb protein can be solubilized and purified. Thus, our results my lead to the purification of the type IIb protein in amounts large enough for future structural studies. Institute of Physiology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland

S 286 P 23-9 ROLE OF NAPI-CAP1 IN THE REGULATION OF TYPE IIA NA/ PI COTRANSPORTER N. Deliot, N. Hernando, S. Gisler, J. Biber, H. Murer The type IIa Na/P i cotransporter plays an important role in the reabsorption of inorganic phosphate (Pi) in the renal proximal tubule. This cotransporter is localized in the proximal brush border and its expression is regulated by several hormones and factors. Using the yeast two hybrid system Gisler et al (J.B.C. 2001; 276:9206-9213) identified several proteins that specifically interact with the C-terminal tail of the type IIa Na/Pi cotransporter. We are now focused on one of these proteins called NaPi-Cap1 that is also localized in the brush border membrane of the proximal tubules. This protein contains 4 PDZ domains and interacts via the PDZ domain 3 with the 3 last amino-acid (TRL) of the type IIa Na/Pi cotransporter. To study the physiological relevance of this protein-protein interaction, we are using a cell line derived from Opossum proximal tubules, the OK cells, in which the type IIa Na/Pi cotransporter also has an apical expression. Recent studies showed that overexpression of the PDZ-domain 3 of NaPi-Cap1 in OK cells induced an inhibition of the apical expression of the endogenous type IIa Na/Pi cotransporter. In addition, we could show that this cell line express the mRNA of NaPi-Cap1. However, further experiments were prevented due to the fact that the only antibody available (derived from the rat form of NaPi-Cap1) does not recognize the Opossum protein (Western blot and immuno-staining). Therefore we have screened an OK-cell cDNA library, using as probe the whole open ready frame (ORF) of the mouse NaPi-Cap1 cDNA. We obtained several clones containing inserts of 2 Kb. These clones contain the complete ORF of the OK protein and partial 5' and 3' untranslated sequences. This sequence information will allow us to raise antibodies specific for the OK protein, in order to investigate weather the interaction of NaPi-Cap1 with the type IIa Na/P i cotransporter plays a role in the regulation (apical expression/ regulated endocytosis) of the cotransporter. Institute of Physiology, University of Zurich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland P 23-10 THE Na/Pi-IIa COTRANSPORTER PROTEIN – PROTEIN COMPLEX S. Pribanic, S. Gisler, D. Bacic, J. Biber, H. Murer The type IIa Na/Pi-cotransporter is localized in the apical membrane of renal proximal tubular (PT) cells and its abundance is under control of a variety of regulators. Increased rates of endocytosis and subsequent lysosomal routing of the type IIa protein due to hormonal stimuli can be observed. To identify proteins involved in above mentioned processes (apical localization, endocytosis), several yeast two-hybrid screens were performed using a whole mouse kidney cDNA library. In this study we describe candidate proteins which potentially interact with the N-terminus (Nt) and an intracellular domain (3rd IC loop) containing a motif necessary for the parathyroid hormone response of the cotransporter. In a recent screen with the Na/Pi-IIa Cterminus, several PDZ-protein interactors were identified, such as NHERF-1 and Na/Pi-Cap1 (Gisler et al. JBC 276: 9206-, 2001). In the Na/Pi-IIa–Nt screen interacting protein candidates were identified as MAP17 (membrane associated protein of 17kDa), VILIP-3 (visinin like protein) and Ser/Thr phosphatase catalytic subunit 2α. Only one protein, HP33 (hepatocellular carcinoma enriched 33 kDa protein), was found to interact with the 3rd IC loop. Immunostainings performed on mouse kidney slices indicated that MAP17 was exclusively localized in the microvilli of PT, whereas VILIP-3 had an apical PT as well as a distal tubular localization. HP33 (in PT cells only) and Ser/Thr phosphatase catalytic subunit 2α (throughout the whole nephron) were detected in the cytoplasm. Interestingly, MAP17 has been described to interact with PDZK1, a human homologue of Na/Pi-Cap1 (Kocher et al., Lab. Invest. 78:117-,1998), indicating the existence of a protein complex comprising Na/Pi-Cap1, Na/Pi-IIa and MAP17 in PT microvilli. GST pull-down experiments and yeast trap assays indicated that MAP17 interacted with Na/Pi-Cap1 only and not with any other microvillar PDZ proteins such as NHERF-1. In summary, our findings suggest that the type IIa Na/Pi - cotransporter is part of at least two different heteromultimeric protein complexes. Such complexes are supposed to play major roles in the apical positioning of Na/Pi-IIa, and in scaffolding proteins involved in signal transduction pathways.

Furthermore, several proteins were identified that might be implied in the regulation (internalization, lysosomal routing) of the type IIa cotransporter. Institute of Physiology, University of Zürich, Winterthurerstr. 190, CH – 8057 Zürich, Switzerland P 24-1 PROTEIN KINASE C MEDIATES CLATHRIN-DEPENDENT ENDOCYTOSIS OF THE HUMAN NONGASTRIC H+,K+-ATPASE IN TRANSFECTED MDCK CELLS H. Bertram, M. Kosch, M. Lerner, D. Lemke, H. Oberleithner, J. Reinhardt In epithelial cells of the kidney collecting duct two subgroups of H+,K+ ATPases are expressed. The gastric H+,K+-ATPase which also mediates H+ secretion in stomach parietal cells and the non gastric H+,K+-ATPases which reabsorb K+ in exchange to H+ or Na+ ions. Both ion-pumps are localized to the luminal plasma membrane, both in vivo and in transfected renal epithelial cells. It is presumed that renal H+,K+-ATPases can be regulated by endocytosis. However little is known about the molecular mechanisms which control plasma membrane expression of renal H +,K+-ATPases. We show that apical expression of the human H+,K+-ATPase (ATP1AL1) in stably transfected MDCK cells can be regulated by the Protein Kinase C (PKC) pathway whereas Protein Kinase A does not influence ion-pump expression in our transfected cell system. Activation of PKC with phorbol-esters (PMA) or with the synthetic diacylglycerol analog 1,2 Dicapryloyl-sn-glycerol (DOG) leads to fast ion-pump internalization. After 20 minutes of PMAstimulation (100 nM, 37°C) the ATPase is almost completely internalized as shown by confocal immunofluorescence and surface biotinylation experiments. Functional inactivation of the H+,K+-ATPase due to PKC activation was additionally shown by intracellular pH measurements. Proton extrusion capacity of ATP1AL1 transfected cells is drastically reduced after PMA incubation. Ion pump internalization and inactivation could be totally prevented with the PKC blocker bisindolylmalenimide. In contrast activation of the PKApathway does not influence the H+-extrusion capacity of transfected MDCK cells as shown with pHi measurements. Immunofluorescence analysis indicates that PKC mediated H+,K+-ATPase internalization is based on Clathrin-dependent endocytosis. Our results show that the non gastric H+,K+-ATPase is a specific target for the PKC pathway. Therefore PKC mediated phosphorylation is a potential regulatory mechanism for apical non gastric H+,K+-ATPase plasma membrane expression. University of Muenster, Institute of Physiology, Robert Koch Str. 27a, 48149 Münster, Germany P 24-2 MEMBRANE CAPACITANCE INCREASES IN XENOPUS OOCYTES DUE TO EXPRESSION OF THE PEPTIDE TRANSPORTER PEPT1 M. Mertl, H. Daniel, G. Kottra Transport-induced current and surface-proportional membrane capacitance were measured in voltage-clamped Xenopus oocytes expressing the intestinal peptide transporter PEPT1 as well as in waterinjected (WI) oocytes. The average membrane capacitance of WI oocytes was 215 ± SEM 10 nF (n=6) suggesting a surface enlargement by a factor of 6.8±0.5 as compared to the smooth oocyte surface calculated for the diameter of 1.01±0.03 mm. PEPT1-expressing oocytes showing a mean dipeptide-induced inward current of 505±75 nA (n=9) displayed an increased membrane capacitance of 342±6 nF corresponding to a 9.6±0.6-fold surface enlargement (oocyte diameter 1.08±0.04 mm). Incubation for 5 hours in a solution containing 5mM of the dipeptide Gly-L-Gln had no effect on WI oocytes, but reduced the transport current and the membrane capacitance of PEPT1oocytes to 149±16 nA (31±3%) and 223±13 nF (66±4%), respectively. Shorter (2 h) incubation reduced only the current (from 451±21 to 226±33 nA, n=3), but not the membrane capacitance. Incubation of oocytes with the peptide transport inhibitor Lys[Z(NO2)]-Pro caused only minor changes suggesting that the transport step per se and not occupation of the substrate binding site is required for capacitance effects. The mechanism underlying the observed changes was studied in oocytes exposed to substrate and 1µM cytochalasin D (CD). CD

S 287 markedly blunted the response to substrate exposure (current reduction to 70±6%, capacitance reduction to 81±2%, n=13) suggesting that the cytoskeleton is involved in these processes. Stimulation of PKC with the phorbol esther PDBu (known to induce endocytosis in other cell systems) reduced dipeptide-induced currents to 50±3% and the membrane capacitance even to 45±5%. These effects could completely be blocked by co-application of CD. However, the PKC-inhibitor bisindolylmaleimide failed to suppress the dipeptide-induced current and capacitance changes leaving the question concerning the role of PKC in the dipeptide-evoked changes open. Nutritional Physiology, Hochfeldweg 2, D-85350 Freising-Weihenstephan, Germany P 24-3 AMPLIFICATION OF INWARD AND OUTWARD TRANSPORT CURRENTS BY THE SIMULTANEOUS PRESENCE OF DIPEPTIDES ON BOTH SURFACES OF PEPT1-CONTAINING XENOPUS OOCYTE GIANT MEMBRANE PATCHES G. Kottra, H. Daniel We have recently shown that the intestinal peptide transporter PEPT1 transports neutral and charged dipeptides not only in its normal (outside to inside), but also in the reverse direction (J. Physiol 536: 495, 2001). In this study, we characterized transport currents in giant patch clamp experiments (inside-out configuration) when equally or differently charged substrates were simultaneously present on both membrane surfaces. Patch pipettes were filled with solutions containing neutral (Gly-Gln, GQ), positively (Lys-Gly, KG) or negatively (Gly-Asp, GD) charged dipeptides, and the induced inward transport current and its voltage dependency (between -160 and +60 mV) were repeatedly measured, with the help of the non-competitive inhibitor glibenclamide added to the cytosolic solution (CS), at increasing cytosolic concentrations of a dipeptide of the same or different charge. In the absence of substrates in the CS, currents driven by the membrane potential and by the transmembrane chemical gradient were inwardly directed in the whole potential range investigated. Addition of a substrate to the CS induced near to +60 mV an outward current component the size of which depended on the charge of substrates present on both sides of the membrane, and became maximal with extracellular GD, and cytosolic KG. The inward current measured near to –160 mV decreased when the cytosolic concentration of GQ was increased in the continued presence of extracellular GQ. By contrast, when KG or GD were present on both sides of the membrane, inward currents increased compared to the currents measured in the absence of cytosolic substrates. The enhancement of inward transport varied with the different substrates added to both sides of the membrane. These observations suggest an increase in the turnover rate of the transporter in the presence of transportable substrates on both sides of the membrane and thus extend the results of earlier trans-stimulation experiments. Nutritional Physiology, Hochfeldweg 2, D-85350 Freising-Weihenstephan, Germany P 24-4 LOCALISATION OF PEPT2 mRNA AND PROTEIN IN DORSAL ROOT GANGLIA D.A. Groneberg, F. Döring, H. Daniel, A. Fischer The peptide-transporter PEPT2 mediates electrogenic uphill transport of di- and tripeptides, selected peptidomimetics and delta aminolevulinic acid. PEPT2 was recently cloned from rat central nervous tissue and its mRNA was localized to astrocytes of the central nervous system. In the present studies the expression of PEPT2-protein and -mRNA in rat dorsal root ganglia was investigated. Immunohistochemistry revealed PEPT2-immunoreactivity in satellite glia cells surrounding the ganglionic neurons. There was no expression in neuronal cells. In-situhybridization studies colocalized the expression of PEPT2-mRNA to satellite cells. This is the first report on the expression of PEPT2protein in the peripheral nervous system where PEPT2 may serve as uptake system for products of protein degradation, for removal of biologically active short-chain peptides and non-peptides such as delta aminolevulinic acid. Dept. of Pediatric Pneumology and Immunology, Charite Campus Virchow, BMFZ Forum 4, Augustenburger Plate 1, D-13353 Berlin

P24-5 CLONING AND FUNCTIONAL CHARACTERIZATION OF A PROTON/AMINO ACID TRANSPORTER FROM MURINE SMALL INTESTINE M. Foltz, M. Boll, I. Rubio-Aliaga, G. Kottra, H. Daniel Although H + -dependent transport of amino acids across apical membranes of mammalian epithelial cells of the small intestine and kidney tubules have been functionally described, these amino acid transporters are not identified at the molecular level. Starting from murine small intestinal RNA, by using RT-PCR we isolated a cDNAclone (mPAT1, proton/amino acid transporter). The cDNA encodes a membrane protein with 475 amino acids and 11 putative transmembrane domains. The only closely related protein found as yet in mammals is the recently cloned rLYAAT1. Northern blot analysis revealed that the mPAT1 mRNA is expressed in the small intestine, kidney, brain and colon. For functional characterization we used the X. laevis oocyte expression system and employed flux studies and electrophysiological analyses. Oocytes injected with mPAT1 cRNA showed a strong increased uptake of Gly, L-Ala and L-Pro. Two electrode voltage studies revealed an electrogenic transport mode, that proved to be independent on Na+, Cl- and K+ ions. Lowering the extracellular pH increased substrate induced currents in mPAT1 expressing oocytes and induced a pronounced decline in pHin. Out of the 20 proteinogenic amino acids the transporter recognizes only small amino acids, such as Gly, Ala, Pro and to a smaller degree also Ser. The Km-values for these substrates varied from 2.8 mM (L-Pro) to 7.5 mM (L-Ala). Distinct structural analogues of these amino acids, such as D-isomers (D-Ala, D-Pro) or derivatives (taurine, betaine) interact as well with the transporters substrate binding site. In conclusion we have cloned mPAT1, an H +/ amino acid transporter, whose functional properties are very similiar to the H+/amino acid transport process described in the apical membrane of Caco-2 cells and renal brush border membrane vesicles. Institute of Nutritional Sciences, Technical University of Munich, Hochfeldweg 2, 85350 Freising, Germany

P 24-6 LOCALIZATION AND CHARACTERIZATION OF THE ORPHAN TRANSPORTERS XT2 AND XT3 THAT BELONGS TO THE FAMILY OF Na+,Cl- -DEPENDENT NEUROTRANSMITTER AND AMINO ACID TRANSPORTERS Z. Ristic1, S.M.R. Camargo1, J. Loffing2 and F. Verrey1. Functionally characterized members of the Na + and Cl --dependent neurotransmitter and amino acid transporter family have been shown to transport monoamines, amino acids or organic osmolytes. We investigate the role of two additional related family members (XT2 and XT3) with unknown function that are expressed at the mRNA level in kidney and small intestine. In vitro translation and Western blots from mouse kidney brush border membranes probed with affinitypurified antibodies showed that XT2 and XT3 migrate at ~70 kDa on SDS-PAGE. Imunohistological experiments on rat kidney sections showed that XT2 and XT3 were localized to the brush border membrane of the proximal tubule. XT2 was predominantly expressed in the S3 segment, whereas XT3 was more prominent in the S1 and S2 segments. However, expression of these orphan transporters in oocytes did not induce any amino acid transport activity, even in combination with several other gene products, suggesting that some other epithelial or kidney cell-specific factor might be required for their surface expression and / or function. Alternatively, these proteins transport substrates that have not yet been identified. 1 2

Institute of Physiology, Winterthurerstr. 190, CH-8057 Zürich and Institute of Anatomy, University of Zurich, Switzerland

P 24-7 SHORT-TERM EFFECT OF ALDOSTERONE ON Na,K-ATPASE CELL-SURFACE EXPRESSION IN KIDNEY COLLECTING DUCT CELLS V. Summa 1*, D. Mordasini*, F. Roger, M. Bens, P.-Y. Martin, A. Vandewalle, F. Verrey1, E. Féraille (*contributed equally) Aldosterone plays an important role for body Na + and volume

S 288 homeostasis by regulating the reabsorption of Na+ across tight epithelia, such as that of the aldosterone-sensitive distal nephron (ASDN). Its major early effect (within 0.5 to 3 h) is the stimulation of luminal Na+ entry into target cells which is mediated by an increase in apical cellsurface expression of the amiloride-sensitive epithelial Na+ channel (Loffing et al., 2001. Am J. Physiol. 280: F675). The Na,K-ATPase (Na + pump), that extrudes Na+ across the basolateral membrane, has been known since two decades to be activated in the short term by aldosterone as well (enzymatic activity and ouabain binding) in the ASDN. We have now addressed the question of the mechanism by which aldosterone stimulates Na,K-ATPase using filter cultures of a mouse kidney cortical collecting duct cell line (mpkCCD). The function of cell surface Na,K-ATPase was assessed at controlled intracellular Na + concentrations by measuring the ouabain-sensitive pump current (IP) on epithelia selectively permeabilized apically to monovalent ions with amphotericin B. Aldosterone increased the maximal IP by 30 % within 2 h. The apparent Na+ affinity was unchanged by aldosterone (8.5 and 10.1 mM, respectively). The increase in IP was inhibited by actinomycin D (dependent on ongoing transcription) but not by amiloride (not secondary to apical Na + influx). To assess whether aldosterone increased the amount of surface-expressed Na,K-ATPase and/or its total cellular amount, basolateral cell surface proteins were biotinylated and immunoblotting of Na,K-ATPase α subunit was performed with and without a prior streptavidin-precipitation. Na,KATPase α subunit increased at the cell surface by approximately 30% within two hours, whereas the total α subunit pool increased at a later time point (late effect). Similarly to the functional results, the early increase in cell surface Na,K-ATPase was not inhibited by amiloride. Interestingly, ex-vivo studies indicate that aldosterone similarly acts on the surface expression of rat cortical collecting duct Na,K-ATPase, but to a larger extent. In conclusion, aldosterone increases the cellsurface expression of Na,K-ATPase and thus, independent of an increase in intracellular Na , coordinately acts onto the surface expression of Na transport molecules on the apical (ENaC) and basolateral (Na,KATPase) membrane domains of its renal target cells. Furthermore, since a qualitatively similar effect of aldosterone was detected in the mouse collecting duct principal cell line mpkCCDcl4, this cell line proved to be a useful in vitro model for further investigations. +

P 24-9 EVIDENCE FOR COLOCALIZATION OF TWO CALCIUM TRANSPORTERS IN PIG DUODENUM T. Hinterding, R.J.M. Bindels*, K. Huber, B. Schroeder BACKGROUND: It has been shown in laboratory animals as well as humans that active Ca2+ absorption in intestine involves Ca2+ transport across the brush border membrane of the enterocytes via Ca2+ influx channels. Originally, in rabbit intestine an epithelial calcium channel (ECaC1) was identified which has approximately 80% homology with the rat calcium transport protein 1 (CaT1, ECaC2). OBJECTIVE: The aim of the present study was to investigate the existence of ECaC1/2 homologues in pig intestine to upgrade the facilities to study Ca metabolism in this animal model. METHODS: RT-PCR with primers derived from conserved regions of known ECaC cDNA sequences was used to identify respective homologues in pig duodenum. Studies of ECaC expression level along the longitudinal axes of the intestine were performed by Northern analyses. In preliminary experiments relative ECaC-mRNA quantification (based on β-actin) was performed comparatively in duodenum from suckling and weaned piglets by real-time PCR with the SYBR green, AmpliTaq Gold DNA polymerase and GeneAmp 9700 thermal cycler PCR system. RESULTS: Data from RT-PCR revealed ECaC homologues with high homology to ECaC1 and CaT1. ECaC expression was higher in the duodenum than in the jejunum and was completely absent in the ileum. Small but significant expression levels were observed in the proximal and distal colon. Data from real-time PCR point to an age-dependency with respect to different ECaC expression. CONCLUSION: The identification of partial sequences of ECaC homologues that are likely to be responsible for Ca2+ entry at the apical membrane in pig duodenum provides the tools for further studies, i.e. adaptation of active Ca2+ absorption during early postnatal life or the role of calcitriol in regulating apical Ca2+ influx. Supported by DFG Schr 342/5-1

+

1

Institute of physiology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland

P 24-8 CLONING OF A NOVEL Na + AND Cl - -DEPENDENT HCO 3TRANSPORTER FROM SQUID GIANT FIBER LOBE L. V. Virkki, I. Choi, B. A. Davis, C. Franke, W. F. Boron The squid giant axon is a classic preparation for the study of neuronal electrophysiological phenomena. Historically, it has been shown that Na+ -driven Cl-/ HCO3- exchange is the main pH-regulatory mechanism in the axon. It was our goal to clone this transporter from squid. Using a degenerate primer approach on Loligo pealei giant fiber lobe (GFL) reverse-transcribed cDNA, we identified a novel, 300-bp sequence related to the bicarbonate transporter superfamily. A full-length coding region was obtained by a combination of library screening and PCR. The amino acid sequence showed 46% identity to mammalian NBCs and 33% identity to the anion exchanger. Northern blot analysis showed strong hybridization to a ~7 kb transcript in GFL, optic lobe and heart, whereas weaker signals were obtained in gill and stellate ganglion. We expressed the novel clone in Xenopus oocytes, and assayed pHregulatory function using pH-sensitive microelectrodes. Superfusion with 5% CO2/33 mM HCO3- induced a fall in intracellular pH, followed by a slow recovery of pH i . No initial rapid hyperpolarization was observed, indicating that the pHi recovery mechanism is electroneutral. The pHi recovery required extracellular Na+ and was blocked by DIDS (50 µM). When Na +o was removed, the transporter reversed and the cell acidified. If extracellular Cl - was removed at this time, the acidification stopped, indicating that the transporter requires Cl- for activity. In conclusion, the characteristics of this novel clone fit those of the Na+ -driven Cl-/ HCO3- exchanger. Yale University School of Medicine, Department of Cellular and Molecular Physiology, 333 Cedar St, New Haven, CT 06520 USA

Department of Physiology, School of Veterinary Medicine, Bischofsholer Damm 15, D-30173 Hannover, Germany; *Cell Physiology, University of Nijmegen, The Netherlands P 25-1 ABSENCE OF REGULATORY VOLUME INCREASE (RVI) IN ISOLATED BOVINE ARTICULAR CHONDROCYTES M.J. Kerrigan, I. Matskevitch, P.W. Flatman, A.C. Hall Chondrocytes, the only cell type of articular cartilage, are responsible for the turnover of the extracellular matrix in response to signals from the physico-chemical environment. Matrix synthesis is influenced by chondrocyte volume, being decreased by hyper-tonicity (Urban et al., 1993). It is not clear if chondrocytes are able to regulate their volume in response to a hyper-osmotic challenge, and in this study chondrocyte RVI and changes in intracellular calcium ([Ca2+]i), a possible regulator of the response were studied. Load-bearing, articular cartilage explants were excised aseptically from bovine metacarpal phalangeal joints (from the local abattoir) and chondrocytes isolated using a standard technique into DMEM (Hall et al., 1996). For measurements of cell volume and [Ca2+]i, chondrocytes were incubated with fura-2 AM (37 oC; 30mins) and experiments performed as described (Kerrigan & Hall 2000). A 43% increase in osmolarity was delivered to cause cell shrinkage. Data were expressed as mean ± s.e.m. with number of joints (n) and cells ([N]). In response to the hyper-osmotic challenge, there was no difference (p>0.05) in the extent of shrinkage between cells isolated in either 280 or 380 mOsm. There was no difference (p>0.05) in either the % of cells showing RVI, or the rate of RVI with <5% responding (n=8, N=138). The post RVD-RVI protocol did not significantly increase in the number of cells showing RVI despite robust RVD (81.6±8.9% (380mOsm); 58.1±16.6% (280mOsm) data are % of cells showing RVD) in either of the isolating osmolarities (n=7, N=211). In some cells, both the hypo- and hyper-osmotic challenges resulted in a rise in [Ca2+]i that subsequently returned to basal levels. There did not appear to be a correlation between these changes and the capacity for volume regulation (n=15, N=349). Western blot analysis of whole chondrocyte membrane proteins using a specific mouse monoclonal antibody (T4; Developmental Studies Hybridoma Bank, Univ. Iowa) to the Na-K2Cl co-transporter showed a band at ~200KDa, consistent with the presence of the co-transporter. These results show that freshly-isolated bovine chondrocytes possess a Na-K-2Cl transporter but are not able to undergo RVI either as a result of a 43% hyper-osmotic challenge or

S 289 by the post RVD-RVI protocol. It is possible that the capacity for RVI is lost as a result of cell isolation. However, it may be that due to the diurnal cyclic hydration and dehydration of cartilage, the capacity for RVI is unnecessary, as the extracellular osmolarity will return towards the ‘resting’ value as cartilage re-hydrates when it is unloaded. Hall, A.C., Starks, I., Shoults, C., Rashidbigi, S. (1996). Am. J. Physiol. 270: C1300-C1310; Kerrigan M.J & Hall A.C. (2000). J. Physiol., 527P; Urban, J.P.G., Hall, A.C., Gehl, K.A. (1993). J. Cell Physiol. 154: 262-270. Supported by the MRC, Arthritis Research Campaign (H0621) and the Wellcome Trust. Membrane Biology Group, Department of Biomedical and Clinical Laboratory Science, The University of Edinburgh, Edinburgh EH8 9XD, UK P 25-2 IMPAIRED RESTITUTION OF SINGLE-CELL LESIONS BY TNFα AND IFNγ IN NATIVE COLON EPITHELIUM P. Florian, M. Fromm, J.D. Schulzke1, A.H. Gitter During inflammatory bowel diseases (IBD), the mucosal barrier of the colon is damaged by erosion/ulcer type lesions and epithelial apoptosis causing local leaks in the intestinal epithelium [1]. In order to block the entry of pathogenic substances, it would be advantageous if the leaks were sealed again as fast as possible. Our results show, however, the opposite: proinflammatory cytokines delay the restitution of small epithelial defects. Methods: By intravital time-lapse video microscopy, by confocal fluorescence microscopy, and by conductance scanning, we measured the conductive leak associated with a single-cell lesion and the time course of morphological and functional repair in the surface epithelium of the mouse colon. Results: Conductance scanning revealed that, under control conditions, single-cell defects caused a leak of 0.8 ± 0.1 µS (n = 18) at the earliest point of recording (1 min post lesion). The leak sealed with an approximately exponential time course (time constant τ = 0.23 ± 0.02 min). Morphological closure of the defect involved flattening and extension of the basal ends of adjacent cells. (This process involves actin-dependent ‘purse string’ contraction.) Addition of TNFα alone did not change the restitution significantly, but together with IFNγ it slowed down the repair to τ = 0.52 ± 0.04 min (n = 10). Conclusions: For the first time, the change in ion permeability associated with a single-cell lesion was quantified in native intestinal epithelium. Recovery comprises formation of new tight junctions which seal the paracellular pathway. It turned out that TNFα and IFNγ hamper the closure of small defects. Thus, these proinflammatory cytokines not only support epithelial cell death by apoptosis, but they also impair the restitution. On the other hand, this means that therapy of IBD by TNFα antibodies additionally works by accelerating the repair mechanisms in the inflamed colon. [1] Gitter AH, Wullstein F, Fromm M, Schulzke JD (2001) Epithelial barrier defects in ulcerative colitis: characterization and quantification by electrophysiological imaging. Gastroenterology 121(6): 1320-1328. Dept. of Clinical Physiology and 1Dept. of Gastroenterology, Universitätsklinikum Benjamin Franklin, Freie Universität Berlin, 12200 Berlin, Germany P 25-3 CULTURED RUMINAL EPITHELIAL CELLS EXPRESS POTASSIUM CHANNELS BELONGING TO THE K ATP AND INWARD RECTIFIER FAMILIES F. Stumpff, M. Schweigel, H. Park, H. Martens Objective of study: Potassium channels have long been postulated as playing an important role in transport phenomena across the ruminal epithelium. In this study, the patch clamp technique was used to confirm the existence of potassium channels in cells derived from this tissue. Method used: Using established methods (Am. J. Phys. 278:G400G408, 2000), ruminal epithelial cells of the sheep were isolated and cultured for a maximum of two passages. Cells were filled with a physiological high potassium, low chloride pipette solution with calcium buffered to 10-7M. Results: When cells with an average capacitance value of 44 ± 4 pS (n=23) were superfused with low potassium Ringer’s solution, reversal potential was –21 ± 2 mV (n =13). Superfusion with high potassium solution led to a depolarization to –3 ± 2 mV (n = 10) (p<0.01). Inward current at –120 mV rose sharply to 391 ± 108 % of the original level (p=0.02), while outward current at 100 mV also

increased slightly to 136 ± 17% (p=0.05). TEACl (10 mM) significantly reduced outward current both in Ringer’s and in high potassium solution to 89 ± 2% (n=4, p=0.04) and 66 ± 18% (n=4, p=0.01) of the original level, respectively, while a reduction of inward current (to 58 ± 20% (n=4, p=0.05)) was only observed when external potassium was elevated. BaCl2 (10mM) reduced potassium-induced inward current to 61 ± 16% (n=4; p=0.04); no statistically significant effect on outward current level could be observed. While iberiotoxin (10-7M) had no impact on current (n=3), potassium-induced outward current was sensitive to the KATP blocker glibenclamide (10-5M) and reduced to 79 ± 13% of the original level (n=4; p=0.02). Conclusions: Cultured ruminal epithelial cells express both K ATP channels and a barium and TEA -sensitive inwardly rectifying potassium channel. Supported by DFG grant Ma 699/14-1. Institut für Veterinärphysiologie, Freie Universität Berlin, Oertzenweg 19b, 14163 Berlin, email: [email protected] P 25-4 BICARBONATE SECRETION BY PORCINE TRACHEAL EPITHELIUM M. Constable, S.K. Inglis Previous studies have shown that HCO3- is an important component of acetylcholine-stimulated submucosal gland secretion in porcine airways (Inglis et al, Am. J. Physiol 272:L372, 1997). The aim of the present study was to identify the mechanisms involved in HCO 3 secretion both under basal conditions and following stimulation with ATP. The epithelium was dissected from porcine trachea and mounted in an Ussing chamber containing HCO 3- and CO 2 buffered solution (37ºC). Replacement with Hepes-buffered (HCO3- and CO2 free) solution significantly reduced basal short circuit current (ISC) (135±12µAcm-2 to 91 ±11µAcm-2), suggesting that HCO3- was secreted at rest by these epithelia. Incubation of paired tissues with acetazolamide (10mM, 60min) had a similar inhibitory effect (136±11 to 103±7µAcm-2), and subsequent replacement with Hepes-buffered solution caused no further inhibition, suggesting that the majority of secreted HCO 3- was produced intracellularly in a reaction catalysed by carbonic anhydrase. To test whether a NaHCO3- basolateral HCO3- uptake on a Na+HCO3cotransporter (NBC) also contributed to basal HCO 3- secretion, the effect of the NBC inhibitor DNDS was tested. DNDS significantly inhibited basal I SC (88±8.7 to 78±8µAcm -2) but did not reduce the inhibitory effect of replacement with Hepes-buffered solution, suggesting that the effect of DNDS did not result from NBC inhibition. HCO3secreted under basal conditions is therefore predominantly generated metabolically rather than taken up across the basolateral membrane on a NBC. Previous studies (Inglis et al, Brit. J. Pharm. 130:367-374, 2000) showed that luminal ATP significantly stimulated anion secretion in porcine tracheal epithelium, presumably by activating P2Y receptors on the apical membrane of surface epithelial cells. To test whether ATP stimulated HCO 3- secretion we compared the effects of ATP (100µM luminal) on paired tracheal epithelia bathed either in HCO3and CO 2 buffered or Hepes buffered solution that contained acetazolamide. The net charge transfer stimulated by ATP was significantly lower in the absence of HCO3- and CO2 (3.9±1.1mCcm-2) than in its presence (8.3±1.5mCcm-2). Addition of bumetanide to inhibit Cl - secretion (10µM submucosal) to Hepes buffered solution further reduced charge transfer (0.2±0.7mCcm-2). ATP therefore stimulates both Cl- and HCO3- secretion across tracheal epithelium. Dept. Child Health, University of Dundee, Scotland, DD1 9SY P 25-5 REUPTAKE OF SMALL ELECTROLYTES INTO A6-CELLS FOLLOWING HYPOTONIC STRESS T. Große, I. Mühlhahn, I. Öztürk, F.X. Beck, A. Dörge Evidence from a previous study suggested that cellular losses of Na, K and Cl contribute substantially (about 70 %) to the regulatory volume decrease and isovolumetric volume regulation of A6-cells following hypotonic stress (Große et al., Pflügers Arch.442: 297-303; 2001). To determine whether these electrolytes are restored when, following an acute hyptonic stress, isotonic conditions are reestablished, cellular element concentrations and dry weight contents were determined by electron microprobe analysis. Hypotonic stress was accomplished by decreasing basolateral osmolarity of A6-epithelia incubated in Ussing

S 290 chambers from 260 to 140 mosmol/l. During the regulatory volume decrease initiated by hypotonic stress the Na, K and Cl concentrations were reduced to 11.7, 81.7 and 11.3 mmol/kg wet weight, respectively. After reestablishing isoosmotic conditions for 30 min by returning basolateral osmolarity to 260 mosmol/l the Na, K and Cl concentrations increased to about control values of 15.5, 145.3 and 40.0 mosmol/kg wet weight, respectively. Especially the increase in K and Cl concentrations was much more pronounced than would have been expected had the osmotic active volume of A6-cells behaved like an ideal osmometer. The cellular element concentrations referred to the dry weight content of K and Cl increased during isotonic conditions by 148.5 and 112.0 mmol/kg dry weight, indicating cellular uptake of these ions. Since this increase in cellular K and Cl could be almost inhibited completely, by changing from hypotonic to isotonic conditions in the presence of basolateral 10 -4 M bumetanide, the cellular ion uptake seems to be accomplished by a Na-K-2Cl-cotransporter. This interpretation is also supported by the finding that during isotonic incubation equilibrium conditions are achieved for the Na-K-2Clcotransporter. The lack of an accompanying increase in Na content might be explained by the action of the Na-K-pump, which exchanges the Na taken up by the cotransporter for extracellular K. Physiologisches Institut der Ludwig-Maximilians-Universität München, Pettenkoferstr. 12, 80336 München

P 25-6 BIOLUMINESCENCE DETECTION OF ATP RELEASE IN POLARIZED A6 EPITHELIA D. Jans, E. Larivière, S.P. Srinivas, A. Segal, W. Van Driessche Extracellular ATP released during cell swelling, functions as an important autocrine and paracrine signal, resulting in the modulation of a broad range of cellular processes. Particularly, in renal epithelial A6 cells, this autocrine mechanism could be involved in volume regulation and be responsible for the increase in intracellular Ca2+ during hypotonicity. Polarized renal A6 epithelia were cultured on permeable supports. ATP release at each border of the epithelium was measured in a stop flow system with a bioluminescence assay using luciferase-luciferin. Solutions at each side of the epithelium could be changed without causing mechanical artifacts. Extracellular osmolality was reduced from 260 to 140 mOsm/kg H2O by the removal of 65 mM NaCl. Reduction of the apical osmolality did not change cell volume, but caused a small and transient release of ATP across the apical membrane. A subsequent reduction of the basolateral osmolality was without effect on apical ATP release. Repetitive application of exogenous ATP at the apical membrane caused small transient increases of the intracellular Ca2+ concentration that were accompanied with transient increases of the apical permeability to Cl-. Apical hyposmolality did not affect basolateral ATP release. On the other hand, basolateral hyposmolality resulted in cell swelling and released a 10-fold amount of ATP across the basolateral membrane. Repeated activation of basolateral purinergic receptors with ATP resulted in transient rises in intracellular Ca 2+ that had a refractory behaviour. The Ca2+ transients depended on the IP3 pathway and gated apical Cl - channels. Our results suggest that hyposmolality causes an ATP release across both borders of the A6 epithelium. ATP or its metabolites may activate local purinergic receptors linked to the activation of apical Cl- channels. Laboratory of Physiology, KU Leuven, Campus Gasthuisberg, B-3000 Leuven, Belgium

P 25-7 THE WNT PATHWAY IS INVOLVED IN TIGHT JUNCTION PROTEIN CLAUDIN-2 GENE REGULATION J. Mankertz, B. Hillenbrand, S. Tavalali, O. Huber, M. Fromm, J.-D. Schulzke Objective of study: The formation of tight junctions as a selective barrier in epithelial cell sheets is tightly regulated during development and in certain physiological conditions. A crucial role in barrier formation plays claudin-2, because this integral membrane protein is a constitutive part of the tight junctional strands and furthermore serves as

a permeant transjunctional cation pore. The Wnt signal transduction pathway is involved in the regulation of cell fate and differentiation. We investigated nuclear effectors of the Wnt signaling pathway (LEF, lymphoid enhancer factor; TCF, T-cell factor) for their ability to interact directly with genomic elements essential in cis for claudin-2 gene expression. Methods: The promoter of human claudin-2 was identified by screening genomic libraries. Transcription activity was measured with luciferase reporter gene constructs in the presence of LEF/TCF expression plasmids. Band shift assays were performed with Cy5-labeled oligodeoxynucleotides and recombinant GST-LEF-1. Results and Conclusions: Analysis of the nucleotide sequence of the human claudin-2 promoter revealed two putative binding sites for LEF/TCF in close proximity. Electrophoretic mobility shifts with promoter specific oligodeoxynucleotides showed binding of recombinant GST-LEF-1 protein to the wild-type promoter fragment containing both LEF/TCF binding sites, but not to a fragment in which the first LEF/TCF motif was mutated. The functional relevance of this result was corroborated by reportergene assays. Claudin-2 promoter activity was increased after elevation of intracellular LEF-1 and TCF4 levels by co-transfection of appropriate expression plasmids as expected for a Wnt-regulated target gene. The regulatory functions of Wnt proteins are linked to cell differentiation and polarization. Tight junction formation is a key event in epithelial cell development. Our data indicate that gene expression of the tight junction protein claudin2 is under control of the Wnt signal transduction pathway which in this way is capable to affect epithelial barrier function, differentiation and transformation. Med. Klinik Gastroenterologie, Infektiologie u. Rheumatologie, UK Benjamin Franklin, FU Berlin, 12000 Berlin, Germany

P 25-8 THE INTESTINAL ANION EXCHANGER DRA (DOWN REGULATED IN ADENOMA) BINDS TO THE SECOND PDZ DOMAIN OF THE CFTR-ASSOCIATED PROTEIN CAP70 G. Lamprecht, H. Rossmann, A. Heil, S. Baisch, H. Kalbacher, J. Biber, M. Gregor, U. Seidler Introduction: dra (down regulated in adenoma) is an intestinal anion exchanger acting in concert with Na+/H+- exchanger 3 (NHE3) in ileal and colonic electroneutral NaCl absorption and maybe also with CFTR in duodenal HCO 3- secretion, dra has a C-terminal PDZ interaction motif (ETKF). Aim of the study: To characterize the interaction of dra with ileal brush border PDZ domain proteins. Methods: 1) Constructs: the C-terminus of human dra (C-dra, aa 567-764) was expressed as a His-S-tagged fusion protein (pET30) and as a biotinylated fusion protein (pinPoint), mutations of the wild type PDZ interaction motif (ETKF) were generated: ETKFminus, ETKI, ETKL, ETKV, ETKG, ETKA. Full length CAP70 and its 4 PDZ domains were expressed as His-S-tagged fusion proteins (pET30). 2) Far western: apical and basolateral membrane preparations of rabbit ileum were separated by PAGE, blotted and overlaid with C-dra or C-dra-ETKFminus, which were detected by anti-dra antibody. CAP70 was detected by anti-CAP70. 3) In vitro interaction ELISA: CAP70-constructs were bound to Nickel covered wells and interacting C-dra was detected using the biotin-tag. A biotinylated 20 amino acids long C-terminal peptide of dra was also used. Results: 1) Only in apical but not in basolateral ileal membrane preparations a 70 kDa protein was detected by C-dra but not by C-dra-ETKFminus in the far western experiment. The 70 kDa band ran at exactly the same size as CAP70 suggesting that the dra construct had bound to CAP70. Subsequently rabbit CAP70 was cloned by PCR using a combination of heterologous primers and RACE. 2) In vitro wild type C-dra and wild type dra-peptide but not C-draETKFminus bound to full length CAP70 and its second PDZ domain. The interaction is specific since full length CAP70 and all 4 PDZ domain constructs bound to C-dra-ETKL but only the second PDZ domain bound to wild type C-dra (ETKF motif). The other mutations did not bind to any of the CAP70 constructs. Conclusion: In rabbit ileum dra binds to the second PDZ domain of CAP70. Other transporters interacting with CAP70 (CFTR, NaPi-II) have been shown to bind to other PDZ domains. These data suggest that CAP70 may cluster several intestinal ion transporters and possibly also signal transduction proteins. 1 st Medical Dept. Univ. Tuebingen, Otfried-Müller-Str. 10, 72076 Tuebingen

S 291 P 25-9 MEMBRANE TRANSPORTERS REGULATING VECTORIAL BICARBONATE TRANSPORT BY HUMAN AIRWAY SEROUS CELLS A. Waltz1, R.J. Bridges2, M.J. Hug1 We have previously shown that the airway serous cell line Calu-3 transports bicarbonate (HCO 3-) ions upon stimulation with agonists increasing cAMP. In our present study we investigated the transporters required for the vectorial transport of HCO3-. Calu-3 cells were cultured on glass coverslips and subjected to simultaneous measurement of the intracellular pH (pHi) by BCECF microfluorimetry and whole cell patch clamp. When cells were exposed to a solution containing HCO3- / CO2 a rapid acidification accompanied by a depolarization and a decrease in membrane conductance (G m) were observed. pH i recovered in the presence of the buffer to slightly more alkaline values and membrane voltage (Vm) hyperpolarized. Clamping Vm to hyperpolarized values using agonists increasing cytosolic Ca 2+ or with an activator of IK1 channels, DCEBIO, prevented pHi recovery in the presence of HCO3/ CO2 indicative for electrogenic HCO3- transport. RT-PCR experiments detected mRNA for both, the pancreatic and the kidney isoform of the electrogenic Na + - dependent HCO 3 - transporter (NBCE1). Immunocytochemical experiments demonstrated expression of both NBCE1 isoforms in the basolateral membrane of polarized Calu-3 monolayers. To understand HCO3- exit, cells were exposed to gradients for Cl- and HCO3-. Replacement of extracellular Cl- by gluconate in the presence of HCO3- resulted in a modest increase in pHi suggesting the presence of a Cl-/ HCO3- exchanger. Stimulation with Forskolin decreased pHi, depolarized the membrane and increased Gm. However, when Clwas replaced in the presence of Forskolin, a marked alkalinization was observed that could not be inhibited by a classical inhibitor of anion exchangers, DIDS. In contrast, addition of Glibenclamide, an inhibitor of the CFTR Cl- conductance, abolished the alkalinization induced by Cl- replacement. We conclude that Calu-3 airway serous cells utilize a basolateral NBC for HCO3 - import that is dependent on V m. HCO3extrusion is stimulated by Forskolin and requires the activity of CFTR. Supported by IMF HU 11 01 03. 1

Department of Physiology, University of Münster, Robert-Koch-Str. 27a, D-48149 Münster; 2Department of Cell Biology and Physiology, University of Pittsburgh School of Medicine, Pittsburgh PA 15261 P 26-1 DURING DESENSITIZATION TO ACh IN RAT ATRIAL CELLS, THE M2 MUSCARINIC RECEPTOR IS INTERNALISED BY CLATHRIN-COATED VESICLES AND NOT CAVEOLAE T.T. Yamanushi, Z. Shui, H. Dobrzynski and M.R. Boyett On application of ACh to atrial muscle, ACh binds to the m2 muscarinic receptor and thereby activates the muscarinic K+ current. During a continued application of ACh, the muscarinic K + current fades as a result of desensitization (Shui et al., this meeting). In large part, the desensitization is the result of an internalisation of the receptor, but there is disagreement concerning how the receptor is internalised, with reports saying that it is internalised by clathrin-coated vesicles (Boyett et al., 1999, J. Physiol. 521, 23P), caveolae (Feron et al., 1997, J. Biol. Chem. 272, 17744-17748) or neither (Roseberry and Hosey, 2001, J. Cell Science, 114, 739-746). To address this issue, immunocytochemistry was used: atrial cells isolated from rats were labelled with monoclonal antibodies to the m2 muscarinic receptor, clathrin and caveolin 3 (followed by appropriate secondary antibodies conjugated to fluorescent markers) and then viewed using a confocal microscope. In untreated cells, the receptor was present in the outer cell membrane, but in cells treated with 10 µM CCh for 2 h, the density of the receptor in the outer cell membrane was significantly (P<0.005) reduced (labelling was 18 ± 2 %, n=54, of control). In untreated cells, caveolin was present in the outer cell membrane and, in CCh-treated cells, the density of caveolin in the outer cell membrane was unchanged (labelling was 97 ± 6 %, n=29, of control; P=0.64). In both untreated and CChtreated cells, clathrin was present either next to the outer cell membrane or deep within the cytoplasm as 1.2 ± 0.07 µm (n=5 cells) diameter bodies. In CCh-treated cells, the receptor was also found either next to the outer cell membrane or deep within the cytoplasm as 1.2 ± 0.07 µm (n=5 cells) diameter bodies. In double-labelling experiments, in 0/18 CCh-treated cells, the internalised receptor was colocalised with caveolin and, in 24/24 CCh-treated cells, the internalised receptor was co-localised with clathrin. It is concluded that during

desensitization the m2 muscarinic receptor is internalised by clathrincoated vesicles and not caveolae. School of Biomedical Sciences, University of Leeds, Leeds LS2 9JT, UK P 26-2 β-ADRENERGIC STIMULATION OF THE Ca2+ TRANSIENT IN DETUBULATED RAT VENTRICULAR MYOCYTES F Brette, K Komukai, C.H. Orchard We have used detubulated rat cardiac myocytes to study the role of the t-tubules in the response to β-adrenergic stimulation. We have shown previously that isoprenaline causes a greater fractional increase in Ltype Ca 2+ current in control cells than in detubulated cells. In the present study we investigated whether this was reflected in the Ca2+ transient. Wistar rats were killed humanely using a UK Home Office Schedule 1 method. Ventricular myocytes were enzymatically isolated, and detubulated as described by Kawai et al. (1999, Amer J Phys, 277; H603-609). Cells were field stimulated and the Ca 2+ transient monitored using fura-2. Detubulation significantly decreased the amplitude of the Ca2+ transient (Mann-Whitney rank sum test) and prolonged its time to peak and time for 50% decline (T 1/2 ; table 1). Isoprenaline (Iso, 0.5 µM) increased Ca2+ transient amplitude and decreased T1/2 in control and detubulated cells; however in the presence of iso, the amplitude and T 1/2 were not significantly different in control and detubulated cells (table 1):

Ca Transient (Ratio Unit) Con (n=15)

Det (n=19)

T1/2 (ms) Con (n=15)

Det (n=19) 398 ± 42*

0.072 ± 266 ± 24 0.009* 0.196 ± 0.209 ± Iso 150 ± 7+ 144 ± 9+ 0.028+ 0.022+ Table 1: Effect of detubulation and iso on the Ca2+ transient. Data are expressed as mean ± S.E.M. * indicates P<0.001 between columns, + indicates P<0.001 between rows. Con: control cells, Det: detubulated cells. Control

0.124 ± 0.017

These data suggest that the t-tubules play an important role in the stimulation of the L-type Ca current by iso, but that their role in the response of the Ca2+ transient to iso is less marked. Thus it appears: (i) that iso increases the gain of Ca2+ release; (ii) that there is differential regulation of the L-type Ca current and other iso-dependent mechanisms; (iii) that beta-adrenergic pathways are still active in detubulated myocytes. This work was supported by the British Heart Foundation and the Wellcome Trust. School of Biomedical Sciences, University of Leeds, Leeds, LS2 9NQ United Kingdom P 26-3 SELECTIVE INHIBITION OF ARRHYTHMOGENIC BUT NOT INOTROPIC EFFECTS OF Ca LOADING IN CARDIAC MUSCLE L. Venetucci, A. Trafford, M. Diaz, S. O’Neill, D.A. Eisner. The use of positive inotropic agents that increase systolic [Ca 2+]i is limited by their tendency to produce spontaneous Ca release from the sarcoplasmic reticulum (SR) and thence delayed after-depolarisations. It would obviously be useful to be able to inhibit spontaneous Ca release without affecting normal release. One difficulty with this is that both the normal and the spontaneous Ca release occur through the ryanodine receptor (RyR) via the process of calcium induced calcium release (CICR). In previous work, however, we have shown that the local anaesthetic tetracaine decreases the frequency of spontaneous release in unstimulated cells (Overend et al.,1997). In separate experiments we showed that tetracaine only had a transient depressant effect on systolic Ca and contraction (Overend et al.,1998). The purpose of the present experiments was therefore to investigate the effects of tetracaine on [Ca2+] i in cells which were overloaded with calcium to the point where they displayed spontaneous release between stimuli. This was achieved in three different ways: (i) elevating external Ca concentration from 1 to 5 mM; (ii) the addition of 500 µM ouabain; (iii) addition of isoprenaline (1 µM). In all cases tetra-

S 292 caine (50-100 µM) abolished spontaneous Ca release without decreasing the amplitude of the systolic Ca transient. These results show that depression of RyR open probability can be used to selectively remove the untoward effects of inotropic agents. Overend, C. L, Eisner, D. A., & O’Neill, S. C. (1997). Journal of Physiology 502, 471-479. Overend, C. L, O’Neill, S. C., & Eisner, D. A. (1998). Journal of Physiology 507, 759-769 Unit of Cardiac Physiology, University of Manchester, 1.524 Stopford Building, Oxford Rd, Manchester M13 9PT P 26-4 EFFECTS OF KETAMINE ON THE INDUCTION OF LTP-LIKE CHANGES IN HUMAN PAIN PERCEPTION T. Klein, W. Magerl, H.C. Hopf, J. Sandkühler, R.D. Treede High-frequency electrical nerve stimulation elicits input-specific and heterosynaptic longterm potentiation-like enhancement of pain perception (nociceptive LTP) in conscious human subjects (Klein et al, Pflügers Arch., Suppl. Vol. 441(6), R239 (2001)). LTP of spinal nociceptive transmission is believed to be a primarily NMDA-R dependent process. Here we tested the effect of ketamine, an NMDA receptor antagonist, on nociceptive LTP in humans. High frequency electrical stimulation (5 x 1 s trains of 100 Hz at 10 s intervals; HFS) were applied to forearm skin in human volunteers by an array of 10 punctate electrodes under ketamine (0.25mg/kg body weight). Input-specific perceptual changes were tested at the site of conditioning stimulation by electrical test stimuli at 10 x detection threshold. All tests were approved by the local ethics committee.Compared to controls (open circles), ketamine (closed circles) abolished the input-specific LTP of pain perception to electrical test stimuli in humans (see Fig.).

(n=6) at 10 nM and 0.031±0.004 a.u. (n=6) at 1 µM melatonin. The increase of [Ca] i induced by 1 µM melatonin was attenuated by 78.5±8.0% (n=5) by 30 µM 2-APB, an inhibitor of IP3-induced calcium release, by 63.5±4.3% (n=6) by 1 µM nimodipine, a blocker of voltage-operated calcium channels and by 78.2±9.1% (n=5) by 30 µM SKF 96365, an inhibitor of store-operated calcium influx. Further, Bay K 8644, an opener of voltage-operated calcium channels, was used to evoke the same amount of tension as 1 µM melatonin. This constriction was accompanied by a 2.23±0.64-fold (n=5) larger increase of [Ca]i than that induced by melatonin. In addition, the contraction induced by 1 µM melatonin was attenuated by 77.5±1.4% (n=5) by 1 µM Y-27632, an inhibitor of Rho-kinase, which is known to modulate calcium sensitivity. Thus, the data of the present study suggest that the melatonin-induced constriction of rat tail arteries is mediated 1) by an increase of [Ca]i due to calcium influx via voltageoperated and store-operated calcium channels as well as to calcium release from IP3-sensitive stores and 2) by a change of calcium sensitivity, probably mediated by Rho-kinase. Institute of Physiology, University Rostock, 18055 Rostock, Germany; 1Department of Pathophysiology, Moscow State University, 119899 Moscow, Russia P 26-6

In conclusion, the inhibition of input-specific LTP in human pain perception by the established NMDA-antagonist ketamine is consistent with previous experimental findings on spinal nociceptive transmission in the rat suggesting an analogous mechanism in human nociceptive LTP.

EFFECTS OF HYPEROSMOTIC SHRINKING ON VENTRICULAR MYOCYTE CONTRACTION AND INTRACELLULAR Ca 2+ IN STREPTOZOTOCIN-INDUCED DIABETIC RATS F.C. Howarth, M.A. Qureshi, E. White Hyperglyceamia and hyperlipideamia may alter the osmotic load placed upon diabetic myocardium. Our aim was to investigate the effects of hyperosmotic shrinking on ventricular myocyte contraction and intracellular Ca2+ concentration [Ca2+]i in streptozotocin (STZ)-induced diabetic rats. Diabetes was induced in rats by injection of STZ (60 mg kg -1). Two months later, single ventricular myocytes were isolated, and characteristics of shortening, [Ca2+]i and sarcoplasmic reticulum (SR) Ca2+ transport were measured in electrically stimulated myocytes exposed to either normal Tyrode (NT, 300 mosmol kg -1 ) or hyperosmotic Tyrode (HT, 440 mosmol kg-1) for 1 or 5 min at 35-36 °C. 5 min exposure to HT significantly reduced cell volume to 0.76 ± 0.03 of that in NT (mean ± s.e.m, n = 12 cells). Hyperosmotic shrinking significantly reduced the amplitude of shortening (by 60 % after 5 min. in control cells, n = 13 cells) and prolonged its duration while significantly increasing the amplitude (by 146 % after 5 min. in control cells n = 18 cells) of the [Ca2+]i transient (P < 0.001) but the effects of shrinkage were the same for control and STZ treated cells (P > 0.05) 2-way repeated measures ANOVA. Fractional SR Ca2+ release was assessed by measuring the ratio of electrically evoked [Ca2+]i transients to that in response to 20 mM caffeine. Shrinking significantly increased SR fractional release (P < 0.001, n = 11 cells) but the response of control and STZ cells were not significantly different (P > 0.05). The effects of raised extracellular osmolarity on contractility does not appear to be altered in myocytes from STZ treated rats.

Inst. of Physiology & Pathophysiology, Johannes Gutenberg University, Saarstr. 21, 55099 Mainz

Department of Physiology, Faculty of Medicine & Health Sciences, United Arab Emirates University, P.O.Box 17666, Al Ain, U.A.E.

P 26-5

P 26-7

MELATONIN CONTRACTS RAT TAIL ARTERIES BY STIMULATING CALCIUM INFLUX AND CALCIUM RELEASE AS WELL AS BY CHANGING CALCIUM SENSITIVITY R. Schubert, V. Kalentchuk1 The pineal hormone melatonin is involved in the regulation of seasonal and circadian rhythms. It has found clinical use as an additive in cancer treatment and as a regulator of rhythm disorders. In the circulatory system melatonin-induced contractile reactions have been observed, although the mechanism of vasoconstriction is not known. The hypothesis that an alteration of smooth muscle calcium concentration and/or calcium sensitivity is involved was tested using isometric tension recording and calcium fluorimetry. Melatonin produced a concentration dependent constriction of rat tail arteries with a pD2 of 8.17. This constriction was accompanied by an increase of the intracellular calcium concentration ([Ca]i); the augmentation of [Ca] i, expressed in arbitrary calcium ratio units (a.u.), was 0.017±0.005 a.u.

MODULATION OF THE Na+-Ca2+ EXCHANGE FUNCTION AND Ca 2+ CURRENTS BY LIPID-SIGNALING PATHWAYS IN CARDIAC MYOCYTES M. Keller, C. Pignier, E. Niggli, M. Egger The Na+-Ca2+ exchange mechanism is one of the major processes regulating the intracellular Ca 2+ concentration in cardiac myocytes. Under pathophysiological conditions (e.g. ischemia) Ca2+ overload and arrythmia are known phenomena in which the Na+-Ca2+ exchanger (NCX) could be involved. Interestingly, changes in the sarcolemmal fatty acid and lipd composition have been observed during ischemia, which may affect the Na+-Ca2+ exchange function. Membrane lipids are sources of second messengers that play vital roles in signal transduction processes (e.g. PIP 2 ). Nevertheless, the precise mechanisms by which the lipid metabolites or messengers are targeting the Na+-Ca2+ exchange function have not been determined conclusively. Here we test the hypothesis that the phospholipase C pathway, which

S 293 involves the activation of the protein kinase C (PKC), may change the NCX activity by phosphorylation. Intracellular Ca 2+ concentration jumps were generated in isolated guinea pig ventricular myocytes by flash photolysis of caged Ca 2+ (DM-nitrophen). The resulting NCX currents (INaCa) were recorded in the whole-cell configuration of the patch-clamp technique, while L-type calcium currents (I Ca-L) were obtained by depolarizing steps from -40 mV to 0 mV. In the presence of phorbol-12-myristate-13-acetate (PMA, 100 nM) or phorbol-12,13dibutyrate (PDBu, 100 nM), which both directly activate protein kinase C (PKC), reduced ICa-L current amplitudes were found at all test potentials. Conversely, INaCa was not significantly affected by PMA or PDBu under our conditions. Our results suggest that the NCX activity is not tightly controlled by PKC sensitive phosporylation pathways. Supported by the Swiss National Science Foundation and the Roche Science Foundation. Department of Physiology, University of Bern, Bühlplatz 5, CH-3012 Bern, Switzerland P 26-8 BETA-STIMULATION SHORTENS RECOVERY OF SARCOPLASMIC Ca2+ RELEASE C. Pignier, M. Egger, E. Niggli It has been observed that coherent activation of Ca2+ release by flashphotolysis of caged Ca2+ unmasks a refractoriness of Ca2+ -induced Ca2+ release (CICR) in cardiac myocytes. Here we examined the effect of βadrenoceptor stimulation on CICR recovery. For this purpose, we induced global Ca2+ release from the sarcoplasmic reticulum (SR) by using UV-laser flash photolysis of caged Ca2+ (DM-nitrophen) in combination with the whole-cell patch clamp technique in guinea pig ventricular myocytes. Resulting Na-Ca exchange currents (INa/Ca) were recorded at –40 mV. The rising phase of I Na/Ca can be separated in two components, where the first corresponds to the photo-release of Ca2+ and the second component, the slower one, is due to the subsequent Ca2+ release from the SR. Pairs of UV-flashes were applied at various intervals (200-1500 ms) to follow the time-course of recovery from refractoriness. Application of Isoproterenol (1µM Iso) significantly shortened the refractoriness, such that complete recovery of Ca 2+ release could be observed earlier. Recovery of Ca2+ release was fitted with a monoexponential function with a mean τ of 352 ± 45 ms (ctrl) and 195 ± 21 ms (Iso). Recovery from refractoriness may depend on the SR Ca2+ content. In order to test this hypothesis, a protocol to asses SR Ca2+ load was adopted . The results suggest that under resting conditions (-40 mV) the SR Ca2+ load could be modulated by treatment. with Iso (4 min). We conclude that β-adrenoceptor stimulation modulates global CICR refractoriness. This might result from modifications of SR Ca2+ content and/or from a change of the steady-state Ca2+ sensitivity and/ or gating kinetics of the SR Ca2+ release channels. Supported by the Swiss National Science Foundation. Institute of Physiology, University of Bern, 3012, Bern, Switzerland P 26-9 Ca 2+-INDUCED Ca 2+ RELEASE SENSITIVITY IN CARDIAC MUSCLE DETERMINED BY TWO-PHOTON PHOTOLYSIS OF CAGED Ca2+ N. Lindegger, E. Niggli In cardiac muscle Ca2+-induced Ca2+ release (CICR) from the sarcoplasmic reticulum (SR) occurs via ryanodine receptors (RyRs) as Ca2+ sparks, usually triggered by L-type Ca2+ channels. The efficiency of excitationcontraction (EC) coupling may be determined by the probability to trigger Ca2+ sparks. For example, the Ca2+ sensitivity of the RyRs may be modulated by phosphorylation after ß-adrenergic stimulation. In addition, many other factors may affect the Ca 2+ sensitivity of the RyRs, including diseases. Two-photon photolysis (TPP) of caged Ca2+ (DM-nitrophen) was used to trigger CICR in Guinea pig and mouse ventricular myocytes. Ca2+ signals were recorded using Fluo-3 in patchclamped cells (whole-cell configuration) with laser-scanning confocal microscopy. To determine the efficiency of the CICR in-vivo, we rapidly elicited Ca2+ signals with trains of increasing photolytic power (doseresponse curve). In-vitro tests showed a nearly quadratic relation between Ca 2+ signals and TPP power while in-vivo experiments showed a sigmoidal curve with saturation at high photolytic power (90 mW, Hill coefficient: 8±0.9). Under high SR Ca2+ load conditions Ca2+ release

signals had larger amplitudes and the dose-response curve of TPP was shifted to the left, suggesting an increase in the Ca2+ sensitivity of CICR. Similarly, after ß-adrenergic stimulation, TPP elicited larger Ca2+ signals. Thus, dose-response curves of Ca2+ signals triggered by TPP provide a rapid measurement of CICR sensitivity. In conclusion, changes of SR Ca 2+ load and/or enhanced Ca2+ sensitivity after RyR phosphorylation could account for the observed changes of EC-coupling efficiency. Supported by the Swiss National Science Foundation. Institute of Physiology, University of Bern, Bühlplatz 5, 3012 Bern, Switzerland P 26-10 GADOLINIUM DILATES RAT SMALL ARTERIES POSSESSING PRESSURE- AND POTASSIUM-INDUCED TONE AND INHIBITS VOLTAGE-OPERATED CALCIUM CHANNEL CURRENTS S. Lewandrowski, H. Gagov1, R. Schubert In the last time there is growing appreciation that stretch-activated cation channels may contribute to arterial contractions. These findings are often based on the effect of gadolinium, a potent blocker of these channels. Arterial contractions do also depend, at least partly, on calcium influx via voltage-operated calcium channels. Hence, it is important to know the effect of gadolinium on calcium channels. Therefore, the hypothesis was tested that gadolinium interacts with calcium channels and with calcium channel-mediated contractile reactions of rat small arteries using isometric force measurements and videomicroscopic diameter determinations as well as ion current measurements with the patch-clamp technique. Gadolinium dilates isobaric preparations of small tail arteries possessing a spontaneous pressure-induced tone with an IC50 of 84±18 nM (n=6). Gadolinium and nimodipine, both at 1 µM, but not 1 µM cromakalim prevent the effect of 30 mM KCl, a contractile reaction mediated mainly by calcium channels. Further, isometric vessel preparations preconstricted with 30 mM KCl were used in order to have conditions free of pressureinduced tone. Gadolinium relaxes small tail, coronary and cerebral arteries with an IC50 of 160±33 nM (n=6), 492±110 nM (n=5), and 340±195 nM (n=5), respectively, which are not different (p=0.14, one-way ANOVA). In addition, gadolinium at 10 µM, the most commonly used concentration in functional studies, markedly attenuated the current-voltage relationship of whole-cell calcium currents. Thus, the data of the present study show that gadolinium at micromolar concentrations blocks calcium currents and inhibits pressure-induced and KCl-induced vessel tone, functional responses known to depend on calcium channels. Consequently, the limited selectivity of gadolinium questions its use as a tool in functional studies. Institute of Physiology, University Rostock, 18055 Rostock, Germany; 1 Institute of Biophysics, Bulgarian P 26-11 L-TYPE Ca 2+ CURRENT AND TRANSCRIPTIONAL DOWNREGULATION IN A RABBIT MODEL OF RAPID ATRIAL PACING S. Wöhrl, C.R. Scherer, R.F. Bosch, N. Rüb, H. Haase, K. Steinmeyer, V. Kühlkamp Introduction: Studies in humans with atrial fibrillation and in animal models have demonstrated that chronic electrical remodeling reduces the L-type Ca2+ current by transcriptional downregulation of the underlying ion channel. However, electrical remodeling starts early after the onset of rapid atrial rates. The time course of channel alterations in the early phases of remodeling is not known. Methods: Rapid (600ppm) right atrial pacing via a transvenously implanted lead in rabbits. 5 groups: 0(sham), 6, 12, 24, and 96 hours pacing, n=5 for each. Ionic currents were recorded by the patch clamp technique in left atrial myocytes. mRNA measurements were done by one step RTPCR and protein measurements by Western Blot analysis in right atrial specimen of the same animals. Results: I Ca,L started to decline after 12h (by 47%) and progressively decreased up to 96h. The reduction was paralleled by decreased mRNA expression of the Ca2+ channel beta-subunits, whereas significant reductions in the alpha1 subunit (by 24%) were observed 12h after pacing onset. A similar reduction of alpha1 subunit protein was detected after 12h (by 15%) of pacing. Stimulation for 96h caused a further decrease of the alpha1 channel protein by 23%. Conclusions: I Ca,L is reduced in very early phases of

S 294 electrical remodeling. The mechanism is a transcriptional downregulation of underlying ion channels which partially precedes current changes.

Max-Planck-Institute for Molecular Physiology, Otto-Hahn-Str. 11, 44227 Dortmund, Germany

Department of Cardiology, University of Tübingen, 72076 Tübingen, Germany; Aventis Pharma GmbH, DG Cardiovascular Diseases, 65926 Frankfurt am Main, Germany

P 27-1

P 26-12 DIFFERENTIAL SECRETION OF VASOACTIVE VESICULAR COMPONENTS FROM VASCULAR ENDOTHELIAL CELLS EVOKED BY HIGH INTRACELLULAR Ca2+ CONCENTRATION ψ G. Zupancic , D. Ogden, C.J. Magnus, C. Wheeler-Jones+, T.D. Carter* Blood clotting and vascular immune responses are regulated by proteins secreted from endothelial cells. In many cases these proteins have opposing functions, for example, some promote coagulation (von Willebrand factor (vWf), and others prevent coagulation (tissue plaminogen activator). They are secreted from separate storage granules by exocytosis in response to elevations of intracellular free calcium ion concentration ([Ca2+]i). To understand the regulation of endothelial secretions it is necessary to determine the kinetics of secretory granule exocytosis in single endothelial cells. Here we report the first high resolution membrane capacitance (C m) measurements of exocytosis from single isolated human umbilical vein endothelial cells (HUVEC). Total Cm and the amplitudes and times of discrete femtoFarad (fF) Cm steps, due to granule exocytosis and endocytosis, were monitored simultaneously. [Ca2+]i was elevated by flash photolysis of Ca2+DM-nitrophen to evoke secretion and monitored with the low affinity Ca2+ indicator furaptra. Sustained elevation of [Ca2+]i to >20 µM evoked slow net increases in endothelial Cm of up to 5 pF, comprising exocytotic and endocytotic steps of amplitude 0.5 to 110 fF. Cm steps of 2.5-9 fF were ascribed to exocytosis of Weibel-Palade bodies (WPb) following comparison with morphometric estimates of the distribution of WPb Cm, and supported by the absence of Cm steps of 2.5-9 fF in cells lacking WPb. The time course of WPb secretion included a delay of >20 s after [Ca2+]i elevation in contrast to immediate secretion of smaller, non-WPb granules. These results show a temporal difference in exocytosis from WPb and non-WPb granules that may provide a basis for differential release of pro- and anticoagulants from vascular endothelial cells. ψ

NIMR, London NW7 lAA. University of Ljubljana, 1001 Ljubljana, FOB 2995, Slovenia. +Department of Veterinary Basic Sciences, RVC, London NWl OTU. *Department of Pharmacology, UC, London WCIE 6BT P 26-13 HISTAMINE AND BRADYKININ INDUCE CYTOSOLIC CALCIUM INCREASE IN CAPILLARY ENDOTHELIUM M.A. Delpiano, M. Cortés, R. Alvarez, R. Vinet The stimulatory effect of histamine and bradykinin on intracellular Ca2+ concentration ([Ca2+]i) was studied in fresh bovine adrenal medulla capillary endothelial cells (BAMCECs), by using the fluorescent probe fluo-3. Histamine and bradykinin induced an increase in [Ca2+]i in 86% (n=121) and 68% (n=49) of the tested cells with a IC50 of 0.64±0.32 and 0.082±0.01 µM, respectively. This [Ca2+]i increase exhibited a fast monophasic time course. In the case of histamine, this component was abolished by previous treatment with thapsigargin (1 µM), but it was not affected neither by caffeine (100 µM), zero external Ca2+ nor by Ca2+ channel blockers, like Co2+ and Ni2+ (1 mM). The histamine response was additionally blocked by the H 1 -histamine antagonist triprolidine (100 nM). In contrast, the [Ca 2+] i increase induced by bradykinin was inhibited by caffeine (100 µM), but was not affected by zero Ca 2+ nor by Co 2+ or Ni2+. It was blocked, however, by the B2bradykinin receptor antagonist B1650 (100 nM). Taken together, these results demonstrate that the increase in [Ca2+]i produced by these peptides is carry out by the liberation of Ca2+ from intracellular stores rather than the entry of external Ca2+ and suggest that a thapsigarginsensitive Ca 2+ store may be involved in the [Ca 2+]i rise induced by histamine, while a caffeine-sensitive one participates in the Ca 2+ response induced by bradykinin. The increase in [Ca2+]i, induced by both agents, could be associated with the formation and liberation of vasoactive substances from BAMCECs, that might modulate the adrenal gland’s microcirculation.

PROPRANOLOL TREATMENT AFTER MYOCARDIAL INFARCTION IN RATS: EFFECT ON HEART FUNCTION AND CYTOKINE EXPRESSION A. Deten, H.C. Volz, H.-G. Zimmer Heart function is severely depressed after acute myocardial infarction (MI) in rats. MI further leads to an extensive myocardial remodeling and congestive heart failure. In the first 2 days after MI, the expression of the pro-inflammatory cytokines interleukin (IL)-1 and IL-6 is strongly upregulated in the infarct area as well as in the non-infarcted myocardium. This increase is similar to that induced by catecholamines. We tested the hypothesis that beta-blockade with propranolol may prevent the cytokine expression after MI. Rats were treated either continuously by miniosmotic pumps (s.c. implanted 24 h prior to surgery, 1 mg/kg.h, MI-P1) or by two single s.c. injections (15 min prior to and 4 h after surgery, 2 mg/rat, MI-P2) and compared to placebo treated and sham-operated controls (MI-C and S, resp.). Both treatment groups had a decrease in left ventricular end-diastolic pressure when compared to MI-C (7, 6, 14 and 4 mmHg in MI-P1, MI-P2, MI-C and S resp.). Heart function was further improved by an increase in left ventricular developed pressure (86, 81, 69 and 142 mmHg, resp.). However, cardiac output was decreased (43, 45, 56 and 89 ml/ min, resp.) due to the lower heart rate (308, 298, 333 and 386 bpm, resp.) and the increased total peripheral resistance (0.49, 0.44, 0.33 and 0.39 mmHg.min.kg/ml, resp.). Despite these hemodynamic alterations, cardiac expression of IL-1 and -6 was not significantly affected in the non-infarcted myocardium as measured by ribonuclease protection assay. In the infarct area, however, cytokine expression was attenuated by about 40% by peri-operative high-dose propranolol treatment. Therefore, we suggest that locally elevated NE in the ischemic area contributes to the observed increase in cytokine expression after MI in rats. Carl-Ludwig-Institute of Physiology, Liebigstr. 27, D-04103 Leipzig

P 27-2 ISOFLURANE PRECONDITIONS MYOCARDIUM AGAINST INFARCTION VIA RELEASE OF FREE RADICALS R. Huhn, *J. Müllenheim, D. Ebel, *J. Fräßdorf, *W. Schlack, V. Thämer The volatile anaesthetic isoflurane exerts cardioprotective effects that mimic ischaemic preconditioning (PC) (1). Release of free radicals is involved in triggering PC (2). We investigated if the two structurally different radical scavengers N-(2-mercaptoproprionyl)glycine (MPG) and Mn(III)-tetrakis(4-benzoic acid)porphyrine chloride (MnTBAP) block isoflurane induced PC. 61 α-chloralose anaesthetized rabbits were subjected to 30 min of occlusion of a major coronary artery and 120 min of subsequent reperfusion. Rabbits of 6 groups underwent a treatment period consisting of either no intervention for 35 min (CON,n=11) or 15 min of isoflurane inhalation (1 MAC end tidal concentration) followed by a 10-min washout period (isoflurane group, n=12). 4 additional groups received MPG (1 mg kg-1 min-1) and MnTBAP (100 µg kg -1 min -1) during the treatment period with (isoflurane+MPG, n=11; isoflurane+MnTBAP, n=9) or without isoflurane inhalation (MPG, n=11;MnTBAP, n=7). Left ventricular pressure (LVP, tip manometer), cardiac output (CO, ultrasonic flowprobe), and infarct size (TTC staining) were determined. Haemodynamic baseline values were not significantly different between groups [LVP 97±17 mmHg (mean±SD); CO 228±61 ml/min]. During coronary occlusion, LVP was reduced to 91±17% of baseline and CO to 94±21%. After 120 min reperfusion, recovery of LVP and CO was not significantly different between groups (LVP 83±20%; CO 86±23% of baseline). Infarct size was reduced from 49±17% of the area at risk in controls to 29±19% in the isoflurane group (P=0.04). MPG and MnTBAP themselves had no effect on infarct size (MPG: 50±14%; MnTBAP: 56±15%), but both abolished the preconditioning effect of isoflurane (isoflurane + MPG : 50 ± 24%, P=0.02; isoflurane+MnTBAP 55±10%,P=0.001). Statistics: ANOVA, Dunnett’s post-hoc test. Isoflurane-induced preconditioning depends on the release of free radicals. (1) Anesthesiology 1997; 87:361-70 (2) Circ. Res. 2000; 87:460-6

S 295 Institutes of Physiology and *Anaesthesiology of the Heinrich-HeineUniversity Düsseldorf, Postfach 101007, D-40001 Düsseldorf, Germany

Department of Physiology, University of Münster, Robert-Koch-Str. 27a, 48149 Münster, Germany P 27-5

P 27-3 EFFECT OF CHRONIC HYPOXIA ON HENT1 TRANSPORTER EXPRESSION AND ENDOTHELIAL NITRIC OXIDE SYNTHASE IN HUMAN FETAL ENDOTHELIUM R. Rojas, P. Casanello, C. Aguayo, L. Sobrevia Hypoxia down-regulates expression of endothelial nitric oxide synthase (eNOS), and increases adenosine release from several tissues. We have shown that nitric oxide (NO) modulates adenosine transport via hENT1 (human Equilibrative Nucleoside Transporters) in human umbilical vein endothelial cells (HUVECs). We studied the involvement of NO in the effect of hypoxia on the expression of hENT1 in HUVECs. Cells were isolated from umbilical veins (Ethics committee approval obtained) and transport of [3H]adenosine (10 µM, 20 s, 22ºC) was measured in Hepes-buffered Krebs solution in the presence or absence of NBMPR (10 µM, 30 min) in cells cultured in normoxia or chronic hypoxia (24 hrs, 2% O2). NO synthesis was estimated by formation of L-[ 3H]citrulline from L-[3H]arginine (30 min, 37°C), and hENT1 and eNOS mRNA detected by reverse transcription-polymerase chain reaction (RT-PCR). Transport of adenosine (38 ± 4 pmol (10 6 cells)-1 s-1, means ± S.E.M., n=16) was significantly increased (unpaired Student’s t test, P<0.05) by hypoxia (74 ± 9) and inhibited by 10 µM NBMPR (10 ± 5). Hypoxia decreased hENT1 (35 ± 5 %) mRNA levels, and L-[3H]citrulline formation (50 ± 5 %), but increased eNOS protein (25 ± 5 %) and mRNA (40 ± 5 %) levels. The effects of hypoxia were blocked by NG-nitro-L-arginine methyl ester. Hypoxia-induced up-regulation of adenosine transport in human fetal endothelium could result from a higher activity of hENT1 as a response to the hypoxia-induced reduction of eNOS activity. Supported by FONDECYT 1000354 & 7000354, DIUC 201.084.003-1.0 (Chile), and The Wellcome Trust (U.K.). P. Casanello and C. Aguayo hold Beca Docente University of Concepción- and CONICYT-PhD fellowships (Chile). CMPL, Department of Physiology, Faculty of Biological Sciences, and Department of Obstetrics & Gynecology, Faculty of Medicine, University of Concepción, P.O. Box 160-C, Concepción, Chile P 27-4 INTRACELLULAR pH RECOVERY FROM ACID LOAD TRIGGERS VON WILLEBRAND FACTOR RELEASE IN ENDOTHELIAL CELLS V. Huck, A. Niemeyer, P. Rogge, R. Ossig, H.J. Lang, H. Oberleithner, S.W. Schneider Reperfusion after ischemic conditions is followed by massive endothelial cell activation and may cause reperfusion injury. We investigated the influence of acidosis, an inevitably concomitant of ischemia, and reperfusion on exocytosis of von Willebrand factor (vWF) in human umbilical venous endothelial cells (HUVEC). vWF, stored in cone-like vesicles called Weibel Palade bodies, is involved in platelet aggregation and the intrinsic pathway of haemostasis. In previous studies we showed that extracellular acidosis (pHex=6.4) is accompanied by parallel changes of intracellular pH (pHin). Moreover, acidosis reduces constitutive and histamine stimulated vWF exocytosis to 46 ± 1.1 % (n=19) and 52 ± 2.5 % (n=9), respectively, within minutes compared to control cells (pH ex =7.4). Reperfusion of preacidotic cells (pH ex=6.4 for Ih, 3h or 48h) with pCO2 /HCO3--buffered cell medium (pHex =7.4, pCO 2=5%, supplemented with l0µM histamine) was followed by an enhanced increase of vWF release (193 ± 8 %, n=8) related to control cells. In parallel to vWF secretion pHin recovered from 6.53 ± 0.02 to 7.23 ± 0.02 (n=5) within 10 minutes. Application of l0uM Cariporide, a NHEinhibitor, during reperfusion delayed intracellular realkalinization (initial slope: 0.27 ± 0.02 ∆pH/min compared with 0.97 ± 0.04 ∆pH/min under control conditions, n=5). This retarded intracellular pH recovery was accompanied by reduced histamine stimulated vWF release of 57 ± 6.6 %, n=10. By using atomic force microscopy we detected upon reperfusion exocytotic pores as a morphological correlate of vWF secretion on the surface of endothelial cells. In summary our data show that i) reperfusion of acidotic cells stimulates vWF secretion, ii) reperfusion is followed by an immediate intracellular realkalini zation and iii) inhibition of intracellular realkalinization attenuates reperfusion induced vWF exocytosis.

THORACIC EPIDURAL ANESTHESIA ATTENUATES HEMORRHAGE -INDUCED IMPAIRMENT OF INTESTINAL PERFUSION IN RATS J. Adolphs1, D.K. Schmidt1, B. Kamin1,2, M. Schäfer1, M. Welte1, A. Pries2, H. Habazettl2,3 Introduction: Injury to the intestine due to hemorrhagic shock is closely associated with bacterial translocation and the development of sepsis and multiple organ dysfunction syndrome. Thoracic epidural anesthesia (TEA) has recently been shown to increase perfusion in ileum of rats (Anesthesiology 2000; 93:844-51). Using in vivo fluorescence microscopy, we analyzed the effect of TEA on intestinal perfusion in rats, subjected to hemorrhagic shock. Methods: Anesthetized rats were equipped with a thoracic epidural catheter. A loop of terminal ileum was exteriorized and either muscularis or mucosa was investigated by microscopy. TEA was initiated by a bolus of 30 µl lidocaine 2% or normal saline (CTRL), followed by a continuous infusion of 60 µl/hr. Hemorrhagic shock was induced by withdrawal of blood to a mean arterial pressure of 30 mmHg. Capillary density of both, the mucosal villi and the muscularis, was determined at baseline, 30 min after onset of TEA, after 60 min of hemorrhagic shock, and 60 min after retransfusion of shed blood. Non-perfused capillaries were assessed and displayed as percentage of total capillary density. Results: In the mucosa, non-perfused capillaries increased significantly during hemorrhagic shock in both groups (CTRL: 10% [3/ 17%]), P < 0.01; TEA: 18% [6/33%], P < 0.05). After retransfusion of shed blood, non-perfused capillaries further increased in the CTRL (14% [2/25%]), but decreased in the TEA (5% [2/9%]) group. In the muscularis, the percentage of non-perfused capillaries during shock was higher in CTRL (40% [18/59%]) than in TEA animals (1% [0/ 4%]) (P < 0.001). After volume resuscitation non-perfused capillaries declined in CTRL (1% [0/4%]) and disappeared in TEA. Conclusion: TEA improves micro vascular perfusion in the muscularis during hemorrhagic shock and attenuates impairment of perfusion in the mucosa during reperfusion. Thus it might have a preventive effect on bacterial translocation. 1 Dept. of Anesthesiology, UKBF, Freie Universität Berlin; 2Dept. of Physiology, Freie Universität Berlin, Arnimallee 22, D - 14195 Berlin; 3Dept. of Anesthesiology, Deutsches Herzzentrum Berlin

P 27-6 DISSOCIATION OF INTRACELLULAR [Ca 2+] AND [NADH] DURING DIFFERENT ISCHEMIC REGIMES USING SURFACE FLUORESCENCE OF ISOLATED HEARTS M. Brachmanski, M.M. Gebhard, R. Nobiling Interaction between mitochondrial energy production and [Ca2+]i have been claimed to explain important aspects of the control of cardiac energetics under various energy demands [Brandes and Bers, Circ Res 1997; 82-87]. We tested this hypothesis for two different types of ischemic situations that occur upon sudden stop of nutritive perfusion or cardioplegia. We used optical methods for the analysis of the surface of isolated beating hearts under a modified Langendorff perfusion with respect to the important intracellular parameters [Ca 2+] i and NADH. [Ca2+]i was measured using the ratiometric dye fura 2, NADH - determinations were performed by measuring the surface fluorescence of the myocardium. Triple wavelength excitation at 340, 360 and 380 nm and measurement at 510 nm emission enabled us to separate the signals appropriately. The results show a basic correlation between the two parameters under the conditions of ischemia (stop of nutritive perfusion with oxygenated Tyrode solution) and a dissociation of the signals upon beginning of cardiac arrest during cardioplegia. A late, slow rise of [Ca2+]i has been reported earlier [Brachmanski et al, Eup J Physiol 2001; Supp 6, R 190]. On reperfusion after stopflow, again some correlation in concentration changes and kinetics was found between NADH and [Ca2+]i , whereas no correlation in both parameters could be seen under the condition of reperfusion following cardioplegia. We conclude from our results that under artificial ischemic situations such as cardioplegia a complete decoupling of cardiac energetics from [Ca2+]i is observed. Hence, [Ca2+]i may have an independent regulative role under artificial pathophysiological and/or clini-

S 296 cal conditions.

Dis.1996;7:732-43. [2] Kersten et al. Am.J.Physiol 2001;280:H174450

Abt. f. Exp. Chirurgie, INF 365, D69120 Heidelberg, Germany P 27-7 REPEATED CYCLES OF ISCHAEMIA ARE NOT MORE EFFECTIVE IN INDUCING LATE PRECONDITIONING THAN ONE 5-MIN CYCLE J. Müllenheim, °D. Ebel, W. Schlack, °V. Thämer Short periods of myocardial ischaemia protect the heart against infarction (ischaemic preconditioning).1 Ischaemic late preconditioning (LPC) can be induced by one 5-min cycle of preconditioning ischaemia.2 We investigated, if cardioprotection can be enhanced by multiple cycles of ischaemia. Thirty-one rabbits were chronically instrumented with a coronary artery occluder and divided into three groups. Chloraloseanaesthetized animals were subjected to 30 min of coronary artery occlusion and 120 min of reperfusion (I/R). Controls (CON, n=10) were not further treated. In the second group, LPC was induced by one 5-min cycle of myocardial ischaemia in awake animals 24 h before I/R (LPC1x5, n=9), and in the third group by five 4-min cycles (LPC5x4, n=12). We measured cardiac output (CO, flow probe) and left ventricular peak systolic pressure (LVPSP, tip manometer) during I/R. Infarct size (IS) was determined by TTC-staining. Statistics: Student’s t-test with Bonferroni correction of the P value. There were no differences in CO and LVPSP during baseline conditions (mean±SD; CO: CON 265±83, LPC1x5 267±83, LPC5x4 244±77 ml min -1, P=1.0; LVPSP: CON 96±24, LPC1x5 93±23, LPC5x4 79±33 mmHg; P=1.0) and after 120 min reperfusion (CO: CON 191±67, LPC1x5 170±27, LPC5x4 163±55 ml min-1, P=1.0; LVPSP: CON 76±25, LPC1x5 56±14, LPC5x4 59±16 mmHg, P=1.0) between the groups. In both LPC groups, IS of the area at risk size was significantly reduced compared with CON (CON 42±14, LPC1x5 22±8, LPC5x4 25±12 %, P<0.05 vs. CON), but there was no difference between the two LPC groups (P=1.0). We conclude that repeated cycles of preconditioning ischaemia do not further increase LPC-mediated cardioprotection provided by one 5min cycle of ischaemia. [1] Bolli, R. Circ.Res. 2000;87:972-983 [2] Müllenheim, J., et al. Pflügers Arch. 2001;442:178-187 Department of Anaesthesiology, °Institute of Physiology, HeinrichHeine-Universität Düsseldorf, Postfach 10 10 07, D-40001 Düsseldorf

Institute of Physiology, and °Departments of Anaesthesiology, Heinrich-Heine-Universität Düsseldorf, Postfach 10 10 07, D-40001 Düsseldorf P 27-9 ACTIVATION OF ATP-DEPENDENT POTASSIUM CHANNELS IS TRIGGER BUT NOT MEDIATOR OF ISCHEMIC PRECONDITIONING IN PIGS P. Gres, R. Schulz, A. Skyschally, G. Heusch Blockade of ATP-dependent potassium channels (KATP) with glibenclamide starting before the preconditioning ischemia and continued until the end of the index ischemia inhibits the reduction of infarct size by ischemic preconditioning (IP) in pigs (Am J Physiol 267: H1341-H1352, 1994). Studies in isolated buffer-perfused rabbit hearts suggested that activation of mitochondrial KATP may act as trigger rather than mediator of IP (Circ Res 87: 460-466, 2000). However, pigs are much closer to humans with respect to coronary blood flow and infarct development. Whether activation of KATP acts as trigger or mediator of IP in pigs in situ, is not yet known. In 27 enflurane-anesthetized pigs the LAD coronary artery was perfused from an extracorporal circuit. Subendocardial blood flow (ENDO, ml/ min/g, microspheres) and infarct size (IS, %, TTC) were measured. After 90 min of severe ischemia and 120 min reperfusion IS was 22.7±4.5 [SEM] (n=6 ENDO:0.05±0.01). IS was reduced to 5.8±1.7 (p<0.05, n=8, ENDO:0.05±0.01) by IP with one cycle of 10 min ischemia and 15 min of reperfusion. Glibenclamide given intravenously (0.5 mg/kg as a bolus followed by 50µg/min continuous infusion until the end of the index ischemia) did not change IS per se (18.2±2.4, n=6, ENDO:0.05±0.01). Whereas glibenclamide starting before the preconditioning ischema inhibited the infarct size reduction of IP (see above), its application starting just prior to the index ischemia did not affect infarct size reduction by IP (9.1±5.5, n=7, ENDO:0.04±0.01). Heart rate, left ventricular pressure and area at risk were comparable among the groups. Conclusion: Activation of KATP acts as trigger and not mediator of infarct size reduction by ischemic preconditioning also in pigs in situ. Department of Pathophysiology, University of Essen, Hufelandstr. 55, 45122 Essen

P 27-8 ISCHAEMIC LATE PRECONDITIONING IS BLOCKED BY ACUTE HYPERGLYCAEMIA IN THE RABBIT HEART IN VIVO D. Ebel, J. Müllenheim°, T. Bohlen, J. Ferrari, H. Südkamp°, R. Huhn°, M. Bauer°, W. Schlack°, V. Thämer Hyperglycaemia is a predictor of mortality after myocardial infarction in patients with and without diabetes mellitus [1]. Hyperglycaemia blocks ATP-sensitive potassium (K ATP) channels [2] which play an important role in ischaemic late preconditioning (LPC). We investigated the hypothesis that acute hyperglycaemia blocks the cardioprotection of LPC. Thirty-eight rabbits were chronically instrumented with a coronary artery occluder and divided into four groups. Anaesthetized animals were subjected to 30 min of coronary artery occlusion and 120 min of reperfusion (I/R). Controls (CON, n=10) were not further treated. In two groups, LPC was induced by a 5-min period of myocardial ischaemia 24 hours before I/R (LPC, H600+LPC, n=9 both). In the H600+LPC group and in a fourth group (H600, n=10), hyperglycaemia of approximately 600 mg dl1 was induced by dextrose infusion starting 30 min before I/R ending at the end of coronary occlusion. Left ventricular peak systolic pressure (LVPSP, tip manometer) and infarct size (IS, TTC-staining) were measured. Statistics: Student’s t-test with Bonferroni correction of the P value. There was no difference in LVPSP during baseline conditions (mean±SD; CON 93±28, LPC 96±25, H 600 90±28, H600+LPC 79±24 mm Hg; P=0.731.0) and after 120 min reperfusion (CON 72±23, LPC 57±13, H 600 53±21, H 600+LPC 48±23 mm Hg; P=0.28-1.0) between the groups. LPC reduced IS from 42±14 (CON) to 22±8 % of the area at risk (P=0.008). Hyperglycaemia itself tended to increase IS compared with controls (H 600 57±16 %, P=0.14), but completely blocked the cardioprotective effect of LPC (H600+LPC 59±19 %; P=1.0 vs. H600; P=0.0004 vs. LPC). An acute hyperglycaemia before and during IS inducing ischaemia blocks the cardioprotection by LPC, most likely by a direct blockade of KATP channels. [1] Jelesoff et al. Coron.Artery

P 27-10 ROLE OF CALPAINS IN REPERFUSION INJURY AND VASCULAR PERMEABILITY OF ISOLATED GUINEA PIG HEARTS V. Gonscherowski*, S. Zahler*, L. Moroder#, B.F. Becker* Changes in myocardial function, e.g. impaired contractility and coronary leak, are observed during reperfusion of ischemic hearts. Due to the increase in intracellular Ca2+ upon reperfusion, calpains (neutral cysteine proteases needing Ca2+) might be activated and contribute to reperfusion injury by cleavage of cellular proteins such as fodrin, which forms part of the sub-membranous cytoskeleton. We conducted experiments on isolated guinea pig hearts and cultured human umbilical vein endothelium (HUVEC), since EC contain large amounts of calpain and play an important role in vascular permeability. After 30 min of perfusion, hearts underwent 15 min of global ischemia with following 60 min reperfusion, or were perfused as nonischemic controls. Heart homogenates were analysed for degradation of fodrin by Western blots. In another set of experiments hearts received additional perfusion with two calpain inhibitors (calpain inhibitor I and calpastatin peptidepenetratin) 10 min prior and 10 min after ischemia. To evaluate coronary vascular permeability, transudate (interstitial fluid appearing at the epicardial surface) and perfusion pressure were measured. The degradation of fodrin was also assessed in confluent monolayers of HUVECs that were or were not incubated with calpain inhibitor I under shear stress conditions.Western blots of fodrin revealed a significant degradation of this protein in the membrane fraction of postischemic heart tissue (n = 11). Unexpectedly, transudate formation increased significantly in hearts perfused with either calpain inhibitor (n = 3-6). This increased coronary leak was even seen in hearts that did not experience ischemia. According to Western blots, fodrin decreased in HUVECs under shear stress. This degradation could be abolished by

S 297

*Institute of Physiology, University of Munich, Schillerstr. 44, 80336 Munich, Germany; # MPI, Munich - Martinsried

used as internal standard. Expression of NHE1, NHE5, rbNBC, AE2b (short form) and AE3 was clearly detectable. Our results are consistent with the in vivo results of Ennis et al and indicate further that Na+ and HCO3- dependent re-acidification occurs via a NBC transporter possibily in the abluminal membrane. The absence of any effect of removal of Cl- makes it unlikely that C1-/HCO3- exchange occurs under the conditions of our experiments.

P 27-11

Department of Pharmacology, University of Cambridge, Cambridge UK CB21PD

incubation of the cells with calpain inhibitor I. These results point to a physiological function of calpains in endothelial cells and suggest a role for this enzyme in maintaining endothelial barrier function. This project is supported by SFB 469.

REPERFUSION-INDUCED GAP-FORMATION IN CORONARY ENDOTHELIAL CELLS S. Walther, M. Schäfer, C. Schäfer, Y. Abdallah, B. Rother, H.M. Piper During ischemia and reperfusion the endothelial barrier function is impaired. The aim of the present study was to investigate wether reperfusion itself can increase endothelial gap formation and if this can be due to an increase in activitiy of the endothelial contractile machinery. As experimental model we used coronary endothelial cells of the rat. During anoxia (pH 0 6.4) and reoxygenation (pH 0 7.4) the time course of cytosolic Ca 2+ (Fura-2) and the intercellular gap formation (planimetrical analysis) were determined. In order to investigate if the intercellular gap formation can be due to an increase of contractile activity, the cells were reoxygenated in presence of NaCN (2mM) or ML-7 (10µM), an inhibitor of the myosin light chain kinase (MLCK). During anoxia EC developed cytosolic Ca 2+-overload (Fura-2-ratio: end-anoxic: 1.21±0.01; normoxic: 1.02±0.01; *; p<0.05). Anoxia induced intercellular gap formation, however reperfusion itself increased this effect (gap formation normoxia (a.u.) 29.5±5.2; anoxia 75.4±12.0; reoxygenation: 110.7±9.2,p<0.05 vs normoxic, end-anoxic). The reperfusion-induced gap formation under control conditions were related to 100%. In presence of ML-7 or inhibition of metabolic recovery with NaCN significantly reduced the reperfusion induced gap formation (Ctr.: 100%; ML-7: 67%±6; NaCN: 33%±8; n=5; p<0.05 vs Ctr.). During 30 min reperfusion the cytosolic Ca2+ remained still elevated under all experimental conditions. Reperfusion itself increases intercellular gap formation and therefore impaires the endothelial barrier function. The reperfusion-induced gap formation is dependent on activation of the contractile machinery, which is mediated by the reenergetisation of the cells at elevated cytosolic Ca2+. Institute of Physiology, Justus-Liebig-University Giessen Aulweg 129, D-35392 Giessen, Germany P 27-12 ENDOTHELIAL CELL ION TRANSPORTERS AT THE BLOODBRAIN BARRIER C.J. Taylor. G. Bennett, S.B. Hladky, M.A. Barrand Secretion of fluid across the blood-brain barrier accounts for the formation of oedema in the initial hours following focal ischaemia/ reperfusion. It is known that the secretion is driven by the abluminal Na,K,Mg-ATPase operating in conjunction with luminal transporters for Na+. Ennis et al (1996, J. Neurochem. 66: 756-763) demonstrated saturable transport inhibitable by dimethylamiloride and suggested that this Na+/H+ exchange together with Cl-/HCO3- exchange would support secretion of NaCl. The present study seeks to identify the transporters mediating the movements of acid or base equivalents in endothelial cells isolated from rat brain microvessels and maintained in primary culture. Internal pH (pHi) was measured using BCECF. Cells in buffer containing 10 mM HEPES and 5% CO2/HCO3-, pH 7.4 maintain pHi = 7.32±0.03 (n=32). Removal of CO 2 /NaHCO 3 - leads to a rapid alkalinisation followed by re-acidification. For cells in HEPES, addition of CO2/HCO3- leads to acidification followed by re-alkalinisation. Replacement of all the Cl- in the medium with gluconate- did not change pHi or the responses to additions or removals of HCO3- (n= 3 out of 3). To investigate the mechanism of re-alkalinisation, pHi was reduced by exposing cells for 200 s to HEPES buffered solution with 20 mM NaCl replaced by NH4C1 (pH 7.4). For HEPES buffered solutions the rate of the subsequent realkalinisation (initial rate 0.17±0.06 mmol l-1 s-1, n=5) could always be reduced by 5 µM EIPA (0.027±0.004, n=7). For cells in CO2/HCO3- buffer, 5 µM EIPA reduced the rate to 0,06±0.01 (n=3) while 5 µM EIPA + 250 µM DIDS reduced it to 0.01±0.0l (n=3). Total replacement of Na+ with NMDG+ produced comparable inhibition. Primers for use in RT-PCR were designed for the rat Na +/H+ exchangers, NHEl, NHE2 NHE3 &NHE5, C1-/HCO3-exchangers, AEl, AE2 (a,b&c) & AE3, and Na +, HCO 3- cotransporter rbNBC. β-microglobulin was

P 28-1 POSTSYNAPTIC SECRETION OF BDNF-GFP AFTER HIGH FREQUENCY STIMULATION OF GLUTAMATERGIC SYNAPSESS M. Hartmann, R. Heumann, V. Lessmann The protein brain-derived neurotrophic factor (BDNF) has been postulated to be released synaptically as a possible retrograde messenger in long-term potentiation (LTP). However, although crucial for this concept, direct evidence for activity-dependent synaptic release of BDNF was lacking. Here we investigate secretion of BDNF labeled with green fluorescent protein (BDNF-GFP), by monitoring the changes in fluorescence intensity of BDNF-GFP secretory vesicles at glutamatergic synapses. Hippocampal neurons were transfected at 8 DIV with a BDNF-GFP fusion construct, and investigated at 11-14 DIV. Using FM 4-64 labelling of active synapses, we selected for synaptic BDNF-GFP vesicle clusters in dendrites, and monitored changes in intracellular fluorescence intensity (reflecting release of BDNFGFP) by time lapse video microscopy. Stimulation with 50 mM K+ resulted in a graded decrease of fluorescence intensity of BDNF-GFP vesicles in postsynaptic structures (tau= 300s +/- 24s), yielding 59 +/ - 2% (mean +/- s.e.m., 17 cells) remaining fluorescence intensity after 10 min, compared to 83 +/- 2% in unstimulated controls. This fluorescence decrease was critically dependent on the influx of extracellular Ca 2+ and was abolished by preincubation with tetanus toxin (20 h, 1 nM) which blocks Ca 2+ dependent vesicular release processes. Importantly, synaptic release of BDNF-GFP upon high frequency electrical presynaptic stimulation (16x 1s bursts at 50 Hz) was strictly dependent on postsynaptic depolarization via ionotropic glutamate receptors, supporting the view that secretion of BDNFGFP takes place from postsynaptic structures. Taken together, these results provide direct evidence that BDNF is a retrograde synaptic messenger in activity-dependent plasticity of glutamatergic synapses. (Supported by DFG grant SFB 509 TP C9) Molecular Neurobiochemistry, Ruhr-University Bochum NC7-132, 44780 Bochum, Germany P 28-2 TARGET-SPECIFIC APPLICATION OF BDNF MEDIATES LONG-TERM POTENTIATION OF SYNAPTIC TRANSMISSION IN GRANULE CELLS OF DENTATE GYRUS Y. Kovalchuk, E. Hanse, K.W. Kafitz, A. Konnerth The neurotrophin (NT) brain-derived neurotrophic factor (BDNF) is involved in numerous aspects of development in the CNS. In addition to this chronic action, BDNF is responsible for the relatively fast enhancement of synaptic transmission at various central excitatory synapses, including the hippocampal perforant path-granule cell synapse. Furthermore, there is increasing evidence that exogenous BDNF facilitates the induction process of hippocampal long-term potentiation (LTP) by enhancing transmitter release, suggesting the action of NT at presynaptic site. To overcome some limitations of earlier studies we have now (i) took the advantages of rather strict laminar organization of the perforant path, (ii) used applications of brief (10 – 600 ms) and targeted pulses of BDNF in slice preparations from mature mice, and (iii) monitored intracellular Ca2+ in different compartments of granule cells and/or in presynaptic structures. We now found that brief BDNF pulses produced an increase in granule cell excitability and a concomitant transient [Ca2+]i increase in spines and dendrites but not in axons of the perforant path. As had been shown for CA1 pyramidal neurons (Kafitz et al., Nature 1999), the rapidly evoked BDNF responses were also induced by neurotrophin-4/5, but not by NT-3 or NGF. BDNF activation was blocked by K252a, a protein kinase blocker. Further, pairing a weak burst of synaptic stimu-

S 298 lation with dendritic BDNF application within a narrow time-window caused an immediate induction of LTP. The LTP induction was prevented by blocking postsynaptic Ca2+ transients and was occluded by tetanic stimulation-evoked LTP. Thus, we identified dendrites and spines but not presynaptic terminals as highly responsive compartments for the rapid, TrkB receptor-mediated Ca2+ signaling. Our results strongly suggest that the BDNF-mediated LTP is induced postsynaptically. Institut für Physiologie, LMU, D-80802 München, Germany

strengthened the β rhythm . Bioelectrical activity was got poor and rhythmic activity was inhibited by the test-substance (15-30 mg/kg). This effect continued more powerfully in 30 mg/kg. After 2.5 hours of the applications, spectral activity of the θ rhythm was getting poor, whereas the spectral activity of the β rhythm was getting more powerful. Conclusions: The effect of the tested substance was found extremely different than those of Fenazepam. Benzodiazepins can get strong both β and especially θ activities. The effect of the test-substance was stronger than Befol so that the effect of that looked as if Befol. Finally, the test- substance did the psychotropic effect on the bioelectrical activity of the brain that was rather different from typical effect of tranquilizer.

P 28-3 LONG-TERM POTENTIATION IS ENHANCED IN NEOCORTICAL FIELD POTENTIALS OF THE TRANSGENIC SYN-RAS MOUSE. G. Barmashenko 1, M. Huemmeke1, T. Mittmann1, R. Heumann2 Recently the intraneuronal protein Ras has been linked to functional processes in the CNS related to synaptic plasticity (Brambilla et al., Nature, 1997; Manabe et al., J.Neurosci., 2000). Here we directly studied the effects of constitutively activated V12Ha-Ras on longterm-potentiation (LTP) in a synRas-transgenic mice model (Heumann et al.; J.Cell Biol.; 2000). Ten adult heterozygous synRas-mice and their wild-type littermates (n=11) at the age of >40 days were anaesthetized with ether and decapitated. Slices of 350µm thickness were prepared and stored in normal ACSF at room temperature. Single slices were transferred to a submerged type chamber and extracellular field potentials (FPs) were recorded from cortical layers II/III, while electrical stimulation was performed from ascending inputs in cortical layer IV. The input-output relation of synaptically evoked FPs revealed significantly higher mean FP amplitudes at stimulus durations of 100200µsec in transgenic mice compared to wild-type animals (p<0.05). For LTP experiments synaptically evoked FPs were collected 40 minutes before and 60 minutes following a theta-burst stimulus (TBS). Compared to recordings in wild-type littermates the strength of LTP was significantly (p=0.003) enhanced in transgenic animals (syn-Ras: 134.2 ± 6.4 %, n=15; wild-type: 112.8 ± 2.3 %, n=25). The induction of LTP could be blocked by bath application of 25µM D-AP5 in both experimental groups. However, application of the MEK inhibitor U0126 selectively reduced LTP in transgenic mice (112 ± 5.2 %, n=10, p<0.01). These data indicate, that activation of Ras in cortical neurons of transgenic syn-Ras mice enhances an NMDA-receptor mediated type of LTP and that MAPkinase activity is critically involved in these Ras-induced alterations of long-term synaptic plasticity. Supported by the Deutsche Forschungsgemeinschaft (SFB 509, C7 and C9). 1 Dept. Neurophysiology, Medical School & 2Dept. Neurobiochemistry, Ruhr-University Bochum, MA4/149, D-44780 Bochum, Germany

P 28-4 THE EFFECT ON BIOELECTRICAL ACTIVITY OF THE BRAIN OF 5-ETHOXY-2-[2-(MORPHOLINO)-ETHYLTHIO]-BENZIMIDASOLE IN RATS Y. Yetkin¹,H. Sabiroglu² Objectives: This study is intended to investigate the effect on the bioelectric activity of 5-Ethoxy-2-[2-(Morpholino)-Etylthio]Benzimidasole in brain. Methods: The study was done on Wistar rats into 5 groups (n=30; weight from 180-220 g). The effects of the testsubstance, Fenazepam and Befol were investigated by electromyography in sensory-motor cortex, dorsal hippocampus, lateral hypothalamus, thalamus and septum , respectively. Serum physiologic was used only for control group (i.p). However, for 2nd and 3rd groups the test-substance were used in 15 and 30 mg/kg, and for 4th and 5th groups 0.5 mg/kg of Fenazepam and Befol were also used to compare with the effects of the test-substance. The changes of the bioelectrical activities of the brain were observed for 5 hour in the same units of the day. The electrodes with the dimensions of 120-150 mkm for cortex and of 70-90 mkm for hippocampus were placed within the brain by stereo-taxis equipment under depth anesthesia (Nembutal: 50 mg/kg; i.v.). After 3-5 days of processes, the bioelectrical activities were recorded. The electrodes were controlled morphologically whether they reached within the areas which are asked for recording or not. Results: After 20 minutes of procedures, Fenazepam strengthened the θ and β activities and occurred non-rhythmic activity, but Befol weakened θ rhythm, whereas

Medical School of Yüzüncü Yýl University, Department of Physiology¹ and Pharmacology ², 65200, Van, Turkey P 28-5 HIPPOCAMPAL MOSSY FIBER LONG-TERM POTENTIATION REQUIRES PRESYNAPTIC Cav2.3 CALCIUM CHANNEL SUBUNITS T. Kirschstein1, A. Pereverzev2, J. Hescheler2, T. Schneider2, C.E. Elger1, H. Beck1 Induction of mossy fiber long-term potentiation (MF-LTP) is thought to be dependent on presynaptic Ca2+ entry, but the route of Ca2+ entry into pre-synaptic terminals has been a matter of debate. Prior work suggested that either N- or P/Q-type Ca2+ channels are not responsible, with the role of presynaptic kainate receptors being controversial. Here, we demonstrate the involvement of presynaptic Ca v2.3-mediated R-type Ca2+ channels in MF-LTP using pharmacological blockers of R-type currents and knock-out mice lacking Cav2.3 subunits. The R-type antagonists NiCl 2 (100 µM) or SNX-482 (500 nM) did not affect mossy fiber excitatory postsynaptic field potentials (fEPSPs), but significantly reduced the level of MF-LTP at 1 hr after tetanization (100 Hz, 2x1 s; 110±20% and 111±9%, vs. 149±41% in controls, mean±SD, p<0.05). This effect was most probably due to block of presynaptic R-type channels since SNX-482 had only small effects on postsynaptic R-type Ca2+ channels in CA3 neurons. In mice lacking the Cav2.3 subunit, MF-LTP was also significantly reduced compared to wild-type mice (110±23%, p<0.05). To exclude that a disturbance of the downstream signalling cascades mediating LTP expression might account for reduced LTP in Cav2.3 knock-out mice, we have pharmacologically activated the key enzymes in this process. Direct stimulation of adenylate cyclase (AC) by forskolin and isobutylmethylxanthine (50 µM each) caused significantly less potentiation after 4 hrs in knock-out compared to wild-type mice (109±26% vs. 141±20%, p<0.05). However, LTP induced by direct activation of the protein kinase A using the cAMP analogue Sp-cAMPS was intact in knock-out mice, indicating that LTP expression mechanisms downstream of AC are intact. These results demonstrate that MF-LTP induction, but not fast synaptic transmission requires presynaptic Ca v2.3-mediated Rtype Ca2+ channels. 1

Department of Epileptology, Sigmund-Freud-Str. 25, 53105 Bonn, Germany 2Department of Neurophysiology, 50931 Köln, Germany

P 28-6 DIFFERENTIAL CONTRIBUTION TO LTP OF HIPPOCAMPAL NMDA RECEPTOR SUBTYPES G. Köhr, V. Jensen, J.K. Utvik, H. Köster, A. Kvello, R. Sprengel, P.H. Seeburg, O.P. Ottersen, Ø. Hvalby Hippocampal CA1 pyramidal cell synapses contain different NMDA receptor subtypes assembled from NR1 with NR2A and/or NR2B subunits. The distinct intracellular C-terminal domains of the NR2A and NR2B subunits predict that they interact with different postsynaptic proteins and hence, that NMDA receptor subtypes may play different roles in synaptic plasticity. Here we investigated the contribution of the hippocampal receptor subtypes to LTP induction by employing NR2A∆C/∆C mice that express NR2A only in its C-terminally truncated from. Postembedding immunogold labeling showed that relative to NR2A in wild-type mice NR2A∆C was reduced to about 50 % in postsynaptic densities of NR2A∆C/∆C mice at P35 but NR2B was not affected. Nevertheless, NMDAR-mediated spinous calcium transients were comparable between genotypes. Hippocampal LTP is reduced in

S 299 adult NR2A∆C/∆C mice but repetitive tetanic stimulation which did not change the extent of LTP in adult wild-type mice restored LTP in NR2A∆C/∆C mice. The NR2B-specific antagonist CP-101,606 blocked both impaired and restored LTP in NR2A∆C/∆C mice without affecting the LTP in adult wild-type mice. Thus, NR2A-containing NMDA receptors mediate and saturate LTP in adult wild-type mice but NR2Bcontaining NMDA receptors mediate LTP restoration in NR2A∆C/∆C mice. These results indicate that differential NMDA receptor signaling in adult mice results from different signaling complexes associated with the C-termini of NR2A or NR2B. Supported by DFG Ko 1064/3-2. Max-Planck-Institute for Medical Research, Department of Molecular Neurobiology, Jahnstr. 29, 69120 Heidelberg, Germany P 28-7 INTERACTION OF L-TYPE CALCIUM CHANNEL BLOCKERS Β [23-35] ON LONG TERM POTENTIATION IN THE AND AΒ HIPPOCAMPAL CA1 IN VITRO D.A. Costello, C.E. Herron Hippocampal Long Term Potentiation (LTP) is an increase in synaptic efficacy induced experimentally in vitro by trains of high frequency stimuli (HFS, 100Hz). Induction of LTP in the CA1 region depends on an increase in the post-synaptic calcium concentration that may be mediated via activation of the NMDA receptor/channel. A form of NMDA receptor-independent LTP has also been identified induced by higher stimulus frequencies (200Hz; Grover & Teyler, 1990). This form of potentiation is mediated in part by activation of L-type voltagedependent Ca2+ channels. Alzheimer’s disease (AD) is a neurodegenerative disorder characterised by senile plaques containing beta-amyloid peptide (Aβ). The Aβ fragment [25-35] is neurotoxic and can also cause an increase in activation of L-type Ca2+ channels (Ueda et al., 1997). LTP is impaired significantly by Aβ [25-35], in vitro and in vivo (Chen et al., 2000; Frier et al., 2001). Here we have investigated the interaction of the L-type Ca2+ channel blockers verapamil and diltiazem with Aβ [25-35] on CA1 LTP in vitro. Transverse hippocampal slices were prepared from male Wistar rats (50–100g) and perfused with artificial cerebrospinal fluid (~30 oC). The Schaffer-collateral / commissural pathway was stimulated (0.033 Hz) and field excitatory post-synaptic potentials (EPSPs) were recorded from of the CA1 region. Verapamil (20µM) or diltiazem (50µM) had no effect on baseline synaptic transmission. Substantial LTP was induced by HFS (3 periods of 10 trains of 10 stimuli at 200Hz with an inter-train interval of 20 seconds) under control conditions (EPSP slope: 161±5%, n=15, mean±SEM, recorded 1 hour post-tetanus). LTP was reduced significantly by verapamil (132±4%, p<0.05, n=6, ANOVA) and / or diltiazem (119±3%, p<0.05, n=6, ANOVA) applied 20 minutes prior to HFS. Application of Aβ [25-35] (500nM) 1 hour pre-tetanus had no effect on baseline synaptic transmission, however it reduced significantly the level of LTP (141±8%, p<0.05, n=6, ANOVA). Co-application of verapamil (20µM) and Aβ (500nM) reduced the attenuation of LTP observed previously in the presence of either agent alone (160±8%, n=4). However, co-application of diltiazem (50µM) and Aβ did not appear to reduce the level of attenuation. These results support a previous observation in vivo. The interaction of verapamil and Aβ [25-35] is under further investigation. Chen, Q., Kagan, B., Hirakura, Y. and Qie, C. (2000). J. Neurosci. Res. 60, 65-72. Grover L.M. and Teyler, T.J. (1990). Nature 347, 477-479. Frier, D.B., Holscher, C and Herron, C.E. (2001). J. Neurophysiol. 85: 708-713. Ueda, K, Shinohara, S., Yagami, T., Asakura, K. and Kawasaki, K. (1997). J. Neurochem. 68, 265-271. Supported by The Health Research Board of Ireland and Enterprise Ireland. Department of Physiology, National University of Ireland, Dublin 2, Ireland.

processes local synaptic input we used two-photon calcium imaging of individual spines and spino-dendritic compartments in coronal slices from primary visual cortex of P14-19 mice. Layer 5 pyramidal neurons were patched and filled with the calcium indicator dye CalciumGreen1. Calcium signals in spines and dendrites were imaged with a custommade two-photon microscope and were triggered with subthreshold synaptic stimulation. We find that strong synaptic stimulation induces a local calcium response in a small spino-dendritic compartment, which was accompanied by a complex EPSP containing a prominent second depolarizing component. These local spikes involve the activation of NMDA receptors and dendritic voltage-sensitive calcium channels and induce long-lasting synaptic depression (LTD). This LTD does not require action potential firing and is restricted to coincident inputs in discrete dendritic segments. The synaptic depression is input-specific, because a second stimulation pathway in the same cell is not affected. On the single synapse level we find that after conditioning stimulation, the calcium response of individual dendritic spines located at the same part of the dendrite is reduced or eliminated. We hypothesize that pyramidal neurons use this position-dependent learning rule to implement input sparsification, a strategy used in artificial neural networks, particularly in relation with associative memory and mapping tasks. Supported by the NEI and the HFSP. 1 Institut für Physiologie der LMU München; 2Dept. of Biological Sciences, Columbia University, New York, NY 10027

P 28-9 THYROID HORMONE INCREASES VOLTAGE-ACTIVATED Na+-CURRENTS IN ACUTELY ISOLATED NEURONS FROM POSTNATAL RATS G. Hoffmann, I.D. Dietzel* In previous experiments we have shown that the biologically most potent component of thyroid hormone, triiodothyronine (T3), causes an upregulation of Na+ currents in cultured hippocampal neurons (O. Potthoff, I.D. Proc. R. Soc. Lond. B, 264, 367-373, 1997). To investigate whether this increase in Na +-current density correlates with changes in action potential upstroke velocity and firing frequency we compared action potentials recorded from cultures treated with 30 nM T3 as compared with sister cultures incubated in the absence of thyroid hormone. Cultures were prepared from 3-7 day old rat pups and recordings performed after 2-4 days in T3- free B18 medium or B18 supplemented with 30 nM T3. Hippocampal cells treated with T3 showed a shortening of action potential rise time (10-90%) by 78 +/12 % compared with control cells (45 treated and 44 untreated cells investigated). In cortical neurons treatment with thyroid hormone decreased the rise time to 83 +/- 12 % (33 test cells, 37 control cells). In addition T3-treatment decreased the minimal interval between successive action potentials by 77 +/- 10% in hippocampal cells and by 85 +/- 12 % in cortical cells. Action potential amplitudes were about 10% larger in T3-treated hippocampal and cortical cells compared with those recorded from control cells. To verify whether the regulation of Na+ currents observed in cell culture also occurs in vivo we treated pregnant rats from the 18th day of gestation with the thyreostatic drug propylthiouracil (PTU) or pups with daily doses of 2 µg T3 and 2 µg thyroxine. T3-treated rats showed significant increases in Na+ current densities and Na+/K+ current ratios as compared with PTU-treated rat pups. Hence the regulation of sodium currents by thyroid hormone could result in functional changes that contribute to the severe neurological symptoms induced by thyroid hormone deficiency in the postnatal period. Lehrstuhl für Molekulare Neurobiochemie, Ruhr-Universität Bochum, NC7- 170, Universitätsstraße 150, D-44780 Bochum, Germany P 28-10

P 28-8 LOCAL DENDRITIC SPIKES INDUCE SYNAPTIC DEPRESSION K. Holthoff1,2, A Konnerth1, R. Yuste2 Mammalian dendrites are active structures capable of regenerative electrical activity. The function of these dendritic spikes is still unclear, but could underlie the computational properties of neurons. Besides sodium-based action potentials, which can propagate throughout the dendritic tree, neocortical pyramidal neurons can also sustain dendritic spikes that are spatially localized. To investigate how the dendritic tree

DIURNAL EFFECTS OF MELATONIN ON SYNAPTIC TRANSMISSION IN RAT HIPPOCAMPAL SLICES A. Wortmann, T. Kühn, E. Berger, E.-J. Speckmann, U. Mußhoff Melatonin has been found to modulate the excitability of CA1 pyramidal cells in a diurnal manner. The aim of the present experiments was to in-vestigate possible diurnal effects of melatonin on the transmission at the Schaffer collateral/CA1 synapse. In addition, we examined the expression of clock genes in the hippocampus. Rats

S 300 (150-250 g) were entrained to a 12/12 hrs light-dark cycle (zeitgeber time ZT0 = lights on). In order to perform subjective night experiments during the daytime the rats were accustomed to a reversed light-dark cycle for a minimum of 4 weeks. Hippocampal slices (500 µm) were prepared at ZT 2-3. The experiments from these slices took place between ZT 5-9 and pooled to form a day group (n=21). As a comparison, slices were prepared at ZT 13-14, the experiments taking place between ZT 16-21 and pooled to form a night group (n=13). After stimulation of the Schaffer collaterals, field potentials (fEPSP) were recorded from the str. radiatum of the CA1-region. Gene expression of several components of the molecular clockwork (clock, Bmal1, period 1-3, cryptochromes 1-2, timeless) was determined in hippocampal tissue by RT-PCR. Bath-applied melatonin (10 nmol/l) has no effects on slices of the night group but increased the amplitude of fEPSP in slices of the day group more than 2 times of the initial level. This effect was completely suppressed through simultaneous administration of the melatonin receptor antagonist luzindole (100 nmol/l). Furthermore, the RT-PCR data revealed that the transcripts of clock genes are present in the hippocampus. The data indicate that various forms of hippocampal activity are related to circadian time. Institut für Physiologie, Robert-Koch-Str. 27a, 48149 Münster P 28-11 REGENERATIVE POTENTIAL OF ORGANOTYPIC SLICES CULTURED ON MICROELECTRODE ARRAYS F. Hofmann, H. Hämmerle, C.S. Leibrock Evaluation of the regenerative potential of neural tissue requires a model system that allows for long-term monitoring of physiological activity. By combining organotypic co-cultures with extracellular multielectrode recording technology the development of new connections can be repeatedly monitored on a functional level in the same culture for days to weeks. Cortical slices prepared from 3 to 7 days old rats or slices of entorhinal cortex (EC) and dentate gyrus (DG) prepared from mice were fixed on microelectrode arrays (MEAs) and maintained in vitro for up to 62 days. Electrophysiological recordings of evoked and spontaneous activity monitored simultaneously in both explants was used to identify regeneration of functional connections between the explants. 69% of the cortex/cortex (n=36) and 95% of the EC/DG co-cultures (n=40) formed functional connections within the first two weeks in vitro. After cultures had reached an adult-like state in vitro, the co-cultures with functional connections were mechanically lesioned. A lesion was considered successful when a few hours later electrical stimulation in one slice only resulted in a response in the same and not the other slice. Both directions were tested. Co-cultures were tested for viability and for regenerated connections for up to 21 days after lesion. In the EC/DG co-cultures, no spontaneous regeneration was observed (n=5). In contrast, 2 of the 7 cortex/ cortex co-cultures showed transmission of the signal to the other slice after at least 6 days after lesion. While the cortical co-cultures grew connections going both ways when first put into culture in a juvenile state, later, after lesions, functional connections were only found for one direction. The difference in regenerative potential displayed by the two tissue systems might be explained by cellular specificity and differing extracellular environments. Further, it would be interesting to investigate whether specific treatments were able to enhance the regeneration of functional connections in the cortex/cortex and also in the EC/DG preparation. NMI Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen, Markwiesenstraße 55, D-72770 Reutlingen, Germany (e-mail: [email protected]) P 29-1 ACTION-POTENTIAL EVOKED MINIATURE-LIKE SYNAPTIC EVENTS REVEALED BY SPATIAL RESTRICTION OF TRANSMITTER RELEASE IN STRIATAL AUTAPTIC CULTURES J.C. Behrends A synaptic connection often involves several contact sites, so that their individual contributions to impulse-dependent transmission overlap in time and cannot be determined directly. Direct electrical stimulation of synaptic boutons in cell culture has been used to overcome this problem. Since action potentials have to be blocked to avoid activation of multiple sites, this method relies on the passive depolarization of

terminals. However, the physiological shape of the terminal depolarization is critical for the resulting presynaptic Ca2+-influx and, therefore, for transmitter release. We here show that responses evoked by normal action potentials at single synaptic contacts can be resolved if, by selectively supplying extracellular Ca 2+ ions to a sufficiently small area, transmitter release is restricted to one or very few terminals. This is achieved by superfusion of an autaptic neuron with nominally Ca2+-free extracellular solution containing 5 mM Mg2+ and appropriately timed focal application of Ca2+ by submillisecond current pulses from a iontophoretic electrode to presumed autaptic boutons. The resulting postsynaptic responses have kinetics closely resembling those of miniature IPSCs in the same preparation, strongly suggesting that they originate from single or few synaptic contacts. Physiologisches Institut der Universität München, Pettenkoferstr. 12, 80336 München P 29-2 PRESYNAPTIC CALCIUM CURRENTS AND TRANSMITTER RELEASE IN CSP KNOCK-OUT MICE R. Schneggenburger, R. Fernández-Chacon, M. Wölfel, R.E. Hammer, T.C. Südhof The vesicle-associated protein cysteine string protein (CSP) has been implicated in presynaptic Ca2+ channel (ICa,pre) modulation, and in the coupling of ICa,pre to neurotransmitter release. To verify its proposed role, we measured ICa,pre and transmitter release at the calyx of Held synapse in CSP knock-out mice (-/-) and their wild-type littermates (+/+). Knock-out mice were viable at birth, but they showed early lethality at 4-5 weeks of age. In slices from P9 – P13d mice, ICa,pre recorded from calyces of Held was 1.1 ± 0.37 nA (n=18) and 1.26 ± 0.33 nA (n=13) for CSP -/- and +/+ mice, and thus, not significantly different. The apparent voltage-dependence of ICa,pre , as well as its sensitivity to ω-Aga-IVa (200 nM) were unchanged (92 ± 4% and 93 ± 7% block for CSP -/- and +/+), showing that P/Q type Ca2+ channels contributed a large fraction of ICa,pre in both genotypes. In paired wholecell recordings with 0.5 mM EGTA in the presynaptic pipette, a fast and a slow phase of transmitter release can be distinguished, indicating heterogeneity of release probability prel (Sakaba & Neher, 2001). Similar fast and slow phases of transmitter release were found in CSP -/- vs +/ + mice (tfast 1.8 ± 1.0 vs 1.6 ± 0.4 ms; tslow 30 ± 9 vs 25 ± 8 ms; n = 6). Our results indicate that I Ca,pre , and the heterogeneity of p rel at the calyx of Held synapse are not significantly changed in CSP -/- mice. Possibly, CSP plays a role in synaptic function at later stages of development, but it does not seem critical for the function of a calyceal synapse during early postnatal stages. MPI for biophysical Chemistry, 37077 Göttingen; and Center for Basic Neuroscience, U.T. Southwestern Med. Ctr., Dallas, TX-75390 U.S.A. P 29-3 THE CALYX OF HELD AS A MODEL SYSTEM OF NEUROTRANSMITTER RELEASE: VIRUS-MEDIATED PERTUBATION OF PROTEIN FUNCTION IN A GIANT PRESYNAPTIC TERMINAL V.C. Wimmer, B. Sakmann, T. Kuner The calyx of Held, which functions as a relay synapse in the auditory pathway, is a large presynaptic terminal located in the medial nucleus of the trapezoid body (MNTB). Simultaneous pre- and post synaptic patch-clamp recordings provide unique access to synaptic function. We attempted to study the molecular physiology of neurotransmitter release by expressing in the calyx presynaptic proteins fused to green fluorescent protein (GFP). Gene transfer was achieved with replicationdeficient sindbis virions produced from pSinRep5 containing the gene of interest and a neurotropic DH(26S) helper virus. Projection neurons in the anteroventral cochlear nucleus were infected by targeted stereotaxic injection of virus solution in anesthesized P21 rats. After recovery, rats were kept for 24 to 48 h to allow expression of the protein and diffusion over about 3-4 mm into the calyx located in the contralateral MNTB. GFP-expressing calyces were examined in sections prepared from paraformaldehyde-fixed tissue or in acute slices. Robust expression of GFP was found in about 5-25 calyces per injected animal. Confocal fluorescence microscopy revealed the structure of the calyx with several main fingers carrying multiple spine-like

S 301 protrusions. Complexin-GFP fusion protein was expressed in the terminal although to a lesser degree than GFP alone, which gave the brightest fluorescence signal. These results suggest that the controlled expression of proteins in the calyx of Held is possible, opening a new approach to study presynaptic function with electrophysiological and optical techniques. Max-Planck-Institute for Medical Research, Dept. of Cell Physiology, Jahnstrasse 29, 69120 Heidelberg P 29-4 PRESYNAPTIC RESTING POTASSIUM CHANNELS OF HIPPOCAMPAL MOSSY FIBER BOUTONS J.R. Schönberger, J.R.P. Geiger, P. Jonas We examined the molecular basis of the resting potential in hippocampal mossy fiber boutons (MFBs). Presynaptic recordings in hippocampal slices were obtained as described previously (Geiger and Jonas, Neuron 28, 927, 2000). The majority of outside-out patches isolated from MFBs contained a large-conductance, voltage-independent K+ channel. Single-channel activity occurred in bursts, with very brief open and closed periods within the burst. In 100 mM external K+, the apparent single channel slope conductance at -90 mV was 87 pS (n = 4; 23°C; determined by levels analysis, Patlak, J Gen Physiol 92, 413, 1988). A subset of patches also contained flickering channels with smaller conductances. The reversal potential of the single-channel current was dependent on external K+ concentration, suggesting that the channel was selective for K+ over Na+. External arachidonic acid (AA, 20 - 60 µM) led to a 22 ± 12-fold increase in the apparent open probability (n = 5), whereas external cpt-cAMP (2 mM) in the presence of AA decreased the open probability (n = 3). Both effects were largely reversible. These data indicate that the open probability of the channel is regulated bidirectionally by AA and cAMP. Our results suggest that the resting potential of MFBs is at least partially set by a functionally specialized K+ channel, presumably a member of the 2P domain K+ channel family. Furthermore, recordings in the whole-bouton configuration showed that the half-duration of an action potential evoked from an holding potential of -60 mV was 1.4-fold longer and that the action potential amplitude was 16 mV smaller compared to an holding potential of -80 mV (n=14; 34°C). As both factors influence transmitter release, presynaptic resting potassium channels could contribute to the modulation of synaptic strength at the mossy fiber-CA3 synapse. Physiologisches Institut der Universität Freiburg, Hermann-HerderStr. 7, D-79104 Freiburg P 29-5 ANALYSIS OF FOREBRAIN RESTRICTED NR1(N598R) EXPRESSION IN MICE B.J. Schupp, F.N. Single, P.H. Seeburg, G. Köhr High Ca2+ permeability and its control by voltage-dependent Mg2+ block are defining features of NMDA receptors (NMDARs). These features are lost if the principal NR1 subunit carries an asparagine (N) to arginine (R) substitution in a critical channel site at amino acid position 598. NR1(R) expression from a single allele in gene targeted NR1+/R mice is lethal soon after birth (Single et al., Journal of Neuroscience, 2000), precluding analysis of altered synaptic functions later in life. Therefore, we employed a Cre-lox system to activate the NR1(R) allele at a later stage in development and to restrict NR1(R) expression to forebrain regions (Cre4/NR1(R) mice). We chose the CA1 region in acute hippocampal slices to look for functional NR1(R)-containing NMDARs in synapses by minimally stimulating the Schaffer collaterals and recording NMDAR-mediated postsynaptic currents (NMDA EPSCs) at -60 mV in the presence of Mg2+. In the absence of functional AMPA receptors in young Cre4/NR1(R) mice (P3) we recorded using the same input NMDA EPSCs at +40 mV but not at -60 mV, indicating incomplete induction of the NR1(R) subunit at this age. Therefore, we investigated Cre expression in Cre4/NR1(R) mice in fixed hippocampal sections by immunohisto-chemistry and revealed significant mosaic expression at P3 but not at P12. Consistently, in older Cre4/NR1(R) mice (P14P30) we recorded AP5 sensitive NMDA EPSCs at -90 mV in the presence of NBQX and Mg2+. Additonally, decreased Ca2+ permeability of NR1(R)-containing NMDARs was indicated by a more negative reversal potential of NMDA EPSCs in 2 mM Ca 2+. Thus, NR1(R)containing NMDARs are expressed in hippocampal synapses. However,

the functional consequences for synaptic transmission can reliably be studied only in older Cre4/NR1(R) mice. Supported by DFG Ko 1064/ 3-3 Max-Planck-Institute for Medical Research, Department of Molecular Neurobiology, Jahnstr. 29, 69120 Heidelberg, Germany P 29-6 FORMATION OF GLUTAMATERGIC SYNAPSES IN NCADHERIN DEFICIENT MOUSE EMBRYONIC STEM CELLS AFTER NEURONAL DIFFERENTIATION IN VITRO A. Copi, K. Jüngling, R. Moore*, R. Kemler*, K. Gottmann Cell adhesion molecules, such as classical cadherins, have been proposed to play an important role in synaptic function, synaptic plasticity and, in particular, in synapse formation. N-cadherin is a homophilic adhesion molecule that has been shown to be expressed in the perisynaptic region selectively at glutamatergic synapses. However, despite its well characterized synaptic localization, the functional role of N-cadherin at synapses is poorly understood. Homozygous Ncadherin ”knock-out” mice show malformations at neurulation stages and die on day 10 of gestation (1). Thus, synapse formation cannot be analysed in N-cadherin ”knock-out” mice. We have now studied synapse formation in cultured homozygous N-cadherin-deficient ES-cells (2), that were differentiated to neurons in vitro. Heterozygous and homozygous N-cadherin-deficient ES-cells derived from the ES-cell line D3 (3) were treated with retinoic acid during embryoid body (EB) formation in hanging drop-culture. EBs were further cultured on polyornithine-coated culture dishes and were replated after two weeks. EBs showed strong fiber outgrowth and MAP-II positive cells with typical neuronal morphology. These cells generated action potentials upon depolarization. To analyze synapse formation we recorded AMPA receptor- and GABAA receptor-mediated miniature PSCs using the patch-clamp technique. We found a selective decrease in the mean amplitude of AMPA receptor-mediated mPSCs in homozygous as compared to heterozygous N-cadherin deficient ES-cells, while GABAA mPSCs were unaltered. This finding can be explained by a reduced number of postsynaptic AMPA receptors, suggesting a role of Ncadherin in AMPA receptor synaptic delivery. Alternatively, the above result might be caused by a reduced glutamate concentration in the synaptic cleft due to a diminished adhesion. In addition to effects on synaptic currents, we observed a significantly increased amplitude of the spike afterhyperpolarization in N-cadherin-deficient ES cells, suggesting changes in the properties or expression of ion channels determining the shape of the action potential. (1) Radice et al., 1997. Dev.Biol. 181, 64-74. (2) Moore et al., 1999. Int.J.Dev.Biol. 43, 831834. (3) Doetschmann et al., 1985. J.Embryol.Exp.Morphol. 87, 2745. Dept Cell Physiology, Ruhr-University Bochum, D- 44780 Bochum; * Dept Molecular Embryology, MPI Immunbiologie, Freiburg P 29-7 TRANSLAMINAR SYNAPTIC CONNECTIONS FROM LAYER 4 SPINY NEURONES TO LAYER 2/3 INTERNEURONES IN RAT BARREL CORTEX M. Helmstaedter, D. Feldmeyer Sensory processing in neocortex is believed to be strongly controlled by inhibitory pathways. Peripheral sensory stimuli are conveyed via thalamic nuclei to neocortical layer 4 spiny neurones, from where the signal is distributed to supragranular layers. Here, we report first direct evidence for translaminar synaptic input from layer 4 spiny neurones to interneurones located in supragranular layers of rat somatosensory cortex. Dual whole-cell voltage recordings from 30 synaptically coupled pairs of layer 4 spiny neurones and layer 2/3 interneurones were performed in acute brain slices from barrel cortex of P20-P29 Wistar rats. For morphological identification, cells were biocytin-filled during experiment, and only neurones exhibiting clear physiological and morphological interneurone characteristics were included in the analysis (n=20). Single action potentials in layer 4 spiny neurones evoked by brief current injections lead to unitary depolarizing postsynaptic potentials in layer 2/3 interneurones. These PSPs were observed after a mean onset latency of 2.0 ± 0.56 ms with a mean peak amplitude of 1.6 ± 1.3 mV (range 0.46 - 5.1 mV), a 20-80% rise time of 0.77 ± 0.38 ms, and a monoexponential decay time constant of 11 ± 3.7 ms (all

S 302 mean ± s.d.). Transmission reliability varied considerably among cell pairs (failure rates: 19 ± 20 %, range 0 - 65 %). According to their axonal projections and physiological properties, layer 2/3 interneurones with synaptic input from layer 4 spiny neurones represent a heterogeneous population, allowing both for feed-forward and lateral inhibition with respect to functional barrel columns. Our data suggests that at least a subclass of layer 2/3 interneurones is strongly driven by layer 4 principal neurones and thus can represent a cellular correlate for earlystage inhibition in the cortical microcircuitry.

of the NR1 and NR2B subunits. In wild-type rat no spontaneous synaptic activity was observed at the ages investigated; however, in reeler mice, spontaneous events were clearly apparent. Taken together, the axonal projection and input-output relations of CR cells in rat and mice suggest that they are an integral part of an early cortical network that is profoundly altered in the reeler mutation. Supported by the DFG (SFB 488, Lu 593/3-1) and the Max-Planck-Society 1

Max-Planck-Institut für medizinische Forschung, Abteilung Zellphysiologie, Jahnstr. 29, D-69120 Heidelberg, Germany

Max-Planck-Institut für Medizinische Forschung, Abt. Zellphysiologie, Jahnstraße 29, D-69120 Heidelberg; 2Anatomisches Institut der Albert-Ludwigs-Universität Freiburg, Albertstraße 17, D79104 Freiburg, Germany

P 29-8

P 29-10

EFFECTS OF SENSORY DEPRIVATION ON CIRCUITRY OF RAT BARREL CORTEX A. Schierloh, K. Becker, W. Zieglgansberger, H.-U. Dodt To examine the effects of sensory deprivation on intracortical connections in rat barrel cortex we used infrared guided laser stimulation and infrared-darkfield imaging. Patch clamp recording was used to register the postsynaptic potentials of pyramidal layer 2/3 and 5 neurons evoked by laser induced release of caged glutamate. Borders of barrels were identified in unstained thalamocortical slices and single neurons were visualized by infrared videomicroscopy. In untreated rats the input probability from layer 2/3 to layer 5 was periodic with more synaptic connections originating from the „sides“ of barrel columns and less from the „centers“. When the whiskers were clipped during the critical period of intracortical synaptogenesis this periodic input probability disappeared. Layer 2/3 pyramidal neurons receive most of their synaptic input from layer 2/3 and 4. In deprived rats a greater part of lateral input originate in the adjacent barrel column. To visualize the spatial spread of intracortical excitation we used a combination of darkfield techniques and infrared videomicroscopy. This method makes use of the „intrinsic optical signal“ as an indicator of neuronal excitation. An averaged darkfield image was stored and subsequently substracted in real time from the incoming image during extracellular stimulation. The contrast of the resulting difference image was digitally enhanced and colour coded. Intracortical stimulation of the barrel cortex results in different patterns of excitation depending on the localization of the electrode. Stimulation of barrel column center i.e. in layer 3 activates only the stimulated barrel column, whereas intercolumnar stimulation activates the two adjacent columns. In contrast intercolumnar stimulation of layer 5 results in an excitation pattern that is restricted to the barrel borders. The dependence of this neuronal excitation patterns on sensory deprivation is under investigation.

EVIDENCE FOR PROTEIN TYROSINE KINASE/PHOSPHATASE REGULATION OF THE PARALLEL FIBRE SEPSP OF RAT CEREBELLAR PURKINJE NEURONES. M. Canepari, D. Ogden. Brief tetanic stimulation of parallel fibres produces a slow EPSP (sEPSP) or s-EPSC in Purkinje neurones (PN), mediated by type 1 metabotropic glutamate receptors (mGluRl) activating a cation selective channel. The sEPSP is not mimicked by photolytic release of intracellular InsP3 or Ca2+ , or inhibited by the PLC inhibitor U73122 (Canepari et al (2001) J.Physiol. 533, 765-72) indicating that PN sEPSP-mGluRl coupling does not involve PLC activation. The coupling reactions were investigated with inhibitors of kinases or phosphatases. The PN mGluR sEPSC in 20 day rat cerebellar slices was evoked by photolytic release of 10-100 µM L-glutamate in the presence of high concentrations of ionotropic glutamate receptor antagonists. The phosphatase inhibitors Na orthovanadate (1 mM) and BpV(Phen) (10 µM) produced reversible inhibition of the sEPSC to 10% of control. Okadaic acid (10 µM) produced a non-reversible inhibition to 42%. The src tyrosine kinase inhibitor PP1 (10 µM) applied in the patch pipette produced potentiation to 299% of control in 10 minutes. The non-selective kinase inhibitors K252a and staurosporine (10 µM externally, 5 mins) potentiated to 146% and 141 % of control respectively. Inhibitors of PKA KT5720 (10 µM) and PKC (chelerythrine, 20 µM) produced no increase over control. The results suggest that protein phosphorylation by tyrosine kinases inhibits coupling of mGluRl to the sEPSP channel and that tyrosine kinase/phosphatases may have a postsynaptic role in regulating the parallel fibrePurkinje neurone sEPSP.

Max-Planck-Institute of Psychiatry, Kraepelinstr. 2, 80804 Munich, FRG P 29-9 CAJAL-RETZIUS CELLS IN THE DEVELOPING RAT AND REELER MOUSE NEOCORTEX G. Radnikow1, J. Koch2, J. Lübke2. D. Feldmeyer1 Neocortical Cajal-Retzius (CR) cells are early generated, transient neurones in the future layer 1. They have been suggested to play an essential role in neuronal migration and corticogenesis; however, their role in the early neocortical microcircuit is still a matter of debate. Here, investigated the dendritic and axonal morphology and input and output synapses of CR cells in rats and reeler mice. In contrast to rat and wild-type mice these mutants show severe alteration in cortical layer formation. CR cells (P5-P15) were investigated using patchclamp recordings combined with intracellular biocytin-filling in acute brain slices. Subsequently, they were processed for light- and electron microscopy and morphologically reconstructed. CR cells in rat are characterised by a long stem dendrite running parallel to the pial surface with several short dendritic processes that were covered with spine-like appendages. The main axon forms a dense horizontally oriented plexus confined to layer 1 spanning ~2 mm of cortical surface. The axon established synaptic contacts preferentially with dendritic spines or shafts of excitatory neurones. In contrast, CR cells in reeler mice project deep into the cortical layers towards the white matter. CR cells receive dense synaptic GABAergic and non-GABAergic input, which is mediated through GABAA and NMDA receptors, respectively, but not AMPA receptors. NMDA receptors are composed

National Institute for Medical Research, Mill Hill, London NW7 1AA.

P 29-11 IMPAIRED SPINE CALCIUM SIGNALING IN PURKINJE CELLSPECIFIC CALBINDING-DEFICIENT MICE J. Hartmann, C.R. Rose, M. Noll-Hussong, M. Meyer, A. Konnerth Calbindin-D28k (calbindin) is a calcium-binding protein that is abundant in the central nervous system. The cell-specific role of calbindin is largely unclear. We used Purkinje cell-specific calbindin nullmutant mice to study the role of calbindin in synaptic calcium signaling and plasticity in these cells. By combining whole-cell patch-clamp recordings with two-photon imaging or conventional confocal microscopy we measured synaptic calcium transients in spines and dendrites of Purkinje cells in cerebellar slices of control and mutant mice. AMPA-receptor-mediated fast synaptic calcium transients produced by single-shock activation of parallel fibers were found to be strongly altered in spines and adjacent dendrites of calbindin-deficient Purkinje cells. In mutant mice, amplitudes of fast calcium signals were enhanced up to 300 % whereas decay constants were reduced by 70 % as compared to control animals. By contrast, the delayed calcium signals, which are due to metabotropic glutamate receptor-mediated calcium release from intracellular calcium stores in Purkinje cell dendrites, were not modified by the absence of calbindin neither regarding their amplitude nor their decay rates. Surprisingly, the induction and maintenance of synaptic long-term depression (LTD) was not altered in calbindin-deficient Purkinje cells. In both control and mutant mice, LTD induction produces a synaptic depression of parallel fiber responses by 40 % (n = 4 for each genotype).Thus, our results show that calbindin regulates fast calcium transients but does apparently not influence delayed synaptic calcium signaling in dendrites

S 303 and spines of Purkinje cells. Furthermore, calbindin-deficiency and the resulting altered amplitude and kinetics of fast synaptic calcium signals do not affect the induction and expression of cerebellar LTD. Institut für Physiologie, LMU München, Biedersteiner Str. 29, 80802 München P 29-12 RELATIONSHIP BETWEEN pH AND CALCIUM MICRODOMAINS IN RAT CEREBELLAR PURKINJE NEURONES D. Willoughby, C.J. Schwiening Electrical stimulation of neurones results in an intracellular acidification. There is evidence that the counter-transport of H + by the plasmamembrane calcium ATPase is responsible for such acid shifts (1,2). The size and kinetics of depolarization-induced intracellular pH (pHi) shifts can vary throughout the cell, giving rise to pH microdomains (3,4). In cerebellar Purkinje neurones electrical activity induces large dendritic acid shifts with little change of pH i in the cell soma (3). Similar differences in activity-induced cytosolic calcium increases have also been reported (5). It is possible that the relative changes in pHi and cytosolic calcium are linked. Here we report that we can simultaneously monitor pHi and calcium in cerebellar Purkinje neurones using a mixture of fluorescent indicators. Whole-cell patch-clamp recordings were made from Purkinje neurones in 200µm thick sagittal cerebellar brain slices from 11-21 day old Wistar rats killed by cervical dislocation. Slices were superfused with HEPES or 5% CO2 / 25mM HCO3- buffered saline at ~24°C. Neurones were loaded with the pH indicator 8hydroxypyrene-1,3,6-trisulphonic acid (HPTS, 125µM) and calcium indicator Fura Red (62.5µM) via the patch pipette (3-5MΩ). Both dyes were excited at 488nm and emitted fluorescence collected between 505-550nm for HPTS or >560nm for Fura Red using a Zeiss LSM 510 confocal microscope. Neurones were held at –70mV and periodically depolarized to 0mV for 1s. Depolarization was accompanied by a fall in the HPTS and Fura Red fluorescence indicating a simultaneous fall in pH i and rise in calcium. The intracellular acid shifts and increases in calcium levels were greatest in the dendrites, whilst changes in the soma were smaller and slower (n=6). These results suggest that HPTS and Fura Red can be used to monitor simultaneous changes in pH and calcium. Furthermore, there is a clear relationship between pH and calcium microdomains in cerebellar Purkinje neurones. The larger, faster acid shifts may result from rapid extrusion of the large calcium loads seen in dendritic regions. We are grateful to the MRC for funding. 1. Schwiening CJ, Kennedy HJ & Thomas RC (1993) Proc. R. Soc. Lond. B. 253,285-289; 2. Trapp S, Luckermann M, Kaila K & Ballanyi K (1996) Neuroreport 7,2000-2004; 3. Willoughby D & Schwiening CJ (2001) J. Physiol. 536.P,122P; 4. Schwiening CJ & Willoughby D (2002) J. Physiol. (in press); 5. Tank DW, Sugimori M, Connor JA & Llinás RR (1988) Science 242,773-777. Department of Physiology, University of Cambridge, Downing Street, Cambridge, UK, CB2 3EG

P 30-1 VESICLE CYCLING IN A SYNAPTOSOMAL PREPARATION STUDIED BY TOTAL INTERNAL REFLECTION FLUORESCENCE MICROSCOPY (TIRFM) D. Loerke, A. Bundy, J. Klingauf Total internal reflection fluorescence microscopy (TIRFM), also termed evanescent wave (EW) microscopy, has been used to visualize single granules in neuroendocrine cells and, more recently, synaptic vesicles in giant non-mammalian terminals. Here we report on a rat synaptosomal preparation, which allows us to monitor vesicle cycling in isolated nerve terminals with both conventional epifluorescence and EW microscopy. Synaptosomes were prepared from rat hippocampus or cortex; they adhere to glass coverslips, and electron microscopy reveals that a sufficient fraction faces the glass with the active zone. When stained with the styryl dye FM2-10 and illuminated to a depth of ~100 nm by the EW, exocytic events could be detected as a decrease in epifluorescence (reflecting dye departitioning out of the membrane), which is preceded by a transient increase in the TIRF signal (corresponding to lateral diffusion of dye into the plasma membrane on the coverslip and subsequent departitioning). Synaptosomes can be partially destained with hypertonic sucrose solution, which specifically

depletes the docked or readily-releasable pool of vesicles. This constitutes ~18% of the total cycling pool compared to ~15% in synapses in culture. Thus, relative contributions of pool sizes appear to be conserved in synaptosomes, making it a useful preparation for visualizing vesicle transport, docking, and exocytosis. Max Planck Institute for Biophysical Chemistry, Microscopy of Synaptic Transmission, Dept. Membrane Biophysics, Am Fassberg 11, 37077 Göttingen, Germany P 30-2 MONITORING CLATHRIN-MEDIATED ENDOCYTOSIS IN HIPPOCAMPAL SYNAPSES V. J. Mueller, M. Wienisch, R. B. Nehring, J. Klingauf While several kinetic components of membrane retrieval have been observed at synapses, it has remained controversial whether all of these involve clathrin. In order to measure clathrin-mediated endocytosis as directly as possible we have cloned the clathrin light chain out of a rat cDNA library and fused it with EGFP. When this was overexpressed in neuronal cultures from hippocampus using the SemlikiForest virus system, we could detect highly mobile sub-resolution particles in dendrites, axons, as well as in glial cells. No such particles were visible, however, when expressing EGFP alone. Superfusion with high potassium leads to a transient increase of the total fluorescence in synaptic boutons only, often accompanied by a slight transient fluorescence decrease in the adjacent axonal segments. The bouton fluorescence rises after the stimulus onset with a lag phase of a few seconds and peaks shortly after stimulus cessation, reaching amplitudes of 15-30% above resting value, depending on stimulus duration and strength. Subsequently, the signal decays exponentially with a τ of ~20 sec’s. Presumably, these transients reflect the fraction of membranebound clathrin in the boutons at any time point, provided that clathrinGFP diffusion from the axon into the bouton and vice versa is not rate-limiting. Thus, the signal appears to be a direct measure of coat assembly and disassembly, respectively. Max Planck Institute for Biophysical Chemistry, Microscopy of Synaptic Transmission, Dept. Membrane Biophysics, Am Fassberg 11, 37077 Göttingen, Germany P 30-3 A SLOW ENDOCYTOSIS PATHWAY IN HIPPOCAMPAL SYNAPSES M. Wienisch, J. Klingauf Recently, a slow mode of endocytosis has been uncovered in the frog neuromuscular junction by making use of the different departitioning kinetics of the styryl dyes FM2-10 and FM1-43 (Richards et al. 2000, Neuron 27:551). To test whether a similar pathway also exists in hippocampal boutons, we compared total recycling pool (TRP) and readily-releasable pool (RRP) sizes using weak (150 action potentials; APs) or strong stimulation (45mM K+) for dye loading. We find that losses of absolute fluorescence for brief hypertonic sucrose pulses, i.e. RRP sizes, are similar for both protocols and dyes, while the relative contributions (RRP/TRP) are not (% ± SEM; FM2-10 and K+ stim.:14.0 ± 0.6; FM1-43 and K+ stim.: 7.1 ± 0.4; FM2-10 and AP stim.:13.0 ± 0.7; FM1-43 and AP stim.:14.7 ± 0.5). Comparison of the absolute TRP values for both protocols by normalisation with respect to the RRPs reveals a significantly larger TRP size for FM1-43 than for FM2-10 after high K + loading. This points to an additional slow endocytic pathway after strong stimulation that retrieves ~ 30 % of the exocytosed membrane, probably via large infoldings from which FM2-10, but not FM1-43 can be rinsed, and from which new vesicles are formed eventually. However, when leaving FM2-10 for an additional 15 min after K+ stimulation, the absolute TRP size rises dramatically by 106 % with little change in the RRP. This suggests that the putative infoldings have been closed and pinched off during this time period, thereby trapping also the more rapidly departitioning dye FM2-10. Varying systematically the washout time, we find that the endocytic compartments are pinched off from the plasma membrane with a half time of 5 - 10 min. Max-Planck Institute for Biophysical Chemistry, Microscopy of Synaptic Transmission, Dept. Membrane Biophysics, Am Fassberg 11, 37077 Göttingen, Germany

S 304 P 30-4

P 30-6

SUBCELLULAR TARGETING OF SAP97 IN MAMMALIAN CNS NEURONS DOES NOT RELY ON THE SH3 AND GK DOMAINS N. Klöcker, R.C. Bunn, E. Schnell, G. Caruana, A. Bernstein, R.A. Nicoll, D.S. Bredt Membrane-associated guanylate kinases (MAGUKs) are multidomain proteins which play a key role in organizing synapses. They provide a scaffold for protein-protein interactions and are involved in subcellular targeting of their binding partners. Recently, a murine truncation mutant of the MAGUK synapse-associated protein 97 (SAP97) was isolated which lacks the SH3 and GK domain. Mice homozygous for this gene insertion show severe craniofacial and palatal dysmorphogenesis. Here we analyzed this mouse mutant to ask whether the SH3 and GK domains are involved in subcellular targeting of SAP97 and potentially interacting partners in CNS neurons. In cultured cortical neurons, wildtype SAP97 was predominantly expressed in perinuclear regions, in a pattern suggesting association with the endoplasmic reticulum. Weaker immunoreactivity was found in neurites colocalizing with both dendritic and axonal markers. Subcellular localization of truncated SAP97 was indistinguishable from wildtype protein. As SAP97 has been involved in intracellular trafficking of the glutamate receptor GluR1 and in interacting with the guanylate kinase anchoring protein GKAP, we further studied the subcellular localization of GluR1 and GKAP. Both in cortical cultures of wildtype and homozygous mutant mice, GluR1 and GKAP clustered at synapses, the latter colocalizing with the MAGUK protein PSD95. In summary, our data demonstrate that the SH3 and GK domain of SAP97 are neither critically involved in subcellular targeting of the protein in mammalian CNS neurons nor does their truncation cause any obvious neuronal phenotype as opposed to the epithelial/mesenchymal phenotype in these mice.

THE LATERAL SPREAD OF EPILEPTIFORM DISCHARGES IN RAT NEOCORTICAL SLICES: EFFECT OF LOCAL PHENCYCLIDINE (PCP) APPLICATION A. Gorji 1, D. Scheller 2, E-J. Speckmann 1,3 Testing the effect of PCP as a classical psychomimetic drug on cortical integration, the substance was applied locally in rat neocortical slices after epileptiform discharges had been induced by 0-Mg2+ artificial cerebrospinal fluid superfusion. The in vitro brain slice technique was used to examine PCP effects on the lateral propagation of epileptiform field potentials (EFP) across adjacent areas of rat frontal neocortex. Simultaneous recordings of EFP were obtained from four microelectrodes placed 2-3 mm apart across coronal slices in the third layer. PCP (1-500 microM) applied locally between recording sites, blocked rapid propagation across treated areas and resulted in the emergence of spatially seperate, independent pacemakers. The same asynchronized pattern of EFP was observed after local application of the N-methyl-D-aspartate (NMDA)-receptor antagonist, DL-2-amino5-phosphono-valeric acid (200 microM. Local administration of haloperidol (100 nanoM) as well as of NMDA (10 microM before PCP application reversibly prevented appearance of multiple pacemakers. Focal application of dopamine (2.5-10 miliM) produced an abnormal pattern of lateral conduction of EFP in 50% of tested slices. Pacemaker failure as an indicator of functional impairment of cortical integration is the proposed mechanism for developing of schizophrenia-like psychosis associated with epilepsy.

Depts. of Physiology and of Cellular and Molecular Pharmacology, University of California, San Francisco, CA 94143, USA. Supported by the Deutsche Forschungsgemeinschaft SFB430 and Kl 1168/2-1 to N.K.

P 30-5 SYNAPTOGENESIS BETWEEN SPINAL CORD AND DORSAL ROOT GANGLION NEURONS IN CO-CULTURE S.A. Fedulova, Y.V. Medvedeva, N.S. Veselovsky (Introduced by D.A. Brown) Experiments were performed on rat spinal cord (SC) and dorsal root ganglion (DRG) neurons in dissociated co-cultures. Neurons form synapses after 12-20 days, and GABAergic transmission was isolated using 20 µM DNQX, 20 µM AP5, 1 µM atropine and 20 µM tubocurarine. Visually-identified SC neurons were used as presynaptic and DRG neurons as postsynaptic cells for recording evoked currents. Evoked inhibitory postsynaptic currents (eIPSCs) were recorded in DRG neurons using whole-cell techniques after extracellular local stimulation of the presynaptic SC neuron by short (0.4-0.6 ms) current pulses. Parameters of spontaneous IPSC (sIPSC) were investigated at different times of co-culturing. The s/eIPSP current-voltage relationship was linear and reversed at –15.4±2.4 mV (n=7) close to the Nernst potential for Cl- under the recording conditions. Bicuculline (20 µM) completely and reversibly blocked spontaneous and eIPSCs. Latency, the time from stimulus start to eIPSC onset, was 4.7±0.29 ms (n=8). Decay time courses could be fit by single exponential functions, and were for sIPSC 21.8±5.4 ms (n=13) and eIPSC 17.4 ± 6.5 ms (n=15). Times to peak were also similar for the three classes of events. Latency and decay of eIPSC, as well as time to peak of sIPSC did not change in the range of standard errors during co-culture. The mean amplitudes and times to peak of eIPSC showed significant enhancement during development in co-culture. During development in culture sIPSC frequency increased. We conclude that these postsynaptic currents recorded in DRG neurons represent the opening of GABA-activated Cl - channels. The results appear to be the first direct demonstration that spinal cord neurons can form functional inhibitory synapses onto DRG neurons. The increase of functional synaptic contacts between SC and DRG neurons during co-culture might simulate real physiological process of inhibitory synapse formation onto DRG neurons in vivo. Bogomoletz Institute of Physiology, Ukraine Nationale Academy of Scinece, 01024 Kiev, Ukraine

1. Institut für Physiologie, Universität Münster, Robert-Koch-Strasse 27a, 48149 Münster, Germany; 2. Janssen-Research Foundation, Raiffeisenstrasse 8, 41470 Neuss, Germany; 3. Institut für Experimentelle Epilepsieforschung, Universität Münster, Hüfferstrasse 68, 48149 Münster, Germany P 30-7 SIGNAL PROPAGATION IN THE CAUDAL MEDULLA OF MOUSE: FLUOROACETATE PROLONGS DEPOLARIZATION AS REVEALED BY RECORDINGS WITH VOLTAGE SENSITIVE DYES S. Hülsmann, D. W. Richter, H. Straub, E. –J. Speckmann Fluoroacetate is known to block metabolism and passive potassium conductances in glial cells. We aimed to investigate the effect of such functional elimination of astrocytes on the signal propagation in the brainstem. Transverse 400 µm slices were prepared from the caudal medulla of neonatal mice (P3-8) containing the ventral respiratory group (VRG) and the hypoglossal nucleus (XII). Propagation of excitation was investigated fluorimetrically using the voltage sensitive dye RH 795. Stimulus-evoked depolarizations propagated from the VRG and peaked after 7.2 ± 0.6 ms (n=6) in the hypoglossal nucleus. During peak EPSPs the whole hypoglossal nucleus was activated. Depolarizations were measured as changes relative to resting light intensity. Mean changes of the fluorescence signal in the hypoglossal nucleus were 0.0021 ± 0.0005 (∆F/F). With application of 1 mM fluoroacetate the maximal amplitude of the depolarization was reduced by 69.1 ± 9.8 % (n=6, p=0.034). Additionally, repolarization was diminished after application of fluoroacetate. Propagation speed and spatial distribution of the voltage signal, however, was not effected by fluoroacetate. While it is known that fluoroacetate generally reduces synaptic transmission we assume that slower repolarization is due to slower decay of EPSPs rather than caused by astrocytic depolarization. Supported by: DFG Abt. Neuro- und Sinnesphysiologie, Zentrum Physiologie und Pathophysiologie Georg-August-Universität, Humboldtallee 23, 37073 Göttingen, Germany P 30-8 SUPPRESSION OF EPILEPTIFORM ACTIVITY BY GABAB RECEPTORS Y. Yanovsky, U. Misgeld We investigated the role of GABAB receptor-mediated mechanisms in suppressing epileptiform activity in the CA3 area of hippocampus disinhibited by GABAA receptor blockade. To this end, we compared inhibitory effects of GABA B receptor activation in wild type and

S 305 weaver mice. In the latter, GABAB receptors cannot activate potassium currents. The GABAB receptor agonist R-baclofen reduced the frequency of epileptiform discharges in disinhibited wild type and weaver CA3 neurons, but much higher concentrations of the agonist were required in weaver (5 µM) as compared to wild type mice (0.1 µM). Concentration/response curves for the reduction of EPSCs evoked by stimulation of CA3 axon collaterals were identical in wild type and weaver mice. This suggests that inhibition of bursting by the low Rbaclofen concentration is associated with only a slight reduction of EPSCs, while EPSCs are strongly reduced by the high agonist concentration. In both, +/+ and wv/wv CA3 neurons endogenous GABA took only a subtle influence through GABAB receptors on the probability of occurrence of epileptiform activity. The impairment of GABAB receptors on weaver GABAergic neurons supported this action possibly by increasing endogenous GABA levels. Our data indicate that activation of GABAB receptors suppresses epileptiform activity more efficiently through activation of K+ conductance than through inhibition of Ca2+ conductance. This requires, however, that GABAB receptor activation is properly timed. Supported by the Deutsche Forschungsgemeinschaft. Institut für Physiologie und Pathophysiologie, Universität Heidelberg, Im Neuenheimer Feld 326, D-69120 Heidelberg, Germany P 31-1 ONLY MINOR CONTRIBUTION FROM TTX-RESISTANT GROUP IV MUSCLE AFFERENTS TO SPINAL MOTOR CONTROL IN THE CAT E.D. Schomburg, H. Steffens A large fraction of muscle afferents originates from free nerve endings. These afferents belong to group III (23% of myelinated fibres in the lateral gastrocnemius-soleus (GS) nerve) or group IV nerve fibres. The latter distinctly outnumber the group III afferents (1300 group IV versus 110 group III in the lateral GS nerve). Despite the great number of group IV muscle afferents, their specific spinal motor actions, compared to those evoked by group III afferents, have not yet been investigated. Therefore, we now separately analyzed the contribution of group III and group IV muscle afferents (chemically activated by injection of KCl or bradykinin into the muscle artery of GS) to spinal motor reflex pathways. In high spinal cats, after corresponding controls group I-III fibres were completely blocked by TTX, leaving group IV-fibre conduction intact. Thus, effects from intra-arterial KCl or bradykinin injection persisting after TTX were attributed to TTX-resistant group IV fibres, while the contribution of group III fibres was approximately defined by the difference between those effects and the control effects before TTX. The chemical activation of group III and IV muscle afferents induced distinct reflex facilitation of the flexor posterior biceps semitendinosus and inhibition of the extensor quadriceps. After the block of all myelinated fibres by TTX the same stimuli induced only minor reflex effects mediated by the persistently conducting TTX resistant group IV afferents. It is concluded that the main spinal motor reflex effects from chemically activated muscular free nerve endings are evoked by group III afferents while the majority of group IV muscle afferents, which respond preferentially with a higher threshold to mechanical stimuli, are probably of greater importance for a reflex adjustment of circulation and respiration during physical activity, than for reflex motor control. The technique of blocking all myelinated nerve fibres by application of TTX at the peripheral nerve proved to be a reliable method for a long lasting selective investigation of group IVfibre actions. Institute of Physiology, University of Göttingen, Humboldtallee 23, D-37073 Göttingen, Germany

common targets in the spinal cord. Conduction delays were estimated from known peripheral delays and published central delays to group II interneurones in the intermediate zone (eg Edgely & Jankowska, 1987, J. Physiol., 389, 647-674). These interneurones are premotor to αmotoneurones. In 2 cases an estimated 20-25% of group II EPSPs would summate with Ia EPSPs. It was of interest to discover whether this type of γ-activity could provide group II afferent spikes with a lead over Ia spikes also during muscle stretches. Group Ia and II discharges from single spindles of tenuissimus muscles were recorded in dorsal root filaments in decerebrated cats with intact ventral roots. Anaesthesia before decerebration was with pentobarbitone sodium 45mg kg-1 intraperitoneally, maintained i.v.; all animals were killed with an overdose of anaesthetic. In one case, when natural γ-activity coupled the afferents as described above, random stretches were applied immediately to the muscle close to the spindle (mean stretch amplitude 0.55mm; mean frequency 29Hz). Although, in response to each stretch there were fewer group II spikes than Ia spikes, the initial spike was a group II rather than a Ia spike in the majority of stretches. The condition was stable over the 180s of recording. The estimated summation of group II/Ia EPSPs was appreciably greater than at constant length. The implication of this finding is that, by co-operative action, static γ-motoneurones may change the gain of the segmental stretch reflex. Institute of Biomedical and Life Sciences, West Medical Building, University of Glasgow, Glasgow G12 8QQ, UK P 31-3 THE EFFECTS OF PROLONGED MUSCLE VIBRATION ON MAXIMAL VOLUNTARY KNEE EXTENSION PERFORMANCE IN THE IPSILATERAL AND CONTRALATERAL LIMB S.W. Jackson* and D.L. Turner It was hypothesised that prolonged vibration of the right Rectus Femoris muscle will cause attenuation in the muscular activation of: a) the vibrated muscle; b) the homologous contralateral muscle; c) a synergist muscle in the vibrated limb; d) the homologous contralateral synergist muscle. Subsequently, it was hypothesised that maximal performance will be reduced in the vibrated limb and in the contralateral limb. Measurements involved the performance of maximal voluntary isometric knee extension contractions (MVC) with both legs, and were taken before (PRE) and after (POST) vibration treatment. Subjects performed an MVC, as rapidly as possible, with the right leg, which was followed by an MVC with the left leg, 30 seconds later. This was repeated three times. Immediately following the final MVC, vibration was applied for 30 minutes to the right Rectus Femoris, during which the subject remained relaxed and passive. 30 Hz vibration induced a significantly greater reduction in maximal force than Control measures (p = 0.018) and 120 Hz vibration (p = 0.041), in the right (vibrated) leg. Also, maximal rate of force development was significantly reduced in the right leg (p = 0.005) and the left (contralateral) leg (p = 0.005) following 30 Hz vibration. Muscle activation (iEMG) was significantly reduced only in the vibrated muscle following 30 Hz vibration (p = 0.015). These results support the work of Bongiovanni et al. (1990) and Kouzaki et al. (2000), who reported a reduction in maximal performance in the vibrated limb muscle. It is also concluded that muscle vibration may act through spinal reflex pathways to influence the motoneuron pool that modulates the activity of contralateral muscles. Exercise Neuroscience Research Group, South Bank University, London, SE1 0AA, UK. P 31-4

P 31-2 FUSIMOTOR-GENERATED COUPLING BETWEEN GROUP Ia AND II AFFERENTS DURING MUSCLE STRETCH M. H. Gladden, H. Matsuzaki Recently it was shown that, by co-operative action at low (<20Hz) frequencies, static γ-motoneurones can achieve stable coupling of spikes in group Ia and II afferents from single muscle spindles. Surprisingly, the group II afferent spikes left the muscle spindles with a lead of several milliseconds ahead of group Ia spikes (Gladden & Matsuzaki, 2001, J. Physiol., 536.P, 51-52P). Thus the slower-conducting group II spikes and faster-conducting Ia spikes might arrive simultaneously at

DEVELOPMENT OF DIFFERENT PROJECTION CLASSES OF INTERSEGMENTAL COMMISSURAL INTERNEURONS IN THE LUMBAR SPINAL CORD OF RAT AND MOUSE U.V. Nissen, J.C. Glover Spinal interneurons are the key players in spinal cord function, yet surprisingly little is known about their anatomical and functional organization. We have used differential retrograde tracing to characterize the distribution and axonal trajectories of intersegmental commissural interneurons (CINs) in the lumbar spinal cord of embryonic rats and mice. Different fluorescent dextran-amines were applied to the ventral funiculus rostral versus caudal to specific lumbar segments.

S 306 After appropriate incubation the preparations were fixed, sectioned, and CIN positions were photographed and plotted. We distinguished between intersegmental CINs with ascending axons (aCINs), with descending axons (dCINs), and with both ascending and descending axons (adCINs). aCINs and dCINs were the predominant classes. aCINs (60%) outnumbered dCINs (35%) on average, while adCINs represented only a minor fraction. The bulk of CINs were located in the ventromedial region throughout development in both species, and both aCINs, dCINs and adCINs could be distinguished from as early as E12.5 in mouse. Our results demonstrate the presence of CINs of different projection classes in a regionally specific pattern from early stages. Supported by Jahres Fund, Danish MRC. Department of Physiology, University of Oslo, Sognsvannsveien 9, P.O.Box 1103, Blindern, N - 0317 Oslo, Norway P 31-5 TOPOGRAPHICAL EMG PATTERNS OF THE HUMAN M. QUADRICEPS FEMORALIS INFLUENCED BY CONTRACTION INTENSITY AND KNEE JOINT ANGLES N.P. Schumanna*, H. Hegerb, V. Wankb, R. Grassmea, H.Ch. Schollea Although the myoelectrical activation patterns of the M. quadriceps fem. was extensively examined in time-course there are only few hints to the topographical aspect of the EMG distribution patterns. But in medicine and sports science it is important to know whether and how the spatial activation patterns depend on the functional conditions of extensor-contraction. Therefore, in 13 male students of sports sciences (23-35 years old) a 32-channel surface EMG of the left M. quadriceps fem. was monopolarly recorded during constant-force contractions (Biovision-EMG-System, Wehrheim; 10-700 Hz, A/D converted at a rate of 3000 samples/s at 12 bit, resolution 2,44 µV/bit). Electrodes (16 mm, ARBO) were fixed on the skin in a rectangular 8x4 scheme (8 along and 4 across the thigh). The interelectrode distance was fitted on thigh length (centre-to-centre: 50-55 mm). For muscle power estimation of M. quadriceps fem. isometric torque of the left knee joint was registered. The subjects sat on leg extensor test apparatus. Hip (60°, 0°) and knee joint angle (90°, 60°, 45°, 30°, 10°) were adjusted. In each of these angle position at first the test persons were asked to perform a maximum voluntary isometric knee extension (MVC). A break followed after each task. The torque threshold for lower muscle activation intensity (20 %, 40 %, 70 %) were determined relating to MVC torque (visual feed back via bar display). As expected, EMG amplitudes increased significantly with increasing muscle force. The spectral EMG interference-maps show individually different topographical activation patterns. The relation between M. vastus medialis, rectus femoris and vastus medialis changes between subjects. Within the muscle heads there are regional activation-shifts influenced by the contraction-force and the knee joint angle. Supported by DFG, Innovationskolleg “Bewegungssysteme”, FSU Jena a

Motor Research Group, Institute of Pathophysiology, FriedrichSchiller-University Jena, Erfurter Str. 35, D-07740 Jena and bInstitute of Sports Science, FSU Jena, Seidelstr. 20, D-07749 Jena

P 31-6 MODIFICATION OF A FUNCTIONAL MOTOR TASK BY NONCONSCIOUSLY PERCEIVED SENSORY STIMULI S. Fellows1, R. Tabaza1, M. Heumann2, W. Klotz2, O. Neumann2, M. Schwarz3, J. Noth1, R. Topper1 Information from the visual cortex can reach the motor cortex over two pathways (Goodale and Milner, 1992). A direct ventral pathway is held to be responsible for object recognition and conscious processing of object properties, while an indirect, dorsal pathway allows automatic integration of visual information into a motor task. Lesions of the ventral system lead to a failure to consciously recognise or describe objects, but spare the ability to integrate visual information into ongoing motor tasks such as grasping an object (Goodale et al., 1991). Such lesions are rare, but an experimental dichotomy can be created between conscious object perception and object manipulation in intact man. This uses the phenomenon of metacontrast masking (Klotz and Neumann, 1999): a visual stimulus followed after a short interval (45 ms) by a similar but larger stimulus is not consciously perceived. If, however, the smaller stimulus provides information in a choice reac-

tion task, reaction times are shortened although the stimulus is completely masked (Taylor and McCloskey, 1996). It remains unclear if masked visual information can qualitatively modify motor performance. We have combined metacontrast masking with performance of a natural motor task, the lifting of an object in a precision grip. This requires close co-ordination of grip force developed by the fingers and the muscles of the wrist, with forces scaled to match object weight. We provided normal subjects (n=13) with visual cues which indicated the load (variable) of an otherwise identical object. These cues were visible or masked and could be correct or incorrect with regard to object load. Grip force is normally closely matched to object load. In the absence of information about load, the grip force used in a preceding lift is adopted: that is, the subject assumes that load will remain unchanged. If the object load has changed from heavy to light, this leads to a characteristic overshoot in the grip force before sensory feedback allows correction to an appropriate level. We found that when a subject received a visible cue that object weight would be heavy, but was, in reality, light, a similar overshoot occurred. If a cue warned of a change from a heavy to a light load, this overshoot was abolished and grip force parameters were optimised for the light load. This was the case for both the visible and the masked cue. Thus we have shown, for the first time, that non-consciously perceived visual information can be utilised to adapt a functional motor task to actual conditions. Given the highly abstract nature of our weight cues, this adaptation clearly involved cognitive processes, indicating, in contradiction of current theory, that either the dorsal visual system is capable of object recognition and parameter selection, or that the ventral system can access sensorimotor processing at a subconscious level. Goodale, M.A. & Milner, A.D. (1992) Trends Neurosci 15, 20-25. Goodale, M.A., Milner, A.D., Jakobson, L.S. & Carey, D.P. (1991) Nature 349, 154156. Klotz, W. & Neumann, O. (1999) J Exp Psychol 25, 976-972. Taylor, J.L. & McCloskey, D.I. (1996) Exp Brain Res no, 62-66 1

Neurologische Klinik, Universitätsklinikum der RWTH Aachen, 52057 Aachen, 2Abteilung fur Psychologie, Universität Bielefeld, 33501 Bielefeld, 3Neurologische Klinik, Städtische Kliniken Dortmund, 44137 Dortmund, Germany P 31-7 MAXIMUM FINGER TAPPING FREQUENCY IS INDEPENDENT OF FINGER DISPLACEMENT AMPLITUDE G. Sjøgaard, C.T. Hye-Knudsen, D. Ekner, A.K. Blangsted, H.B. Olsen, K. Søgaard The aim of our study was to measure the time interval of double clicking during computer mouse work and to compare this value to the maximum frequency for mouse clicking in order to identify the fraction of maximum performance. Maximum frequencies of finger tapping have been reported in which significant finger movements occur. During computer mouse clicking, however, the displacement of the finger is minimal. It was hypothesized that maximum clicking frequency would be significantly larger than normal tapping frequency on the table and therefore the fraction of maximum performance during double clicking would be lower when expressed relative to maximum mouse clicking frequency. Twelve subjects (5 females and 7 males) performed ten double clicks at their own pace. Further, they performed table tapping and mouse clicking, respectively, for 10 s at maximum frequency. All sequences were video recorded for calculation of movement frequency and amplitude. EMG from surface electrodes was recorded from the antigravity muscle extensor digitorum communis. The maximum table tapping frequency, mean (SD), was 5.5 (0.56) Hz and the movement amplitude was 15.0 (7.7) mm. The maximum mouse clicking frequency was similarly 5.5 (0.84) Hz but with a significantly smaller movement amplitude of 0.89 (0.15) mm. When performing mouse double clicking the interval between the two clicks was 229 (0.028) ms resulting in the overall mean of the instantaneous frequencies to be 4.4 (0.50) Hz, and the amplitude to be 1.39 (1.3) mm. The fraction of maximum frequency performance was thus 80 %. The relative EMG activity was similar during the maximum tapping: and clicking tasks: 7.8 (2.9) %EMGmax and 7.0 (3.2) %EMGmax, respectively, while during the double clicking the activity was 4.2 (1.6) %EMGmax. The low % of EMGmax activity during the maximum tapping and clicking tasks indicates that only a limited fraction of MU is activated, which in turn may be activated at high rates. The ten double clicks were spaced over 19 s (range 7-40 s) that may account for the somewhat lower relative EMG load of around 50 % compared to the 80 % maximum frequency performance. In conclusion, the movement amplitude was more than 10 times larger

S 307 during table tapping than mouse clicking, but maximum frequency and relative EMG activity were similar. This supports that motor control and not muscle mechanics limits the maximum repetition frequency of finger activation. The instantaneous frequency during double clicking is closed to the maximum performable. National Institute of Occupational Health, Department of Physiology, Lersø Parkallé 105, DK 2100 Copenhagen Ø, Denmark P 32-1 THE EFFECT OF PERIPHERAL NERVE LESIONS ON 8 - 10Hz TREMOR DURING FINGER MOVEMENTS J.J.A. Scott, F.A. Proudlock Slow movements of the digits are characterised by 8-10 Hz discontinuities (2) which appear to be centrally generated since coherent oscillations are also evident in the eye when performing a related task (1). We have examined the nature of these oscillations in a group of 12 subjects subsequent to reinnervation of intrinsic hand muscles following a complete nerve lesion.The frequency content of abduction and adduction movements was recorded, using an optical position laser transducer, in 12 index fingers and 9 little fingers subsequent to a complete ulnar transection. Concurrently, surface electromyograms (EMG) were also recorded from first dorsal interosseus and abductor digiti minimi brevis muscles for index and little fingers, respectively. The maximum voluntary contraction (MVC) was measured using an isometric strain gauge. The subjects attempted to perform a 7.5° amplitude, 5°s-1 velocity abduction – adduction movement task and a position holding task with the digit abducted at 10° from the resting position. Significant peaks in the power spectra of acceleration and rectified integrated EMG traces were identified. Although background noise was higher in the spectral curves of acceleration traces from movements performed by nerve lesioned subjects in comparison to controls, 8-10Hz oscillations were evident in 87.5% of index fingers and 75% of little fingers during slow movements and 91.7% of index fingers and 94.4% of little fingers during a position holding task. The occurrence and frequency of the oscillations was not related to the MVC, even in those subjects too weak to perform active movements. These data demonstrate the robust nature of these 8-10Hz oscillations even following the dramatic changes in peripheral connectivity of muscle and nerve resulting from nerve transection and reinnervation. (1) McAuley J & Marsden CD (2000) Brain 123: 1545-1567 (2) Wessberg J & Vallbo AB & (1996) J. Physiol 493: 225-233 Dept of Pre-Clinical Sciences, University of Leicester, PO Box 138, Leicester, LE1 9HN UK

physiological and biochemical as well as topographically oriented characterisation of the dynamics of muscle fatigue during the load situation. During fatiguing the muscles the distinct decrease of SEMG median frequency correspond with a decrease of phosphocreatin and pH and an increase of inorganic phosphor. During the recovery phase the initial values were reached again. Supported by ”Kompetenzzentrum für Interdisziplinäre Prävention” Univ. of Jena – BGN; 1Motor Research Group, Inst. of Pathophysiology and 2Inst. of Diagnostic and Interventional Radiology, Univ. of Jena, 07740 Jena, Germany P 32-3 INFLUENCES OF MUSCLE FIBRE DEPTH ON THE PHASE OF CORRELATIONS BETWEEN SURFACE EMG CHANNELS – A MODEL R. Grassme, Ch. Anders, H.Ch. Scholle In previous experiments (Anders et. al 2000 and Schack et al. 2000) the cross coherence and phase relationships between bipolar surface EMG (SEMG) channels was investigated. There was found a linear behaviour of the phase relationship along superficial muscle fibres using isolated M. latissimus dorsi activation as an example. This is a clear sign of a time shift between SEMG-channels due to action potential propagation. In contrast, deep muscles fibres (M. erector spinae) did not show systematic phase behaviour. It is shown that symmetric transfer functions have a phase angle of zero or 180°, whereas nonsymmetric transfer functions are able to distort the phase of the signals. Therefore, we investigated the influence of the finite length of muscle fibres on the phase properties of action potentials measured at the surface. We performed simulations with finite fibre models also including realistic end effects which cause field components with a longer range than propagating action potentials. The resulting phase angles were calculated as functions of frequency and the center position z of the correlated bipolar SEMG channels. These functions resulted in approximately linear frequency relationships for superficial fibres whereas they had a very complex behaviour for deep fibres. Two types of nonlinearity, which often could be seen in the experiment, occupy a great area of the z-axis. Therefore, deep muscle fibers are characterised by nonlinear phase dependencies on frequency. Larger distances between correlated channels often result in phase angles swaying around ϕ=±180°. Anders C., et al., Coherence and phase characteristics of deep and superficial muscles measured with surface EMG. Clin.Neurophysiol. 2000;111(1):S88; Schack B. et al. Quantification of synchronization processes by coherence and phase and its application in analysis of electrophysiological signals. Int J Bifurcat Chaos 2000: 10 (11): 2565-2586. Supported by „Kompetenzzentrum für Interdisziplinäre Prävention“, University of Jena – BGN.

P 32-2 METHODICAL APPROACH FOR SIMULTANEOUS SEMG AND 31P-MR-SPECTROSCOPY CHARACTERISATION OF LOW BACK MUSCLE FATIGUE H.C. Scholle1, R. Grassme1, R. Rzanny2, M. Rottenbach1, W.A. Kaiser2 Objective of study: Loading of back region effects strain which depends on individual properties of loaded persons. Therefore, individual morphofunctional characteristics of back muscles are discussed to play an important role for posture and movement (dys)functions of spine (e.g. Panjabi 1992) and could be a relevant reason in back pain pathogenesis. The extent and velocity of back muscle fatigue are considered as significant factor for the strain magnitude in the back region (e.g. Ng et al. 1997). Thus, the aim of our study was the characterisation of individual dynamics of muscle fatigue during a sustained isometric back muscle contraction. Method: In 5 healthy volunteers (aged 22–26) 31P-MR-Spectroscopy (MRS) was performed by the MRT Vision (Siemens, 1,5 T) using the standard heart surface coil (30x30cm). By means of a 2D-CSI-sequence a transversal layer with a thickness of 100mm was analysed which cut the back muscles on the level L4/L5 (voxel size: 40x40x100mm; matrix8x8; TR/TE/ α:465ms/3ms/45°). With a temporal resolution of 30s, 5 MRS measurements before, 5 during, and 10 after the load situation were performed. Following parameter were analysed: phosphocreatin, inorganic phosphor, and pH. Simultaneously, bipolar SEMG on L4/L5 level (2cm lateral) was recorded. Spectral SEMG parameter as the median frequency and spectral power were calculated. Results and conclusions: Bipolar SEMG recordings and 31P-MRS of back muscles (L4/L5 level) could be simultaneously performed which allow the electro-

Institute for Pathophysiology, Motor Research Group, University of Jena, D-07740 Jena P 32-4 INTERACTIONS BETWEEN MACAQUE VENTRAL PREMOTOR (F5) AND PRIMARY MOTOR (M1) CORTEX ON CORTICOSPINAL OUTPUTS TO UPPER LIMB MOTONEURONS H. Shimazu, G. Cerri, M.A. Maier*, P.A. Kirkwood, R.N. Lemon The grasp of objects under visual control depends upon activity in many different motor structures, and includes the interaction of neurons in the ventral premotor area (F5) and the corticospinal outputs from the primary motor cortex (M1). We demonstrate here that stimulation of F5 can produce marked modulation of the corticospinal outputs from M1 and that the effects of this interaction can be observed in intracellular recordings from upper limb motoneurones. Two adult macaque monkeys (M. fascicularis) were prepared for stimulation of the hand representations of F5 and M1 using implanted intracortical microwires (see Cerri et al., this meeting). All procedures were carried out according to UK Home Office regulations. Subsequently, monkeys were prepared for a terminal experiment under isoflurane/50:50 N2O:O2 anaesthesia. During recording anaesthesia was maintained with chloralose (50-70 mg.kg -1 i.v. supplemented when necessary with barbiturates); paralysis was maintained with pancuronium bromide (0.3 mg.kg-1.hr-1 i.v.; see Maier et al., 1998). Single M1 stimuli (100-200 µA) evoked a sequence of descending volleys (D, I1, I2 and I3), recorded

S 308 from the contralateral dorsolateral funiculus, which reflected direct (D) and indirect (I) excitation of corticospinal neurones. In contrast, F5 stimulation (double shocks, 3 ms separation) evoked only a small I wave, but no D wave. The same F5 stimuli delivered before the M1 shock (3 ms between second F5 and M1 stimuli) strongly facilitated the I2 and I3 waves from M1, but had no effect on either the D or the I1 waves. Responses to F5 were greatly reduced by local injection of muscimol in M1, which also completely abolished the I2 and I3 waves from M1 and their facilitation from F5. Intracellular recordings were made from 45 contralateral motoneurones (MNs) identified from the median, ulnar or radial nerves. M1 stimulation evoked monosynaptic EPSPs after the D-wave in 29/45 (55%) MNs, whereas little or no postsynaptic effects were recorded after a single F5 stimulus. Postsynaptic effects considered to result from the I-waves (I-EPSPs and/or I-IPSPs) were recorded in all MNs after M1 stimulation, compared with only 21/45 (46%) for a double shock to F5. In the combined F5+M1 condition, there was robust facilitation of I-EPSPs considered to result from the I2 and I 3 waves in 16/45 (36%) MNs. Facilitation of I-wave associated IPSPs was seen in two MNs. These results are consistent with cortico-cortical inputs from F5 to M1 producing short-latency facilitation of corticospinal outputs from M1. Such a pathway might be important in allowing inputs from F5 to modulate grasp-related outputs originating in M1. Reference: Maier, M A. Illert, M. Kirkwood, P A. Nielsen, J. Lemon, R N. (1998) J. Physiol. 511: 191-212. Supported by The Wellcome Trust, EU COSPIM project and an EU Marie-Curie Fellowship. Sobell Department of Motor Neuroscience and Movement Disorders, Institute of Neurology, UCL, Queen Square, London UK and * INSERM U483, Université Pierre et Marie Curie, Paris, France P 32-5 EXCITATORY INTERACTIONS BETWEEN MACAQUE VENTRAL PREMOTOR (AREA F5) AND PRIMARY MORTOR CORTEX: AN EMG STUDY G Cerri, H Shimazu, MA Maier*, RN Lemon The ventral premotor area (F5) is thought to play an important role in visuomotor control of the hand. Stimulation of F5 evokes hand and finger movements. However, since there are few corticospinal projections from F5 to the cervical enlargement, these are probably mediated via other structures, and a likely candidate is the primary motor cortex (M1) which has massive reciprocal cortico-cortical connections with F5. Here we demonstrate robust, short-latency effects of conditioning stimulation of F5 upon EMG responses evoked by single shocks to M1, consistent with activation of these connections. The study was performed on two adult, purpose-bred monkeys (M. fascicularis), according to UK Home Office regulations. The hand representations of F5 and M1 were first identified using structural MRI scans and intracortical microstimulation (ICMS). Small arrays of fine low-impedance elgiloy microwire electrodes were implanted intracortically at the centre of these representations under full anaesthesia (isoflurane, 50:50 N2O:O2). Subsequently, effects of ICMS through these arrays was investigated under light sedation using either ketamine or ketamine plus small doses of medetomidine HCl; 4-5 recording sessions were carried out in each monkey. EMG responses were recorded from needle electrodes positioned in intrinsic hand muscles. With some background activity in the muscle, test stimulation of M1 (single bipolar shocks, duration 0.2 ms, 70-100 µA) evoked short-latency (7.3-9.5 ms) EMG responses consistent with corticospinal action. These responses were conditioned by F5 stimulation of similar intensity. At least two F5 shocks (at 3 ms interval) were needed to show any effects. While such stimuli rarely produced any EMG effects alone, they strongly facilitated M1 effects for intervals of between 2.5-6 ms between second F5 shock and the test M1 shock. Maximum facilitation (120-230% of test response amplitude) was obtained at 3 ms interval. No signs of suppression were observed. These effects were specific to particular electrode pairs, suggesting a rather focal interaction. No such interactions were observed from other prearcuate or postcentral locations. The results are consistent with a facilitatory effect exerted by the ventral premotor cortex (F5) on the corticospinal outputs from the primary motor cortex (M1) to hand motor nuclei. These facilitatory actions could be of considerable importance for the coupling of grasprelated neurons in F5 and M1 during visuomotor tasks. Further studies are needed to ascertain whether interactions occur at other sites. Supported by The Wellcome Trust, EU COSPIM project and an EU Marie-Curie Fellowship.

Sobell Department of Motor Neuroscience and Movement Disorders, Institute of Neurology, UCL, Queen Square, London UK and * INSERM U483, Université Pierre et Marie Curie, Paris, France P 32-6 HUMAN BRAIN ACTIVITY DURING MOTOR IMAGERY AND REAL FINGER MOVEMENTS J.P. Kuhtz-Buschbeck, S. Ulmer2, M.Gölge, A. Grosskopf’, S. Wolff3, R. Baron3, O. Jansen2 To examine the potential involvement of human primary sensorimotor areas during motor imagery, brain activity was studied in normal volunteers by functional magnetic resonance imaging on a Siemens Vision 1.5 Tesla scanner. Ten healthy male right-handed subjects either performed or imagined finger movements with their dominant and their non-dominant hand. In the real movement condition, the subjects rhythmically (at 2 Hertz) compressed a flexible cube between the thumb and the tips of the index and middle finger. The peaks of the exerted grip force, measured outside the scanner, reached values of about six Newton. During the motor imagery condition, the same rhythmical movement was vividly imagined. To compare brain activity in these two conditions with the baseline during rest, the statistical program SPM99 was used. Real finger movements consistently engaged the contralateral primary sensorimotor cortex (M1/S1), but additional significant (p<0.01) activation of the ipsilateral M1/S1 was found in two subjects during movements of the non-dominant hand. During motor imagery we observed activation of the contralateral Ml/S1 in six participants, regardless of the hand used. Besides this, the supplementary motor area was consistently activated during motor imagery. The activated volume was, however, smaller during motor imagery than during the real movement condition. The preliminary results suggest that primary sensorimotor areas may be engaged during motor imagery in normal subjects. ‘Physiologisches Institut, 2Sektion Neuroradiologie der Neurochirurgie, 3 Klinik fur Neurologic, Christian-Albrechts-Universität, Olshausenstrasse 40, D 24098 Kiel, Germany P 32-7 fMRI-BRAIN MAPPING OF THE MOTORIC FUNCTION OF THE EXTERNAL ANAL SPHINCTER H. Hinninghofen, B. Wietek, S. Anders, N. Birbaumer, W. Grodd, P. Enck Objectives: The cortical representation of motor function of the anorectum and pelvic floor may be important to understand the varying effectivness of biofeedback training of the anal sphincter performance in patients with fecal incontinence. Method: We investigated 7 healthy volunteers (33 to 55 yrs) by means of whole brain functional Magnetic Resonance Imaging (fMRI) during voluntary external anal sphincter contraction. A water-filled ballon probe was positioned in the anal canal, and the subjects were instructed to squeeze maximally on a visual signal provided. As control condition subjects squeezed a little ball with the right hand. Each condition was repeated 32 times in an event-realated design. Images were aquired with a commercial 1.5 Tesla tomograph (Siemens, Vision) using echoplanar imaging (EPI; 28 axial slices, 4 mm slice thickness, 64 matrix, acquisition time 3 sec, 448 scans per subject). Spatial preprocessing (realignment, smoothing, normalisation to a standard brain) and statistical analysis were performed with SPM99 (Dept. Cogn. Neurol., London). Results: During sphincter contraction significant (p<.001) activation was seen bilaterally in the senso-motor cortex (BA 4 and 5) inbetween body trunk and extremities representation and in the right supplementary motor cortex (BA 6) (all z> 7.59), the right basal ganglia (putamen) (z =5.79), and the superior cerebellum (CE) (z=6.66) as well as bilaterally in the neo-cerebellum (z=6.8 and z=7.59, for left and right, resp.). For hand contractions significant activity in the left primary motor cortex (MI) and the right cerebellum were found, as expected. For both conditions the activated CE areas corresponded with recently described sensorimotor mapping (homunculus) of the human CE. Conclusion: These results show that similar areas are involved in processing of voluntary anal sphincter contractions as in processing of anal sensations. The paradigm described here will allow to examine the mechanisms of biofeedback training in incontinence and thus, increase the effectiveness of training. Supported by a grant from DFG, En 50/18-1

S 309 Dept. for General Surgery, University of Tuebingen, Hoppe-Seyler-Str. 3, 72074 Tuebingen, Germany P 32-8 ORGANIZATION OF COMMISSURAL INTERNEURONS IN THE CHICKEN EMBRYO SPINAL CORD E.A. Jensen , J.C. Glover Sensorimotor processing in the spinal cord is dependent to a large extent on intersegmental connections that originate from interneurons. Still, little is known about the anatomical organization of commissural interneurons (CINs) in chicken embryos. We have used retrograde labeling to characterize the distribution of CINs in the cervical, brachial, thoracic and lumbar spinal cord. Fluorescent dextran- amines were applied to the cut surface of ventral midline. After appropiate incubation the preparations were fixed, sectioned and CIN positions were photographed and plotted. We have also used differential retrograde tracing to characterize the distribution and axonal trajectories of ascending commissural interneurons (aCINs) and descending commissural interneurons (dCINs) in the lumbar spinal cord. The majority of aCINs and dCINs were located in the ventromedial region at this stage, and aCIN outnumber dCINs. Our results also show intersegmental CINs with both ascending and descending axons (adCIN), but these represented only a minor fraction. The ongoing studies are focusing on development of CINs of different projection classes in the chick spinal cord. Department of Physiology, University of Oslo, Songsvannsveien 9, PB 1103, Blindern, N – 0317 Oslo, Norway P 32-9 MODULATION OF THE LOW THRESHOLD REFLEXES IN HUMAN LOWER LIMBS BY CHANGES IN KNEE POSITION K.K. Kalantari, R.H. Baxendale In man, stimulation of group I afferents in the common peroneal nerve evokes non-monosynaptic excitation of quadriceps (Fournier et al 1986). This is attributed to Ib afferents Brooke et al (1990) acting through an oligosynaptic pathway. This abstract describes experiments performed to investigate the effect of the knee joint position on these reflexes. Experiments were performed in 14 volunteers. The protocol was approved by the local Research Ethics Committee. Single pulses of 200 microseconds duration were delivered to the common peroneal nerve at 1.3-1.5×motor threshold for tibialis anterior. The responses were recorded in the averaged EMG of rectus femoris during ongoing contractions at 20% of maximum. The average latency and duration of the responses was 29.4 ± 2 and 15.7 ± 2 milliseconds respectively. Reflexes were elicited at intervals of 10o between 180o and 130o. The mean magnitude and area of the response showed a very significant drop (P≤0.005) between 180 o and 170 o followed by a more gradual decrease at more flexed positions. At 130o of flexion, no response was detected in any of the subjects.

Figure 1. A series of reflexes elicited in the averaged EMG of rectus femoris by stimulation of the common peroneal nerve at 1.3x motor threshold. Reflexes were elicited at a series of knee positions between 130o and 180 o. The reflex is present only near full extension. The modulation of the reflex gain by knee position may be of functional significance during gait since the strongest effects occur when the knee is extended, that is in the position at about heel strike. Brooke JD. McIlroy WE, (1990) EEG & Clin Neurophysiol, 75: 401-409. Fournier E et al (1986) J Physiol, 377: 143-169 Div. of Neuroscience, IBLS, The University, Glasgow G12 8QQ, Scotland UK

P 33-1 THE NATURAL VISUAL ENVIRONMENT OF CATS C. Kayser, B. Betsch, W. Einhäuser, K.P. Körding, P. König To understand the operations performed by sensory areas it is important to know the statistical structure of the natural input these systems are exposed too. Interested in the visual system, we developed a unique setup to record natural videos as seen by freely behaving cats. An micro-CCD-Video-camera is reversibly mounted on the head of three cats and connects to a backpack containing a VCR carried by the human experimenter. Videos were recorded when the animals were exploring different environments like forests and grasslands. We find a number of remarkable features of these videos. (1) The spatial power spectrum of the cat-cam movies closely matches the control, showing that a 1/f^2 slope is indeed natural. (2) The statistical analysis of these videos reveals a predominance of horizontal orientations. (3) Analysing the spatial distribution of contrast we do not observe higher values in the centre of the images. This contrasts with results from humans, which tend to fixate areas of higher contrast (Reinagel & Zador). (4) Analysing a sequence of images we find that the precise position of contours varies faster than their orientation. (5) Furthermore, collinear contours prevail over parallel shifted contours. (6) Finally, in the temporal domain we find a good agreement between the second order statistics of cat-cam and control videos. The distribution of velocities in the videos on the other hand shows a more kurtotic distribution for the cat-cam compared to the control. These findings demonstrate the rich structure of natural visual stimuli and directly relate to results of anatomical and physiological studies. Institute of Neuroinformatics, Uni/ETH Zürich, Winterthurerstr. 190, 8057 Zürich, Switzerland. Supported by ZNZ, Boehringer Ingelheim & SNF (31-65415.01)

P 33-2 LIGHT-ADAPTATION OF MAMMALIAN RETINAL RODS REVISITED K. Donner1, S. Nymark2, C. Haldin1, H. Suominen2, A. Koskelainen2 Mammalian rods have been reported to light-adapt much less efficiently than rods of ”lower” vertebrates. Our hypothesis is that (when corrected for differences in quantum catch) the differences are due to the different temperatures at which the experiments have been done. To test this, we studied light-adaptation in rat rods as a function of temperature. Rod photoresponses to brief flashes of light were recorded by ERG across the isolated, aspartate-superfused rat retina. Intensity-response functions were recorded in darkness and under steady adapting background lights over the temperature range 12-36 °C. Light-adaptation was measured by changes in the number of flash-induced photoisomerizations per rod (Rh*) needed for a half-saturated response, as function of background intensity (Rh*s-1). The reciprocal of this value provides a sensitivity measure (”fractional sensitivity”), where effects of response compression (caused by changes in the circulating rod current) have been factored out. Background intensities were increased in 0.5 log unit steps from 0.003 to 10 3 Rh*s -1. At all temperatures, fractional sensitivity stayed constant under the weakest backgrounds, then began to decrease with increasing background intensity. The background intensity needed to reduce fractional sensitivity by 1 log unit (denoted I10) was taken as an index of lightadaptation capacity. At 36°C, I 10 was 500-800 Rh*s-1, similar to the values 500-1000 Rh*s-1 found in the intact human eye as well as in single rods of cat retina (Friedburg et al. 2001, J. Physiol. 534:217; Tamura et al. 1989, Science 245:755). As temperature was lowered, I10 decreased as follows: to ca. 200 Rh*s-1 at 28 °C, 40 Rh*s-1 at 20 °C and 7 Rh*s-1 at 12 °C. In frog (Rana temporaria), I10 is about 20 Rh*s-1 at 16.5 °C and 7 Rh*s-1 at 9 °C (Donner et al. 1995, Vision Res. 35:2255). Thus, at 36°C, I10 of rat rods was similar to that of human or cat rods. At lower temperatures, it was similar to that of amphibian rods at the same temperatures. The results contradict the idea that there are basic differences in rod light-adaptation between mammals and lower vertebrates. 1 Dept. of Biosciences, University of Helsinki, FIN-00014 Helsinki, Finland; 2Lab. of Biomedical Engineering, Helsinki Univ. Technology, Espoo, Finland

S 310 P 33-3 MECHANISMS OF COMPENSATION AND CORRECTION OF VISUAL INFORMATION ARE SPATIALLY ORGANIZED M.Sharikadze, A.Kezeli (introduced by D.A.Brown) According to the hypothesis we proposed earlier (Kezeli et.al. 1991), the McCollough-type contingent after-effects (McCollough, 1965) arise from central mechanisms of compensation and correction. Many details of this hypothesized mechanism are still unknown. Specifically, it is not clear whether, during elaboration of compensatory shifts, information from the whole visual field is processed, or whether information from different regions is processed separately and integrated later. Experiments were carried out in five subjects, functional state of the visual system of which was assessed with standard tests (Landolt Acuity test, Farnsworth-Munsell D-15 test, Ishihara’s Pseudoisochromatic Plates, and Nagel Anomaloscope). The subjects were presented with horizontal black-red and vertical black-green gratings (spatial frequency-5 cycle/grad), displayed on a computer screen (30° x 40°) in the standard paradigm for acquisition of the McCollough effect. All five subjects acquired clear-cut McCollough effects. Then, in the middle of display, a fixation image was presented (different words were written, with consecutive presentation of letters at the same point). While reading these words, at the periphery of the visual field (with 7° of eccentricity) the subject was presented with the same spatial frequency gratings of 4° x 5° angular size, the orientation and colour of which was a counterpart of the grating presented at the first stage. Two types of combined stimuli were used for testing: Full-screen Test and Local Test, containing achromatic gratings of both orientations. The phantom colours, evolving during viewing the gratings of the same orientation, depended on the location of the gratings: the colours on the Full-screen Test-grating and on the Local Test were opposite. The data obtained indicate that central mechanisms of compensation and correction are elaborated separately for different regions of the visual field. Laboratory of Physiology of Vision, I.Beritahvili Institute of Physiology, Georgian Academy of Sciences, 14, L.Gotua Str., Tbilisi, 380060, Georgia P 33-4 THE DOUBLE ILLUSION: A NEW TYPE OF COLOR AFTEREFFECT M.Malania, A.Kezeli, T.Kezeli, M.Roinishvili We have shown that when a subject is consecutively presented with horizontally oriented red- and vertically oriented green surfaces, than the surface-oriented contingent McCollough type color aftereffect could be elaborated. Following 20-25 such presentations the horizontally oriented white surfaces are perceived by the subject as greenish, while the vertically oriented ones - as reddish. It is known that during observation of the Mach Cards the illusory alterations of their spatial orientation are evident - an observer believes that sometimes they are vertical and sometimes transit into the horizontal surface. In our experiments we endeavored to determine if would the illusory alteration of the orientation affect the surface-oriented contingent McCollough type aftereffect. Experiments were carried out in eight individuals with normal color vision (aged 20-58). After we had determined that the subjects had clearly well acquired surface-oriented contingent McCollough type aftereffect, they were presented with white Mach Cards. Five subjects out of eight clearly observed illusory color alteration in the cases when the likewise illusory displacement of the Mach Cards was the case - i.e. horizontally placed card was perceived as greenish, while illusorily vertical its location induced alteration of the illusory color as well - the subject perceived it as reddish. Therefore, we encountered interesting case of visual illusion, which may be entitled as the Double Illusion - an orientation-induced alteration of illusory color during illusory alteration of orientation as such. Lab. of Physiology/Vision, I.Beritahvili Institute of Physiology, Georgian Academy of Sciences, 14, L.Gotua St., Tbilisi, 380060, Georgia P 33-5 INTERACTION OF DIFFERENT TYPES OF EYE-MOVEMENTS IN HUMANS AND CATS C. Kayser, G. Möller, F. Knecht, P. König

Eye-movements are an important tool we use to analyze our visual environment. Three different types of eye movements are distinguished: the vestibulo-occular reflex (VOR), scanning movements and stabilizing movements (OKN). Strong evidence indicates that these are controlled by different systems in the brain. But not much is known how these different types of eye movements interact. Here we study their interaction in humans and cats using gratings and natural movies as stimuli. The rich structure of the natural movies is optimal to stimulate target-searching movements. Stabilizing eye-movements were induced by superimposing gratings and natural stimuli with global movements of different velocities. Eye movements of all subjects were measured using a Dual Purkinje Imaging eye-tracker. For cats moving natural images induced a stronger optokinetic response than gratings. Also the velocity with the strongest effect was greater for natural images. This argues for a synergistic interaction between scanning and stabilizing movements; both systems are affecting each single saccade. For humans there was no difference between these two kinds of stimuli, both eliciting a strong optokinetic response. Notably, the saccades for moving stills show a bimodal distribution with some saccades being purely optokinetic and others scanning the image structure. This shows that in humans the OKN and scanning movements do not add linearly at each saccade, but both types of saccades appear, separated in time. Surprisingly, natural movies with superimposed global movement show a weaker effect than stills moving with the same speed; both in humans and cats. The OKN due to the global movement is partially cancelled by the motion inherent to the video. These results demonstrate qualitative differences in the interaction of eye-movement systems in cats and humans. Supported by ZNZ & SNF (31-65415.01). Inst. Neuroinformatics, Uni/ETH, Winterthurerstr. 190, CH-8057 Zürich P 33-6 LARGE FLUCTUATIONS IN THE RESTING MEMBRANE POTENTIAL OF FROG SACCULAR HAIR CELLS F. Jørgensen, A.B.A. Kroese The resting membrane potential (Vm) of hair cells provides the main driving force for the apical mechano-electrical transduction current. Another function of Vm, however, is the regulation of the basolateral liberation of a excitatory transmitter. We have investigated spontaneous fluctuations of Vm, with the aim to determine which ionic conductances are involved in control of Vm.Vm of isolated hair cells from the sacculus of Rana esculenta (bath in mM: 2 KCl, 110 NaCl, 1 CaCl2, 0,5 MgCl2, 5 HEPES, 5 Glucose) was examined at room temperature with the whole cell patch clamp technique, with 110 KCl, 5 NaCl, 2 MgCl2, 5 EGTA in the pipette. Vm of 32 hair cells was -57.0 + 1.6 mV (mean + sem) and the input membrane conductance determined from the IV curve was 1.0 + 0.1 nS (range -90 to -50 mV). This low conductance range was flanked by the inward rectifying GK,IR (21.2 + 2,6 nS) and the combined outward rectifying K+ conductance, GK,Ca and GK,DR (29,3 + 2.4 nS). Vm was found to be stable in 56% of the cells. In the other 44%, Vm showed spontaneous fluctuations between two distinct levels, up to 30 mV apart, determined by the low conductance range. These fluctuations could be reversibly inhibited by increasing [K+]o from 2 to 8 mM. The [K+]o of 8 mM also induced an increase in and shift of GKIR along the voltage axis, which resulted in a reduction of the size of the low conductance range. The finding that lowering [K+]o from 2 to 0.5 mM did not provoke voltage fluctuations in stable cells, excludes an instability of GK,IR as the main cause of the spontaneous fluctuations. Blocking the transduction channels with 100 µM amiloride hyperpolarized Vm by 10.8 + 4 mV (n=13), and induced voltage fluctuations in previous stable cells (n=4). These results show that both low membrane conductance and current through the transduction channels contribute to the instability of Vm. IMB, Physiology & Pharmacology, University of Southern Denmark, Odense, 5000 Winsløwparken 21, Denmark P 33-7 DEFECTS IN PHYSIOLOGICALLY RELEVANT ION CHANNELS OF THE COCHLEA LEAD TO SYNDROMIC AND NONSYNDROMIC HEARING IMPAIRMENT B. Haack, M. Ebauer, T. Toth, M. Wróbel, M. Pfister, N. Blin, S. Kupka Non-syndromic severe to profound neurosensory hearing impairment

S 311 (NSHL) is one of the most common human sensory disorders affecting one in 1000 children with at least 60% of cases being inherited. The mode of inheritance of non-syndromic hearing disorders can be distinguished in autosomal dominant (10-15%), autosomal recessive (70%), X-linked (1-3%) and mitochondrial forms (less than 1%). Syndromal hearing defects are categorized according to the underlying defects and contain more than 700 syndromes. In daily practice, nonsyndromic hearing impairment (NSHL) is predominant and accounts for up to 70% of all inherited sensorineural hearing defects. Both types of hearing impairment were linked to defects in genes coding for ion channels and ion channel subunits, respectively (e.g. GJB2, KCNQ1, HERG, CLDN14, SLC26A4). Our molecular genetic investigation of the type of mutation within affected individuals (sporadic cases and families) will aid in understanding the physiological malfunction of the concerned proteins. Department of Anthropology and Human Genetics, University of Tübingen, Germany P 33-8 2 PHOTON IMAGING OF FM 1-43 REVEALS MEMBRANE UPTAKE, TRAFFICKING AND FUSION IN MAMMALIAN INNER HAIR CELLS C.B. Griesinger, C.D. Richards, J.F. Ashmore Sound-triggered synaptic transmission of inner hair cells (IHCs) in the mammalian inner ear is locked to the stimulus phase at high frequencies. High release rates necessitate efficient membrane recycling for vesicular pool replenishment. We have studied apical membrane retrieval in guinea pig IHCs by 2-photon imaging of the fluorescent membrane marker FM 1-43 (5µM) applied to scala media in the intact epithelium in situ. Internalised signal first appeared in an apical endosome. Subsequently, labelled membrane was transported to the basal synaptic zone by kinesin-dependent trafficking and targeted to membraneassociated hotspots (≤ 1 µm), candidates for release sites. The fluid phase endocytosis marker Lucifer Yellow produced a labelling pattern similar to FM 1-43, consistent with FM 1-43 labelling the membrane component of endocytosis. Using 2-photon voxels we estimate the apical endocytosis rate at 3.2 µm2 s-1, equivalent to 1200 vesicles of 30 nm diameter. Depolarisation (40 mM K+) of IHCs in excised epithelial strips accelerated the decay rate of basal fluorescence and doubled the rate of apical endocytosis. Blockers of voltage-gated calcium channels abolished these effects suggesting that increased exocytosis triggers apical endocytosis. Transepithelial electrical stimulation of IHCs in situ (20 Hz stimulus for 50 s) selectively de-stained hotspots by about 40 % of their pre-stimulus fluorescence. The data suggest that apical endocytosis contributes to membrane pools in the basal synaptic zone of IHCs. It is likely that it represents a form of ‘rapid endocytosis’ described in neurones and secretory cells, which compensates for phases of increased release. Supported by the MRC and BBSRC Department of Physiology, University College London, Gower Street, London WC1E 6BT, UK. P 33-9 ELECTROPHYSIOLOGICAL AND MORPHOLOGICAL CHARACTERIZATION OF SPINY STELLATE, STAR PYRAMIDAL AND PYRAMIDAL CELLS IN LAYER IV OF RAT BARREL CORTEX I. Flagmeyer1, J.F. Staiger2, H.J. Luhmann1 Within the barrel cortex, layer IV resembles the initial stage of sensory processing of tactile information acquired by the whiskers. In order to study the cellular determinants of this processing, we performed wholecell current-clamp recordings from layer IV neurons (n=56) in 500 µm thick thalamocortical slices prepared from male wistar rats (P19–21). Cells were labelled with biocytin and the exact laminar location and morphology was identified for each cell. Morphologically, layer IV neurons could be classified as pyramidal (n=10), star pyramidal (n=17) and spiny stellate cells (n=29). All types of neurons showed similar passive membrane properties (resting potential –81 ± 3.52 mV, input resistance 175 ± 55.34 MOhm). Stimulating the cells for one second with depolarising current pulses and measuring the intervals between the first two spikes (ISIs) characterised the majority of cells (n=46) as intrinsic bursting cells (IB) with ISIs of about 9 ms. A second group of cells (n=10), termed regular spiking cells (RS), showed ISIs larger than

20 ms. With increasing current intensity RS cells also revealed burst firing. We found no correlation between the cells morphology and the firing pattern. Our results indicate that layer IV neurons possess active intrinsic membrane properties which enable them to boost the afferent sensory input into the barrel cortex. The firing pattern of these neurons is not invariable but strongly dependent on the extent of membrane depolarisation. This qualifies layer IV neurons to respond to incoming information from the thalamus in a state-dependent manner. Supported by DFG grant Lu375/3-3 to HJL and STA431/2-2 to JFS 1 Institute of Neurophysiology, 2Vogt-Institute for Brain Research, Heinrich-Heine University, P.O.B. 101007, 40001 Düsseldorf, Germany

P 33-10 SPATIO-TEMPORAL DYNAMICS OF ODOR REPRESENTATIONS IN THE MAMMALIAN OLFACTORY BULB H Spors, T Margrie, A Grinvald, B Sakmann Odors evoke spatial patterns of activity on the level of the olfactory bulb. Our previous studies in rats showed that these spatial patterns change substantially within one respiration cycle and that individual glomeruli are activated in sequential fashion. In mammals rhythmical sampling i.e. breathing and sniffing determine odor stimulation. We examined the influence of repetitive odor sampling and odor concentration on the evoked spatio-temporal patterns with the new voltage sensitive dye (RH 1838) in anesthetized rats and mice. The use of a computer controlled olfactometer and synchronization of data acquisition and respiration allowed measurements with highly precise timing of all related events. We found that increasing odor concentrations reduced response latencies, increased response amplitude and recruited new glomerular units. However the sequence of the activation of the glomeruli was independent of odor concentration. We also found small but robust membrane potential oscillations locked to the nasal respiration cycle with and without odor stimulation, consistent with recent results obtained by in vivo intracellular recordings (Margrie, SFN 2001). Sensory input specifically and locally changed the amplitude and phase of these membrane potential oscillations. These findings in the rodent olfactory bulb suggest that both spatial pattern and timing of activation contribute to coding of odor identity and its concentration. Using the glutamate antagonists NBQX and AP5 we demonstrate that the voltage sensitive dye reports neuronal activity downstream of the olfactory receptor neurons. It remains to be examined to what extend the observed spatio-temporal dynamics are directly determined by sensory input from the olfactory receptor neurons and how they are shaped by computation in the bulbar circuitry. Max-Planck-Institut für medizinische Forschung, Jahnstr 29, 69120 Heidelberg, Germany P 34-1 DISCHARGE BEHAVIOR OF PYRAMIDAL CELLS OF RAT SOMATOSENSORY CORTEX IN RESPONSE TO NOISY INPUT CURRENT A. Rauch, G. La Camera, W. Senn, H.-R. Lüscher, S. Fusi Cortical phenomena observed in vivo, like working memory, can be related to the properties of individual neurons, in particular to the current-to-frequency relationship. In this study we measured the response of layer 5 pyramidal cells in rat somatosensory cortex of acute brain slices to noisy currents imitating the barrage of hundreds of inhibitory and excitatory synaptic inputs. If these inputs are random and statistically independent, the total synaptic current is a stochastic process (an Ornstein-Uhlenbeck process) characterized by its mean, variance and time correlation length. These parameters in turn can be expressed in terms of the presynaptic activity and the synaptic efficacy. We injected noisy currents with short time correlation length (1ms) and with different mean and variance into pyramidal cells and measured the discharge rate. The current-to-frequency relationship could be fitted by the response function of two simple integrate and fire model neurons described by only five independent parameters including adaptation. These models provide a synthetic description of the response of the pyramidal cells to a large variety of realistic input currents. This allows for quantitative predictions on the behavior of large populations of interconnected neurons and it is a first step in

S 312 bridging the gap between in vitro and the in vivo experiments. Institute of Physiology, University of Bern, 3012 Bern CH P 34-2 CORTICAL LAYER IV SPINY NEURONS SHOW CELL TYPE SPECIFIC DIFFERENCES IN THEIR SYNAPTIC INPUTS D. Schubert*, J. F. Staiger, R. Kötter, K. Zilles, H. J. Luhmann Spiny neurons in layer IV of the neocortex are a morphologically heterogeneous group of excitatory neurons. The spatial distribution of the intracortical synaptic inputs onto these spiny neurons is still not completely understood. We used a combination of whole-cell patchclamp recordings and caged glutamate photolysis to obtain detailed maps of the presynaptic fields delivering monosynaptic excitatory or inhibitory input to visually identified and biocytin labeled spiny neurons in layer IV (i.e. barrels) of rat somatosensory cortex. Caged glutamate photolysis was limited to fields of 50 µm x 50 µm in a region containing all cortical layers and at least two barrel-related columns. The cells were morphologically categorized into spiny-stellate, star-pyramidal and pyramidal neurons. Activation of presynaptic neurons in the same barrel elicited extensive and strong excitatory inputs onto all neurons. In contrast to most spiny stellate cells, all star-pyramidal and all pyramidal cells received additional excitatory synaptic inputs from neurons located in non-granular layers within the same column. A small number of cells received distinct excitatory inputs from the neighboring barrel. Excitatory transcolumnar inputs from nongranular layers were altogether rare. The extent and origins of inhibitory inputs were more variable. All spiny neurons received inhibitory inputs from neurons located within the same barrel. Additional inhibitory inputs predominantly originated from layers II, III and V of the barrel-related column. These results show that spiny neurons in layer IV are strongly interconnected with other neurons in their respective barrel and, depending on the cells’ morphology, receive also excitatory and inhibitory synaptic inputs from non-granular layers. Thus spiny stellate cells are predominantly involved in early signal processing within their barrel, while star-pyramidal and pyramidal cells receive additional modulatory inputs from neurons in non-granular layers. A small subset of spiny neurons may also integrate transcolumnar inputs via barrelto-barrel connections. Institute of Neurophysiology, Heinrich Heine University, Universitätsstraße 1, 40225 Düsseldorf, Germany P 34-3 FMRI-BRAIN MAPPING OF ANORECTAL SENSORIC STIMULATION B. Wietek, M. Lotze, H.Hinninghofen, F.Schick ,W. Grodd, P. Enck Introduction: Clinically, the mapping of sensory stimulation of the pelvic floor and rectum is highly important since loss of sensory information of this compartement leads to incontinence. If any sensory input from this region is available continence training would be promising. Therefore, we established a method to map the sensory stimulation of the rectum using fMRI. Methods: Five healthy subjects were investigated during non-painful stimulation of the distal rectum by means of pneumatic stimulation with a rectal balloon probe. Functional Magnetic Resonance Imaging (fMRI) of the whole brain was acquired with a commercial 1.5 Tesla tomograph (Siemens Vision) using echo planar imaging (EPI; event related: 28 axial slices of 4 mm slice thickness, 64 matrix, acquisition time 3 sec). Measurement was performed in an event related design. Preprocessing and statistical analysis was performed with SPM96 (Dept. Cogn. Neurol., London). For group analysis individual datasets were normalized using the Montreal template. Images from 3 to 9 seconds after end of stimulus, lasting 3 seconds, were assumed to be activated with a statistical threshold of p <0.01 (extent threshold 10 voxels). Results: In group statistics anal stimulation revealed significant activation in in the insular bilaterally and the secondary sensory cortex, the whole cingular gyrus (CG), the mediosagital primary and sensory cortex, the dorsolateral prefrontal lobe bilaterally, the anterior and posterior cerebellar lobe and the right amygdala. Rectal stimulation showed less activation intensity and activation sites in SII, insula, anterior and medial CG, dorsolateral prefrontal and inferior parietal lobe. Cerebellar activation sites were seen in the left anterior and posterior lobe with rectal representation maxima approx. 1 cm lateral to anal in the anterior and approx. 1.5

cm mediocranal to anal in the posterior lobe. Discussion: Although stimulation below the pain threshold was performed, the activated regions during anal and rectal stimulation were similar to those observed during painful stimulation using a group design (3). This includes SII and the insula, the anterior and medial cingular gyrus and the dorsolateral prefrontal cortex. Primary motor cortex (M1) activation, present in images 6 seconds after anal stimulation but absent after 3 sec., may indicate a reflectory increase in anal sphincter tonus. Cerebellar activation in the anterior lobe was observed inferior to the hand and inferiolateral to the lip representation far distinct to the foot representation. The activation in the posterior lobe is much more prominent than during movements of the extremities. The activation of the right amygdala was previously observed during aversive stimuli. Rectal stimulation showed less activation of limbic structures such as the amygdala and the cingular gyrus. Supported by DFG, En 50/18-1 B.M. Wietek (e-mail: [email protected]), Section on experimental Radiology, Radiologic University Hospital, D-7276 Tübingen, Hoppe-Syler-Str. 3, Germany P 34-4 ANESTHESIA-INDUCED EEG BURSTS AFFECT TACTILE INFORMATION PROCESSING IN SOMATOSENSORY CORTICAL BUT NOT THALAMIC AREAS IN RATS C. Vahle-Hinz, O. Detsch‡, M. Siemers, E. Kochs‡, B. Bromm The electroencephalograhic (EEG) pattern of burst-suppression consists of alternating periods of high voltage bursts and periods of isoelectricity. It may be reversibly induced by anesthetics, e.g. the volatile anesthetic isoflurane. The neuronal mechanisms underlying the EEG burst-suppression pattern are not yet understood, however, they are generally attributed to interactions within thalamocortical networks. We studied the influence of this peculiar EEG activity on processing of tactile information in single neurons of the thalamic ventrobasal complex (VB) and the primary somatosensory cortex (S1) in rats. The electrocorticogram (ECoG) was recorded above the barrel field of the S1 cortex. To induce the EEG burst-suppression pattern, end-tidal isoflurane concentrations between 1.6 and 2.2 vol.% were used. Mechanosensory neurons in VB and S1 were selected which responded with sustained discharges to vibration of whiskers under baseline isoflurane anesthesia (~1.0 vol.%). The response activity of VB neurons (n = 21) did not differ during ECoG isoelectricity and bursts. In contrast, the response activity of S1 neurons (n = 18) was found to be 3 fold higher during bursts (3.5 ± 0.6 spikes/s) compared to isoelectric periods (1.2 ± 0.3 spikes/s). Thereby, during EEG bursts, the responsiveness of S1 neurons was similar to that of VB neurons during both EEG phases. The ongoing activity of VB neurons did not differ between both EEG phases and was not different from ongoing activity of S1 neurons during isoelectricity; during bursts, however, S1 ongoing activity increased markedly. The results suggest that the profound suppression of cortical sensory information processing by isoflurane is suspended during EEG burst-associated elevated cortical excitation. This effect appears to be independent from the activity of the sensory thalamus and vice versa does not affect thalamic sensory processing. Institut für Physiologie, UKE, Martinistr. 52, 20246 Hamburg and ‡Klinik für Anaesthesiologie, TU München, 81675 München, Germany P 34-5 EVIDENCE OF SOMATOSENSORY EVOKED 600 Hz ACTIVITY IN THE RABBIT’S BRAIN L. Flemming, M. Eiselt. F. Gießler, J. Haueisen, U. Zwiener The electrical stimulation of somatosensory nerves evokes a high frequency activity (around 600 Hz) within the time interval of the first cortical answer. This phenomenon has been demonstrated in humans, monkeys and rats but not in rabbits. Five White New Zealand rabbits with an age of at least 6 months were anaesthetised with a combination of Ketamin (24 - 30 mg/kg/h) and Xylazine (2.4 -3.0 mg/kg/h). Initially, the anaesthetic agents were injected in the muscle and later in the femoral vein. The cannulation of the femoral artery was used for monitoring of physiological parameters. Animals were kept normothermic and were allowed to breath spontaneously. Stimulation electrodes were implanted and placed on the right median and right tibial nerve. The scalp and skull bone of the left hemisphere

S 313 were removed in a rectangular area above the somatosensory cortex SI (from -5 to 4 mm relative to the Bregma) to place the ECoG electrode on the surface of the cortex. ECoG (grid of 4x 4 electrodes (interelectrode distance 1.25 mm)) and 16 channel MEG (array of 4 x 4 pick-up coils (distance between adjacent coils 8.4 mm)) were recorded simultaneously. ECoG/MEG-data evoked by electric nerve stimulation (current: 0.2-0.7 mA, stimulus duration 200 µs, interstimulus interval 253 / 503 ms) were recorded (sampling rate: 5 kHz, high pass filter: 0.3 Hz, low pass filter: 1.5 kHz) and averaged (5000 trials). A band pass filter of 450 to 750 Hz were used for the 600 Hz activity. The somatosensory evoked potentials (SEP) and the somatosensory evoked fields (SEF) consisted of several components. The latency of the first component (first cortical answer) was for the median nerve 16 ± 2.5 ms and for the tibial nerve 25 ± 2.5 ms. The 600 Hz activity appeared in the time interval of the first component of the SEP of the median nerve. The main burst of this activity had a latency of 12.8 ± 2.5 ms and was before the main part of the first component. The duration of the 600 Hz activity and of the first component were about 9 ± 2 ms but the strength of the high frequency was approximately 30 times lower than the signal of the first component. Therefore, rabbit cortical cells generate high frequency activity which is correlated to the first arrival of afferent somatosensory information in the cortex. Biomagnetic Center, FSU-Jena; supported by EU-project SimBio

Three abduction-adduction move-ment models were selected for their different innervation patterns: 1) the metacarpophalangeal (MCP) joint of the little finger, 2) the carpometacarpal (CMC) joint of the thumb and 3) the MCP joint of the index finger. Local anaesthetic blocks were effected by injection of 1% lignocaine and confirmed by EMG recordings and cutaneous testing using Semmes-Weinstein filaments. Proprioception was tested by imposed ramp movements of the anaesthetised digit (1), which the subject tried to match with the normal digit. For the little finger, ulnar nerve block anaesthetised the muscles, skin and joint. Four subjects still detected some movement but reported it as extension-flexion and neither amplitude nor velocity could be matched. For the thumb, median nerve block anaesthetised the bulk of the intrinsic muscles and skin, only excepting the radially innervated aspect overlying the MCP and CMC joints. Subjects still detected more than 90% of the movements and graded them accurately. Addition of a radial block led to 40% of the moves being undetected. The index finger shows extensive mixed innervation, the musculature being innervated by the ulnar nerve and the skin by the median and radial nerves. Following median or ulnar block, >85% of the moves were matched; combined median and radial blocks only reduced this to 80%. These results show that, in the hand, a high degree of proprioceptive function can persist not only after loss of specific modalities of sensory feedback but also after significant loss of combined modalities, with information being derived from limited intrinsic and/ or extrinsic sources.1. Proudlock F and Scott JJA. (1998) J. Neurosci. Meth. 85: 153-159.

P 34-6 SOMATOSENSORY EVOKED POTENTIALS IN RABBITS DEPEND ON SHORT LASTING FUNCTIONAL STATES OF THE BRAIN M. Eiselt, L. Flemming, F. Gießler, J. Haueisen, U. Zwiener Stimulation of somatosensory nerves causes characteristic potentials (SEP) at the cortical surface. We suppose that the characteristics of individual cortical potential depend on the functional micro-state of the brain. We investigated 8 White New Zealand rabbits using 16channel ECoG recordings with high spatial resolution. Animals were anaesthetized by continuous infusion of Ketamin 24-30 mg/kg*h and Xylazin 2.4-3 mg/kg*h. Skull bone was removed over an area from -5 to 4 mm relativ to the Bregma over the left hemisphere. A 4*4 electrode grid (interelectrode distance 1.25 mm) was placed over the left somatosensory cortex (S1). Direct stimulation of the right medial or tibial nerve was performed (0.2-0.4 mA; interstimulus interval 503 ms; stimulus duration 200 µs; 1000 stimulations). The ECoG was continuously recorded (filtered 0.3-300 Hz), digitized (2 kHz) and stored on disk. SEPs were averaged depending on the background activity. Two states of background activity were distinguished. State 1: ECoG activity in the frequency band 8-14 Hz >30 µV; State 2: ECoG activity in the same frequency band <30 µV. The SEP consisted of fast (P1 and N1) and slow components. The latency of potential P1 depended on the nerve stimulated (medial nerve: 17±2 ms, tibial nerve: 21±3 ms). The potential N1 was recorded at 27±2 ms (median nerve) and 31±3 ms (tibial nerve), respectively. The maximum of the slow positive potential appeared at about 100 ms. State 1 occurred in short (< 30 ms) and long duration episodes (190±5 ms) and the mean duration of state 2 was about 850±400 ms. Approximately 17±10% of the whole recording time was occupied by state 1. There were no significant differences concerning the amplitude and the latency of P1 and N1 between state 1 and 2. However, slow potentials were significant only during state 1. According to our results characteristics of SEP partially correlate with cortical functional states. Early SEP components seem to be independent from these states but slow positive potentials were only detected if 8-14 Hz cortical oscillations exceeds a threshold. Pathophysiol., FSU Jena, 07740 Jena; supp. by TMWFK, EU IST10378 P 34-7 PROPRIOCEPTION IN THE HAND FOLLOWING PERIPHERAL NERVE BLOCKS J.J.A. Scott, F. Proudlock, J. Dias, R. Sharma Following peripheral nerve lesions, the patterns of proprioceptive loss in the hand are complex due to the distributions of the median, ulnar and radial nerves. As part of a study into recovery following nerve injuries, we have investigated proprioception in the hand following peripheral nerve blocks. The experiments were performed on 7 informed volunteers, using protocols approved by the LUHT ethical committee.

Dept of Pre-Clinical Sciences, University of Leicester, PO Box 138, Leicester, LE1 9HN UK P 34-8 SPATIAL DISCRIMINATION OF ONE VERSUS TWO TEST STIMULI IN THE HUMAN SKIN I. Martikainen, A. Pertovaara Previous studies indicate that localisation of painful heat and mechanical stimuli in the skin may be as accurate as that of tactile stimuli. In the present study we determined whether the good ability to localise painful stimuli also applies to cold pain and whether the ability to localise a single test stimulus is associated with the discriminability of two concurrent signals from each other. Therefore, accuracy of localisation and two-point discrimination ability were assessed in the forearm of healthy humans with innocuous mechanical and noxious cold stimuli. Accuracy of localising single test stimuli was significantly better in the transverse direction than in the longitudinal direction, independent of modality. Moreover, the localisation was slightly better for noxious cold than for innocuous mechanical stimuli. In contrast, the two-point discriminability was an order of magnitude more accurate with innocuous mechanical than with noxious cold stimuli. According to the present results localisation of single cold pain stimuli in the skin is at least as good, if not better, as that of single tactile stimuli. In accordance with earlier results, the localisation of both pain and touch is better in radial-ulnar than proximal-distal direction. In contrast to localisation of single stimuli, the discrimination of two stimuli from each other is considerably worse for cold pain than for tactile stimuli. The similar capacity to localise single pain and tactile stimuli contrasts with a large difference in the ability to discriminate two tactile versus pain stimuli from each other. This dissociation indicates that localisation of single cutaneous stimuli and discrimination of two cutaneous stimuli from each other are based on different mechanisms. From the evolutionary point of view, both a good localisation and an accurate information about the spatial features of the external object are important for the tactile sense, whereas the localisation of the harmful stimulus is sufficient spatial information for the pain sense. This corresponds with the present findings. Dept.Physiol., Univ. Turku, Kiinamyllynkatu 10, FIN-20520 Turku, Finland P 34-9 NO EVIDENCE FOR DORSAL ROOT REFLEXES IN HUMAN SKIN M. Klede*, L. Kohllöffel, M. Schmelz The existence of dorsal root reflexes in human skin are controversially discussed. Under the condition of spinal sensitization, activation

S 314 of low threshold mechanosensitive afferents cause pain instead of touch (allodynia) which might be attributed to excitation of unmyelinated nociceptors by presynaptical coupling. This excitation would lead to axon reflex vasodilation via antidromic conduction. We have reevaluated this hypothesis in a human model of electrically induced allodynia using the laser Doppler imaging technique. Transdermal electrical stimulation (1 Hz, 50 mA) was used to provoke large areas of allodynia (diameter 16 ± 1.5 cm) in 11 subjects. An anesthetic strip applied via intradermal microdialysis fibers (2% lidocaine, 5µl/min) close to the stimulation site prevented the spread of the concomitant axon reflex erythema beyond the strip, whereas allodynia developed symmetrically. In allodynic areas, which were not included in the axon reflex erythema, allodynia was elicited by gently stroking with a cotton swap, either once or for 10 s. The vasodilatory effect was quantified by laser Doppler imager for 15 s before and after stimulation at a distance <5mm. The subjects gave a pain ratings on a numeric scale from 0 to 10 for each mechanical challenge. Mechanical stimulation with one stroke or for 10 s induced pain ratings of 1.45 ± 0,09 out of 10, but no increase in vasodilation, on either side of the lidocaine strip. We found no correlation between the pain rating during stimulation and the difference between superficial blood flow before and after stimulation. The electrical stimulation in combination with the lidocaine strip is a suitable model to test vasodilation in an allodynic area without interference of the flare reaction. We conclude that even with very sensitive methods of laser Doppler imaging we could not confirm the existence of dorsal root reflexes in human skin in an electrically induced allodynic areas. Dept. Physiology, University of Erlangen, 91054 Erlangen, Germany P 35-1 GALANIN-IMMUNOREACTIVE RAT DRG NEURONS RESPONSIVE TO HEAT BUT NOT TO CAPSAICIN UNDER CONDITIONS LACKING NGF AND GDNF S. Engert, J.R. Wendland, M. Petersen Behavioral studies in galanin (GAL) knock-out mice suggest an involvement of GAL in thermal hyperalgesia. In rat, after nerve injury, GAL is upregulated while the heat transducing receptor VR1 is downregulated in DRG neurons. Here, we investigated on the cellular level GAL expressing neurons with regard to capsaicin and heat sensitivity. Isolated DRG neurons from adult rats were cultured for up to 6 d in supplemented F12 medium lacking NGF and GDNF. In doublelabeling experiments, capsaicin (1µM) or heat (44°C) sensitivity was functionally identified using cobalt-influx method while GAL was detected immuno-cytochemically. With time in culture, the proportion of capsaicin as well as heat responsive neurons decreased, whereas the proportion of GAL immunoreactive neurons increased. We then investigated colocalization of GAL immunoreactivity with heat- or capsaicin-induced cobalt uptake. Of the heat responsive neurons, at day 1 in culture 17% also were galanin immunoreactive. This proportion increased to 30% at day 2 and remained at this level until day 6. Of the capsaicin sensitive neurons at days 1 and 2, 15% also were GAL immunoreactive. In contrast to the heat experiments, the proportion decreased to 5% at day 3 and remained at this level until day 6. The soma size distribution profile of heat responsive neurons was different from the one of capsaicin responsive neurons. Co-stimulation with the PKC activator PMA at day 6 yielded an increase in the proportion of heat responsive neurons but not of capsaicin responsive neurons. In summary, we found several indications for a population of DRG neurons responsive to noxious heat but not to capsaicin: (i) differences in colocalization with GAL, (ii) different soma size profiles and (iii) differential activation by PMA. This population was present under cell culture conditions in absence of NGF and GDNF. (Supported by SFB 353 to MP.) Institute of Physiology, 97070 Würzburg, Röntgenring 9, Germany P 35-2 EFFECTS OF GABAPENTIN ON FAST SYNAPTIC TRANSMISSION IN THE MOUSE SPINAL CORD DORSAL HORN K. Bayer, H.U. Zeilhofer Gabapentin is a rather novel lipophilic analog of γ-aminobutyric acid, which is therapeutically used against certain forms of epilepsy and

neuropathic pain. Despite its structural similarity to GABA, it does not bind to GABA A or GABAB receptors. Instead, it exhibits high affinity binding to α2δ subunits of voltage-activated calcium channels. It is at present still unclear whether and how this binding relates to its antinociceptive action. Here we have investigated the effects of gabapentin on fast synaptic transmission in the mouse spinal cord dorsal horn, a site, which is believed to be critically involved in the antinociceptive action of gabapentin. Whole-cell patch-clamp recordings were made from visually identified neurons in the superficial layers of the dorsal horn. Transverse 250 µm thick slices of the lumbar spinal cord were prepared from 7 to 14 day old BalbC mice. Excitatory glutamatergic and inhibitory glycinergic postsynaptic currents (EPSCs and IPSCs) were evoked by extracellular electrical stimulation and recorded in isolation using appropriate combinations of APV (20 µM), CNQX (10 µM), bicuculline (10 µM) and strychnine (0.3 µM). Gabapentin reversibly inhibited EPSCs mediated by α-amino-3-hydroxy5-methyl-4-isoxazole propionic acid receptors (AMPA-EPSCs) and N-methyl D-aspartate receptor-mediated EPSCs (NMDA-EPSCs). At a saturating concentration of gabapentin (1 µM), the average reduction in PSC amplitudes was 39.9 ±3.5%, n=7 (mean±SEM) and 35.7 ±5.7%, n=7 (mean±SEM), for AMPA-EPSCs and NMDA-EPSCs, respectively. Glycinergic IPSCs were reduced by 41.6 ±4.9%, n=7 (mean±SEM). Inhibition of NMDA-EPSCs was unaffected by addition of D-serine (100 µM), an agonist at the glycine binding site of the NMDA receptor, indicating that inhibition of NMDA-EPSCs was not secondary to a reduction in glycine release. To distinguish between a pre- and a postsynaptic site of action we investigated the effects of gabapentin on spontaneously occurring miniature AMPA-EPSCs and on glycinergic miniature IPSCs. Gabapentin increased the frequency of both types of miniature PSCs (mPSCs). It had no effect on mAMPA-EPSC amplitudes, but caused a slight, statistically insignificant increase in the amplitudes of glycinergic mIPSCs. Neither of the effects was prevented by addition of the GABAB receptor antagonist CGP55845 (1 µM). Furthermore currents elicited by exogenous application of aspartate (250 µM), which was used to activate NMDA receptors, glutamate (1 mM) and glycine (1 mM) were unaffected by gabapentin (1 µM). There results suggest that gabapentin affects synaptic transmission in the spinal cord dorsal horn via a presynaptic effect independent of GABAB receptor activation. The effects of gabapentin on transmitter release may possibly be mediated via a direct interaction of gabapentin with the α2δ subunits of presynaptic calcium channels. Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Universität Erlangen-Nürnberg, Fahrstrasse 17, D-91054 Erlangen, Germany P 35-3 RAT GROUP III AND IV MUSCLE RECEPTORS BUT NOT GROUP I RECEPTORS ARE RESPONSIVE TO ATP AND CAPSAICIN J. Reinöhl, U. Hoheisel, S. Mense Two types of free nerve endings processing nociceptive information are postulated to exist: peptide-containing nerve endings with capsaicin- and proton-sensitive vallinoid receptors, and peptide-free endings with ATP-sensitive P2X3 purinoreceptors. The present study investigates, whether both groups of neurone are present among group III and group IV muscle afferents. In deeply anaesthetized rats the activity of single nerve fibres innervating the gastrocnemius-soleus muscle was recorded, and ATP (concentration 7.6 mM) and capsaicin (concentration 655 µM) were injected intramuscularly into the receptive field of the units (25 µl or 50 µl). ATP was dissolved in tyrode, capsaicin in 70% ethanol and then diluted to its final concentration with 0.9% NaCl (1:10). 25 % of the group III and IV units showed clear responses to ATP, 33,3 % were capsaicin-sensitive, and 33,3 % were excited by both substances. In all cases, the response following ATP injection was more distinct than the response to capsaicin. ATP- and capsaicin-sensitive fibres were found among nociceptive and nonnociceptive neurones. After administration of capsaicin most of the units did no longer respond to mechanical stimulation. In contrast to group III and IV units, group Ia units (primary muscle spindle endings) did not react to ATP or capsaicin injection. The data indicate that nociceptive and non-nociceptive group III and IV muscle receptors are responsive to ATP and/or capsaicin while group I receptors lack any responsiveness to ATP or capsaicin. There is some indication that 1. there is some overlap between ATP- and capsaicin-sensitive endings and that 2. following administration of capsaicin the ending is

S 315 desensitized to further stimulation. Institut fur Anatomie und Zellbiologie, Universität Heidelberg, Im Neuenheimer Feld 307, D-69120 Heidelberg P 35-4 TEMPORAL SUMMATION IN REFLEXES EVOKED BY SELECTIVE STIMULATION OF SURAL NERVE C-FIBRES IN THE DECEREBRATED RABBIT R.W. Clarke, J. Harris, P. Kasher, K. Brown-Reid Treatment of rabbit sural nerve with tetrodotoxin (TTX) 30 nM abolishes transmission in myelinated axons while preserving most Cfibre conduction (R.W. Clarke, J. Harris and K. Brown-Reid, J. Physiol., in the press, 2001). After TTX, stimulation of sural C afferents in decerebrated rabbits evoked reflex responses in the ipsilateral medial gastrocnemius (MG) muscle nerve at latencies > 100 ms that were enhanced after spinalization. Reflexes were measured from averages of 8 stimuli applied at 1 Hz, repeated at intervals of 20 min. We have now analyzed responses to individual stimuli from these experiments to show the temporal evolution of C-fibre-evoked reflexes before and after TTX application, and subsequent spinalization. Experiments were performed on 7 rabbits decerebrated under halothane (2 - 3%): nitrous oxide (70%) anaesthesia. Stimuli were applied to the sural nerve and reflexes recorded from MG as described previously (Clarke et al., 2001). Reflex responses were measured in 3 time bands: 5 - 12 ms (phase 1); 12 - 100 ms (phase 2); and 100 - 260 ms (phase 3). Temporal evolution was assessed by calculating the linear regression line for sweep number against voltage-time integral for each phase of the reflex over 8 stimuli. The protocol was as described previously (Clarke et al., 2001). In the control state, all reflexes showed significant temporal facilitation (slopes ± s.e.m. 35 ± 5, 27 ± 2 and 41 ± 3 µV ms stimulus-1 for phases 1 - 3 respectively). After TTX 30 nM, no significant responses remained in phases 1 and 2, but the phase 3 response still showed significant temporal facilitation (slope of regression line 38 ± 5 µV ms stimulus-1) that was not significantly different from that seen in the control state. After spinalization, the slope of the regression line became significantly negative (-6 ± 2 µV ms stimulus-1). Thus, temporal facilitation of reflexes evoked by selective stimulation of C-fibres is unaffected by activity in A-fibres. Division of Animal Physiology, School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough LE12 5RD, UK P 35-5 CLASSICALLY CONDITIONED WITHDRAWAL REFLEXES DURING STANCE W. Föhre, Th. Kaulich, D. Timmann, F.P. Kolb Our previous studies on the classically conditioned withdrawal reflex were performed with the subject laying on a daybed. The responses were found primarily in the stimulated leg. After applying a painful stimulus however, the complete withdrawal reflex pattern should be observable in all extremities. There is some evidence that the withdrawal reflex pattern may depend on the posture of the subject. The aim of the current study was to investigate the reflex under more natural conditions in a group of young volunteers. To elicit the withdrawal reflex, a train (100 ms) of current pulses was applied via surface electrodes to the subject’s medial plantar nerve as an unconditioned stimulus (US). Coterminating with, but preceding this by 450ms, a tone was presented via headphones as the conditioned stimulus (CS). Subjects stood on a platform, equipped with strain-gauges to record vertical ground forces. The center of vertical pressure was communicated to the subject via an oscilloscope screen. Physiological reactions were recorded biomechanically by measuring the vertical ground forces separately for each leg and electrophysiologically by EMG measurements of different leg muscles. The positions of the ankle and the knee joint were recorded by an ultra sonic system (Zebris). A session consisted of 50 US-alone trials, 80 trials with paired stimuli, and a final block of 10 US-alone trials. Normal subjects could be conditioned within one session. CRs were established between trials 25-60 with amplitudes of up to 60% of the normalized UR that was derived from the averaged value. CR were found in muscles of both legs with different latencies. The process of establishing the CR generally started fairly abruptly and was followed by a continuous optimization of the responses, i.e. an

increase of CR incidence, a decrease of CR latency, and an increase of CR amplitude. Our data suggest that the plastic process responsible for conditioning is not focused primarily on the leg stimulated but rather on the whole body, to permit optimal reaction to the perturbation. Supported by W. Sander Foundation (94.090.1-3). Institute of Physiology, University of Munich; Pettenkoferstr. 12, 80336 München P 35-6 SPINAL CORD STIMULATION INHIBITS LONG-TERM POTENTIATION OF SPINAL WDR NEURONS J. Wallin, A. Fiskå1, A. Tjølsen1, B. Linderoth, K. Hole1 Objective of study. Long-term potentiation (LTP) of dorsal horn neurons has been suggested to be a phenomenon that contributes to the development of chronic pain. Spinal cord stimulation (SCS) may be an effective tool in alleviating neuropathic pain. The aim of this study in rats was to examine if SCS suppresses LTP of dorsal horn neurons. Method. Monopolar SCS was applied via a miniature epidural disc electrode (frequency 50 Hz, pulse width 0.2 ms, stimulation intensity individually set to 90% of the motor threshold). Following implantation of the SCS electrode, recording of activity of single dorsal horn WDR neurons evoked by stimulation of the sciatic nerve was performed. The baseline response was calculated when 8 test stimuli given every 4 min for 28 min had been given. Spikes appearing 0-80 ms and 80-300 ms after stimulus were defined as A- and C-fiber responses, respectively. After the baseline was established, LTP was induced using a conditioning stimulus given directly to the sciatic nerve (6 x C-fiber threshold, 20 trains of 2 s duration, 100 Hz, 0.5 ms rectangular pulses, 10 s interval between trains). Results. The conditioning stimulus induced a significant increase in the C-fiber response to test stimuli (LTP). SCS reduced this response significantly (p<0.05) 32-48 min after cessation of SCS. A-fiber-induced firing responses were at no time point significantly different from baseline. Conclusions. SCS inhibits the increased C-fiber response following high-intensity electrical stimulation of the sciatic nerve in the rat. These data suggest that SCS affects the C-fiber component of dorsal horn central sensitization. It is possible that the inhibitory effect of SCS is related to changes in chemical transmission in the spinal dorsal horn. We believe that there may be two different mechanisms through which SCS exerts this effect: by antidromic activation of the dorsal columns and by orthodromic activation of a dorsal column-brainstem-spinal loop. Department of Clinical Neuroscience, Section of Neurosurgery, Karolinska Institutet, S-17177 Stockholm, Sweden and 1Department of Physiology and Locus on Neuroscience, University of Bergen, 5009 Bergen, Norway P 35-7 SENSITIZATION OF THE LOW THRESHOLD HEAT-GATED CURRENT IN CHICK SENSORY NEURONS BY ACIDIFICATION OF THE EXTRACELLULAR MEDIUM A. Klusch, A. Schmalfuß, M. Petersen VR1 is a multifunctional receptor transducing noxious heat, capsaicin and protons in mammals. Some animals like birds are insensitive to capsaicin but sensitive to heat, suggesting the expression of a heat transducing receptor different from the VR1. Recently, we described a low threshold heat response in chick DRG neurons. The proposed receptor showed both similar and distinct properties compared to the VR1. Here, we investigated whether the heat response in isolated chick DRG neurons can be sensitized by protons. Experiments were performed on dissociated DRG neurons from chick kept in culture for up to 3 days in supplemented F12 medium with NGF. Whole-cell patch-clamp and cobalt uptake methods were used. In patch-clamp experiments, in a subpopulation of neurons heat stimuli of up to 52°C elicited inward currents in a temperature-dependent manner. Acidification of the bath solution augmented the heat-evoked current. The sensitization of the heat response to a stimulus of 50°C plateaued at pH 6.2 with an approx. 2.5fold increase in current amplitude compared to pH 7.3. Repeated application of heat in an acidic solution resulted in a small decrease in current amplitude. In cobalt uptake experiments, the proportion of neurons responsive to heat increased with time in culture. Acidification of the medium (pH 6.6, 6.2, 5.8) at different temperatures (RT, 37°C, 44°C) caused a change in the pro-

S 316 portion of responsive neurons. The maximum increase was reached at 37°C in an extracellular medium of pH 6.2. At 44°C the increase in proportion with respect to pH 7.3 was much smaller and it was not different between pH 6.6 and 5.8. In summary, the heat response in chick DRG neurons can be sensitized by extracellular acidification of the medium, suggesting a protonation of the proposed low threshold heat receptor. Supported by SFB 353 to MP. Institute of Physiology, 97070 Würzburg, Röntgenring 9, Germany P 35-8 EARLY ONSET OF PAIN LIKE BEHAVIOUR AND ABNORMAL ACTIVITY IN INTACT FIBRES AFTER THE INJURY OF NEIGHBOURING NERVES C. Constantin, N. Gorodetskaya, W. Jänig To investigate the mechanisms of neuropathic pain following nerve injury, animal models with experimental nerve lesions are used. It is presumed that signals from both lesioned and intact afferent neurones contribute to the development and maintenance of abnormal pain behaviour. Here we used the Spared Nerve Injury (SNI) model (Decosterd & Woolf, 2000) in order to answer the following questions: 1) Which components of pain-like behaviour are exhibited by rats with SNI? 2) Do intact sural afferent nerve fibres develop abnormal discharge properties? 3) Do the abnormal discharge properties change with time after the nerve lesion (1day vs 1week )? Nerve lesions (ligation and cutting of tibial and common peroneal nerves, sparing sural and saphenous nerves) and neurophysiological experiments were done under general anaesthesia (Pentobarbital, 60 mg/kg ip or iv) on 7 male Wistar rats. For neurophysiological experiments rats were immobilised (Pancuronium, 1mg/kg, iv) and artificially ventilated. Single A- and Cfibers in fine filaments teased from the intact sural nerve were identified electrically and tested for spontaneous activity (SA) mechano- (MS), cold- and heat-sensitivity (CS, HS). All animals were tested for their pain behaviour to the same stimuli applied to the sural nerve territory 3 times before SNI and every second day within one week post lesion. After SNI lesion all rats tested developed strong mechanical and cold allodynic behaviour in less than 24 hours. The abnormal pain-like behaviour was maintained over the whole period of testing. One hundred twenty two A-fibres and 46 C-fibres which could be activated from the skin and/or exhibited SA were investigated. Among these 22% of the A- and 87% of the C-fibres had SA. Some A-fibres from hair follicles which normally are silent exhibited SA and afterdischarges following mechano stimulation. There was no significant difference concerning SA, MS and CS between both time periods. These results show that, following SNI, SA and abnormal evoked activity developed in some intact A- and C-fibres starting within the first day post-lesion and were maintained for one week post-lesion. Because the early onset of SA in intact nerve fibres after lesion of the neighbouring nerves occured within the same time window in which mechanical and cold allodynic behaviour appeared we suggest that the abnormal activity in these intact nerve fibres is involved in triggering and maintenance of abnormal pain behaviour.

male, 4 female) only on normal skin and subtracted PET scans during repeated rapidly-ramped painful heating to 50 ºC from scans during non-painful heating to 40 ºC. Results: Despite similar pain intensities across skin conditions in experiment 1, the Z-score map shows that heat allodynia preferentially activates the medial thalamus, anterior insula, perigenual anterior cingulate, orbital, medial and dorsolateral prefrontal cortices and putamen of both hemispheres, and the dorsomedial midbrain. In contrast, the Z-score map in experiment 2 shows that normal heat pain activates primarily the lateral thalamus, lenticular nucleus, mid/posterior insula/SII, mid/anterior cingulate cortex, vermal and paravermal cerebellum bilaterally, and the contralateral sensorimotor and premotor cortex. Conclusion: Dissociation of heat pain, represented by a lateral thalamic pathway to the parietal lobe, from heat allodynia, represented by a medial thalamic pathway to the frontal lobe, suggests forebrain anatomical differentiation of intensity and psychophysiological context of a pain state. Supported by Max Kade Foundation. *Institute of Physiology, University Clinic Eppendorf Hamburg, University of Hamburg P 35-10 ATTENTIONAL MODULATION OF PAIN PROCESSING IN THE SECONDARY SOMATOSENSORY CORTEX Y. Nakamura, R. Paur, R. Zimmermann, B. Bromm The influence of attention on the processing of phasic pain in the secondary somatosensory cortex (SII) was analysed by means of multilead magnetoencephalograms (MEGs) in response to painful infrared heat stimuli applied to the left hand back in 7 male healthy subjects, 22 - 28 y.o.a. Three paradigms were chosen which required different levels of attention in a mixed task design: normal attention was achieved by our standard protocol with randomized intensities and interstimulus intervals which keeps the vigilance constantly high (Scharein and Bromm, Pain Rev. 5, 216-26; 1998); a weak tone given 3 s after each heat pulse prompted the subject to rate the strength of induced pain. In the no attention task the subjects were asked to ignore the stimuli. High attention was achieved by presenting tones of different frequencies which preceded the heat pulses, and subjects had to decide, whether the matched frequency fits pain intensity; correct decisions were paid to enlarge the compliance. This strategy was thought to deliver attention dependent results under comparable levels of vigilance. Single fixed dipole sources for the pain-evoked responses were calculated in the individual cortex anatomy which was determined by magnetic resonance tomography. Though the pain stimuli followed the same intensity pattern in all paradigms, pain-induced SII activity increased markedly from the no attention to the normal attention task (p < 0.001). In contrast, further increase of attention from normal to high was not accompanied by an additional enhancement of SII activity. These findings support the view that activation patterns of SII follow a saturation function which soon reaches a maximum which cannot be enlarged by increasing the relevance of the pain-inducing stimuli. Supported by the Deutsche Forschungsgemeinschaft (Br 310 20-2).

Physiologisches Institut der CAU, Olshausenstraße 40, 24118 Kiel

Institut für Physiologie, Universitätsklinikum Hamburg-Eppendorf, Germany

P 35-9

P 35-11

DISSOCIATION OF HEAT PAIN AND HEAT ALLODYNIA IN THE HUMAN BRAIN J. Lorenz*, D.J. Cross, S. Minoshima, T.J. Morrow, P.E. Paulson, K.L. Casey Objective: Pain after innocuous stimuli is called allodynia. Heat evoked allodynia within inflamed tissue is generally believed to be caused by a lower response threshold of the same peripheral nociceptors and central projection systems that mediate heat pain. By adjusting intensities of a heat stimulus on normal and capsaicin-treated skin to equal pain intensities we aimed to test whether the forebrain responses of both phenomena differ. Methods: Experiment 1: To image heat allodynia, we subtracted positron emission tomography (PET) scans obtained during slowly-ramped (0.9 ºC/s) painful heating of normal skin to 2 ºC above the individual’s heat pain threshold (HPT) from scans during equally intense but normally innocuous heating (HPT – 2 ºC) of capsaicin-treated skin in 14 healthy male volunteers. Experiment 2: To image heat pain, we tested another group of 14 healthy volunteers (10

CORTICAL REPRESENTATION OF NOCICEPTION FROM VEINS IN HUMANS H. Holthusen*, M. Ploner, P. Noetges, A. Schnitzler Deep pain and pain from skin differ with respect to their affective component reflecting different subcortical processing. It is not clear, however, whether nociceptive information arising from superficial (ectodermal) or deep (mesodermal) tissue is processed in different cortical areas in humans. We therefore evoked pain from skin and underlying veins by laser stimulation and compared the respective cortical representation by whole-head magnetoencephalography. Stimulation: In six volunteers 40 laser stimuli (Tm:YAG; 1 ms, 2000 nm) at each were delivered to the skin of the dorsum of the hand (interstimulus intervals 10-14s) or intravenously to the inner wall of a dorsal hand vein (intervals 30-60s); stimulation sites were <5 cm apart. Stimulus intensity was adjusted to evoke pain intensities between 60-80% VAS on a visual analogue scale (0% = no pain, 100% = maximally tolerable pain). Measurements: Cortical activity was

S 317 recorded with a Neuromag-122 whole-head neuromagnetometer. Only sources of responses with 95% confidence limits of source localisation <10 mm were accepted. Sources were superposed to individual MRI scans to determine anatomical position. Both, cutaneously (600 mJ) and intravenously (500 - 800 mJ) applied laser stimuli evoked sharp, well-localised pain of intended intensities between 60 and 80% VAS. Three cortical sources were found to be activated: two contralaterally (parietal, SI; temporoparietal, SII), and one ipsilaterally (temporoparietal, S II). Field patterns did not differ between the stimulation sites. However, responses from veins arrived significantly earlier (p<0.05 two-tailed student’s t-test), which might be due to differences in peripheral conduction velocity. Venous nociception shares the cortical representation of cutaneous nociception in human somatosensory cortices. Thus, the cortical representation of nociceptive processing from tissues of mesodermal and ectodermal origin appears to be similar. [email protected], Universitätsklinikum Düsseldorf, Klinik für Anaesthesiologie, Moorenstraße 5, 40225 Düsseldorf, Germany P 35-12 TNF-α α INDUCES ACTIVITY IN AXOTOMIZED AFFERENTS IN THE FIRST HOURS AFTER SPINAL NERVE L5 LESION S. Eschenfelder, D. Zöllner, H.-J. Häbler Ectopic activity which develops in some primary afferent neurones about 7 to 24 hours after spinal nerve L5 injury in rats is thought to trigger and maintain neuropathic pain behaviour. Mechanisms leading to these spontaneous discharges are unclear. Here we studied whether TNF-α influences onset and time course of spontaneous activity immediately after nerve lesion. In 12 male Wistar rats under pentobarbital anaesthesia a lumbosacral laminectomy was performed, the dura was opened and activity was recorded from fine filaments isolated from the left dorsal root L5. While recording the left ventral ramus of the spinal nerve L5 was ligated and cut. Immediately after cessation of the injury discharge, TNF-α (5 ng, 100 ng/ml, group 1, n=5 or 2.5 ng, 50 ng/ml, group 2, n=4) was applied locally at the lesion site. The remaining rats received Tyrode’s solution and served as control. Nerve filaments were recorded for up to five hours after TNF-α application to test for the generation of spontaneous activity. In the first group ectopic discharges occurred 5-15 minutes after application of TNF-α in 4 of 10 initial filaments. In 159 other filaments (63.3 %, n=250) spontaneous activity with a discharge rate of 9.5±0.4 imp/s (mean±SEM, n=198) was recorded. In the second group only 2 of 10 initial filaments showed active neurones which started to fire ectopic impulses 12-20 minutes after TNF-α. In another 84 filaments (43.3 %, n=194) spontaneous activity with a slightly lower discharge rate of only 7.3±0.3 imp/s (mean±SEM, n=98) was found. In both groups neuronal activity which was almost absent in the control animals was recorded only from A-fibres and was rather regular or occurred in bursts. These results show that local application of TNF-α to the lesion site of the L5 spinal nerve can prematurely induce spontaneous activity in axotomized afferents. Therefore, local inflammation may be involved in triggering and maintaining neuropathic pain behaviour in rats. Physiologisches Institut, Christian-Albrechts-Universität, Olshausenstraße 40, D-24098 Kiel, Germany P 35-13 ACTION OF ANANDAMIDE AND NOXIOUS HEAT IN PRIMARY NOCICEPTIVE NEURONS OF RATS W. Greffrath1, T. Fischbach1, H. Nawrath2, R.-D. Treede1 Noxious heat stimuli increase cytosolic calcium ([Ca2+]i) in primary nociceptive neurons of rats by activation of the native vanilloid receptor subtype 1 (VR1). The endogenous eicosanoid derivative anandamide (AEA) was suggested to be the endogenous ligand of VR1, but also of the cannabinoid receptor 1 (CB1) which is coexpressed with VR1 in nociceptive neurons. We thus investigated the effects of both, AEA and noxious heat in acutely dissociated dorsal root ganglion (DRG) neurons of rats. [Ca2+]i was measured microfluorimetrically using FURA2 in small DRG neurons (diameter ≤30 µm) in response to application of AEA (1 nM - 100 µM), noxious heat (about 48 °C for 10 s), or both. About 75 % of the small DRG neurons tested were activated by AEA (≥10 µM). AEA-induced [Ca 2+] i -transients were completely blocked in Ca2+-free solution. Amplitudes of AEA- and of heat-induced

transients were positively correlated in heat-sensitive neurons (r = 0.753, n = 36; P < 0.001). The ED50 of AEA-induced [Ca2+]i-transients was about 10 µM (log 10 ED50 = -4.98 ± 0.001), the maximum effect was an increase of fluorescent ratio by 1.15 (264 ± 69 %). When AEA was co-applied with the CB1-antagonist AM-251 (10 µM), the effect of 3 µM AEA was increased. These results suggest that AEA may cause biphasic effects in small DRG neurons: AEA at lower concentrations activates both, VR1- and CB1 receptors, whereas AEA at higher concentrations predominantly induces excitation of primary nociceptive neurons. Supported by the Deutsche Forschungsgemeinschaft (Tr 236/11-2). 1 Institute of Physiology and Pathophysiology, Saarstr. 21, D-55099 Mainz; 2Institute of Pharmacology, Obere Zahlbacher Str. 67, D-55101 Mainz

P 36-1 ALTERED HIGH-FREQUENCY NETWORK OSCILLATIONS AND ATTENUATED EPILEPTIFORM DISCHARGES IN THE HIPPOCAMPUS OF CONNEXIN36-DEFICIENT MICE N. Maier, M. Güldenagel, G. Söhl, H. Siegmund, K. Willecke, A. Draguhn Many neuronal networks generate coherent oscillations at different frequencies, depending on the functional / behavioural state of the animal. At high (> 20 Hz) frequencies, electrical synapses (gap junctions) have been shown to contribute to the synchronisation of neurones during such rhythms. Here, we made use of a targeted disruption of one neuronal connexin (a constituent of gap junctions) to analyse the role of electrical transmission in high-frequency oscillations (~200 Hz “ripples”) and in pathologically synchronised events (4aminopyridine-evoked field discharges) in the mouse hippocampus. In extracellular recordings from hippocampal slices of both wild-type (WT) and Cx-36-deficient (Cx36 -/-) mice we found spontaneous ~200 Hz oscillations on top of slower potential shifts, reminiscent of sharp waves in vivo. In slices from Cx36 -/- mice, the occurrence of sharp waves was reduced to 36% of controls and ripple episodes could be detected at 65% of the frequency observed in WT mice. The structure within the sharp wave-ripple complexes was largely unaltered. Moreover, epileptiform field activity elicited by addition of the K + channel blocker 4-aminopyridine (100 µM) were less severe in Cx36 /- tissue. Several other functional parameters (population spike amplitudes, paired-pulse behaviour, tetanically evoked gamma oscillations) were unchanged by deletion of Cx36. We conclude that gap junctions containing Cx36 contribute to the generation of network discharges reminiscent of sharp wave-ripple complexes and to 4-AP-induced epileptiform activity in vitro. Johannes-Müller-Institut für Physiologie der Charité, Tucholskystr. 2, 10117 Berlin and Institut für Genetik der Universität, Römerstr. 164, 53117 Bonn P 36-2 EFFECTS OF GLYCINE AND TAURINE ON LOW MAGNESIUM INDUCED EPILEPTIFORM ACTIVITY IN COMBINED ENTORHINAL CORTEX-HIPPOCAMPAL SLICES OF ADULT RATS A Kirchner, J. Breustedt, V. Schmieden, U. Heinemann Epileptic discharges predominantly result from an imbalance of excitation and inhibition. In the CNS inhibition is mediated by ionotropic GABAA and glycine receptors. Many clinically used anticonvulsants act by enhancing inhibition through the GABA A receptor. We were interested, whether activation of the inhibitory glycine receptor by the agonists glycine and taurine is also capable of suppressing epileptic discharges. Epileptic activity was introduced in combined enohinal cortex-hippocampal slices of adult rats by lowering the magnesium concentration in the bathing medium. Extracellular field potentials were recorded in stratum pyramidale of area CA1 and in layer III of the medial entorhinal cortex (mEC). Three distinct patterns of activity were observed: In the mEC seizure like events (SLEs) and late recurrent discharges (LRDs), in area CA1 recurrent short discharges (RSDs). In a concentration of 10 mM both glycine and taurine almost completely (95±5 and 100%) reduced the frequency of the RSDs in a reversible manner in area CA1. In a lower concentration of 5 mM taurine still reduced the RSDs by 30 %, whereas 5 mM glycine had no effect. In the mEC taurine abolished the SLEs with a concentration of

S 318 2mM. In comparison, application of glycine in concentrations of up to 10 mM did not reduce the SLEs. Instead, SLEs were converted into late recurrent discharges upon wash of glycine. The LRDs of the mEC were blocked by glycine and taurine in a concentration dependent fashion. We obtained an IC50-value of 4.5 mM for glycine and 1.9 mM for taurine. Combined application of glycine and 0.2-1 mM sarcosine - an inhibitor of the glycine transporter 1 (GLYT1) which is expressed in the hippocampus - did not lead to an increase of the blocking effect of glycine. We propose that the receptor responsible for mediating the effect of glycine and taurin may be the inhibitory glycine receptor. The slightly greater efficacy of taurin in comparison to glycine may be due to a combined action of taurine on GABAA and glycine receptors. The data presented here point to the possibility that glycine receptors may serve as a target for the development of new antiepileptic drugs. Especially the effects on LRDs seem interesting since this activity pattern represents a model of difficult to treat status epilepticus. Johannes-Müller-lnstitut für Physiologie, Universitatsklinikum Charite, Humboldt-Universitat Berlin, Tucholskystrasse 2, 10117 Berlin P 36-3 EFFECT OF REVERSIBLE HIPPOCAMPAL INHIBITION ON AMYGDALOID KINDLED SEIZURES AND AMYGDALOID KINDLING RATE M. Alasvand Zarasvand, J. Mirnajafi-Zadeh, M. Mortazavi, Y. Fathollahi, M.R. Palizvan In this study effect of reversible inhibition of CA1 region of dorsal hippocampus by lidocaine on amygdaloid kindled seizures and amygdala kindling rate was investigated. In experiment 1, lidocaine (1% and 2%) was infused (1 µl/2 min) into the hippocampus of amygdala kindled rats bilaterally and animals were stimulated at 5, 15 and 30 min after drug injection. 24 h before lidocaine injection, saline was also infused (1 µl/ 2 min) into the hippocampus as control. Obtained results showed that afterdischarge duration was reduced 5 min after lidocaine (1% and 2%) injection. Stage 5 seizure duration was also decreased 5 and 15 min after 2% lidocaine. In experiment 2, rats were divided into four groups. In group 1, animals were implanted only with a tripolar electrode into the amygdala but in groups 2-4, two guide cannulae were also implanted into the CA1 regions of the dorsal hippocampi. Animals were stimulated daily to be kindled. In groups 3 and 4, saline or 2% lidocaine (1 µl/2 min) were also injected respectively into the hippocampus, 5 min before each stimulation. Results obtained showed that amygdala kindling rate and the number of stimulation to receive from stage 4 to stage 5 seizure were significantly increased in group 4. Thus it may be suggested that in amygdala kindling, activation of the hippocampal CA1 region has an important role in seizure acquisition and seizure severity so that inhibition of this region results in decreasing of seizure severity and retards amygdala kindling rate. Department of Physiology, School of Medical Sciences, Tarbiat Modarres University, Tehran, I.R. Iran

P 36-4 EFFECTS OF THE ANTICONVULSANT VALPROIC ACID ON CARDIOMYOCYTE DIFFERENTIATION OF EMBRYONIC STEM CELLS H. Sauer, L. Na, H. Nau, J. Hescheler, M. Wartenberg The antivonvulsant valproic acid (VPA) exerts teratogenic properties and has been demonstrated to cause neural tube defects and malformations of the heart. In the present study the effect of VPA on the differentiaton of cardiomyocytes from pluripotent murine embryonic stem cells (ES cells) was investigated. VPA dose-dependently inhibited the development of spontaneously beating embryoid bodies grown from ES cells as well as of the extension of beating areas of cardiac cells as evaluated by anti α-actinin imunofluorescence. Furthermore, VPA significantly increased reactive oxygen species (ROS) as evaluated by the redox-sensitive dye dichlorodihydrofluorescein diacetate (H 2 DCF-DA). To evaluate whether the inhibition of cardiomyocyte differentiation was owing to oxidative stress exerted by VPA, the compound was coadministered with the free radical scavenger vitamin E. This treatment significantly resumed cardiomyogenic differentiation, indicating that VPA inhibits cardiomyogenesis of ES cells by inducing severe oxidative stress.

Department of Neurophysiology, University of Cologne, Robert-KochStr. 39, D-50931 Cologne, Germany P 36-5 TEMPERATURE DEPENDENT CHANGES OF BIOCHEMISTRY AND ELECTROPHYSIOLOGY IN ISCHEMIC BRAIN SLICES J. Wörfer, C. Greiner, E.-J. Speckmann1, H. Wassmann Objectives: Recent clinical controversies shed an ambigous light on therapeutic hypothermia under conditions of reduced cerebral perfusion. There may be competing effects of different temperatures on neuronal tissue itself and on cerebral vasculature. We tried to establish an in vitro model of electrophysiological and metabolic effects of modulated temperature on ischemic brain tissue which excludes systemic factors, e.g. local cerebral blood flow, which are difficult to account for in other systems. Materials and methods: Experiments were carried out using rat hippocampal brain slices (n=24) in an interface chamber. Electrophysiological detection of direct current (DC) and evoked potentials (EP) was complemented by in vitro microdialysis (rate: 1 sample/5 minutes) within the slices. Dialysates were tested for several amino acids including neurotransmitters (aspartate, glutamate, and GABA) by HPLC. Ischemia was simulated by simultaneous withdrawal of glucose and/or oxygen in both the gaseous and the aqueous phase. Results: To keep the slices stable, control experiments had to be performed at 34°C (n=9). If ischemia simulations were terminated 5 minutes after sudden depolarisation had indicated the breakdown of cellular membrane potential, these slices showed almost no recovery of EP, and microdialysis detected characteristic rises of several amino acids including excitotoxic transmitters. If temperature was reduced (30°C, n=8; 28°C, n=7), latencies of ischemic depolarisation were longer, changes of amino acid levels were smaller or altogether missing, and recovery of EP occurred. Conclusions: Our model demonstrates significant electrophysiological and biochemical effects of temperature modulation in vitro. Hypothermic neuroprotection within isolated neuronal tissue is clearly distinct from systemic side effects which are to be observed in vivo. Klinik und Poliklinik fur Neurochirurgie und 1 1nstitut fur Neurophysiologie, Universitätsklinikum Münster, Albert SchweitzerStraße 33, 48129 Münster

P 36-6 DIFFERENT SUSCEPTIBILITY OF SEGMENTAL SPINAL REFLEX PATHWAYS TO ISCHEMIC CONDITIONS IN THE CAT H. Steffens, H. Kolenda, J. Hagenah, E.D. Schomburg Susceptibility to ischemic conditions play an important role for an impairment not only of supraspinal but also of spinal nervous structures. Therefore, we investigated the susceptibility of different segmental spinal pathways to short-term ischemia by means of testing the transmission in monosynaptic and excitatory and inhibitory polysynaptic pathways. The experiments were performed on anaemically decapitated high spinal cats. Monosynaptic reflexes were recorded from ventral roots L7/S1. The gastrocnemius-soleus (GS), posterior biceps-semitendinosus (PBSt), quadriceps (Q), sural (Sur) and cutaneous branch of the superficial peroneal (SPC) nerve were prepared for electrical stimulation. Monosynaptic reflexes were elicited by electrical stimulation of the PBSt or GS nerve. Facilitatory or inhibitory FRA pathways were tested by conditioning effects from cutaneous afferents (SPC and Sur). Ia inhibition to PBSt was tested by stimulation of the Q nerve, and Renshaw inhibition was tested by antidromic stimulation of a part of the ventral roots L7/S1. Shortterm ischemia (1-10 min) was induced by snare occlusion of the aorta. In the beginning of ischemia, except for the Ia inhibition, all other tested inhibitions and the FRA excitation of PBSt decreased and ended before the monosynaptic reflex vanished. After the end of ischemia, the inhibitions started significantly later to recover than the monosynaptic reflexes. The FRA excitation started earlier to recover than the monosynaptic reflex. The early decrease of the conditioning effects in the beginning of the ischemia together with longer recovery times in the polysynaptic inhibitory pathways suggest, that oligo- or polysynaptic reflex pathways are more sensitive to ischemic conditions than monosynaptic ones. This would indicate a higher ischemic susceptibility of the small interneurones compared to the big motoneurones.

S 319 Institute of Physiology, University of Göttingen, Humboldtallee 23, D-37073 Göttingen, Germany

Department of Molecular Neurobiology, Max Planck Institute for Medical Research, Jahnstraße 29, D-69120 Heidelberg P 37-1

P 36-7 EEG COHERENCES DURING DIFFERENT HYPOXIC HYPOXIA COMPARED TO NORMOXIC CONDITIONS D. Wirth, J. Wendt, , H. Welsch, H.J. Volke, I. Schmitz, U. Buhss, M. Nehring, F. Noack, P. Dettmar, M. Rudolf Objective of study: The aim of the present study was to prepare the online use of intercorrelative criteria of the electroencephalogram (EEG) to employ them for the early recognition of central nervous activation and of an initial decrease of the central nervous performance willingness. The conceptional basis for the choice of 351 coherences was the anatomical connection of both brain hemispheres. Method: A suitable method to investigate these issues seems to be the comparison of reactions during normoxic conditions and conditions of hypoxic hypoxia. Comparative studies before and during strain caused by short term hypoxic hypoxia in simulated altitudes of up to 25000 ft. were conducted. Subjects: Three groups of subjects were employed. The first group were students from whom the EEG was recorded several times to investigate intraindividual stability and reliability of coherences in the spontanous EEG. The second group were young, untrained men who participated in an experimental, step-wise ascent to an altitude of 4000m and the third group were pilot aspirants during a training in the altitude-simulation chamber. Results: The results were variform and showed different dependencies including a fundamental strain dependent behavior. In the spontanous EEG, rhythmic oscillations in the range of minutes which were interindividually different and depend on the electrode site and the frequency band could be shown. Under conditions of a stay at higher altitudes without oxygen respiration the coherences decreased with increasing altitude depending on the electrode site and the frequency band as well as on the basic values. The results will presented in tables and functions of time. Conclusions: This study contributes to the knowledge of temporal and selective derivation modalities of the EEG and could be employed to formulate a more differentiated EEG derivation and analyzing standard depending on the task to discriminate the degree of strain. Technische Universität Dresden, Professur Biopsychologie, Fakultät Mathematik und Naturwissenschaften, Mommsenstr. 13, 01069 Dresden, Germany

P 36-8 D-2- AND 3-HYDROXYGLUTARIC ACID ACTIVATE NMDA RECEPTORS AND CONTRIBUTE TO NEURODEGENERATION V. Pawlak, S. Kölker, B. Ahlemeyer, J.G. Okun, F. Hörster, E. Mayatepek, J. Krieglstein, G.F. Hoffmann, G. Köhr The inherited metabolic diseases D-2-hydroxyglutaric aciduria and glutaryl-CoA dehydrogenase deficiency are characterized by progressive neurodegeneration in early childhood. In this study, we investigated the pathophysiological role of two metabolites, D-2- and 3hydroxyglutaric acid (D-2 and 3-OH), which accumulate in body fluids and brain tissue of affected patients. Since the organic acids D-2 and 3OH are structurally very similar to the main central excitatory neurotransmitter L-glutamate, we tested the hypothesis that excitotoxic mechanisms are involved in the pathophysiology of these diseases. Owing to its high Ca2+ permeability, the NMDA subtype of glutamate receptors plays a key role in mediating excitotoxic damage. Therefore, we expressed NMDA receptor channels in HEK293 cells, applied D-2 or 3-OH via a fast application system and found an activation of NR1/NR2A and NR1/NR2B subtypes. This result was confirmed using a cell viability assay (trypan blue staining) in cultured neurons which showed an induction of neuronal damage via NMDA receptors upon D2 or 3-OH incubation. Furthermore, both organic acids disturb intracellular Ca2+ homeostasis (Ca2+ imaging using fura-2), and they generate reactive oxygen species (rhodamine-123 fluorescence). An additional neurotoxic mechanism of D-2 but not of 3-OH is the reduction of respiratory chain complex V activity. These results reveal some of the mechanisms causing neuronal damage in the cerebral organic acid disorders D-2-hydroxyglutaric aciduria and glutaryl-CoA dehydrogenase deficiency and particularly indicate that excitotoxicity contributes to their neuropathophysiology.

RESPONSES OF XENOPUS LAEVIS TADPOLE OLFACTORY NEURONS TO AMINO ACIDS AND FORSKOLIN I. Manzini, W. Rössler, F. Peters, D. Schild We report on responses of olfactory receptor neurons (ORNs) upon application of amino acids and forskolin using a slice preparation of the olfactory epithelium of Xenopus laevis tadpoles. Both amino acids and forskolin proved to be potent stimuli. Interestingly, a substantial number of ORNs that responded to amino acids did not respond to forskolin. This suggests that some amino acids activate transduction pathways other than the well-known cAMP-mediated one. The differential processing of cAMP-mediated stimuli on one hand and amino acid stimuli on the other was further elucidated by Ca2+imaging of mitral cells in the olfactory bulb. The projection pattern of amino acid-sensitive ORNs to mitral cells differed markedly from the projection pattern of forskolin-sensitive ORNs. Mitral cells activated by amino acids were located laterally compared to those activated by forskolin, and only a small proportion responded to both stimuli. We therefore conclude that sensory transduction of a number of amino acids is cAMP-independent, and amino acid- and forskolin-mediated responses are processed differentially at the level of the olfactory bulb. Physiologisches Institut, Abteilung Molekulare Neurophysiologie, Universität Göttingen, Humboldtallee 23, D 37073 Göttingen, Deutschland

P 37-2 COMPLEX INTRACELLULAR REGULATION OF AN INSECT NEURONAL LOW-VOLTAGE-ACTIVATED CHANNEL PERMEABLE TO SODIUM AND CALCIUM VIA DARPP-32-MEDIATED INHIBITION OF PP1/2A A. Defaix, S.N. Schiffmann, B. Lapied Identification of the different intracellular pathways that control phosphorylation/dephosphorylation process of ionic channels represents an exciting alternative approach for studying the ionic mechanisms underlying neuronal pacemaker activity. In the cockroach Periplaneta americana central nervous system, octopaminergic neurones, called dorsal unpaired median (DUM neurones), generate spontaneous repetitive action potentials. Short-term cultured adult DUM neurones isolated from the terminal abdominal ganglion of the nerve cord were used to study the regulation of a tetrodotoxin-sensitive low-voltage-activated channel permeable to sodium and calcium, under whole cell voltage- and current-clamp conditions. A bell-shaped regulation of the amplitude of this maintained current versus [ATP]i was observed, suggesting the existence of phosphorylation mechanisms. Using the specific PKA inhibitor, H89 (200 µM) and elevating [cAMP]i (100 µM), increased and decreased the current amplitude, respectively. This indicated a regulation of the current via a cAMP/PKA cascade. Furthermore, intracellular application of PP2B inhibitors, cyclosporin A (100 nM), FK506 (5 µM) and PP1/2A inhibitor, okadaic acid (2 µM) decreased the current amplitude. From these results and because octopamine regulates DUM neurone electrical activity via an elevation of [cAMP]i, we wanted to know if, like in vertebrate dopaminergic neurones, octopamine receptor stimulation could affect the current indirectly via PKA-mediated phosphorylation of DARPP-32 known to inhibit PP1/2A. Consequently, experiments were performed using intracellular application of phospho-DARPP-32 and non-phosphoDARPP-32 (0.3 mg/ml). Phospho-DARPP-32 strongly reduced the current amplitude whereas non-phospho-DARPP-32 did not affect the current. All together, these results confirmed that DARPP-32mediated inhibition of PP1/2A regulated the maintained sodiumcalcium current. Finally, we demonstrated that this current contributed to the development of the predepolarizing phase of the DUM neurones pacemaker activity via the modulation of the mLVA calcium current previously described in this preparation. Laboratoire de Neurophysiologie (RCIM) UPRES EA 2647, Université d’Angers, UFR Sciences, 2 Bd Lavoisier, F-49045 Angers Cedex, France

S 320 P 37-3

P 37-5

SIGNALING EVENTS IN MUSCARINIC SUPPRESSION OF THE SLOW AFTERHYPERPOLARIZATION (SAHP) IN CA1 PYRAMIDAL NEURONS IN HIPPOCAMPAL SLICE A.O. Volynets, W. Müller* The sAHP is an important Ca2+-dependent negative feedback on cellular excitability and discharge and is mediated by small conductance Kchannels. Among several neurotransmitters acetylcholine, by acting on muscarinic receptors, is an important blocker of the sAHP, thereby enhancing intracellular Ca2+-increases and lowering the threshold for induction of synaptic plasticity. This effect specifically depends on Ca-Calmodulin dependent protein kinase II (Müller et al. 1992). Here we use infrared video-guided whole-cell patch-clamp recording to study the signaling chain mediating this effect. The stable agonist carbachol (CCh, 2-5 µM) blocked the sAHP following directly evoked repetitive firing within 3 min. The effect was reversible within 15 min washout of CCh. The muscarinic antagonist atropine (1-10 µM) reversed this effect within 10 min. With intracellular application of the GDP analog GDPßS (2 mM) blockade of the sAHP by CCh was strongly reduced (n=3). When the non-hydrolizable GTP analog GTPγS (100 µM) was applied intracellularly, CCh irreversibly blocked the sAHP and depolarized neurons, and even a small afterdepolarization became apparent. In this situation, atropine no longer had any effect on the sAHP (n=3). The calmodulin antagonists chlorpromazine (100 µM, n=5) and calmidazolium (10 µM, n=1) inhibited the effect of CCh on the sAHP. We conclude that muscarinic inhibition of the slow AHP is mediated by a G-protein and calmodulin.

HIF-1 DNA BINDING AND VEGF SYNTHESIS IN HUMAN PROXIMAL TUBULAR CELL CULTURES REQUIRE PI3 KINASE ACTIVITY W. Jelkmann, D. Stiehl, D. Katschinski, B. Kreft, T. Hellwig-Bürgel Recent evidence suggests that tubular cell-derived cytokines play a role in inflammatory and degenerative processes in the kidney. F.e., vascular endothelial growth factor (VEGF) is thought to initiate permeabilisation of blood vessels, extravasation of plasma proteins and leukocytes, but also to protect tubular cells from apoptosis. VEGF gene expression is controlled by transcription factors such as hypoxia-inducible factor-1 (HIF-1). The aims of the present study were to investigate: (a) the capacity of primary human proximal tubular epithelial cell cultures (PTEC) to produce VEGF and express VEGF receptor mRNA, (b) the influence of hypoxia and interleukin-1β (IL1β) on HIF-1 stability and DNA-binding, and (c) the effects of inhibitors of phosphatidylinositol 3-kinase (PI3K: LY 294002) and p42/ p44 mitogen-activated protein kinases (MAPK: PD 98059) signaling pathways in HIF-1 activation and VEGF synthesis. PTEC were grown in monolayers from human kidneys. Hypoxia was induced by incubation at 3% O2. HIF-1α was demonstrated by Western-blot analysis and HIF-1 DNA-binding by gel shift assay and mRNAs by RT-PCR. The results show that significant amounts of VEGF mRNA and VEGF protein were present in extracts and culture media of normoxic PTEC. Soluble VEGF receptor protein was not detected in culture media, although VEGF receptor mRNA was present at low level. VEGF synthesis was increased on hypoxic incubation or treatment of normoxic cells with IL-1β. Nuclear HIF-1α protein levels and HIF-1 binding to DNA were also increased under these conditions. Western-blot analysis showed HIF-1α accumulation on hypoxia exposure or IL-1β treatment. This response was abolished in presence of LY 294002. Treatment with PD 98059 did not lower HIF-1α levels. The synthesis of VEGF was lowered on either LY 294002 or PD 98059 treatment. We conclude that PI3K signaling is not only important in the hypoxic induction of HIF-1 and VEGF synthesis but also in the response to IL1. MAP-kinases appear to affect VEGF synthesis independent of HIF1 activation.

AG Molekulare Zellphysiologie des ZNS, Neurowissenschaftliches Forschungszentrum der Charité, Schumannstrasse 20/21, 10117 Berlin, Germany

P 37-4 INHIBITION OF PI3-KINASE ABOLISHES THE ANGIOTENSIN II-INDUCED ACTIVATION OF NADPH-OXIDASE AND P38 MAPKINASE IN ADULT VENTRICULAR CARDIOMYOCYTES S. Wenzel, H.M. Piper, K.-D. Schlüter An activated renin-angiotensin-system plays an important role in the hypertrophic growth of cardiomyocytes and their variabilities of transcription. In vivo and in vitro angiotensin II increases the expression of transforming-growth-factor-beta (TGF-beta) which is able to modulate other mediator systems, e.g. beta-adrenergic stimulation. It is further known that NADPH-oxidase, proteinkinase C (PKC) and p38 MAP-kinase are involved in the angiotensin II-induced induction of TGF-beta expression. The aim of this study was to demonstrate further signaltransduktion steps between the AT1-receptor and NADPHoxidase activation. In particular it was investigated wether this step is PI3-kinase-dependent. Methods: Freshly isolated cardiomyocytes were used. To investigate NADPH-oxidase activation the consumption of NADH/NADPH was determined spectrophotometrically. The activation of p38 MAP-kinase was determined as ratio of phosphorylated p38 MAP-kinase versus total p38 MAP-kinase proteins by western blotting. Results: Angiotensin II (100 nM) increases the activity of NADPH-oxidase (1,30±0,12-fold; n=8, p<0,05 vs. control) and p38 MAP-kinase (1,51±0,17-fold; n=10, p<0,05 vs. control) within 60 minutes. This increased activation of NADPH-oxidase is abolished after pre-incubation of cells with wortmannin (100 nM, 0,8±0,11fold; n=13, p<0,05 vs. control) and Ly 29004 (100 nM, 0,65±0,21fold; n=13; p<0,05 vs. control), two inhibitors of the PI3-kinase. The activation of p38 MAP-kinase ist blocked significantly by pre-incubation of cells with wortmannin (100 nM, 0,8±0,17-fold; n= 8, p<0,05 vs. control) and Ly (100 nM, 0,7±0,15-fold; n=6; p<0,05 vs. control) either. To define wether PKC activation is upstream or downstream of PI3-kinase activation cells were pre-incubated with those inhibitors and following stimulated with phorbolester (100 nM). No effect could be shown, which point out that the activation of NADPH-oxidase in fact is PI3-kinase-dependent but not the activation of proteinkinase C. Conclusion: These data demonstrate the agreement of cardiomyocytes and smooth muscle cells in respect to their siganlling pathways after angiotensin II-induction. Sibylle Wenzel, Physiologisches Institut, Aulweg 129, 35392 Giessen, Germany

Institute of Physiology, Medical University of Lübeck, Ratzeburger Allee 160, 23538 Lübeck, Germany

P 37-6 DOES ALDOSTERONE INTERACT WITH EPIDERMAL GROWTH FACTOR SIGNALING IN MDCK-CELLS? M. Gekle, R. Freudinger, S. Mildenberger, K. Drumm, A. Krug, S. Silbernagl It has been suggested that steroid hormones modulate the action of peptide hormones, in part by rapid mechanisms. Epidermal growth factor (EGF) regulates cell proliferation and electrolyte transport using the mitogen-activated protein kinases ERK1/2 as downstream signals. Furthermore, the EGF-receptor is involved in signaling by Gprotein-coupled receptors and cytokines. Here, we tested the hypothesis that aldosterone interacts with EGF-receptor (EGF-R) signaling using collecting duct derived Madin-Darby canine kidney cells. Cellular Ca2+ and pH were measured by videoimaging using fluorescent dyes, ERK1/2-phosphorylation was determined by Western Blot and ELISA, EGF-R-phosphorylation was determined after immunoprecipitation. Nanomolar aldosterone induced rapid increases of ERK1/2 phosphorylation, cellular Ca2+-concentration and Na+/H+-exchange activity of similar extend as 10 µg/l EGF. Substantial increases in cAMP, IP3 or reactive oxygen species could not be observed. Analysis of the doseresponse curve indicates that aldosterone sensitized the cells for EGF without alteration of the maximum effect. Changes in Ca2+ resulted from La3+-sensitive Ca2+-influx and were prevented by inhibition of ERK1/2 activation. By contrast, prevention of Ca 2+-influx did not abolish ERK1/2 activation. Furthermore, aldosterone induced a rapid increase of EGF-receptor Tyr-phosphorylation. Moreover, inhibition of EGF-R-kinase reduced aldosterone-induced signaling significantly. These data indicate that aldosterone can modulate the EGF-receptorERK1/2 signaling pathway. Possibly, the EGF-receptor represents a membrane target for rapid effects of aldosterone. Supported by the DFG (Ge 905/4-1 and SFB487/A6).

S 321 Physiologisches Institut, Universität Würzburg, Röntgenring 9, 97070 Würzburg, Germany, Phone: ++49 931 312739,FAX: ++49 931 312741, E-Mail: [email protected]

These results suggest a competitive cross talk, presumably at the expression level, between two receptors converging on an identical signalling pathway. (1) Wellner-Kienitz, M.-C. et al. (2000) Circ.Res. 86, 643-648; (2) Bünemann, M. et al. (1997) J.Physiol.(Lond.) 501, 549-554. Supported by DFG (Po-212/9-2).

P 37-7 MANIPULATION OF THE RECEPTOR-GIRK SIGNALLING PATHWAY IN ATRIAL MYOCYTES BY HIGHLY EFFICIENT GENE TRANSFER USING A SIMPLIFIED RECOMBINANT ADENOVIRUS SYSTEM K. Bender, L.I. Bösche, M.-C. Wellner-Kienitz, L. Pott Manipulation of protein expression by gene-transfer provides a potent tool for studying various aspects of cellular signalling. Highly differentiated cells, like adult cardiac myocytes, are very resistant to transfection using conventional methods, such as lipofectamine or Ca2+-phosphate. In our hands, transfection rates in terms of reporter (GFP)-positive cells vary from virtually zero in adult rat ventricular myocytes to < 5% in atrial myocytes using lipofectamine. Although this permits electrophysiological recordings from identified atrial cells (e.g. refs. 1, 2), protein assays are not feasible. Moreover, a graded manipulation of expression is impossible. Here we demonstrate that a simplified method to produce recombinant adenovirus particles, based on the Adenovirus5 pAdeasy-system (3) can be used to infect cardiac myocytes with high efficiency. After ligation of the cDNA of interest into a shuttle vector, it is transferred into the genome of Adeno5 virus by homologous recombination in E.coli. The gene of interest is introduced in place of the deleted E1 gene of the virus, which renders the virus defective for replication and incapable of producing infectious particles in target cells. For generating the recombinant adenoviral plasmid a minimum of enzymatic manipulations is required. The recombinant virus are produced in the packaging cell line HEK293, which can complete the missing gene. After 5-7 days, recombinant virus can be harvested. The cell lysate can be standardized and used for transfecting cardiac myocytes. At 48h after infection 80-90% of the myocytes of both atrial and ventricular origin (rat) are GFP-positive. The method is being used for manipulating G-protein-coupled signalling pathways by over expression or expression of mutant components. 1.Bender, K. et al. (2001) J.Biol.Chem. 276, 28873-28880; 2. Wellner-Kienitz, M.-C. et al. (2001) J.Biol.Chem. 276, 37347-37354; 3. He, T. C. et al. (1998) Proc.Natl.Acad.Sci.USA 95, 2509-2514. Supported by DFG (Po212/9-2). Institut für Physiologie, Ruhr-Universität Bochum, D-44780 Bochum, Germany

Ruhr-Universität Bochum, Institut für Physiologie, D-44780 Bochum, Germany P 37-9 G PROTEIN-INDEPENDENT INHIBITION OF ATRIAL GIRK CURRENT BY ADENOSINE IN ATRIAL MYOCYTES STRONGLY OVER EXPRESSING A1 RECEPTORS L.I. Bösche, A. Rinne, K. Bender, M.-C. Wellner-Kienitz, L. Pott Atrial GIRK channels are activated by various G protein–coupled receptors including muscarinic M 2 (M 2 AChR) and purinergic A 1 (A1AdoR). In adult rat atrial myocytes the amplitude of GIRK current evoked by adenosine is limited by the density of A1AdoR to ~30 % of the current in the presence of ACh at a saturating concentration. We have shown previously that over expression of A 1AdoR increases sensitivity to adenosine (1). In the present study we confirmed these results, which were obtained by transfection using lipofectamine, by means of a more efficient adenovirus-based method of gene transfer („pAd-Easy“). In a fraction of A1AdoR transfected myocytes, which showed a particularly high fluorescence-intensity of the reporter protein (GFP), activation of GIRK current by Ado (10 µM) was followed by a rapid decay (t1/2 ~ 300 ms). Upon washout of Ado a rebound of current was observed that was proportional to the degree of the initial decay. In those cells Ado caused an inhibition of M2AChR- induced current in the continuous presence of ACh. In GTP-(-S loaded myocytes stably and irreversibly activated GIRK current was reversibly inhibited by Ado. Upon activation of endogenous M2AChR by a highly saturating concentration of ACh (1 mM) a similar inhibitory component can be detected. These results suggest a novel mechanism of GIRK current inhibition by G protein-coupled receptors, which requires fast activation of a large number of receptors, and which seems to be independent of GTP-cycling of a heterotrimeric G protein (1) WellnerKienitz,M.-C. et al. (2000); Circ. Res. 86, 643-648. Supported by DFG (Po-212/9-2). Ruhr-Universität Bochum, Institut für Physiologie, D-44780 Bochum, Germany P 37-10

P 37-8 ANTISENSE INHIBITION OF M2 RECEPTOR EXPRESSION IN RAT ATRIAL MYOCYTES RESULTS IN AN INCREASED SENSITIVITY OF GIRK CURRENT TO A1-ADENOSINE RECEPTOR STIMULATION L.I. Bösche, K. Bender, M.-C. Wellner-Kienitz, L. Pott Atrial GIRK channels are activated by various G protein–coupled receptors including muscarinic M 2 and purinergic A 1. In rat atrial myocytes the amplitude of GIRK current evoked by adenosine is limited by the density of A 1 receptors to ~30 % of the current in the presence of ACh at a saturating concentration. We have shown previously that over expression of A 1 receptors (A 1AdoR) not only increases sensitivity to adenosine, but reduces sensitivity to ACh, suggesting a reduction of the number of functional of M 2AChR in the membrane (1,2). In the present study we confirmed these results, which were obtained by transfection using lipofectamine, by means of a more efficient adenovirus-based method of gene transfer („pAdeasy“). Moreover, we manipulated expression of M2AChR by treatment with an 18-meric antisense oligonucleotide directed against the rat M2AChR sequence and measured GIRK currents induced by ACh and Ado respectively. In AsODN-treated myocytes maximum current evoked by ACh at saturating concentrations was decreased by 35%, and half time of activation upon rapid exposure to ACh, a sensitive indicator of changes in receptor density (2), was prolonged by 70 %. This was paralleled by an increase in density of current induced by 100 µM Ado by 30% and a reduction in activation half time by 60%. Decreasing the number of functional M2AChR by desensitization / down regulation, following incubation for 3 d with carbachol (500 µM), resulted in a reduction in ACh-induced current and an increased half time of activation without, however, affecting these parameters for the A1AdoR-induced current.

HYPOXIA INDUCES PERMEABILITY OF PORCINE BRAIN DERIVED MICROVASCULAR ENDOTHELIAL CELLS BY Γ ACTIVATION OF PKB/AKT AND PLCΓ S. Fischera, M. Wiesnetb, D. Renza, W. Schaperb In vivo, ischemia and ischemia/reperfusion are known to damage the blood-brain barrier (BBB) leading to the development of vasogenic brain edema. To evaluate signaling pathways involved in these permeability changes, an in vitro model of the blood-brain barrier consisting of porcine brain derived microvascular endothelial cells (BMEC) was used. Hypoxia and H 2O2, one of the agents associated with pathological events induced by ischemia/reperfusion, increasesd the paracellular flux of inulin across the BMEC monolayer significantly. Both conditions also changed the localization of the tight junction (TJ) proteins, zonula occludens 1 (ZO-1), zonula occludens 2 (ZO-2), and occludin to a more diffuse distribution. Accordingly, the expression of ZO-1, ZO-2, and occludin decreased during hypoxia and H 2O 2 treatment in a time-dependent manner. Although the effects of hypoxia and H2O2 on paracellular permeability and TJ protein expression were similar, the mechanisms leading to these changes seem to be different. Hyperpermeability induced by H 2O2 was completely blocked by PD98059, which specific inhibits mitogen-activated protein (MAP) or ERK kinase 1 (MEK1). Hypoxia-induced permeability changes, which were shown to be mediated via vascular endothelial growth factor (VEGF) and NO were not changed by PD-98059 suggesting that this pathway does not involve ERK1/ERK2 activation. However, hypoxia and VEGF-induced permeability changes were blocked by LY294002, a specific inhibitor of the phosphatidylinositol 3-kinase (PI3-K) and protein kinase B (PKB/akt) pathway. Inhibition of phospholipase Cγ (PLCγ ) by U73122 but not of protein kinase C (PKC) by bisindolylmaleimide (BIM) also prevented hypoxia- and VEGF-induced

S 322 permeability changes. Results suggest that hypoxia- and VEGF- induced hyperpermeability is mediated through PI3 kinase-PKB and PLCγ signaling. Max-Planck Institute for Physiological and Clinical Research, a Department of Anesthesiology and Intensive Care, b Department of Experimental Cardiology, 61231 Bad Nauheim, Germany P 37-11 A ROLE OF THE POU-DOMAIN TRANSCRIPTION FACTOR POU4F2 IN EPITHELIAL DIFFERENTIATION? G. Schley1, K.D. Wagner1, N. Wagner2, H. Scholz1 Among the Pou-domain genes, the Pou4f2 factor (formerly Brn-3b) is critical for the survival of ganglion cells in the developing vertebrate retina. Our recent findings suggest that Pou4f2 expression in the embryonic retina is initiated by the Wilms’ tumor transcription factor Wt1. Wt1 was originally identified on the basis of its mutational inactivation in a class of pediatric renal carcinoma (Wilms’ tumor). Its targeted disruption in mice revealed a requirement for Wt1 in the development of the kidneys, gonads, and mesothelium. Wt1 is thought to be necessary for the differentiation of mesenchyme to epithelial tissue in these organs. We wanted to know if interaction of Wt1 and Pou4f2 is also required for the development of non-neuronal tissues, aiming in particular at epithelial differentiation. For this purpose, the expression of both genes was studied by mRNA in situ hybridization and immunohistochemistry. A strikingly similar expression pattern of Wt1 and Pou4f2 was seen in the epithelial layer of the stomach and intestine. The two proteins were also co-localized in the epithelium of the gall bladder and in visceral epithelial cells (podocytes) of the kidneys. Pou4f2 mRNA levels (measured by real-time RT-PCR) were increased more than 8-fold by stable expression of Wt1 in the human embryonic kidney (HEK)293 cell line. Furthermore, we transiently co-transfected approximately 4 kilobases of the 5’-regulatory sequences of the human POU4F2 gene into cultured cells, together with expression plasmids encoding different Wt1 isoforms. Activity of the Pou4f2 promoter was increased more than 4-fold upon expression of the Wt1(-KTS) variant. These findings indicate that Wt1 can activate the Pou4f2 gene in non-neuronal cells. The similar spatial expression of the two genes suggests a role for Pou4f2 in Wt1-dependent epithelial differentiation. 1 Johannes-Müller-Institut für Physiologie und 2Klinik für Innere Medizin I, Humboldt-Universität, Charité, Berlin

P 38-1 MITOCHONDRIAL CONTROL OF CALCIUM SIGNALLING IN MOTONEURONS THAT ARE PARTICULARLY VULNERABLE IN AMYOTROPHIC LATERAL SCLEROSIS (ALS) F. Bergmann , B. U. Keller Motoneurons are selectively damaged in human amyotrophic lateral sclerosis (ALS) and corresponding mouse models of this neurodegenerative disease, a process that involves disruption of intracellular calcium signalling as well as mitochondrial dysfunction. Low calcium buffering capacity KB was identified as a prominent risk factor for selective motoneuron damage. Under physiological conditions, low KB leads to high amplitude Ca2+ transients for a given influx and large microdomains around Ca2+ influx sites. In a first order approximation, low KB should also enhance the role of low-affinity Ca2+ buffers like mitochondria in vulnerable motoneurons.To investigate the role of mitochondria for Ca2+ regulation in vulnerable motoneurons we performed measurements of cytosolic and mitochondrial [Ca 2+] (Ca i, Camit) of hypoglossal motoneurons in mouse brain slices based on CCD camera imaging. Antidromic stimulation of action potentials lead to an immediate rise of Cai. Simultaneous measurements of Cai and Camit revealed rises of Camit without a delay within a few ms after cell depolarisation. This was substantially faster compared to mitochondrial responses previously found in other systems, suggesting that mitochondria are located in close vicinity of voltage dependent Ca2+ channels. As expected, addition of protonophores (CCCP/FCCP) reduced mitochondrial Ca2+ transients (<50%), increased peak Cai responses (113%) and prolonged Ca i recovery times (2-fold prolongation). Recovery times of Camit were faster than those of Cai and occurred within Is after stimulation. A multicomponent time course of Camit recovery indicated that several mechanisms contributed to the decay.Our results

indicate that vulnerable motoneuron populations are characterized by a specialised calcium homeostasis, including i) low cytosolic buffering, ii) high amplitude transients during electrical activity, iii) large microdomains around open channels and iv) fast mitochondrial uptake of Ca i. With respect to ALS, rapid mitochondrial regulation of Cai provides an integrative model to explain the synergistic impact of disrupted Ca2+ signalling and mitochondrial dysfunction for selective motoneuron degeneration. Zentrum Physiologie und Pathophysiologie, Universität Göttingen, Humboldtallee 23, 37073 Göttingen, Germany P 38-2 EVIDENCE FOR CaT1-MEDIATED Ca 2+ INFLUX BY THAPSIGARGIN-INDUCED Ca2+ RELEASE FROM INOSITOL 1,4,5-TRISPHOSPHATE (IP3)-SENSITIVE Ca2+ STORES IN HEK 293 CELLS A. Latas, E. Krause, I. Anderie, I. Schulz Depletion of IP3-sensitive Ca2+ stores leads to activation of the calcium-release-activated calcium channel (CRAC) (Hoth & Penner (1992) Nature 355, 353-356) in different cells. Recently it had been reported that CaT1, a member of the TRP family of cation channels, exhibits features of the CRAC (Yue et al. (2001) Nature 410, 705709). Comparison of endogenous CRAC with heterologously expressed CaT1, however, showed that they manifest several clearly distinct properties, which made it unlikely that CaT1 encompasses the full CRAC pore (Voets et al. (2001) J Biol. Chem. 276, 47767-47770). Here we show that rat CaT1 (Hediger et al. (1999) J. Biol. Chem. 274, 22739-22746) which had been expressed in HEK 293 cells, showed properties of the CRAC. Using Fura2 fluorescence measurements we could show that in HEK 293 control cells Ca2+ store depletion in the presence of the SERCA-inhibitor thapsigargin (200 nM) led to increased Ca 2+ influx into the cell indicating activation of the endogenous CRAC. This store-operated Ca2+ influx was completely inhibited in the presence of 2-aminoethyldiphenylborate (2-APB, 50 µM). In rCaT1-transfected HEK 293 cells cytosolic [Ca2+] ([Ca2+]cyt) was higher than in control cells. Omission of Ca 2+ from the bathing solution resulted in decrease of [Ca 2+]cyt, whereas readdition of Ca2+ brought [Ca 2+]cyt back to the control value. If Ca 2+ stores were depleted by thapsigargin in the absence of Ca2+ and with 2-APB (50 µM), subsequent readdition of Ca2+ resulted in a short lasting increase in [Ca2+]cyt by ca. 2 fold for about 10 sec duration. Subsequently [Ca2+]cyt decreased back to the control value of [Ca2+]cyt before Ca2+ store depletion. We conclude from these experiments, that Ca 2+ store depletion transiently activates CaT1. Probably increased [Ca2+]cyt leads to closure of the CaT1 channel. Physiologisches Institut der Universität des Saarlandes, Geb. 58, 66421 Homburg/Saar P 38-3 DISASSEMBLY OF CYTOSKELETON ACCELERATES SPREADING OF AGONIST-EVOKED Ca2+ SIGNALS IN MOUSE PANCREATIC ACINAR CELLS M. Redemann, I. Schulz, A. Schmid Stimulation of pancreatic acinar cells with acetylcholine (ACh) or bombesin at supramaximal concentrations elicits cytosolic Ca2+ signals that start in the luminal cell pole and then spread throughout the cell. In previous studies we have shown that the analysis of the propagation rate of these secretagogue-induced Ca2+ waves can be used to study the coupling of hormone receptors to intracellular signal cascades which cause activation of protein kinase C or production of arachidonic acid via activation of the cytosolic phospholipase A2. In the present study we examined the role of an intact mircotubule and actin cytoskeleton on the spreading of cytosolic Ca2+ signals produced by activation of bombesin and ACh receptors. Freshly isolated acinar cells were loaded with fluo-3 and the propagation rate of Ca2+ waves was monitored by use of a confocal laser-scanning microscope. Our experiments show that in pancreatic acinar cells disassembly of actin filaments by pretreatment of the cells with either cytochalasin D (5 µM) or latrunculin B (10 µM) for 5 minutes caused a faster spreading of ACh- and bombesin-evoked Ca2+ waves. An acceleration of agonistevoked Ca2+ waves could also be observed when microtubule network was disrupted by preincubation of the cells with nocodazole (50 µM).

S 323 It has been suggested that hormone-induced actin re-assembly in several cell types is inhibited by cGMP. We therefore investigated the spreading of Ca2+ waves in the presence of the membrane permeant cGMP-analogue dibutyryl-cGMP (db-cGMP). However, incubation of pancreatic acinar cells with 3 mM db-cGMP had no significant effect on the propagation rate of neither ACh- nor bombesin-evoked Ca2+ waves. The effect of cytoskeleton disruption on the spreading of cytosolic Ca2+ signals might be due to an altered spatial distribution of intracellular Ca2+ stores. Another explanation could be that a cytoskeleton-dependent interaction of proteins, which are part of a negative feed-back mechanism of the signal cascade, is impaired by cytoskeleton break-down. Institute of Physiology, University of the Saarland, D-66421 Homburg/Saar, Germany P 38-4 CHANGES IN ATP-INDUCED SIGNALING IN OSTEOBLASTLIKE CELLS FROM CONNEXIN 43 NULL-ALLEL MICE M. Wiemann*, B. Gramsch, K. Schirrmacher, D. Bingmann Osteoblast-like cells (OBL) from calvaria of new born rats communicate via gap junctions composed of connexin43 (Cx43) but also via ATP-mediated extracellular signals. While gap junction coupling of OBL is apparently involved in the control of growth and differentiation, the role of ATP-mediated signaling is widely unknown. To elucidate the interplay between these two signaling pathways, OBL from Cx43 null allel (Cx43-/-) mice were compared to wild type OBL (Cx43+/ +). Experiments were carried out on 3-6 weeks old primary cultures grown on collagen coated coverslips (high GEM medium, 10 % (v/v) fetal calf serum). Measurements of intracellular calcium (Cai) and dye injection studies were done in Hanks balanced solution. By means of Lucifer Yellow injection we found that dye coupling was virtually absent in Cx43-/- OBL although a marked over-expression of immunoreactive Cx45 occurred. Most striking, responses to bath applied ATP (100 µM) were potentiated in Cx43-/- OBL cells: Increases in Cai were pronounced and a considerable number of cells showed repetitive Cai waves. Subsets of cells could be distinguished on the basis of partially identical Cai transients. While outgrowth of Cx43-/- OBL cells from calvaria was sparse during the first week of culture, cell density was increased vs. controls after 3 weeks, which might be in line with the altered ATP susceptibility. In Cx43-/- cultures there were also noticeable changes in mineralization (decreased) and expression of cell bound alkaline phosphatase (increased). Heterozygous Cx43(+/-) OBL exhibited an intermediate phenotype except for the ATP responses which were not different from controls. We suggest that (i) the loss of Cx43 precedes changes of the osteoblast phenotype, that (i) changes of the extracellular ATP signaling pathways may contribute to these effects and (iii) that extracellular signaling may partly compensate the lack of gap junction communication. Institut für Physiologie, Universität Essen, D-45122 Essen, Germany P 38-5 INTRA- AND EXTRACELLULAR ATP REGULATES GAP JUNCTIONAL COUPLING OF HENSEN-CELLS IN CORTI‘S ORGAN A. Blödow, R. Junker, A. Ernst; H.-A. Kolb The supporting cells of the organ of Corti are structurally and electrically coupled by gap junctions. It is believed that such gap junctional coupling among supporting cells like Hensen cells provides metabolic uniformity as well as cochlear homeostasis for endolymphatic potassium ions. We applied the dual whole-cell technique to isolated Hensen cells of guinea pig to evaluate the role of extra- and intracellular ATP on gap junctional coupling. In the absence of intracellular ATP spontaneous gap junctional uncoupling occurs. Addition of ATP to the pipette filling solution stabilises gap junctional coupling in a dosedependent manner. 5 mM ATP was found to be sufficient for stable coupling for recording times up to 60 min. Extracellular ATP causes gap junctional uncoupling for concentrations above about 25 µM. The uncoupling effect of external ATP could be completely suppressed by the presence of 1 mM suramine as expected for the involvement of purinergic receptors. To explore the cytosolic regulation mechanism of gap junctional uncoupling specific activators of protein kinase C were applied, but neither addition of 1 nM phorbol-12-myristate-13 acetate (TPA) nor of 200 µM 1-oleoyl-2-acetyl-sn-glycerol (OAG) to

the control bath influenced gap junctional coupling. But addition of a calmodulin inhibitor W7 to the pipette filling solution suppressed the uncoupling effect of externally applied ATP. Therefore, it appears likely that a Ca2+-dependent regulatory role of calmodulin is involved in regulating gap junction coupling of Hensen cells. Institute of Biophysics, University Hanover, Herrenhaeuser Strasse 2, D-30419 Hanover, Germany P 38-6 PURINERGIC RECEPTOR EXPRESSION AND FUNCTION IN NES2Y ß-CELLS R.M Shepherd, A-M Gonzalez, A.T. Lee and M.J. Dunne NES2Y cells are human ß-cell line derived from a patient with Type I Hyperinsulinism of Infancy (HI). Here, we studied the function and expression of purinoceptor subtypes in these cells by microfluorimetry and cell imaging with Fura-2, RT-PCR and by immunofluorescence using a P2X4 selective antibody. ATP (1µM-0.1mM)-evoked rises in [Ca2+]i which were oscillatory and dependent upon both Ca2+ influx and mobilisation of Ca 2+ from intracellular stores, since [1] they were reduced in amplitude and duration in the absence of external Ca2+ (n=109) and [2] absent when thapsigargin (100nM – an inhibitor of intracellular Ca2+ ATPases) was present in a Ca2+-free bathing medium (n=44). The phospholipase C inhibitor, U73122 (2µM), greatly inhibited the magnitude of the ATP-evoked [Ca 2+] i responses (by approximately 80%) and the proportion of responding cells fell to 30% of controls (10/29). These results implicate the involvement of P2Y rather than the P2X subfamily of receptors. Experiments with suramin (100µM, n=21) and PPADS (100µM, n=30), P2Y receptor antagonists with different subtype specificities, suggest the involvement P2Y4 receptor subtypes since suramin had no significant effect on 0.1mM ATP-evoked [Ca2+]i levels, whereas PPADS produced as 40% reduction in the magnitude of these responses. Using specific oligonucleotide probes, RT-PCR analysis confirmed expression of P2Y2, P2Y4 and P2X4 mRNAs. Finally, since we were unable to detect P2X-evoked rises in [Ca2+]i , immunofluorescence was used to identify the site of P2X4 localization. These studies showed that P2X4 was present in NES2Y ß-cells but that the protein was localised in the cytoplasm and not at the plasma membrane. In summary, we have shown that the ATP-evoked [Ca2+]i responses in NES2Y cells are governed by P2Y receptor-mediated events rather than P2X receptors, and that although mRNA for P2X4 can be detected, this plays no significant role in governing ATP-evoked rises in cytosolic Ca2+ . Department of Biomedical Science, Sheffield University, Sheffield, S10 2TN, UK P 38-7 GADOLINIUM DOES NOT INHIBIT STRETCH-INDUCED CFOS EXPRESSION IN NEONATAL AND ADULT RAT VENTRICULAR MYOCYTES B. Husse, W. Briest, U. Mueller-Werdan, G. Isenberg Mechanical stretch has been reported to induce a rapid activation of immediate-early (IE) genes in vivo and in vitro. The IE gene expression was insensitive by a block of stretch-activated ion channels (SACs) [Sadoshima et.al. 1992]. Gd3+ should have been precipitated with the phosphate of the culture media or the SACs should have been blocked by antibiotics, hence the possible involvement of Gd3+-sensitive SACs on stretched-induced IE gene expression was re-investigated in the absence of extracellular phosphate and other constituents interacting with Gd 3+. Neonatal rat heart cells (2 days) and isolated ventricular myocytes of adult rats (20 hrs) were cultured on Bioflex culture plates coated with Laminin. After a culture period of 4 hrs in a Tyrode solution without phosphate and antibiotics, the cyclical mechanical stretch (cms) was applied with the Flexercell Strain Unit (0.25 Hz, 10% elongation) for 30 min. Adult ventricular myocytes: cms enhanced the c-fos mRNA expression by a factor of 2.71±0.40 (Tyrode solution) in comparison to non-stretched controls. Gd3+ (50µM) did not prevent the stretch-induced c-fos expression (2.99±0.86 fold). Neonatal cardiomyocytes: cms induced c-fos expression by a factor of 3.84±2.25 (culture medium) and by 1.36±0.49 (not significant, Tyrode solution). In absence of cms, Gd3+ enhanced c-fos expression by 3.31±0.27 (culture medium) and by 4.00±1.70 (Tyrode solution). Cms and Gd3+ together induced c-fos expression by a factor of 7.16±2.64 (culture medium)

S 324 and by 3.36±1.48 (Tyrode solution). These results suggest that Gd3+sensitive SACs are not involved in the transduction of mechanical stretch into the c-fos expression. We speculate that the cells have taken up Gd3+ and that Gd3+ activates Ca2+-sensitive regulator proteins.

Laboratory of Molecular Signalling The Babraham Institute Cambridge CB2 4AT UK P 39-2

Department of Physiology, Martin-Luther-University Halle, D-06097 Halle P 38-8 cENO-ASSOCIATED Ca 2+ WAVES IN THE INTACT MOUSE BRAIN ”IN VITRO” O. Garaschuk*, C. Stosiek, J. Linn, A. Konnerth Spontaneous oscillations of the intracellular Ca2+ concentration mark the earliest developmental stages of many brain structures and are assumed to control neuronal differentiation and synaptic wiring. In cortical slices from newborn rats we had previously described a new form of such Ca2+ oscillations, which involved almost all neurons and propagated through the entire slice. We named this activity cortical early network oscillations (cENOs). To elucidate the properties of cENOs in the intact cortex we conducted simultaneous fluorometric and field potential measurements in the isolated brain of newborn mice. Neurons were stained by local extracellular application of the Ca2+ indicator Calcium Green-1 AM and visualized by means of twophoton imaging. When applied for 1-5 min the dye stained cells in a region with a diameter of 150-200 µm up to a depth of 200 µm, so that layer 1 and layers 2-3 could be clearly distinguished morphologically. In the entorhinal/temporal cortex the cENOs were detected in layers 2-3 at a frequency of 1-3 times per min. In layer 1 cENO-associated Ca2+ transients were only observed in the neuropil, enriched with dendrites of layer 2-3 pyramidal neurons, but not in cell bodies. cENOassociated field potentials were registered at every depth but had their maximal amplitude (50-150 µV) in layers 2-3. Simultaneous field potential measurements in different cortical areas showed that cENOs occur everywhere except the frontal cortex. They propagate in a wave-like fashion from the posterior to the anterior part of the brain with a speed of 25 mm/s. Finally, cENOs were abolished by the application of AMPA and NMDA receptor blockers, indicating that they require glutamatergic synaptic transmission. Thus, we demonstrate that cENO-associated Ca2+ waves occur during the earliest postnatal stages in the intact brain of different rodent species. They express a region-specific activity pattern which may be important for establishment and initial adaptation of short as well as of long-distance neuronal connections.

METHYL PYRUVATE DECREASES OPEN PROBABILITY OF A 55 pS-CHANNEL IN THE INNER MITOCHONDRIAL MEMBRANE L. Buntinas, P. Krippeit, G. Drews, C. Eikholt, T.F.W. Horn, D. Siemen The inner mitochondrial membrane contains an ATP-sensitive Kchannel, mK(ATP), that may be involved in the basic mechanisms of cell function (volume regulation or apoptosis) as well as in several diseases (neurodegeneration). Activity of this channel might be altered by mechanisms that include NO signaling, insulin secretion, intracellular Ca2+ regulation or ischemia-reperfusion injury. In spite of great interest in the mK(ATP) only a single paper exists describing it on the single channel level (Inoue et al., 1991, Nature 352:244). Methylpyruvate (MPyr) is able to block the plasma membrane K(ATP). Consequently, we tested the effect of MPyr on single ion channels of the inner mitochondrial membrane. Mitochondria were prepared from cultured human T-lymphocytes (Jurkat cells) by homogenization and 4 subsequent steps of centrifugation at different speeds. Patch clamp recordings were made from mitoplasts (i.e. after removing the outer membrane by a hypotonic treatment). The basic solutions in bath and measuring pipette contained 150 mM KCl and 10 mM K-HEPES at a pH of 7.2. Solutions were applied by a pump-driven flow system. The blocking effect of MPyr on the plasma membrane K(ATP) was proven by whole cell experiments in the current clamp mode where 20 mM MPyr decreased the current by 86%. It could be partially reversed by 100 µM of the K(ATP) channel agonist diazoxide. Recording from the mitochondrial membrane, we occasionally observed single channel events of about 55 pS. These events appeared regularly when the mitochondria were pretreated with 50 µM diazoxide. The channel events showed rectification into the positive. They were almost continuously open (Po > 0.89, 200-500 µM Ca2+). In the depolarizing range Po was almost voltage independent. After adding 20 mM MPyr Po was reduced to about 0.5, primarily by increasing the closed times. This effect was reversible after switching back to a KC1 control solution. We conclude that there exists a channel in the inner mitochondrial membrane that can be activated by Diaz and partially blocked by MPyr. However, it was not conclusively shown that this channel is the mK(ATP) channel. Supported by the BMFT. Dept. of Neurology, University of Magdeburg, D-39120 Magdeburg, FRG

Institut für Physiologie, LMU, D-80802 München, Germany P 39-3 P 39-1 MITOCHONDRIA ARE MORPHOLOGICALLY AND FUNCTIONALLY HETEROGENEOUS WITHIN AND BETWEEN DIFFERENT CELLS Tony J. Collins, Peter Lipp, Michael J. Berridge & Martin D. Bootman Mitochondria play key roles in the life and death of cells. We investigated whether mitochondria represent morphologically continuous and functionally homogenous entities within single intact cells. Physical continuity of mitochondria was determined by three-dimensional reconstruction of fluorescence from mitochondrially-targeted DsRedl or calcein. The mitochondria of HeLa, PAE, COS-7, HUVEC, hepatocytes, neocortical astrocytes and neuronal cells all displayed heterogeneous distributions and were of varying sizes. In general, there was a denser aggregation of mitochondria in perinuclear positions than in the cell periphery where individual isolated mitochondria could clearly be seen. DsRedl and calcein were highly mobile within the lumen of individual mitochondria. However, a lack of fluorescencerecovery-after-photobleaching of DsRedl or calcein demonstrated lumenal discontinuity between mitochondria. Mitochondria were not electrically coupled, since only individual mitochondria were observed to depolarise following irradiation of TMRE-loaded cells. Functional heterogeneity of mitochondria in single cells was observed with respect to membrane potential, sequestration of hormonally-evoked cytosolic calcium signals and the timing of permeability transition pore opening in response to tert-butyl hydroperoxide. Our data indicate that the mitochondria within individual cells are morphologically heterogeneous and unconnected, thus allowing them to have distinct functional properties.

LARGE NUCLEAR PORES EXERT LOW ELECTRICAL CONDUCTANCES WHEN PLUGGED BY MACROMOLECULES V. Shahin, C. Schäfer, T. Danker, H. Oberleithner Nuclear pore complexes (NPCs) form the gate through which molecules enter or leave the nucleus by either active or passive transport. To measure the nuclear envelope electrical conductance in isolated Xenopus laevis oocyte nuclei and to evaluate single electrical nuclear pore conductance, we used the recently developed nuclear hourglass technique (NHT; Danker et al., PNAS 96: 13531-5, 1999). Our initial results indicated that NPCs of Xenopus l. oocytes are in an electrically open state under physiological conditions, with a mean conductance of 1.7 nS per NPC. This is in contrast with nuclear patch clamp data which indicate that NPCs are usually electrically closed. Here we present experimental conditions where mean single NPC conductance is reduced 26-fold due to electrophoretic plugging by negatively charged nucleoplasmic macromolecules. In addition, data indicate that under physiological conditions when macromolecules are present in the cytosolic solution, nuclear surface is heavily folded underestimating „true“ nuclear envelope surface by a factor of 2.6. When „true“ nuclear envelope surface area is taken into consideration modified values of mean single NPC conductance of 654 pS for electrically open conditions and 25 pS for electrically plugged conditions can be calculated. We conclude that the large overall NE conductance detected with the nuclear hourglass technique in intact Xenopus l. oocyte nuclei can be explained by the sum of single NPC conductances in the pSrange as long as open probability is high. This confirms previous patch clamp work concerning single NPC conductance but disagrees with the view that mean open probability of NPC channels is usually low.

S 325 Department of Physiology, Univ. of Münster, Robert-Koch Str. 27a, 48149 Münster, Germany, e-mail: [email protected]

Department of Physiology, University of Münster, Robert-Koch-Str. 27a, 48149 Münster, Germany; e-mail: [email protected]

P 39-4

P 39-6

INFLUENCE OF ALDOSTERONE ON NUCLEAR ENVELOPE ELECTRICAL RESISTANCE (NEER) C. Schäfer, V. Shahin, L. Albermann, I. Buchholz, H. Oberleithner The Nuclear Pore Complex (NPC) is a functional barrier between the nucleus and the cytoplasm that mediates the passive and active exchange of macromolecules across the nuclear envelope. With the assistance of the nuclear hourglass technique (NHT; Danker et al., PNAS 96: 13531-5, 1999) the electrical nuclear envelope (NE) resistance of isolated Xenopus laevis oocytes was measured in steroid hormone stimulated conditions, since measurable effects on NE resistance were expected due to import of activated receptors and export of mRNA. To test the hypothesis that import/export of macromolecules through NPCs alters NE electrical resistance (NEER) we microinjected oocytes with 50 fmol aldosterone. Then, nuclei were mechanically isolated between 2 to 30 min after aldosterone injection. Subsequently NEER was measured using the NHT. We observed three NEER stages: an early NEER peak two minutes after injection (increase: 19 ± 5.1 % ± SEM, n=17), and a late NEER peak 18 minutes after injection (increase: 22 ± 4.3 % ± SEM, n=10). By atomic force microscopy we detected at the time of the early NEER peak small molecular structures attached to the NPC surface. At the time of the late NEER peak we detected large molecular structures (plugs) inside the NPC central channels. Ribonuclease treatment eliminated the late NEER peak indicating that plugs resemble ribonucleoproteins. Co-injection of aldosterone (50 fmol) and its competitive inhibitor spironolactone (500 fmol) prevented the NEER changes. We conclude that the early NEER increase represents the translocation of the aldosterone-activated receptor while the late NEER increase represents the translocation of mRNA. Application of the NHT allows us to trace the genomic signal cascade of steroid hormones. Since plugs can be harvested with atomic force microscopy the oocytes could serve as a bioassay for identifying the molecular components of the intracellular hormone signaling cascade.

SIGNAL TRANSDUCTION TRIGGERED BY PROINFLAMMATORY CYTOKINES INHIBITING THE ERYTHROPOIETIN GENE: ROLE OF TRANSCRIPTION FACTORS GATA-2 AND NF-κB K. La Ferla, T. Hellwig-Buergel, C. Reimann, W. Jelkmann Background: Erythropoietin (Epo) stimulates proliferation and differentiation of erythrocytic progenitor cells. Decreased plasma Epo concentrations are reported in the anemia of inflammatory diseases. IL-1β and TNFα were shown to reduce hypoxic Epo gene expression. The human epo promoter contains binding sites for the transcription factors (TFs) GATA-2 and NF-κB. GATA-2 was shown to repress Epo gene expression, whereas the role of NF-κB in the control of Epo production is still unknown. Aim: This study was intented to investigate the role of TFs GATA-2 and NF-κB in the repression of Epo gene expression during inflammation. Methods: In vitro studies were performed on the human hepatoma Epo-producing HepG2 cells. Binding activity for the mentioned TFs was determined by electrophoretic mobility shift assay. Cells were transiently transfected either with a reporter gene construct spanning a sequence from the Epo promoter containing a GATA (pGATA-wt) or a TATA element (pGATA-mut). Cells were also stably transfected with a dominant-negative form of IκBα (pCMV-IκBα and pCMV-IκBαM) and Epo production was assessed. Oligo-decoy experiments were performed for both TFs. Results: EMSAs revealed a strong GATA-2 binding activity when cells were challenged with cytokines under normoxia and hypoxia, whereas hypoxia alone led to decreased binding activity. Reporter gene experiments showed a hypoxic increase of luciferase activity in cells transfected with pGATA-wt as Epo production increased, whereas stimulation with cytokines led to a luciferase activity decrease showing a repression of the Epo promoter activity. Cells transfected with pGATAmut showed increased luciferase activity. HepG2 cells stimulated with cytokines showed also significant NF-κB binding activity. In cells transfected with pCMV-IκBαM NF-κB activation was inhibited and secreted Epo in culture supernatants increased. Conclusion: GATA-2 and NF-κB may be critically involved in the suppression of Epo production induced by proinflammatory cytokines. They may play a crucial role in the development of anemia of inflammatory diseases.

Department of Physiology, Univ. of Münister, Robert-Koch Str. 27a, 48149 Münster, Germany; [email protected] P 39-5 CHARACTERIZATION OF ALDOSTERONE-INDUCED EARLY GENES USING ATOMIC FORCE MICROSCOPY L. Albermann, C. Schäfer, V. Shahin, J. Reinhardt, H. Oberleithner Aldosterone is an important mediator of osmoregulation in vertebrate cells and is thought to be involved in cell differentation. Its signaling pathway starts with the formation of a complex with the mineralocorticoid receptor (MCR) which is then translocated into the cell nucleus where transcription of target genes is initiated. Finally, mRNAs of specifically activated genes appear in the cytoplasm. Using atomic force microscopy (AFM), we were able to visualize the signaling cascade described above. We injected stage VI oocytes of Xenopus laevis with aldosterone and isolated their nuclei after different incubation times. Nuclei obtained 2 min after application of aldosterone clearly exhibited macromolecules (”flags”) attached to the nuclear pore complexes (NPCs) when examined by AFM. The estimated size of the observed particles (80-160 kDa) corresponds with the actual size of the MCR (~120kDa). NPCs of oocytes injected about 20 min prior to preparation showed large macromolecules (”plugs”) within their central channels. We concluded that these plugs contain transcripts induced by aldosterone. Electrical measurements support these findings; appearance of both flags and plugs is linked to a rise in nuclear envelope electrical resistance (NEER), probably due to partial plugging of NPCs. The increase of NEER parallel to occurrence of plugs is prevented by adding RNase A. Alterations of NEER and formation of flags and plugs in answer to aldosterone are suppressed by coinjection of the MCR-inhibitor spironolactone. Therefore we consider plugs to represent the early genomic response to aldosterone stimulation of Xenopus oocytes. AFM not only allows imaging of structures in the submicrometer range, it may also be used to manipulate them. Applying forces to the AFM tip approximately 10 fold higher than those used for imaging, we were able to dislocate plugs from NPCs. The macromolecules sticking to the AFM tip were then used as substrates for further experiments, for example RT-PCR protocols.

Institute of Physiology, Medical University, Ratzeburger Allee 160 23538 Luebeck, Germany P 39-7 IN VITRO OXIDATION MIMICS THE OXIDATIVE STRESS CONFERRED BY PLASMODIUM FALCIPARUM ON THE HOST ERYTHROCYTE V. Brand, A.-C. Uhlemann, P.G. Kremsner, S.M. Huber, F. Lang Plasmodium falciparum parasitizes human erythrocytes and utilizes hemoglobin of the host to provide amino acids. Digestion of hemoglobin produces a strong oxidative stress which is counteracted by the antioxidant defense of the parasite and the host. Oxidative processes are involved in the induction of the new permeability pathway (NPP) of the infected erythrocyte which is permeable for carbohydrates such as sorbitol (Huber et al, EMBO J, in press). Consequently, infected but not control erythrocytes hemolyze in isosmotic sorbitol solution. To mimic the effect of infection on host redox state, non-infected cells were subjected to experimental oxidative stress and redox state and membrane permeability were compared to those of infected cells. The redox state was assessed by reduced/oxidized glutathione [GSH/(GSH+ GSSG)] ratio and metHb formation as estimated by light absorption of the hemolysate at 405 and 415 nm (Soret bands of metHb and oxyHb, respectively). As a measure of membrane permeability, hemolysis in different isosmotic carbohydrate solutions was determined in oxidized and infected cells. Infection by P. falciparum decreased the GSH/ (GSH+GSSG) ratio of human erythrocytes from 0.99 ± 0.02 to 0.91 ± 0.02 (n = 5) and increased 405/415 nm light absorption ratio from 0.758 ± 0.008 to 1.025 ± 0.028 (n = 5) indicating oxidative stress of infected cells. Experimental oxidation of control cells with tbutylhydroxyperoxide (t-BHP; 1 mM for 15 min) and further postincubation (2.5 h) in physiological salt solution induced a GSH/ (GSH+GSSG) ratio (0.93 ± 0.01; n = 12) and 405/415 nm light absorp-

S 326 tion ratio (1.004 ± 0.025, n = 6) which was not different from that of infected cells. In addition, infected and oxidized erythrocytes hemolyzed in different isosmotic carbohydrate solution with the identical permeability rank order of sorbitol > mannitol > myo-inositol ˜ lactose ˜ sucrose ˜ raffinose. Taken together, the data suggest that in vitro oxidation induces in human erythocytes the phenotype of a cell infected with Plasmodium falciparum. Department of Physiology I, University of Tübingen, Gmelinstr. 5, 72072 Tübingen.

Furthermore, endogenous generation of reactive oxygen species (ROS) is downregulated whereas expression of P-glycoprotein is elevated. Inhibition of glycolysis by treatment of tumor spheroids with either deoxyglucose or iodoacetate resulted in increased generation of ROS, an earlier onset of necrosis development as well as a significant downregulation of P-glycoprotein and reversal of the MDR phenotype. This effect could be blunted by preincubation with the free radical scavenger ebselen, indicating that P-glycoprotein expression is regulated by the intracellular redox state. Our data suggest that inhibition of glycolysis may be a promising approach to downregulate intrinsic P-glycoprotein expression in hypoxic tumor tissue and circumvent the MDR phenotype.

P 39-8 REACTIVE OXYGEN SPECIES MODIFY CCK-EVOKED ACTIN FILAMENT POLYMERISATION AND AMYLASE SECRETION IN MOUSE PANCREATIC ACINAR CELLS A. González, J. A. Rosado, G. M. Salido, J. A. Pariente Actin filament reorganisation is a key event in the regulation of a large number of cellular processes, such as modulation of cell shape and morphology, Ca2+ homeostasis and exocytosis. There has been different hypothesis about the putative roles of the actin cytoskeleton in regulated exocytosis. In specialised secretory cells, such as pancreatic acinar cells, the apical actin barrier prevents docking of secretory vesicles with the membrane and, therefore, exocytosis in resting conditions. On the other hand, a contractile structure, mediating granule transport and providing a machinery to force the expulsion of secretory material is also required for regulated exocytosis (Valentijn et al, 1999). Reactive oxygen species (ROS) are known to be pathogenic factors that induce cellular alterations in different cell types, although the cellular mechanisms involved in this process remain unclear. The aim of this study was to investigate the effect of ROS on actin filament reorganisation and its relevance to exocytosis in pancreatic acinar cells. Adult male Swiss mice were sacrificed by cervical dislocation, the pancreas was rapidly removed and the acinar cells were isolated as described previously (González et al, 1998). Actin filament content was determined using FITC-phalloidin (Rosado & Sage, 2000) and filament location was visualised using confocal laser scanning microscopy. Amylase release was measured using the procedure published previously (Jensen et al. 1982). Treatment of pancreatic acini with CCK-8 induced spatial and temporal changes in actin filament reorganisation with an initial depolymerisation of the apical actin barrier followed by an increase in the actin filament content in the subapical area leading to amylase release. ROS such as H2O2 increased actin filament content and potentiated the polymerising effects of CCK-8 in these cells but abolished the disruption of the apical actin layer and amylase release induced by CCK-8. Similar to CCK-8, ROS generated by oxidation of hypoxanthine with xanthine oxidase induced an initial decrease in actin filaments located under the apical membrane followed by a smaller increase in the actin filaments content in the subapical area. Xanthine oxidase generated ROS are able to increase amylase release in pancreatic acini although combination with CCK-8 leads to abnormal exocytosis. The regulation of the actin cytoskeleton by ROS might be involved in radical-induced cell injury in pancreatic acinar cells. González, G., Pfeiffer, F., Schmid, A. & Schulz, I. (1998) Am. J. Physiol. 275, C810-C817. Jensen, R. T., Lemp, G. F. & Gardner, J. D. (1982) J. Biol. Chem. 257, 5554-5559. Rosado, J. A. & Sage, S. O. (2000) Biochem. J. 347, 183-192. Valentijn, K., Gumkowski, F. D. & Jamieson, J. D. (1999) J. Cell. Sci. 112, 81-96. Department of Physiology, University of Extremadura, Cáceres, 10071, Spain. Supported by Consejería de Sanidad y Consumo-Junta de Extremadura. P 39-9 EFFECTS OF INHIBITION OF GLYCOLYSIS ON ENDOGENOUS ROS GENERATION, NECROSIS DEVELOPMENT AND MULTIDRUG RESISTANCE OF MULTICELLULAR TUMOR SPHEROIDS M. Wartenberg, A. Datschew, H. Acker, J. Hescheler, H. Sauer The development of intrinsic P-glycoprotein-mediated multidrug resistance (MDR) in multicellular prostate tumor spheroids is downregulated by either endogenously generated or exogenous reactive oxygen species (ROS). With increasing size multicellular tumor spheroids develop steep oxygen gradients as evaluated by oxygen microelectrode measurements, and switch their metabolism towards anaerobic glycolysis.

Department of Neurophysiology, University of Cologne, Robert-KochStr. 39, D-50931 Cologne, Germany P 40-1 THERMAL DEPENDENCE OF VON WILLEBRAND FACTOR EXOCYTOSIS IN CULTURED HUMAN ENDOTHELIAL CELLS P. Rogge*, T. Görge, R. Ossig, H. Oberleithner, S.W. Schneider Due to their salient location at the interface between blood and tissue endothelial cells play a pivotal role in the initiation of coagulation and inflammation. This function is predominantly effected by the secretion of proinflammatory and procoagulatory mediators, such as the von Willebrand-factor (vWF). Therefore, in experimental setups vWF exocytosis, a highly regulated cellular process, is used as a surrogate marker of endothelial cell activation. Recent publications using vWF as a surrogate marker do not guarantee a meticulous control of physiological conditions in the experimental setup. In this study we investigated the thermal dependence of vWF exocytosis in human endothelial cells and tried to assess its influence on the reliability of experimental data. Furthermore, we investigated whether thermal alterations might be a useful technique to study isolated cellular processes at altered kinetics. Human umbilical venous endothelial cells (HUVECs) were incubated in a Hepes buffered solution at different temperatures. Exocytosis of vWF was induced by stimulation with 50 µM histamine. The different thermal conditions chosen are 4 °C, room temperature (22 °C), hypothermia (33 °C), normothermia (37 °C) and hyperthermia (41 °C). Stable temperatures and a prompt temperature change are obtained by rinsing the bottom of a 12-well-plate using a thermostated perfusion system. Basal vWF release is reduced to 50 % at room temperature and blocked at 4 °C. Between 33 – 41 °C vWF release is not significantly altered. In stimulated cells vWF exocytosis is linearly correlated to incubation temperature. (22 °C: 30 ± 5 %; 33 °C: 75 ± 5 %; 41 °C: 125 ± 5 % compared to normothermia (100 %), n = 6 for each experiment). Prompt increase of incubation temperature from 33 to 37 °C has no significant effect on stimulated exocytosis whereas hyperthermic preconditioned cells display elevated vWF exocytosis upon normothermic stimulation (120 %, n = 4). We conclude that (i) stimulated vWF exocytosis is linearly correlated to incubation temperature between 33 °C and 41 °C. (ii) Incubation temperature of about 4 °C might be used as blocker of exocytosis and thus permits investigation of isolated cellular processes at altered kinetics. Institut für Physiologie, Robert-Koch-Strasse 27a, 48149 Münster, Germany P 40-2 EXTRACELLULAR PHOSPHORYLATION OF ADENINE NUCLEOTIDES ON UMBILICAL ENDOTHELIAL CELLS S. Richter, A. Will, D. Joachim, S. Mattig, A. Deussen Extracellular dephosphorylation of adenine nucleotides is well established. Scarce information, however, is available on the potential of extracellular phosphorylation of adenine nucleotides. The presence of ecto-nucleotide kinase has previously been suggested. The aim of the present study was to gain information on extracellular adenine nucleotide phosphorylation on intact human endothelial cells. Methods: Human umbilical endothelial cells (HUVEC) were isolated using a dispase solution (100mg/20 ml PBS, 50 min) and cultured to confluence on microcarrier beads. Cell covered beads transfered into a chromatography column were perfused with a HEPES-buffer at 2 ml/min. As the primary substrate for extracellular nucleotide metabolism e-ATP (etheno-ATP) was infused. To provide the phosphor moiety for phosphorylation of the resulting e-ADP to e-ATP via the extracellular kinase reaction

S 327 UTP or GTP (10 and 100 µM, respectively) were added to the column perfusate. e-Nucleotides and e-adenosine were analysed using HPLCtechniques in conjunction with fluorescence detection. Results: In the absence of UTP and GTP infusion of e-ATP resulted in the following breakdown products during a single column passage (24 s): e-ADP 29%, e-AMP 3.9%, and e-adenosine 7.6%. However, in the presence of UTP (100 µM) the e-ADP and e-adenosine fractions were decreased to 12.6 and 3.4%, respectively, while the e-ATP fraction was increased from 59.5 to 79.6%. Quantitatively similar results were obtained in the presence of 100 µM GTP. Directionally similar, but quantitatively smaller effects were obtained using UTP and GTP at 10 µM. When isolated HUVEC were incubated with 14C-ADP a significant fraction (14 %) of the label was recovered as labeled ATP after 1.5 min. Conclusions: The results are consistent with extracellular phosphorylation of ADP to ATP. This reaction is most likely mediated by ecto-nucleotidase kinase activity via a phosphor transfer reaction. Institut für Physiologie, Medizinische Fakultät Carl Gustav Carus, TU Dresden, Fetscherstr. 74, 01307 Dresden P 40-3 BREAKDOWN OF ENDOTHELIAL BARRIER FUNCTION IN MARBURG VIRUS INFECTIONS M. Böckeler, U. Ströher, J. Seebach, H. Feldmann, H.-J. Schnittler Marburg virus (MBGV) and Ebola virus (EBOV) cause a severe hemorrhagic fever in humans with lethality up to 89%. The cause of death is a fulminate shock that results from direct and indirect alterations of the vascular system. Pro inflammatory agents such as cytokines, peroxides and proteases, which are released from virus infected monocytes and macrophages, cause endothelial activation as well as fluid distribution problems. In cell culture models, TNF-α has been shown to significantly contribute to the decrease in the barrier function during MBGV infections. With the present study we would like to contribute to the understanding of the mechanisms leading to changes in barrier function. We investigated the tyrosine phosphorylation of the junction proteins of endothelial cells following exposure to culture supernatants of MBGV-infected macrophages. The observed decrease in the endothelial barrier function was completely independent of tyrosine phosphorylation of the VE-cadherin/catenin complex. However, a moderate tyrosine phosphorylation of PECAM-1 could be detected. Comparable results were obtained following treatment of endothelial monlayers with TNF-α/H2O2 (100ng/ml, 10µM). Immunofluorescense investigations further displayed that the TNF-α/H2O2- induced breakdown of the barrier function was associated with a moderate increase in tyrosine phosphorylation of cell-cell junctions and an intensive staining of focal contacts. Stress fiber formation was also observed indicating that focal contacts and the actin filament cytstoskeleton were activated. Thus, tyrosine phosphorylation of PECAM-1 seems to be an important component in the decrease of the endothelial barrier function. The presented data indicate that the break down of the barrier function during filovirus infection and following TNF-α/H2O2- treatment is mediated by the activation of the actin filament system, but is independent of the phosphorylation of the VE-cadherin/catenin complex.

n=4), which was abolished by in the presence of the superoxide scavenger superoxide dismutase (SOD, 500U/mL; p<0.05), and inhibition of endothelial NADPH-oxidase using diphenylene-iodoniumchloride (DPI, 30µM; p<0.01). The activity of endothelial ectonucleotidases, as assessed by release of PO4- and amount of AMP derived from ADP in presence of the 5'-nucleotidase-inhibitor AOPCP, was reduced by chronic depolarisation over 24 hours from 0.54 +/- 0.16 to 0.39 +/0.11 µmol/min/µg protein (AMP derived from ADP, n=6, p<0.05). This was not the case in the continuous presence of SOD (n=6). Viability of cells after 24 hours of depolarisation (trypan blue exclusion) was not affected by depolarisation (n=4). Solutions of ADP (4µM), that had been preincubated with chronically depolarised cells (24h) induced irreversible aggregation. In contrast, the same solutions did not induce aggregation when incubated with depolarised cells in the presence of SOD or with control cells. We conclude, that membrane-potential dependent endothelial O2- production is capable of inhibiting the activity of endothelial ectonucleotidases. Therefore, chronic depolarisation, which seems to accompany situations of pathophysiologically high blood pressure, might favour platelet aggregation. Institute for Physiology, LMU, Schillerstr, 44, 80335 München P 40-5 RESTING POTENTIAL OF ENDOTHEL CELLS IN THIN SLICE OF GUINEA PIG HEART IS LITTLE AFFECTED BY HYPOXIA C. Malik, G. Ettorre, A. Scholz Tonus of vessels in the heart are proposed to be modulated by various factors including low oxygen tension. The vaso-relaxation is usually accompanied by a hyper-polarization of the resting potential. In a thin slice preparation from guinea left ventricle it was found that endothel cells in small vessels (20 – 80 µm) possessed voltage dependent delayed rectifier K+ current. The underlying current was described to be mainly the Kv1.3. We wanted to investigate in endothel cells whether this current or the resting potential might be modulated by hypoxia. We prepared thin slices (100 µm) from the left ventricle of guinea pigs (290–380 g) with a vibratome. Endothel cells in cut-open small vessels on the top of the slices were investigated with the patch-clamp technique. Bicarbonate-buffered control solutions were bubbled with different O2 partial pressure and the O2 tension was measured with a Paratrend 7© directly in the inlet of the bath perfusion. In 55 endothel cells, in identified arterioles, we found no significant changes in resting potential (-46.7 ±20.9 mV vs –42.9 ±24.9 mV mean ±S.D. in hypoxia 6-15 torr). Even prolonged application (15-25 min) we could not observe significant changes in these endothel cells. The majority of endothel cells, in identified venules, displayed a small leakage current and a prominent voltage dependent K + current. The peak amplitude was nearly unchanged in hypoxia (5-40 torr) from 261 ±108 pA to 249 ±137 pA, respectively. The result suggests that neither the resting potential nor the voltage-dependent K + current is largely affected, therefore other factors in parallel with hypoxia might result in vasorelaxation of cardiac vessels. Physiologisches Institut, Justus-Liebig-Universität, Aulweg 129, D35392 Gießen

Institute of Physiology TU-Dresden, Fiedlerstraße 42, 01307 Dresden, Germany. P 40-6 P 40-4 ENDOTHELIAL ECTONUCLEOTIDASE-ACTIVITY IS DECREASED BY DEPOLARISATION-INDUCED SUPEROXIDE PRODUCTION: IMPLICATIONS FOR PLATELET AGGREGATION F. Krötz, T. Riexinger, H.Y. Sohn, T. Gloe, M. Keller, B.F. Becker, U Pohl We could earlier demonstrate, that depolarisation of cultured human endothelial cells (HUVEC) is associated with an enhanced superoxide (O2-) production. Some ectonucleotidases, like NTPDasel (CD39) are known to be regulated in a redox-dependent manner. We investigated, whether depolarisation affects the activity of endothelial ectonucleotidases through O2- and whether this has effects on ADPinduced platelet aggregation. Depolarisation using the cation-ionophor gramicidin (100nM) induced a significant increase of O2- production in endothelial cells (+53 +/- 17% vs. control, DCF-flourescence, p<0.05,

POSTNATAL DEVELOPMENT OF PLASMA CATECHOLAMINE RESPONSES TO HYPOTENSIVE STRESS IN THE FOAL S.J. O’Connor, D.S. Gardner, J.C. Ousey, N. Holdstock, P. Rossdale, A.L. Fowden and D.A. Giussani Newborn foals are particularly susceptible to episodes of acute hypotension, especially if they are premature (Webb et al. 1984). Baroreflex-mediated neural and endocrine responses to acute hypotension in other species include increased sympathetic outflow and release of vasoconstrictor hormones (Nuyt, et al. 2001). However, defence responses to acute hypotension in the foal are unknown. This study investigated the plasma catecholamine response to acute hypotensive stress during the early postnatal period in the foal. Seven newborn foals were surgically prepared with hind limb arterial and venous catheters under general anaesthesia (halothane) at 2-3 days of age. At 1 and 2 weeks of age foals were placed in a custom made sling in close proximity to their mothers. Hypotension was induced in the foals by intravenous infusion of sodium nitroprusside (50µg.kg-1.min-1

S 328 for 10 minutes). Arterial blood pressure was monitored from 10 minutes before until 10 minutes after the end of the infusion. Arterial blood samples were taken every 5 minutes for measurement of arterial blood gases, pH and measurements of plasma catecholamines via HPLC (Silver, et al. 1987). Arterial blood gases and pH remained unchanged from baseline throughout the experimental protocol at both ages. There were no significant differences between basal plasma catecholamine concentrations in week 1 (noradrenaline: 104.3 ± 25.69 pg.ml-1, adrenaline: 202.3 ± 46.0 pg.ml-1) and week 2 (noradrenaline: 99.58 ± 12.18 pg.ml -1, adrenaline: 121.3 ± 20.4 pg.ml -1). Sodium nitroprusside infusion produced falls in arterial pressures, which were of similar magnitude at week 1 (51.4 ± 3.2 mmHg) and week 2 (48.6 ± 12.6 mmHg) of postnatal life. Acute hypotension increased the plasma concentrations of noradrenaline (P < 0.05), but not adrenaline, and decreased the ratio of adrenaline: noradrenaline in plasma (P < 0.05). Although no significant differences were measured in the increment in plasma noradrenaline during acute hypotension between week 1 and week 2, the change in the ratio of adrenaline: nor adrenaline was significantly attenuated at week 2 (-0.56 ± 0.3) relative to week 1 (4.18 ± 1.6; P < 0.05). These data suggest that with advancing postnatal age, greater concentrations of adrenaline relative to noradrenaline are released into the circulation during acute hypotensive stress. Nuyt, A.M., Segar, J. L., Holley, A.T. & Robillard, J.E. (2001). Pediatric Research. 49 (1): 56-62. Silver, M., Fowden, A.L., Knox, J., Ousey, J.C., Franco, R. & Rossdale, P.D. (1987). J. Reprod. Fert., Suppl. 35: 607- 614. Webb, A.L. et al. (1984) Equine Vet. J. 16: 319-323.

renin release and blood pressure regulation. In the recent past, two basic mechanisms have been attracted most attention: 1) Shear stress dependent release of nitric oxide (NO) by endothelial cells has been proposed to act as a local feedback system to control vessel diameter in the kidney. 2) The activity of the neuronal NO-synthase isoform has been suggested to influence pressure dependent renin release at the site of the macula densa. It remains unclear, however, to which extend longer - term (24h) blood pressure oscillations or controlled reductions in renal perfusion pressure (RPP) can modulate renal mRNA levels of the NO-synthase isoforms and NOS activity in vivo. To clarify this question the following protocols were used: Con: 24 hour time control without servocontrol of RPP; P90, P85, P70: RPP was fixed over 24h to 90, 85, or 70mmHg respectively; OSC: RPP was sinusoidally oscillated with a frequency of 0.1 Hz (amplitude: ±10mmHg) around the same mean value used during P85. Maintaining mean RPP at 84±1mmHg increased blood pressure to 143±4mmHg (P<0.01 vs. Con). Superimposing BP oscillations on this RPP attenuated the hypertensive effect of P85 (129±5mmHg, P<0.05 vs. P85). Nonetheless mRNA levels of e-NOS and b-NOS, as determined by means of RNase protection assays, differed only slightly between all protocols (mRNA levels of i-NOS remained at the detection limit). In addition, none of the protocols induced major changes with respect to NADPH diaphorase staining at the site of the vessels, macula densa cells or renal tubules. Therefore, neither 24 hour pressure reductions nor 24h 0.1Hz BP oscillations seem to have lasting effects on the renal NO system in freely moving rats.

Department of Physiology, University of Cambridge, Cambridge CB2 3EG, UK

Medizinische Fakultät der Humboldt Universität (Charité), Johannes Müller - Institut für Physiologie, Tucholskystraße 2, 10117 Berlin, email: [email protected]

P 40-7 NITRIC OXIDE LIBERATION FROM S-NITROSO-N-ACETYLPENICILLAMINE IS LARGELY INFLUENCED BY IRON CONCENTRATION C. Planitz, T. Lauer, A. Deussen The natural metabolism of nitric oxide (NO) is of great interest in biological systems under physiological and pathophysiological conditions. Investigations of its chemical interactions with various compounds have most frequently employed synthetical NO-donors as in vitro sources. In a former study we found a decrease of the NOliberation from S-Nitroso-N-Acetylpenicillamine (SNAP) in the presence of bovine serum albumine (BSA). To elucidate whether this effect represented a NO-quenching or an alteration of the NO-liberating kinetics of SNAP by BSA, we investigated the influence of BSA and several physiologically occuring bivalent metal ions on the NO-release of SNAP. Methods: NO was measured as its first oxidative derivative nitrite by the GRIESS reaction in a flow injection analysis system. Frozen samples of 200 µM SNAP were diluted to 20 µM by addition of salt solutions (control=aqua bidest., NaCl, MgCl2, CaCl2, FeCl2:0,8 mM). In additional experiments we added EDTA resp. BSA in a salt-containing HEPES buffer. Thawed samples were immediately measured for NO2 content (t=0). After incubation for 60 min at 37°C a second NO 2 measurement was performed. Results: NaCl, MgCl2, CaCl2 solutions showed increases in NO2 - concentrations similar to bidistilled water. FeCl2 caused a nearly five-fold increment in the NO-liberation rate. While EDTA induced a marked decrease in NO-production more than 80% at a concentration above 1 µM, BSA in concentration dependently caused a more gradual decrease. Conclusion: The NO-production rate of SNAP does not only depend on the surrounding temperature, but is also influenced by the concentration of iron ions and decreased by BSA and EDTA. These results have implications for the use of SNAP as a NO-donor in protein-containing solutions as well as in biological systems. Institut für Physiologie, Medizinische Fakultät Carl-Gustav Carus, Technische Universität Dresden, Fetscherstr. 74, D-01307 Dresden P 40-8 NEITHER 24H BLOOD PRESSURE OSCILLATIONS NOR GRADED PRESSURE REDUCTIONS HAVE LASTING EFFECTS ON RENAL NOS GENE EXPRESSION OR NOS ACTIVITY IN THE FREELY MOVING RAT B. Nafz*, H. Welp, F. Theilig, S. Bachmann, P.B. Persson Nitric oxide has been proposed to play a major role in modulating renal

P 40-9 A NEW METHOD TO MEASURE HAEMOGLOBIN OXYGEN SATURATION IN MICROVESSELS B. Zegenhagen-Phiong1, A. R. Pries1,2 Introduction: Many efforts have been made to measure local oxygen saturation (SO 2) in microvessels. For intravital microscopy, optical methods are specially suited. These are based on the different absorption spectra of oxygenated and deoxygenated haemoglobin with measurements performed at two or more wavelengths. Conventional methods are complicated due to the need of a respective number of optical pathways and video cameras. In contrast, recent imaging systems allow a fast recording of pictures at preselected wavelengths using a single optical path. The aim of this study was to adapt an imaging system for measuring oxygen saturation in microvessels. Methods: An intravital microscope with an imaging system (T.I.L.L. Photonics, Gräfelfing, Germany) was modified for trans-illumination. Validating experiments were made using glass capillaries filled with human haemoglobin solutions, rat erythrocyte suspension and rat whole blood at physiological pH and different SO2. Absorption spectra from 400 nm to 620 nm were recorded and analysed using the optical theory of Pittman et al. (1975) calculating oxygen saturation from extinction values measured at two isosbestic wavelengths and one wavelength with a maximum difference. Alternatively a four wavelength method was used. Two pairs of wavelengths are chosen so that E(λ1)/E(λ2) ≈ 1 at high SO2 (first pair) and at low SO2 (second pair). Extinction ratios depending on SO2 follow a linear curve fit function with the slope including a scattering factor. SO2 calculation from this function leads to results disregarding scattering. Results: For a given SO2, in all measured spectra the position of maximal/minimal values as well as their relative heights were conserved for all samples. Curve amplitudes are reduced for particle suspensions due to light scattering. Location of isosbestic wavelengths is not conserved rendering Pittman’s three wavelength approach problematic (mean deviation: ∆SO2 = 23%). Estimation at four wavelengths proves reliable at SO2 from 37% to 100% (mean deviation: ∆SO2 = 3,4%), yet problematic at 0% (mean deviation: ∆SO2 = 34%). Conclusions: Modern imaging systems allow flexible access to oxygen saturation in microvessels. Photometric methods based on isosbestic wavelengths are problematic due to optical effects like sieving and scattering. A four wavelength method based on relative heights within a spectrum proves to be suitable for medium and higher SO2 values. Methods based on analysis of spectral shape rather than distinct wavelengths may be more valid and reliable over a large SO2 range. Supported by DFG (FOR 341/1)

S 329 1

Dept. of Physiology, Freie Universität Berlin, Arnimallee 22, D-14195 Berlin. 2Institute of Anesthesiology, Deutsches Herzzentrum Berlin, D-13353 Berlin

Institut für Kardiovaskuläre Physiologie, Klinikum der JWG-Universität Frankfurt, Theodor-Stern-Kai 7, 60590 Frankfurt/Main P 40-12

P 40-10 THE Ca2+ ANTAGONISTS, AMLODIPINE AND NIFEDIPINE ACTIVATE THE ENDOTHELIAL NITRIC OXIDE SYNTHASE BY ALTERING PHOSPHORYLATION ON SER1177 AND THR495 H. Lenasi, B. Fichtlscherer, A. Mülsch, R. Busse, I. Fleming Ca2+ antagonists such as nifedipine (Nif) and amlodipine (Aml) increase the generation of nitric oxide (NO) from native and cultured endothelial cells. As endothelial cells do not express voltage-dependent Ca2+ channels, the cellular mechanisms underlying this phenomenon are unclear. The aim of this investigation was to determine whether or not the activation of the endothelial NO synthase (eNOS) by Ca2+ antagonists is related to alterations in eNOS phosphorylation. In isolated, endothelium-intact porcine coronary arteries, Nif and Aml elicited an NO-mediated relaxation and a leftward shift in the concentrationrelaxation curve to bradykinin. A similar response was observed using the protein kinase C inhibitor Ro 31-8220. Nif increased the generation of NO, as detected by electron spin resonance spectroscopy; the rate of formation of paramagnetic NOFe-diethyldithiocarbamate being 60±20 and 120±20 nM/10 min in control and Nif (3 µM)-stimulated cells, respectively. In cultured endothelial cells, the Ca2+ antagonists elicited an 8-fold increase in cyclic GMP content within 10 minutes. Although EGTA attenuated Nif-induced NO production, neither Nif nor Aml elicited a Ca2+ transient in endothelial cells, or affected the Ca 2+ response to bradykinin. In unstimulated endothelial cells, eNOS was not phosphorylated on Ser 1177 but constitutively phosphorylated on Thr 495 . Aml however, elicited the phosphorylation of Ser1177 and attenuated Thr 495 phosphorylation, with a time course similar to that of eNOS activation. Inhibiting PKC induced similar changes in eNOS phosphorylation i.e., an increase on Ser1177 and a decrease on Thr495.These results indicate that the Ca2+ antagonists, Nif and Aml, enhance endothelial NO generation via a mechanism related to PKC-mediated alterations in the phosphorylation of eNOS. Institut für Kardiovaskuläre Physiologie, Klinikum der J.W.G.Universität, D-60590 Frankfurt am Main

P 40-11 THE EXPRESSION AND ACTIVITY OF CYP 2C9 IN CORONARY ARTERY ENDOTHELIAL CELLS IS ENHANCED BY CYCLIC STRETCH B. Fisslthaler, R. Popp, U. R. Michaelis, I. Fleming, R. Busse Coronary arteries are subjected to pronounced rhythmic variations in diameter and blood flow, stimuli known to modulate the expression of a variety of genes. Epoxyeicosatrienoic acids (EETs) are metabolites of cytochrome P450 (CYP) 2C which play a crucial role in the nitric oxide/prostacyclin-independent relaxation in the coronary circulation. The purpose of this study was to determine the effect of cyclic stretch and fluid shear stress on the expression and activity of the coronary CYP 2C. In cultured porcine coronary endothelial cells, acute application of cyclic stretch (6%, 1 Hz, 10 min) elicited the generation of 8,9-, 11,12-, and 14,15-EET. Prolonged stretch and shear stress (3 to 12 dynes/cm2, 24 hours) increased the expression of CYP 2C mRNA and protein 5- to 10 fold, which was accompanied by a 4- to 8- fold increase in EET-generation. A corresponding increase in CYP 2C mRNA and protein expression was also observed in endothelial cells of pressurized porcine coronary artery segments perfused under pulsatile conditions (calculated circumfential strain 6-8%, 1 Hz) for 6 hours. Inhibition of the epidermal growth factor receptor (AG1478, 500nM) in cultured endothelial cells attenuated the stretch-mediated increase in CYP 2C protein expression, however, inhibitors of tyrosine kinases, phosphatidyl inositol-3 phosphate kinase and p38 mitogen activated protein kinase, all of which are activated by cyclic stretch, had no influence on the stretch-induced increase in CYP 2C expression. These results have identified the coronary CYP 2C as a novel mechanosensitive gene product in native and cultured endothelial cells. As this enzyme generates EETs as well as superoxide anions this finding has wide reaching implications for vascular homeostasis in conditions of manifest endothelial dysfunction.

ELLIPSOMETRY STUDIES OF LIPOPROTEIN DEPOSITION AT THE ENDOTHELIUM AND INTERACTION WITH AN HMG-CoA REDUCTASE INHIBITOR G. Siegel*, C. Abletshauser, M. Malmsten Proteoheparan sulfate (HS-PG) can be adsorbed to a methylated silica surface in a mo-nomolecular layer via its transmembrane hydrophobic protein core domain [3]. Due to electrostatic repulsion, its anionic glycosaminoglycan (GAG) side chains are stretched out into the blood substitute solution representing a receptor site for specific lipoprotein binding through basic amino acid-rich residues within their apolipoproteins [4]. The lipid binding process was studied by ellipsometric techniques. Low density lipo-protein (LDL) was found to deposit strongly at HS-PG, particularly in the presence of Ca2+ thus creating the complex formation PG/LDL/Ca2+ [2]. This ternary complex build-up may be interpreted as arteriosclerotic nanoplaque formation on the molecular level responsible for the arteriosclerotic primary lesion. On the other hand, HDL bound to HS-PG protected against LDL docking and completely suppressed calcification of the proteoglycan-lipoprotein complex [4,5]. Although much remains unclear regarding the mechanism of lipoprotein depositions at proteoglycan-coated surfaces, it seems clear that the use of such systems offers possibilities for investigating lipoprotein deposition at a ‘nanoscopic’ level under close to physiological conditions. Therefore, we tested the system on its reliability in a biosensor application (PCT/EP 97/05212, USPTO 09/319,970) in order to unveil possible acute pleiotropic effects of the lipid lowering drug fluvastatin. Already with a normal blood Ca 2+ concentration, the VLDL/IDL/LDL plasma fraction from a high risk patient with dyslipoproteinemia and type 2 diabetes mellitus showed beginning arteriosclerotic nanoplaque formation, that massively increased at higher Ca2+ concentrations. Fluvastatin, whether applied to the patient (one single 80 mg slow release matrix tablet) or acutely in the experiment (2.2 µmol/l), markedly slowed down this process of ternary aggregational nanoplaque complexation at all Ca2+ concentrations used. This action resulted without any change in lipid concentrations of the patient. In an advanced approach, we wanted to even closer mimic the in vivo-situation, replacing the ‘silica surface membrane’ by the natural membrane of endothelial cells in order to examine whether arteriosclerotic nanoplaques can likewise form at the in vivo-proteoglycan receptor [1]. Ellipsometry measurements illustrated both the stability of the model surface and the Ca2+dependent binding of lipoproteins at the cell-based substrate. [1] Malmsten M, Siegel G, Becker A (2001) Colloid Interface Sci 240:372374; [2] Malmsten M, Siegel G, Wood WG (2000) J Colloid Interface Sci 224:338-346; [3] Siegel G, Malmsten M, Klüßendorf D, Leonhardt W (1999) Atherosclerosis 144:59-67; [4] Siegel G, Malmsten M, Klüßendorf D, Michel F (2001) Biosensors & Bioelectronics 16: 895904; [5] Siegel G, Malmsten M, Michel F (2000) Int J Angiol 9:129134. *Institute of Physiology, Biophysical Research Group, The Free University of Berlin, Arnimallee 22, DE-14195 Berlin, Germany P 41-1 REGULATION OF L-TYPE CA2+ CHANNELS OF HUMAN TRABECULAR MESHWORK CELLS BY UNOPROSTONE K. Steinhausen1, H. Thieme1, M. Wiederholt1, A. Ottlecz2, G. Lambrou2 Purpose: The docosanoid unoprostone has been shown to reduce intraocular pressure (IOP) in patients with both ocular hypertension or primary open angle glaucoma. It relaxes trabecular meshwork (Ophthalmologica 211: 153, 1997), a smooth-muscle like tissue involved in the regulation of ocular outflow in the eye. We tested the effect of unoprostone on L-type Ca2+ currents of human trabecular meshwork cells. Methods: Cultured human trabecular meshwork (HTM) cells were investigated using the perforated patch configuration of the patch-clamp technique. Results: In trabecular meshwork, unoprostone reversibly inhibitied L-type Ca 2+ currents. This effect was dose-dependent. Application of unoprostone 10 -6 M lead to a significant reduction to 90 ± 1.5% (p < 0.01, n= 6; recovery 99.4 ± 0.6 %) of control current, a concentration of 10 -5 M reduced control current to 81 ± 2.6 % ( p< 0.01, n= 11, recovery 97 ± 1%), and

S 330 unoprostone 10-4 M lead to a reduction of 77 ± 4.5% (p< 0.05, n= 4, recovery 94 ± 4%). After preincubation of HTM cells with the tyrosine kinase inhibitor herbimycin (10 -6 M), application of unoprostone (105 M) had no effect on L-type Ca2+ current amplitude. These cells showed a significant (p<0.05) reduced current density (11 ± 1.7 pA/pF, n=11) than control cells (28 ± 4 pA/pF, n= 35) and inactivated much slower: the inactivation time constant (τ/ms) was significantly increased (475 ± 53 τ/ms, n= 8 versus 283 ± 32 τ/ms, n=7, p<0.05). Conclusions: In human trabecular meshwork cells unoprostone inhibits L-type Ca2+ currents. This effect is mediated by tyrosine kinases. An inhibition of L-type Ca2+ currents leads to a decline in [Ca2+]i and contributes to the relaxation observed in TM after application of unoprostone. Support: Novartis Ophthalmics, Basel, Switzerland 1

Institut für Klinische Physiologie, Klinikum Benjamin Franklin, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany. 2 Novartis Ophthalmics, Basel, Switzerland. P 41-2 RAT MYOMETRIAL MYOCYTES: STRETCH AND CELL INFLATION MODULATE ISAC AND ICa,L INDEPENDENTLY T. Shmygol, S. Wray, V. Kazanski#, G. Isenberg# Stretch is a physiological stimulus for the uterus. Myometrial strips respond to mechanical stretch with rhythmical Ca 2+ transients and contractions, presumably, mediated by both Ca2+-entry and Ca2+-release (Kasai et all, 1995). We have investigated whether stretch-induced Ca 2+ entry is caused by stretch-activation of non-selective cation channels (current I SAC) or modulation of the L-type Ca2+ channels (current ICa,L ). Myometrial myocytes isolated from uteri of pregnant (day 21) rats were superfused with Tyrode (37oC) solution and voltageclamped (CsCl-electrode solution) with 175 ms pulses from –70 to 0 mV (0.2 Hz). Mechanical stimuli were a) axial stretch, by increasing the distance between the glass stylus and patch pipette b) cell inflation, by applying positive pressure from the patch pipette and increasing cell diameter by approx. 10%), and c) cell swelling by dialysis with hypertonic pipette solution (250 instead of 150 mM CsCl) . ISAC. Axial stretch induced a negative ISAC with a reversal potential of –5 mV and a modest outward rectification. Inflation induced I SAC increased within one min to a negative maximum (-0.3 nA), and disappeared within 5 min (11 of 16 experiments). Osmotic cell swelling never induced ISAC (n=8). ICa,L. Osmotic cell swelling had no effect on ICa,L (n=8). Axial stretch reduced ICa,L by less then 20% (10 of 14 cells). Inflation increased peak ICa,L by approx. 60% (8 of 14 cells), or it reduced it (5 of 14 cells). Microscopy suggests that the increase in ICa,L correlates with a separation of the cytosol from the surface membrane. When inflation increased both ICa,L and ISAC (4 out of 8 cells), then ISAC followed the increase of I Ca,L with a 30 s time delay. In 4 experiments inflation augmented ICa,L without any effect on ISAC. Conclusions: 1) ISAC is thought to flow through non-selective cation channels because its properties resemble ISAC in other cells. 2) Osmotic cell swelling does not modulate ISAC and ICa,L. 3) The signals by which stretch and inflation increase ISAC and I Ca,L are different, speculatively, we attribute these signals to the subsarcolemmal cytoskeletal cortex and the cytoplasmic stress fibres. Departments of Physiology, University of Liverpool, U.K. and #Martin-Luther-University, D-06097 Halle, Germany P 41-3 CHARACTERISATION OF OUTWARD POTASSIUM CURRENTS IN ISOLATED UTERINE CELLS FROM NEONATAL RATS Y. Li, T. Burdyga, T. Shmigol, R.D. Smith, S. Wray Major changes with development in the expression of ion channels have been reported for many cell types, but have been little studied in smooth muscle. As such changes would be expected to play an important role in controlling excitability and function, we investigated the outward currents present in neonatal uterine cells. 13-15 days old female rats were humanely killed by cervical dislocation following CO 2 anaesthesia .The whole- cell patch clamp technique was used to measure outward K+ currents in enzymatically isolated uterine single cells. Total outward currents (cell capacitance = 5.99 ± 1.15pF,n=50) were evoked by depolarising pulses from –40mV to 70 mV, from a holding potential of –80 mV. The external (bath) normal or 0-Ca Krebs solution contained (mM): 154 NaCl; 5.4 KCl; 1.2 MgSO4; 2 CaCl2; 11 HEPES (pH7.4).

The intracellular pipettes solution contained (mM): 130 KCl; 15 NaCl; 0.3 MgCl2; 0.043 CaCl2; 0.1 EGTA; 1 Na2ATP; and 10 HEPES (pH 7.2). All drugs were applied to the external bath solution. The data are given as means ±S.E.M and were recorded at 32oC or room temperature for kinetic measurements. Three distinct K+ channels have so far been found: 1) A rapidly inactivating transient current(”A” type current) was present in 94% cells (n = 50).It had a half-activation voltage of 8.8mV±1.4mV , half-inactivation –73.5±5.6mV(n=6) and a time constant of activation of ton=1.29±0.58 ms and toff = 8.1±0.2 ms at +70 mV(n=10). It was sensitive to 4- aminopyridine in a dose dependent manner, and its amplitude was reduced by exposure to Ca²+-free, 1mM Cd²+ or 400nM iberiotoxin solutions. 2) A delayed rectifier K+ currents that showed slow activation and noninactivation, was sensitive to tetraethylammonium (TEA), was also found in all cells. It had a halfactivation voltage of 12.2±5mV (n=6)and time constant of activation of ton=60±6ms at 70mV. 3) In 10% cells a Ca²+-sensitive K + current that was blocked by 400nM charybdotoxin, 400nM iberiotoxin and 1mM Cd²+ was found. Interestingly no inward Ca2+ current could be recorded from these cells under our experimental conditions. This observation and the large amount of A type current in these cells, compared to adults, leads us to suggest that ion channel expression is such as to dampen the excitability of the neonatal uterus. Department of Physiology, The University of Liverpool, Liverpool, L69 3BX, U.K. P 41-4 CURRENTS UNDERLYING SPONTANEOUS AND NORADRENALINE-INDUCED ACTIVITY IN ISOLATED SHEEP LYMPHATICS E.A.H. Beckett, K.D. Thornbury, M.A. Hollywood, N.G. McHale Intracellular microelectrode recordings were made from smooth muscle of sheep mesenteric lymphatics to investigate the electrophysiological basis of lymphatic pumping. Sheep mesenteries were obtained from the local abattoir and 1-2 cm lengths of lymphatic were mounted in a sylgard-bottomed bath (1968) and superfused with Krebs solution at 37°C. Smooth muscle cells were impaled with glass microelectrodes filled with 3M KC1 having resistances of 80 - 120 MΩ. Resting membrane potential varied from -49 to -65 mV (mean -57.5 ± 4.1, SD, n=57) Approximately 50% of the vessels showed spontaneous electrical and mechanical activity at frequencies of 2.5 - 8 min-1 (4.8 ± 1.4, SD, n=32). The basic pattern of electrical activity consisted of slow depolarisations (dV/dt 0.33 ± 0.06 V sec-1) of relatively low amplitude (16.3 ± 2.0 mV, n=12) and with duration at half amplitude of 0.79 s ± 0.44, n=9 . Spike complexes generally consisted of one or more action potentials superimposed on this ‘plateau’ phase and were often preceded by a slow diastolic depolarization or ‘pre-potential’. Noradrenaline (1 µM) induced depolarizing events in quiescent preparations. Both agonist induced oscillations and spike complexes were attenuated or completely abolished by the putative IP3 receptor antagonist 2-aminoethoxydiphenyl borate (2-APB). Caesium reduced the frequency of spontaneous firing by approximately 30% by flattening the pre-potential phase. In addition to having a negative inotropic effect, Cs+ also caused gradual membrane depolarization and prolonged the plateau. The importance of ICa(L) in spontaneous action potential generation was highlighted since nifedipine abolished spontaneous events. INa appeared to underlie the rapid upstroke of the action potential as tetrodotoxin (1 µM) either attenuated action potential amplitude or stopped activity completely. Spontaneously active segments of lymphatic vessel were inhibited by the chloride channel blocker, anthracene9-carboxylic acid suggesting that ICl(Ca) plays a significant role in the generation of spontaneous activity in this tissue. Smooth Muscle Group, Department of Physiology, Queen’s University, 97 Lisburn Road, Belfast BT9 7BL P 41-5 ACTIVATION OF IP 3 RECEPTORS TRIGGERS RYANODINE SENSITIVE CALCIUM RELEASE IN SMOOTH MUSCLE C. White, J.G. McGeown The release of Ca2+ stored in the sarcoplasmic reticulum (SR) of smooth muscle is mediated through inositol 1,4,5- trisphosphate (IP 3) and ryanodine receptors. We set out to investigate the interaction between these receptors in isolated rat gastric myocytes. Store depletion was

S 331

Smooth Muscle Group, Dept of Physiology, 97 Lisburn Road, Queen’s University of Belfast, BT9 7BL. U.K.

In vascular smooth muscle cells (SMCs) depletion of intracellular Ca2+ stores has been shown to induce a transsarcolemmal Ca2+ influx. The dependence of this so-called capacitative or store-operated Ca2+ entry (SOCE) on the extracellular pH (pHo) was the objective of this study. During Ca2+ store-depletion upon store Ca2+ uptake inhibition by BHQ (2,5 Di-(tert-butyl)-1,4-benzohydroquinone), cytosolic Ca2+ ([Ca2+] i) was monitored with the Fura-2/AM fluorometry on SMCs isolated from pig coronary arteries. BHQ-dependent isometric tension was recorded on smooth muscle strips. Moreover, SOCE into SMCs was identified with the Fura-2-Mn2+-quenching technique. BHQ (106mol/l) caused a plateau-like elevation of [Ca 2+]i in SMCs indicated by an increase of the Fura-fluorescence ratio (340nm/380nm) from 0.88±0.02 to 1.25±0.04 (mean±SE; p< 0.001; n=20). Moreover, BHQ increased the resting tension of muscle strips from 29.9±1.19 to 49.5±4.51 mN (p<0.001; n=22). With acidosis (pHo= 6.5) the BHQinduced increase in [Ca2+]i amounted to only 46.2±7% (p<0.001; n=9) of control (pH o= 7.4), whereas alkalosis (pHo= 8) increased [Ca2+]i to 144.6±11.4% (p<0.001; n=11). Likewise, the BHQ-induced increase in resting tension of muscle strips could be reduced by acidosis to 54.7±2.1% (p<0.001; n=9) and increased by alkalosis to 139.4±4.9% (p<0.001; n=13). The enhanced Mn2+ quench of Fura-2 fluorescence in the presence of BHQ (-224±16 counts⋅s-1/min vs. -40±7 under control), indicating store-operated entry of divalent cations, could be significantly reduced by acidosis (-46±9.6 counts⋅s-1/min; p<0.001; n=6), whereas alkalosis did not significantly affect the quench rate. Conclusions: the BHQ-induced increases in [Ca2+]i and in resting tension may be partly due to SOCE, which, according to the Mn2+-quenching, can be reduced by extracellular acidosis. Thus, inhibition of store-operated Ca2+ entry may be a component of acidosis-dependent vasodilatation.

P 41-6

Study Group for Ca2+ Antagonism, Department of Physiology, University of Freiburg, Hermann-Herder-Str. 7, D-79104 Freiburg, Germany

directly visualized using confocal microscopy and the low affinity indicator Oregon Green BAPTA-5N AM[2]. Caffeine (10mM) and carbachol (10µM) reversibly depleted identical regions of the store, suggesting that they evoked Ca2+ release from a single SR pool. Double labelling of IP3 receptors (anti-IP3 receptor polyclonal antibody; 1:100 dilution; Calbiochem) and ryanodine receptors (BODIPY® TR-X ryanodine, Molecular Probes) in the same cells revealed a high degree of co-localization. This is consistent with the idea that they may be coupled to the same SR Ca2+ pool. Further experiments were carried out using fura-2 to measure global changes in cytoplasmic [Ca2+] in gastric myocytes. The response to carbachol was inhibited by both ryanodine (100µM) and 2-aminoethoxydiphenyl borate (2APB, 100µM), an inhibitor of IP3 receptors[1], whereas the caffeine response was inhibited by ryanodine but enhanced by 2APB. Local changes in cytoplasmic [Ca2+] were also imaged using fluo-4 and confocal line scans. These demonstrated that responses to caffeine and carbachol were initiated from the same sites within the cell and that these sites often coincided with Ca2+ spark sites. Spontaneous and agonist evoked events were all blocked by ryanodine but only carbachol responses were inhibited by 2APB. From these results we conclude that carbachol responses are IP3 dependent but that the majority of the Ca 2+ response to carbachol depends on release of a ryanodine sensitive store probably as a result of ryanodine receptor activation. 1. Maruyama T, Kanaji T, Nakade S, Kanno T, Mikoshiba K. (1997) J. Biochem. 122:498-505; 2. White, C, Montague, S, McGeown, JG. (2000) J. Physiol. 526P: 106P. Funded by the Wellcome Trust.

EXTRACELLULAR APPLICATION OF THE SCORPION TOXIN, IMPERATOXIN I, INHIBITS CA 2+ STORE RELEASE IN RAT GASTRIC MYOCYTES L.E. Burns, C. White, U. Anderson, J.G. McGeown Imperatoxin I (IPTxI) is a protein from the venom of the scorpion Pandinus imperatur. It inhibits ryanodine binding and channel opening in sarcoplasmic reticulum (SR) from cardiac and skeletal muscle[1], an inhibition which depends on phospholipase A2 activity rather than direct interaction with the ryanodine channel[2]. This suggests that the toxin might have an effect when applied extracellularly and we set out to test this in smooth muscle cells. Gastric myoctes were isolated from the antral region of the rat stomach and loaded with fura-2AM to allow measurement of [Ca2+]i. Superfusion with IPTxI alone elevated [Ca2+]i, the increase being transient at low concentration (0.5nM and 1.0nM) but sustained at higher concentrations (5nM and 20nM). Release from Ca2+ stores was assessed by applying caffeine (10mM in 0Ca2+ solution) and comparing the resulting [Ca 2+]i transients before and during exposure to IPTxI. Caffeine responses were reduced by 76±8% (mean±SEM) in the presence of 20nM IPtxI (P<0.005, n=7, paired t-test), and by 72±16% in the presence of 5nM IPTxI (P<0.05, n=6). In the presence of 1nM IPTxI, however, the average size of the caffeine transient was increased by 21±10% (P<0.05, n=6). These effects were fully reversible. Superfusion with 20nM IPTxI also reduced the mean [Ca2+]i response to carbachol (10µM in 0 Ca2+ solution), but this was not statistically significantly (P>0.5, n=5). We conclude that IPTxI can affect Ca2+ release from smooth muscle SR, even when applied extracellularly. It seems likely that these effects reflect its phospholipase A2 activity, releasing fatty acids which diffuse into the cytoplasm and modify Ca2+-release channel activity[2]. Funded by the Wellcome Trust; LEB is recipient of a studentship from the RDO of HPSSNI. 1. Valdivia H, Kirby MS, Lederer J, Coronado R (1992) Proc Natl Acad Sci 89: 12185-12189; 2. Zamudio FZ, Conde R, Arevalo C, Becerril B, Martin BM, Valdivia HH, Possani LD (1997) J Biol Chem 272: 11886-11894. Address for correspondence: Smooth Muscle Group, Dept of Physiology, 97 Lisburn Road, Queen’s University of Belfast, BT9 7BL. U.K.

P 41-8 A STUDY OF MOUSE SARCOPLASMIC RETICULUM AND MYOMETRIAL CONTRACTILITY A.J.G. Matthew, S. Wray Much has been learnt about uterine physiology and pre term labour from in vitro studies of animal and human myometrium. Despite this, much needs to be elucidated to enable prevention or better treatment of pre-term labours. The aim of this work was to answer questions concerning contractility and the sarcoplasmic reticulum (SR), using the mouse myometrium as a model tissue. Female CD-1 mice were killed by cervical dislocation under CO2 anaesthesia, and the uterus removed. After dissection of the myometrium, small strips (1mm X 3mm) were loaded with the fluorescent Ca 2+ indicator, INDO-1-AM. The SR was inhibited using the SR Ca++ ATPase (SERCA) inhibitor cyclopiazonic acid (10µM). SR release of Ca++ was determined in Ca2+ free EGTA (1mM) solutions and applying oxytocin (100nM) (n=4). The mouse myometrium produced regular, phasic contractions, which were proceeded by intracellular Ca ++ transients (n=20). Inhibition of the SR produced marked effects on force and calcium. There was a significant increase in the frequency of contractions from 1.15+0.155 contractions/min to 6.2+1.53 contractions/min. There was also an increase in the amplitude of contractions by 16.31+3.13 percent. Baseline calcium also increased. There was no spontaneous activity or Ca++ transients in the absence of extracellular Ca++, and baseline Ca++ decreased. The application of oxytocin under these conditions resulted in a small increase in cytosolic Ca++ and force, 7.61+1.24 percent of average spontaneous contractile force amplitude (n=5). These findings suggest that in the mouse myometrium, (1) extracellular Ca++ is essential for spontaneous force production, (2) the SR has the capacity for limiting force production, as both force and Ca++ increase when it is inhibited, and (3) there is an IP3 (agonist) releasable Ca++ store in the SR but it’s capacity is limited. This is consistent with SR Ca++ release onto Ca++ gated potassium channels (vectoral Ca++ release). The Physiological Laboratories, The University of Liverpool, Crown Street, Liverpool L69 3BX, Great Britain

P 41-7 P 41-9 STORE-OPERATED Ca 2+ ENTRY INTO SMOOTH MUSCLE CELLS OF PIG CORONARY ARTERIES IS INHIBITED BY EXTRACELLULAR ACIDOSIS J. Weirich, L. Dumont, G. Fleckenstein-Grün

DEVELOPMENT OF A NOVEL ELECTRODYNAMIC SMOOTH MUSCLE SPHINCTER Th. Noack, D. Karwath, C. Grub, P. Noack, A. Wree, S. Haas, T.

S 332 Wedel, U.T. Hopt, H.J. Schrag Objective: In the treatment of chronic fecal incontinence, surgical methods exist using a dynamic gracilis-construct or an artificial sphincter. The latter are mostly electromechanical valves which have the disadvantage of long term biological incompatibility. This study was performed to overcome these disadvantages by avoiding the use of non-biological mechanical components by far. Methods: The novel artificial electrodynamic smooth muscle sphincter (ESMS) has been developed and was examined in a new in vitro ileostomy model. The ESMS is based on electrorelaxation of an intestinal smooth muscle frill which occludes a partially isolated bowel. In a first step, the isometric tension from isolated circular porcine fundus and colon muscle strips was recorded during pharmacological and electrical field stimulation (EFS). The best suitable tissue was then used for the muscle frill. Experiments were carried out using a Krebs solution, gassed with Carbogen. Results: The relaxation response on prestimulated preparations by acetylcholine (ACh, 10 µM) using EFS (40 Hz, 60 mA, 2-5 ms, train duration 1 s) was significantly higher in fundus muscle than in colon preparations. The tissues responded with a small contraction which was followed by a dilation. In preparations from human subjects these dilations recovered slowly (τ = 27.5 s) . Preparations from porcine circular fundus recovered faster (τ =14.5 s) and the longitudinal fundus was very fast (τ =2.2 s). In colon muscle the relaxing component was not expressed. Tetrodotoxin inhibited all these effects, L-NNA abolished the inhibitory effect and atropin the excitatory effect after EFS. After free transplantation of a fundus muscle flap (6 cm long, 3 cm wide) around the isolated small intestine and simulation of the functional cholinergic activity in response to ACh (1-10 µM) a continent artificial sphincter was achieved. Electrical stimulation of the free nerve fibers of the fundus frill caused muscle relaxation with the concomitant reduction of the occlusion pressure and flow was maintained. This process was inhibited by the dynamic of the dilation. Conclusions: ESMS is a promising innovative experimental model and will be further developed in the surgical treatment of fecal incontinence.

As calyculin A has the largest effects it indicates that phosphatase type I is predominate in human myometrium. The Physiological Laboratory, The University of Liverpool, Crown Street, Liverpool, L69 3BX, U.K. P 41-11

Institut fur Physiologie, Universität Rostock, D-18055 Rostock

GENDER DIFFERENCES IN RELAXATIONS OF RAT AORTA A. Razak, J.R. Docherty We have investigated endothelium-dependent relaxations and relaxations to 17α-oestradiol (Oe) in aortic rings from male and female rats. Isometric contractions were obtained to methoxamine, phenylephrine (Phe) and KCl, and tissues were relaxed by cumulative concentrations of test agents.In methoxamine contracted tissues, acetylcholine produced significantly greater maximum relaxations in tissues from male (85.6±3.8%, n=6) than female (65.2±6.7%, n=6) rats and relaxations were abolished by the NO synthase inhibitors L-NAME or L-NNA. Oe (0.1-30 µM) produced significantly greater relaxations of KCl than Phe evoked contractions in both male and female, but for each contractile agent the maximum relaxation was greater in tissues from female rats. In KCl contracted tissues, L-NAME did not significantly affect relaxations to Oe in tissues from females but significantly reduced the relaxation in tissues from male rats. The guanylate cyclase inhibitor ODQ significantly reduced relaxations to Oe in tissues precontracted with KCl from male, but not female, rats. The calcium entry blocker nifedipine produced significantly greater relaxations of KCl than Phe evoked contractions in both male and female. In experiments carried out in calcium-free solution in which calcium stores were depleted, Oe significantly reduced the contraction to calcium restoration. It is concluded that relaxations to acetylcholine are significantly greater in aorta from male rats, but that Oe produces significantly greater relaxations in aorta from female rats and these relaxations do not involve NO, at least in females, but may involve restriction of calcium entry.

P 41-10

Department of Physiology, Royal College of Surgeons in Ireland, 123 St. Stephen’s Green, Dublin 2, Ireland

THE EFFECT OF INHIBITING MYOSIN PHOSPHATASE ON FORCE, CALCIUM AND MYOSIN PHOSPHORYLATION IN HUMAN MYOMETRIUM S. Kupittayanant, T. Burdyga, S. Wray Smooth muscle force can be regulated by myosin light chain kinase and myosin phosphatase activity. Modulation of myosin phosphatase activity has been shown to greatly influence force in permeabilised and tonic smooth muscles without changing calcium. The objective of this study was to determine if such modulation influences contractions in an intact, phasic smooth muscle; the uterus. The effects of phosphatase inhibitors on force, calcium and myosin phosphorylation were studied. Strips of term pregnant non-labouring human myometrium were dissected from biopsies obtained, following informed consent, from women undergoing elective caesarean sections. Intracellular calcium and force were simultaneously measured in the presence and absence of phosphatase inhibitors. Myosin phosphorylation was measured by urea gel electrophoresis and Western blot analyses. All values are mean ± s.e.m; n is the number of samples, and significance was tested by Student’s paired t-test and significance taken at p<0.05. The results show that application of 10 -8 M okadaic acid, an inhibitor with more specificity for the type II phosphatase, produced no significant change in the amplitude of spontaneous force and calcium transients (n=5). Application of another type II phosphatase inhibitor, cantharidin (108 M), also produced no significant change in the amplitude of spontaneous force and calcium transients (n=5), but in contrast to okadaic acid, produced a small but significant decrease in the rate of relaxation (123±7% slower compared to 100% of control). At 10-7 M, both okadaic acid and cantharidin decreased the amplitude of force and calcium transients, and at 10-6 M they abolished force and calcium. The most marked inhibition was seen on the application of calyculin A; an inhibitor with more specificity for the type I phosphatase. Calyculin A, at 10-8 M, markedly increased the baseline force, but not calcium (n=7), and produced a significantly slower relaxation (115±6% slower compared to 100% control). The level of myosin phosphorylation was increased with all three phosphatase inhibitors; however, with Calyculin A it was apparent that there was a dual phosphorylation band. The results suggest that there is a small modulation of phasic contractions of pregnant, non-labouring uterus, by myosin phosphatase.

P 42-1 EFFECT OF HYPOXIA ON ELECTRICAL AND MECHANICAL ACTIVITY IN VASCULAR SMOOTH MUSCLE (RAT PORTAL VEIN) P. Overhoff, H.H. Hopp, R. Schubert, Th. Noack Objective: Hypoxia induces in most arterial and venous smooth muscle tissues a rapid vasorelaxation. The mechanism for this vasorelaxation is reported as a hyperpolarisation, induced by the opening of ATPsensitive potassium channels. Here it was investigated if this is the only mechanism by which hypoxia induces vasorelaxation. Methods: Portal veins from WKY-rats were mounted in a ,,Bülbring“ capillary to measure both, isometrical mechanical activity and electrical activity and were superfused with a Hepes buffered Krebs solution, aerated with oxygen. Hypoxia was induced by a decrease of the pO2 and an increase of the pN2. The pO2 was measured using an oxygen-sensing system (Licox). Results: The veins showed a rhythmical mechanical activity of 3 contraction cycles per minute. This activity was increased in amplitude but not in frequency when moderate concentrations of noradrenaline (NA, 0.1 to 0.3 µM) were applied. The mechanical activity was preceded of fast deflections in electrical activity which were also increased by NA, indicating the activation of calcium channels. These were identified as L-type calcium channels by inhibiting this activity by nifedipine (NIF, 1 µM). This drug inhibited only the amplitude of the contractions but did not alter the frequency or the time course of the regular dilatations. Below a pO2 of 100 mmHg a slight change of the rhythmic was observed. The intervals became irregular and especially phases of rest disappeared. The phases of activity were restored by Glibenclamide (1µM). Phases of rest were not restored by this drug, indicating a separate effect of hypoxia which could be due to electrical uncoupling of the cells. Using octanol and heptanol (1µM) under normoxic conditions did show the same effect on electrical and mechanical activity as hypoxia in the presence of Glibenclamide. Conclusions: It is concluded, that hypoxia induces opening of Glibenclamide-sensitive potassium channels (KATP) and electrical uncoupling via gap junctions. Both effects are responsible for vasorelaxation.

S 333 Institut fur Physiologie, Universität Rostock, Box 100888, 18055 Rostock P 42-2 PROTEIN AND mRNA EXPRESSION OF S100 CALCIUM-BINDING-PROTEINS IN SMOOTH MUSCLE TISSUES AND SMOOTH MUSCLE CELLS OF GUINEA-PIG B. Daub*, M. Schroeter, G. Pfitzer, V. Ganitkevich Electrophysiological data from guinea-pig smooth muscle cells indicate that up to 98 % of the cytoplasmic calcium ions are buffered by calcium-binding-structures. As potential calcium-binding-structures in addition to calmodulin we investigated the expression of S100 calciumbinding-proteins in guinea-pig smooth muscles. Proteins of the S100 family shows a tissue specific and species specific expression and bind calcium ions by EF-hand motives. The mRNA expression of the S100 proteins S100A1, S100A4, S100A6, S100A10 and S100B was detected by RT-PCR. RNA was isolated from smooth muscle tissues of urinary bladder, ileum, aorta and single smooth muscle cells from urinary bladder of guinea-pig. RT-PCR-products of S100A1, S100A4, S100A6 and S100A10 were detected from the mRNA in the investigated guinea-pig smooth muscle tissues and urinary bladder cells. S100B was not detectable by RT-PCR in mRNA of smooth muscle tissues and cells, but in control mRNA of guinea-pig brain. The homology of PCR-product sequences was 85 - 93 % to human 83 - 88 % to rat and 81 - 87 % to mouse. For the investigation of protein expression polyclonal antisera were raised in rabbits against the immunogene peptid sequences of S100A4, S100A6 and S100A10 derived from RT-PCR products. The protein expression was detected by western-blot analysis. Our data indicates that the proteins S100A4, S100A6 and S100A10 are expressed in guinea-pig bladder, ileum and aorta and may play a role in the calcium homeostasis of smooth muscles. Supported by the Deutsche Forschungsgemeinschaft (Ga 611/1-2) Department of Physiology, University of Cologne, Robert-Koch-Str. 39, 50931 Köln, Germany

P 42-4 A REDOX-SENSITIVE PATHWAY MEDIATES HIF-1α AND HIF2α EXPRESSION IN RESPONSE TO THROMBIN, DESFERRIOXAMINE AND COBALT CHLORIDE IN VASCULAR SMOOTH MUSCLE CELLS A. Görlach, J. Hess, T. Kietzmann Hypoxia-inducible factor-1α (HIF-1α) is a basic-helix-loop-helix (bHLH)-containing member of the PER-ARNT-SIM (PAS) family of transcription factors that plays a central role in the response to hypoxia. In vascular smooth muscle cells, HIF-1α can also be activated under normoxic conditions by thrombin and growth factors. The HIF isoform HIF-2α/endothelial PAS domain protein (EPAS) is mainly expressed in endothelial cells and certain tumor cell lines and is also induced in response to hypoxia. However, regulation of HIF-2α in vascular smooth muscle cells has not been investigated yet. Human aortic or pulmonary artery smooth muscle cells were serum-deprived and stimulated with cobalt chloride (CoCl2) and desferrioxamine (DFO), known to mimick the hypoxic response, as well as with thrombin. HIF-2α mRNA and protein expression was already detectable under normoxic control conditions. Enhanced levels of HIF-2α protein were observed in response to all three stimuli. In vascular smooth muscle cells, HIF-1α expression was shown to be redox-sensitive. Treatment with increasing concentrations of H 2O 2 also resulted in enhanced expression of HIF-2α protein in these cells. In contrast, H2O2 did not modulate basal expression of HIF-2α in the prostate carcinoma cell line ECV304. To further investigate the role of reactive oxygen species in regulating the expression of the HIF isoforms, vascular smooth muscle cells were treated with several antioxidants including vitamin C, pyrrolidine dithiocarbamate and N-acetylcysteine. All antioxidants prevented expression of HIF-1α and HIF-2α in response to thrombin as well as to DFO and CoCl2. Moreover, treatment with the Clostridium difficile toxin B, which inhibits Rac (an activator of the NADPH oxidase), Rho and Cdc42, also decreased thrombin- and CoCl2-induced HIF-1α and HIF-2α expression. These findings indicate that vascular smooth muscle cells show basal expression of HIF-2α protein which is enhanced by thrombin, DFO and CoCl2, and that upregulation of HIF2α as well as of HIF-1α protein in response to these stimuli essentially requires a cell-type specific redox-sensitive pathway.

P 42-3 NITRIC OXIDE MEDIATES S-NITROSOALBUMIN-INDUCED DILATATION OF RAT AORTA N.N. Orie, P. Vallance, K.P. Moore S-Nitrosothiols are formed by the S-nitrosation of a thiol residue by the nitrosonium ion (NO+). The major circulating S-nitrosothiol in plasma is S-nitrosoalbumin (S-NO-albumin), which is present at a plasma concentration of ~20 nM. It is generally believed that S-nitrosothiols are vasodilators, since they can release nitric oxide (NO). However, Kowaluk and Fung1 have suggested that S-nitrosothiols exert an action that is independent of the release of NO. The aim of this study was to examine whether S-NO-albumin causes vasodilatation through an NO dependent pathway or via the foramtion of a trans-nitrosation intermediate. Thoracic aortic segments (with intact endothelium) from male Spraque Dawley rats (200-250g) were mounted in 25ml organ baths containing physiological salt solution (PSS, pH 7.4). The PSS was continously aerated with 95% O2 and 5% CO 2 at 37°C. S-NOalbumin was added cummulatively into the PSS following stable contraction of the tissues with 1 µM phenylephrine to achieve a concentration of 1nM-1µM. S-NO-albumin produced concentrationdependent relaxation of all control segments with full (100%) relaxation achieved by 1 µM of the compound (IC50 = 28±6 nM). To determine whether the S-NO-albumin caused vasodilatation through a transnitrosation reaction at the cell surface, aortic segments were incubated with 1 mM N-iodoacetyltyrosine (IAT), a cell impermeable thiol blocking agent. IAT had no effect on the vasodilatory response to S-NO-albumin (IC50 = 25±6 nM in the presence of IAT). To confirm that S-NO-albumin mediated vasodilatation involves the release of NO, tissues were incubated with 1µM haemoglobin, which scavenges NO, or 10µM ODQ (which inhibits guanylate cyclase). Both of these treatments completely blocked the vasodilatory effect of S-NO-albumin. These results show that S-NO-albumin is an effective vasodilator, and this is secondary to the release of NO rather than through the formation of transnitrosation intermediate at the cell surface. 1. Kowaluk, EA. & Fung, HL. (1990) J Pharmacol. Exp Ther 255: 1256-1264 Centre for Hepatology, Royal Free and University College Medical School, UCL Rowland Hill street, London NW3 2PF, United Kingdom

Experimentelle Kinderkardiologie, Deutsches Herzzentrum München, Klinik an der TU München, Lazarettstr. 36, D-80636 München P 42-5 UROCORTIN RELAXES RAT TAIL ARTERIES BY A PKAMEDIATED REDUCTION OF CALCIUM SENSITIVITY L. Lubomirov 1, H. Gagov 1, P. Petkova-Kirova 1, D. Duridanova1, V. Kalentchuk2, G. Pfitzer3, R. Schubert4 Urocortin is an endogenous vasodilator although the mechanism of vasorelaxation is not completely understood. The hypothesis that an alteration of smooth muscle calcium concentration is involved was tested using isometric tension recording and calcium fluorimetry. Urocortin produced a concentration dependent relaxation (pD 2 8.59±0.06, n=6) of vessels pre-contracted with a physiological salt solution containing 42 mM KCl (42 mM K-PSS). Removal of the endothelium did not alter the effect of urocortin. Urocortin at 100 nM relaxed vessels pre-contracted with 42 mM K-PSS by 59.6±4.6% (n=8) and vessels pre-contracted with 500 nM noradrenaline by 25.2±6.8% (n=6). Both effects were not accompanied by a change in the intracellular calcium concentration. Urocortin at 100 nM produced a significant rightward shift of 0.33±0.07 units of normalised intracellular calcium (n=5) of the relationship between tension and intracellular calcium. The effect of urocortin was accompanied by a reduction of MLC 20 phosphorylation. The urocortin-induced relaxation was considerably reduced in the presence of 0.3 mM Rp-8-CPTcAMPS, a cyclic AMP-dependent protein kinase (PKA) inhibitor. The PKA-activator Sp-5,6-DCl-cBIMPS relaxed 42 mM K-PSS precontracted vessels (pD 2 4.98±0.07, n=6). Sp-5,6-DCl-cBIMPS at 0.1 mM relaxed vessels by 85.3±2.5% (n=5), but did not change the intracellular calcium concentration. In conclusion, the data show that urocortin is a potent, endothelium-independent dilator of rat tail arteries and suggest that this effect is mediated by PKA causing a reduction of the sensitivity of the contractile apparatus for calcium. 1

Institute of Biophysics, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria; 2Institute of Experimental Cardiology, Cardiology Research

S 334 Center, 121552 Moscow, Russia; 3University Cologne, Institute of Physiology, Cologne, Germany; 4University Rostock, Institute of Physiology, 18055 Rostock, Germany P 42-6 STRETCH-DEPENDENCE OF VESICAL RHYTHMIC ACTIVITY E. Neu, M.Ch. Michailov, W. Seidenbusch Vesical physiology is of fundamental importance for pathogenesis of urogenital disturbances (from cystitis up to renal hypertonia) and their therapy. 1989 we suggested that trigonal (T) motor activity (slow motor rhythms: strips and isovolumetric cystotonometry) mediates transition from collecting (sympathetic) to expulsion (parasymp.) phase of urinary bladder which was induced by transformation of electrical spike into a burst-plateau (BP) activity after critical stretching, leading to detrusor (D) excitation [4c] via vesical stretch-dependent ion channels (reported 1991 [2b]); their evidence was confirmed 1993 (incl. human myometrium and ureter) [2c], also via patch clamp [6]. Methods: D and T preparations of guinea pig (incl. a special computer programme) [see 2a, 4a,b]. This hypothesis from 1989 is supported by evaluation of recent and earlier results: After stretching to 50 and 80 mN resp. an increase of contractile amplitudes and frequency of D and T was correlated with an increase of action potentials (AP), frequency, and duration of burst/BP activity. Rate of rise (RR) and fall (RF) were also strongly increased. A multidimensional further investigation about correlations between electrical BP activity, T rhythms and micturition on ionic-molecular [3, 5], cellular [1, 2a,b, 4a,c, 6], organ [2d, 4b, 7], system level [2d, 4c] could open new dimensions in urogenital physiology, pathology and therapy. The authors thank Directors Profs. D.-P. HÄDER/Univ. Erl.-Nürnberg, E. NEHER, W. STÜHMER/Max-PlanckSoc. Göttingen, and D.G. WEISS/Univ. Rostock as well as ICSD e.V.München and Arbeitsamt Nürnberg for their help and interest. Detrusor tension 3 mN 50 mN

(n=69) contraction [mN] 5.3±2.1 17.2±9.3

MOTOR ACTIVITY frequency contraction [per min] [mN] 3.2±0.7 20.2±6.3 3.7±0.4 27.1±6.1

Trigone frequency [per min] 0.28±0.08 0.52±0.12

(n=56) duration [sec] 54.3±9.8 41.6±9.7

Myocytes tension 3 mN 80 mN

(n=70-106) AP [mV] 35.13±9.61 63.25±4.96

ELECTRICAL ACTIVITY frequency duration no bursts 5.53±2.84 1.54±1.18 10.26±2.93

Bursts/BP RR* [V/s] 0.41±0.19 3.27±0.76

(n=33-56*) RF* [V/s] 0.43±0.28 2.32±0.58

Table: Motor (isometrically) and electrical activity (intracellular recording). Initial values: tension 3 mN, MP -34.51±5.27 mV, stretch 80 mN: MP -46.86±3.28 mV. Literature: [1] Kuznetsov SA, Langford GM, Weiss DG, Nature 356: 722-725, 1992. [2a] Michailov MCh, Grindler H, Welscher U, Beitr Urol, Karger/Basel 3: 188-201, 1983. [2b] Michailov MCh, Neu E, Martin D et al, Eur J Physiol, 419 (Suppl): R98, 1991 and [2c] GynGeb Rdsch 33: 333-334, 1993 and [2d] Physiol Res 48 (Suppl): S96, 1999. [3a] Neher E, Neuron 20: 389-399, 1998. [3b] Proc IUPS (20) Christchurch: 2422, 2001. [4a] Neu E, Michailov MCh, Seidenbusch W et al, Eur J Physiol, 420 (Suppl): R99, 1992. [4b] Biomed Tech 39: 312-313, 1994. [4c] Proc IUPS 17 Helsinki: 529, 1989; (19) St. Petersburg: P036.02/3, 1997; (20) Christchurch: 291, 2001. [5] Terlau H, Stühmer W, Naturwiss 85: 437-444, 1998. [6] Wellner M, Isenberg G, J Physiol 466: 213-227, 1993. [7] Welscher U, Neu E, Staehler G et al, Physiol Res 48 (Suppl): S138, 1999.

mdx (n=15) and C57 control (n=15) mice at baseline and in response to atropine and phenylephrine.Results: We demonstrate that mdx mice have significant tachycardia, consistent with observations in Duchenne patients. Heart rate was nearly 15% faster in mdx mice (809 ± 5 bpm) than in control mice (706 ± 13 bpm, P<0.01). ECGs revealed significant shortening of the rate corrected QT interval duration (QTc) in mdx mice compared to control mice (55.5 ± 0.5 ms vs. 61.6 ± 1.5 ms, P<0.05). PR interval durations were shorter at baseline in mdx compared to control mice (27.2 ± 0.4 vs. 29.4 ± 0.5 ms, P<0.05). The muscarinic antagonist atropine significantly decreased heart rate and increased PR interval duration in mdx mice but not in control mice. Pharmacological autonomic blockade and baroreflex sensitivity testing demonstrated an imbalance in autonomic nervous system modulation of heart rate, with decreased parasympathetic activity and increased sympathetic activity in mdx mice. Conclusions: mdx mice have tachycardia and an autonomic nervous system imbalance affecting heart rate modulation, consistent with observations in Duchenne patients. The cholinomimetic effects of atropine in mdx mice but not in C57 control mice suggest a simple test that may be predictive of dystrophin-deficiency. Address: Mouse Specifics, Inc. and The CuraVita Corporation, P.O. Box 200821, Boston MA USA 02120, USA P 42-8 SHORT-TERM FEEDING OF RATS WITH A PALATABLE DIET INDUCES A PROLONGED VASCULAR DYSFUNCTION E.K. Naderali, G. Williams Human and animal studies have shown detrimental effects of obesity on endothelial function resulting in an impaired arterial relaxation. In this study we used Molvany’s myograph to measure the contractility of mesenteric arteries in the rats, to determine whether the functional changes were due to obesity or the diet per se. Male Wistar rats were fed either standard laboratory chow throughout (n=7; controls) or given a highly palatable diet (n=7; diet-fed) for 3 days; or fed the diet for 3-days and then returned to chow for 3 further days before sacrifice, (n=7; diet-to-chow). There were no differences in body weight, fat pad and gastrocnemius muscle masses, and the plasma levels of glucose, insulin and leptin between the 3 groups. However, compared with controls, diet-fed rats had significantly raised plasma non-esterified fatty acids (NEFA) and triglycerides levels (each +100%; t-test, p<0.01). The diet-to-chow group had intermediate plasma NEFA and triglycerides levels that were significantly higher than in controls (ttest, p<0.01). Arterial responses to vasoconstrictors such as noradrenaline and KCl showed no significant differences between the 3 groups, while endothelium-dependent vasorelaxation responses to carbamylcholine (CCh) were significantly (by 18.3%; ANOVA, p<0.01) attenuated in the diet-fed group. Endothelium-independent vasorelaxations induced by sodium nitroprusside (SNP) were also significantly (11%; ANOVA, p<0.05) decreased in the diet-fed group. Despite the removal of palatable diet, these vascular abnormalities were remained persistent in the diet-to-chow group (CCh 11% and SNP 10% less than chow-fed controls). In conclusion, acute feeding of rats with a palatable diet induces marked endothelium-dependent and – independent vasorelaxation, in the absence of any visible obesity. Moreover, these abnormalities largely persisted after the removal of the diet suggesting that the diet per se may have important detrimental effects on arterial function, which may be mediated by raised NEFA and/or triglyceride levels.

Prof. Seidenbusch, Inst. Exp.Physik, Univ. Innsbruck, A-6020 Innsbruck

Diabetes and Endocrinology Research Unit, Department of Medicine, University of Liverpool, UCD, Daulby Street, Liverpool L69 3GA, UK

P 42-7

P 42-9

ECG OBSERVATIONS IN DYSTROPHIN-DEFICIENT (mdx) MICE: A NOVEL TEST FOR DUCHENNE MUSCULAR DYSTROPHY V. Chu, J. Otero, I. Amende, J. Morgan, R. Volk, T. Hampton Objective: Characterize electrocardiograms (ECGs) in conscious mdx mice to see whether they correspond to observations in patients with Duchenne muscular dystrophy (DMD). Methods: A new system for non-invasively recording ECGs in conscious mice (no anesthesia, no implants) and e-MOUSE analyses software in male

PROPERTIES OF THE α1S SUBUNIT OF THE DHP RECEPTOR TAGGED AT EITHER THE N- OR C-TERMINAL WITH DSRED S. Papadopoulos, V. Leuranguer, K. Stroffekova, K.G. Beam Its bright fluorescence and its high resistance against photobleaching have made the newly cloned red fluorescent protein (DsRed) a very attractive probe for tagging-based studies in cell physiology. When used together with its relative, green fluorescent protein (GFP), proteins tagged with the two distinct fluorophores can be used for imaging or fluorescence resonance energy transfer (FRET) studies to analyze co-

S 335 localization and interaction. A major drawback when using DsRed is the proteins tendency to form closely packed tetramers that can interfere with proper expression and targeting of fusion proteins. We used the plasmids pDsRed1 and the newer pDsRed2 (which is supposed to be less prone to aggregate) to prepare constructs encoding the pore forming subunit of the skeletal muscle DHP receptor, α1S, tagged at either the C-terminal (α1S-DsRed1, α1S-DsRed2) or N-terminal (DsRed2α 1S). Bright red inclusions were observed under the fluorescence microscope after expression of the constructs in dysgenic (α1S-null) myotubes, although fewer inclusions were present with DsRed2. However, correctly targeted protein was also produced because all three constructs restored EC coupling and electrophysiological analysis indicated normal-sized (as compared to the currents measured after expression of wild-type α1S) Ca2+ currents for the three fusion proteins. Co-expression with GFP-α 1S together with laser scanning confocal microscopy are being used to determine whether, in addition to the inclusions, red fluorescence can be detected in association with the correctly targeted DsRed-tagged constructs. Supported by NIH (NS24444 & AR44750 to KGB) MDA (KGB & KS) DFG (SP) and FRM (VL).

passive tension. Already at very low passive-tension levels (<1-2 mN×mm-2), Vp could greatly exceed the unloaded shortening velocity measured in fully Ca2+-activated skinned rabbit psoas fibres. Degradation of titin in relaxed myofibrils by low doses of trypsin (5 min) drastically decreased Vp. In intact myofibrils, average Vp was the faster, the smaller the release step applied. Also, Vp was much higher at 30oC than at 15oC (Q 10: 2.0-3.04-6.15, for release steps of 150-250-450 nm/sarcomere, respectively). Viscous forces opposing the shortening are likely to be involved in determining these effects. The results support the idea that the contractile system imposes a braking force onto the passive recoil of elastic structures. However, elastic recoil may aid active shortening during phases of high elastic energy utilisation, i.e., immediately after the onset of contraction under low or zero load or during prolonged shortening from greater physiological SLs.

Department of Anatomy & Neurobiology, Colorado State University, Fort Collins, CO 80523

INACTIVATION AND RESTORATION OF THE SKELETAL MUSCLE DHP RECEPTOR STUDIED IN A γ1 SUBUNIT KNOCKOUT MOUSE D. Ursu, S. Sebille, B. Dietze, D. Freise, V. Flockerzi, F. LehmannHorn, W. Melzer The knockout of the α 1 or the β 1 subunit of the skeletal muscle dihydropyridine (DHP) receptor causes a complete elimination of both slow L-type calcium current and excitation-contraction coupling. Mice deficient of the skeletal muscle-specific γ 1 subunit show no comparably strong effect. Milder alterations of Ca current and Ca release activation and a change in the voltage dependence of Ca current inactivation have been reported. Here we focussed on the kinetics of inactivation and recovery from inactivation. L-type calcium inward currents were measured in primary cultured myotubes derived from satellite cells of adult mice. 15 to 30 s lasting steps to different membrane potentials were applied in the whole cell patch clamp configuration. Recovery from inactivation (30 s, +20 mV) was determined by applying 100 ms test pulses at different times after repolarization to a variety of different holding potentials. The kinetics of inactivation and recovery were found to be slower in the γ1-deficient cells at most of the investigated potentials. Contractures were elicited in extensor digitorum longus (EDL) muscle of adult mice by superfusion with solutions containing 120 mM potassium. The force response was 3-fold higher in the γ1-/- preparations possibly as the result of slowed inactivation of the voltage sensor for EC coupling. Recovery from force inactivation was studied by repetitive tetani of 500 ms duration. The recovery time course was slower and incomplete in the γ 1-/preparations. These results support a modulatory role of the γ1 subunit in slow inactivation and recovery of DHP receptor-controlled events in skeletal muscle.

P 42-10 ELEMENTARY Ca2+ RELEASE EVENTS IN SKINNED FIBERS OF ADULT MAMMALIAN SKELETAL MUSCLE W.G. Kirsch, D. Uttenweiler, M. Both, R.H.A. Fink To date the existence of discretely measurable elementary Ca2+ release events (ECRE) in adult mammalian skeletal muscle has been examined with controversial results, ranging from non-existent in rat cut fiber voltage-clamp experiments (Shirokova et al., 1998) to a scarce appearance of spontaneous events in intact mouse muscle fibers (Conklin et al., 1999). The present study is based on our recent finding of abundant ECRE, spark- and ember-like events, in saponin and mechanically skinned fibers of adult rat and mouse skeletal muscle (Kirsch et al. 2001). Using confocal laser scanning microscopy with the calcium indicator fluo-4, we demonstrate many spontaneously occurring elementary Ca 2+ release events within a single skinned mammalian fiber. The frequency of these events could be further increased by application of 0.2-0.5 mM caffeine or reduced by application of 2 mM tetracaine. In addition to spark-like events, long duration events were observed exhibiting phases of steady fluorescence during their time course and appearing more like embers (González et al., 2000). ECRE in adult mammalian muscle could also be identified at the onset of the global Ca2+ release evoked by membrane depolarisation in mechanically skinned fibres. Furthermore, the two peaks of the frequency of occurrence for ECRE of mammalian skeletal muscle coincided with the expected locations of the transverse tubular system within the sarcomere, suggesting that ECRE mainly originate at triadic junctions. In conclusion, elementary Ca2+ release events are abundant in adult skinned mammalian muscle fibers and comprise two fundamental release types, spark- and ember-like events. Supported by DFG FOR 240/2-2 and BMBF 13N7871.

University of Heidelberg, Institute of Physiology and Pathophysiology, Im Neuenheimer Feld 326, D-69120 Heidelberg, Germany P 43-1

Universität Ulm, Abteilung für Angewandte Physiologie, Albert-Einstein-Allee 11, D-89069 Ulm P 43-2

Medical Biophysics, Institute of Physiology, University of Heidelberg, Im Neuenheimer Feld 326, 69120 Heidelberg, Germany P 42-11 TITIN-BASED CONTRIBUTION TO SHORTENING VELOCITY OF SKELETAL MYOFIBRILS A. Minajeva, C. Neagoe, M. Kulke, W. A. Linke The shortening velocity of skeletal-muscle fibres is determined principally by actomyosin crossbridges. However, these contractile elements are in parallel with elastic elements, whose main structural basis is thought to be the titin filaments. If titin is stretched, it may contribute to sarcomere shortening simply because it can recoil ‘passively’.Here, the titin-based contribution to shortening velocity (Vp) was quantified in single rabbit psoas myofibrils. Non-activated specimens were rapidly released from different initial sarcomere lengths (SLs) by various step amplitudes sufficient to buckle the myofibrils; Vp was calculated from the release amplitude and the time to slack reuptake. V p progressively increased (upper limit of detection, ~60 µm×s -1 ×sarcomere-1) between 2.0 and 3.0 µm SL, albeit much steeper than

CYTOPLASMIC Ca2+ INPUT FLUX DETERMINED IN DYSGENIC MYOTUBES AFTER HETEROLOGOUS EXPRESSION OF DIFFERENT CA2+ CHANNEL α1 SUBUNIT ISOFORMS R.P. Schuhmeier, E. Gouadon, N. Kasielke, B.E. Flucher, M. Grabner, F. Lehmann-Horn, W. Melzer In skeletal muscle, Ca 2+ release from the sarcoplasmic reticulum is controlled by the direct interaction between dihydropyridine (DHP) receptors in the plasma membrane and ryanodine receptors (RyR) in the sarcoplasmic reticulum (SR). The targeting motif of the DHP receptor α1 subunit responsible for its proper incorporation into the junctional membrane region facing the SR has been localized within a 70 aa sequence near the C-terminal end. In the present study we determined the intracellular Ca2+ kinetics after heterologous expression of α1 subunits that differ in their mode of targeting and their mechanisms of interaction with RyR. Four types of GFP-tagged calcium channel α1 subunits were expressed in α1S-deficient GLT myotubes: α1S (skeletal muscle, L-type), α1C (cardiac, L-type), α1A (brain, P/Q-type, lacking the targeting signal) and α1Aas(1592-clip) (a truncated α1A-α1S chimera containing the targeting sequence). We simultaneously measured calcium inward current and intracellular fura-2 calcium signals at 100 ms step

S 336 depolarizations to +30 mV. Based on published parameters for intracellular calcium binding, the global cytoplasmic Ca 2+ input flux (composed of Ca2+ entry and SR Ca2+ release) was calculated. Comparison of the two L-type α1 subunits, which are both targeted to the junction, showed that the Ca2+ inward current density was considerably larger in cells expressing α1C whereas cells expressing α1S showed a faster and more than 2-fold higher Ca2+ input flux. These differences are consistent with the two distinct types of Ca2+ release control mediated by these isoforms: Ca 2+-current-dependent (α 1C ) and direct conformational coupling (α1S). α1A led to a much smaller Ca2+ input flux than α1Aas(1592clip) probably because it is not specifically targeted to the junction.

oid muscle is composed by: Type I = 8 ± 1, type IIA = 25 ± 12, type IID = 54 ± 6, and type IIB = 13 ± 6, and the corresponding MHC content is: MHCI = 10 ± 5, MHCIIA/IID = 77 ± 4, and MHCIIB = 13 ± 6. These data show, that the sternohyoid muscle is a practicable candidate for a transplantation into the bed of the zygomatic muscle.

Universität Ulm, Abteilung für Angewandte Physiologie, Albert-Einstein-Allee 11, D-89069 Ulm, Germany

EFFECTS OF ANTIOXIDANT TREATMENT ON HSP72 AND IL-6 GENE EXPRESSION IN HUMAN CONTRACTING SKELETAL MUSCLE C.P. Fischer, N.J. Hiscock, M.A. Febbraio, B.K. Pedersen Free oxygen radicals are generated during muscle contraction, and are known to induce the expression of heat shock proteins (HSP). HSP are ubiquitous proteins known to regulate the expression of other genes. Of note, it has been shown that HSP72 (the inducible form of the 70 kDa family of HSP) can induce the expression of the cytokine interleukin (IL)-6 in cultured monocytes. Recent work from our group has demonstrated that both HSP72 and IL-6 gene expression are increased in contracting skeletal muscle. Hence, it has been suggested that the increase in IL-6 in contracting skeletal muscle may be mediated via the induction of HSP72. Of note, the increase in HSP72 mediated IL-6 in monocytes only occurred via a CD14-dependant pathway, suggesting that IL-6 can only be mediated by extracellular HSP72. We hypothesized, therefore, that the production of musclederived IL-6 during concentric exercise occurs through an HSP72 independent pathway and that antioxidant treatment would inhibit the exercise-induced increase in HSP72-mRNA, but not the exerciseinduced increase in IL-6 mRNA. To test these hypotheses, 11 healthy, young males ingested either 400 IU of vitamin E and 500 mg of vitamin C (Treatment) or a placebo (Control) daily for 4 weeks before completing 3 h of two-legged knee-extensor exercise at 50% of maximal power output. Muscle biopsies were obtained before, immediately after, and 3 h following exercise. These samples were analysed for HSP72 and IL-6 mRNA using real time PCR and the data were statistically analysed using a two way (group x time) ANOVA. There were no differences in either IL-6 or HSP72 gene expression at rest. However, 3 h of exercise markedly increased (P<0.05) IL-6 mRNA in both Treatment and Control, and while these values decreased (P<0.05) 3 h following exercise, IL-6 mRNA was still elevated (P<0.05) compared with Rest. No differences, however, were observed in IL-6 mRNA when comparing Treatment with Control. Importantly, HSP72 increased (P<0.05) ~4 fold in Control when comparing Post-exercise with Rest, but no such increase was observed in Treatment. These data demonstrate that 3 h of non-damaging exercise was sufficient to induce both an IL-6 and HSP72 response. However, since antioxidant treatment blunted the exercise-induced HSP72, but not IL-6 response, our data suggest that in contracting muscle, IL-6 is not mediated via an HSP72 dependant pathway.

P 43-3 CONTRACTILE PROPERTIES AND THEIR RELATIONS TO HISTOCHEMICAL FIBRE TYPES AND MYOSIN ISOFORM COMPOSITIONS OF THE POSTERIOR BELLY OF RAT DIGASTRIC MUSCLES C. Soukup, G. Asmussen There are only a few informations about mammalian jaw opener muscels. Therfore, the contractile properties, the enzyme histochemical determined muscle fibre composition and the myosin isoforms were studied in the posterior belly of digastric muscle of the rat. Data are presented as mean ± SD. In vitro studies with direct all over stimulation show, that the muscle is a fast contracting one. (contraction time: 15.5 ± 2.1 ms; half–relaxation time: 13.3 ± 2.2 ms, at 35°C). Cooling down the muscle from 35°C to 25°C increase the contraction and halfrelaxation times of the single twitches with a temperature coefficient (Q 10) of 0.43 ± 0.15 and furthermore the twitch tension (cold potentiation). With decreasing temperature the tetanic tension decreases with a Q10 of about 1.45 and consequently also the twitchtetanus ratio decreases. Additionally a post-tetanic potentiation of the single twitches was found. Using enzyme histochemical techniques four fibre types (I, IIa, IIb, IId) were determined. More than 90 % fast contracting fibres (IIa: 37.6 ± 11.9%, IIb/IId: 53.0 ± 9.1%) and Type I - fibres: 8.9 ± 2.8% were found. These data are comparable with the densitometrically evaluated relative concentrations of electrophoretically separated myosin havy chain (MHC) isoforms confirming to our mechanical and histochemical results (MHCIIa: 34.8 ± 2%, MHCIIb/MHCIId: 56.2 ± 3.1%, and MHCI: 8.9 ± 3.7% of total MHC). Our results point out that, the posterior belly of the digastric muscle of the rat is a mixed fast skeletal muscle representing a very close relation between contractile, histochemical and biochemical properties. University of Leipzig, Carl Ludwig Institute of Physiology, 04103 Leipzig, Liebigstrasse 27, Germany

University of Leipzig, Carl Ludwig Institute of Physiology, 04103 Leipzig, Liebigstrasse 27, Germany P 43-5

P 43-4 MUSCLE FIBRE TYPES AND MYOSIN ISOFORM COMPOSITION OF THE ZYGOMATIC AND STERNOHYOID MUSCLES OF ADULT RABBITS F. Hieke, S. S. Keiner, R. Schulz, G. Asmussen One of the most promising methods in the surgical therapy of the facial nerve parasis seems to be the transplantation of the innervated sternohyoid muscle on the denervated zygomatic one Keiner et al., HNO, 48, 598, 2000). However, the functional properties of the donor as well as of the normal or denervated receiver are not well documented. The fibre type composition of the muscles was studied histochemically (myosin-ATPase after acid and alkaline preincubations) and the distribution of the myosin heavy chain (MHC) isoforms was analysed by SDS-gel electrophoresis. The main part of the normal zygomatic muscle (pars auricularis) contains the following fibre types (%, mean ± SD): Type I = 18 ± 1, type IIA = 47 ± 10, type IID = 30 ± 13, and type IIB = 5 ± 3, corresponding to a MHC composition of MHCI = 26 ± 1, MHCIIA/IID = 69 ± 4, and MHCIIB = 5 ± 3. There are small regional differences in the fibre type distribution within the muscle. The percentage of type I fibres in the deeper parts of the muscle is higher than in the superficial ones. Denervation (cutting the facial nerve at the foramen stylomastoideum) leads to a marked atrophy of the fast, and a moderate hypertrophy of the slow fibres combined with an increase of hybrid fibres (type IIC and/or neonatal). The sternohy-

Dept. of Infectious Diseases M7641, Rigshospitalet, Tagensvej 20, DK-2200 Copenhagen N, Denmark P 43-6 AN INTRINSIC SHG-SIGNAL IN ADULT AMPHIBIAN SKELETAL MUSCLE FIBRES M. Both, W.G. Kirsch, D. Uttenweiler, R.H.A. Fink Recently, it has been shown that the non-linear effect of second harmonic generation (SHG) provides a powerful tool to detect membrane potential changes with chiral membrane bound dyes (Campagnola et al., Biophysical Journal 77, 1999). However, when we applied this technique to follow membrane potential changes in skeletal muscle we observed a strong intrinsic SHG-signal. Therefore we studied this new intrinsic SHG-signal in more detail in isolated single amphibian skeletal muscle cells from M. tibialis anterior. In our experiments we used a Leica TC SP2 MP system, a combined CLSM augmented with a ps-pulsed Ti:Sa-Laser for non-linear microscopy. The SHG-signal was generated by an initial laser tuned to 900 nm wavelength and was detected on the transmitted light path via a band pass filter (450 nm, 50 nm FWHM) to discriminate the SHG-signal from the excitation light and fluorescence emission. The intrinsic SHG-signal turns out to be periodically following the striation pattern. It is strongly depend-

S 337 ent on the sarcomer length and can be observed for sarcomer lengths greater than 2 µm. The intensity shows a linear dependency on the sarcomer length and is saturated at sarcomer lengths greater than 3 µm. Additionally, the signal is also strongly dependent on the polarisation of the initial laser beam. Thus, it is vitally important to correct for this intrinsic SHG-signal with respect to the sarcomer pattern when using potential sensitive dyes in skeletal muscle. From the sarcomer length dependency we assume that the intrinsic signal may occur at the level of the proteins involved in contractile activation. Supported by DFG FOR 240/3-1 and BMBF 13N7871. Medical Biophysics, Institut für Physiologie und Pathophysiologie, Universität Heidelberg, Im Neuenheimer Feld 326, 69120 Heidelberg, Germany P 43-7 FOREARM INTRACELLULAR PH AFTER MAXIMAL WHOLE BODY EXERCISE S.Volianitis*, N.H. Secher, B. Quistorff Competitive rowing elicits extreme blood acidosis (6.74; Nielsen, 1999). It is not known to what extent the intracellular muscle pH (pHi) is affected during whole body exercise. Thirteen competitive rowers (age 24.5 ± 2.3 y, height 1.78 ± 3.2 m, weight 72.1 ± 5.4 Kg) performed a simulated maximal 2000 m rowing trial on an ergometer (Concept II, USA). Immediately after maximal rowing the intracellular metabolic response was studied using 31P-MRS on the wrist flexor muscles of the non-dominant arm. Arm venous blood samples were collected at rest and immediately after the trial. Significant differences were identified with Student’s t-test and the level of significance was set at P < 0.05. In response to maximal rowing, muscle venous O2 haemoglobin saturation decreased from 51 ± 4% to 29 ± 9% (mean ± SD) and lactate increased from 1.0 ± 0.1 to 16.8 ± 3.6 mrnol-l-1 with a reduction in pH from 7.43 ± 0.01 to 6.90 ± 0.01 (P < 0.05). The ratio of inorganic phosphate to phosphocreatine (Pi/PCr) at rest was 0.12 ± 0.03 and increased to 1.50 ± 0.49 at the end of the rowing trial (P < 0.05). pHi decreased from a baseline value of 7.05 ± 0.02 to 6.43 ± 0.19 (P < 0.05). The half time for pH recovery to baseline values was 594 ± 253 s. The data are in agreement with the pHi values obtained during exhaustive handgrip exercise (Nielsen et al. 2001) and indicate that extreme metabolic acidosis can be induced in the forearm arm flexor muscles during whole body exercise. Nielsen H. B. Acta Physiol Scand, 1999, 165, 113-114. Nielsen H. B., L. Hein, L. B. Svendsen, N. H. Secher and B. Quistorff. Acta Anaestesiol Scand, 2001 (in press). The NMR Centre, The Panum Institute. *The Copenhagen Muscle Research Centre, Rigshospitalet, AN 2041, Blegdamsvej 9, DK-2100, Copenhagen O, Denmark

P 43-8 FIBRE SIZE AND METABOLIC PROPERTIES OF MUSCLE FIBRE TYPES IN RAT EXTRAOCULAR MUSCLE C. Wenzel, K. Punkt, R. Rätze, G. Asmussen Mammalian extraocular muscles (EOM) are characterized by an extraordinary high shortening velocity connected with a high fatigue resistance. The metabolic basis of this unusual combination is unknown. Cross-sectional area (CSA), succinate dehydrogenase (SDH), and glycero3-phosphate dehydrogenase (GPD) activities were measured in single fibres of adult rat EOM using quantitative histochemistry. The optical density (OD) was determined cytophotometrically. The same fibres were classified in serial sections according to their histochemically assessed myosin-ATPase and SDH activities. All data are presented as mean ± SD. The orbital region of the EOM contains two fibre types: Type 1 (fast) shows a small CSA (226 ± 115 µm²), a medium SDH (OD 0.36 ± 0.07) and a low GPD (0.17 ± 0.04) activities. Type 2 (slow) exhibits an extremely small CSA (109±27 µm²) combined with as well as low SDH (OD 0.14 ± 0.04) and GPD (0.15 ± 0.05) activities. The global region of the EOM contain four fibre types: Type 3 – 5 (fast) form a continuum with an inverse correlation between CSA (46 – 522 µm²) as well as SDH (OD 0.16 - 0.82) and GPD (OD 0.41 - 0.82) activities (p< 0.001). In contrast to the muscle fibre types of skeletal muscles there was a positive relationship between SDH and GPD activities. Type 6 (slow) is characterized by a medium CSA (395 ± 38 µm²) and as well as extremely low SDH (OD 0.10 ± 0.02) and GPD (OD 0.13

± 0.02) activities. The metabolic characterization of the EOM gives a good explanation of their specific mechanical properties. University of Leipzig, Carl Ludwig Institute of Physiology, 04103 Leipzig, Liebigstrasse 27, Germany

P 43-9 PHOSPHATE INDUCED DEPRESSION OF ACTIVE TENSION IN RABBIT SOLEUS (SLOW) MUSCLE FIBRES AT DIFFERENT TEMPERATURES M.E.Coupland, K.W. Ranatunga In a previous study, we examined the active tension depression induced by added inorganic phosphate (Pi, a product of ATP-hydrolysis released in active muscle fibres) at a range of temperatures (5-30 ºC) in rabbit psoas (fast) muscle fibres (Coupland et al, 2001). The results showed that the tension depression became less pronounced at the higher temperatures: 25 mM P i / control tension ratio was ~0.5 at 5-10 ºC, whereas it was ~0.8 at 30 ºC. Experiments of Millar & Homsher (1992) have shown that the Pi-induced tension depression at 10 ºC is less marked in slow than in fast muscle fibres. The aim of the present study was to examine the effect of added Pi on maximal active tension in slow fibres at a range of temperatures. Experiments were done on slow fibres isolated from soleus muscles of rabbits that had been killed by an intra-peritoneal injection of an overdose of sodium pentobarbitone. Muscle fibres were chemically skinned (using 0.5% Brij) and tension recording was done using a trough system and methodologies described previously (see Ranatunga, 1994; Coupland et al, 2001). In one type of experiment (10 fibres), the maximal Ca-activated tension (pCa 4.5, ionic strength 200 mM) in each fibre was recorded at a constant temperature but using a range of added Pi levels (1.6 mM - 48 mM). In a second type of experiment (2 fibres), the active tension with 25 mM added Pi and with no added Pi (control) was recorded at a range of temperatures (10 ºC - 30 ºC). As in the case of fast fibres, tension versus Pi-concentration data at a constant temperature could be fitted to a hyperbolic relation. When normalised to control tension, the estimated minimum tension for excess [Pi] from such analyses was 0.53 at 10 ºC, 0.62 at 20 ºC and 0.67 at 30 ºC (n = 2-8). The mean (± s.e.m.) 25 mM P i / control tension ratios from two types of experiments were 0.63 (± 0.02, n = 6), 0.71 (± 0.01, n = 10) and 0.72 (± 0.02, n = 4) at 10 ºC, 20 ºC and 30 ºC, respectively. These preliminary data show that, compared to fast fibres, the tension depression by Pi at lower temperatures and the decrease of Pi-induced tension depression on heating are less marked in slow fibres. Coupland, M.E. Puchert, E. & Ranatunga, K. W. (2001). J. Physiol., 536, 879-891; Millar, N.C. & Homsher, E. (1992). Am. J. Physiol., 262, C1239-C1245; Ranatunga, K.W. (1994). Biophys. J. 66, 1531-1541. We thank The Wellcome Trust for financial support. Department of Physiology, School of Medical Sciences, University of Bristol, Bristol BS8 1TD, England P 43-10 FAST UPTAKE AND SLOW RELEASE OF CALCIUM BY MITOCHONDRIA IN MOUSE SKELETAL MYOTUBES DEMONSTRATED WITH SHORT-TERM (5 S) APPLICATION OF 4CHLORO-M-CRESOL M.H. Gschwend, H. Brinkmeier, R. Rüdel For the study of the mitochondrial Ca2+ household in mouse skeletal myotubes, we stimulated intracellular Ca2+ release using short-term (5 s) application of 4-chloro-m-cresol (4-CmC, 500 µM), a potent ryanodine receptor (RyR) agonist. Digital fluorescence imaging with either of two Ca 2+-sensitive dyes was used to determine intracellular Ca 2+ changes. The cytosol-specific Ca 2+ indicator fluo-4 monitored cytosolic calcium, [Ca2+]cs, whereas the mitochondria-specific Ca2+ indicator rhod-2 monitored mitochondrial calcium, [Ca 2+]mt. Mitochondria-rich regions were chosen for the recording of mitochondrial Ca2+ transients. The responses to 4-CmC provided by fluo-4 and rhod-2 were similar with respect to delay (1-2 s) and time to peak (2-5 s). This suggests a fast uptake of Ca2+ into the mitochondria. Upon washout of 4-CmC, the rhod-2 signal remained much longer elevated (tens of s) than that of fluo-4 (decay to baseline in 1-5 s). The uptake of Ca 2+ into mitochondria was prevented when the mitochondrial membrane potential was destroyed by the uncoupler FCCP. Blocking the

S 338 mitochondrial Ca2+/Na+ exchanger by the benzothiazepine CGP 37157 resulted in a pronounced inhibition of Ca2+ release from the mitochondria. Finally, we investigated the mitochondrial Ca2+ household in connection with other Ca2+-releasing drugs. Caffeine (5 mM), applied for 5 s, also induced a fast uptake followed by a slow release of Ca2+ from mitochondria. The peak amplitude of the rhod-2 signal was similar as with 4-CmC, although in most experiments repetitive short-term application of caffeine was required. Ca2+ release induced by depolarization with KCl (50 mM) had to be performed twice or more often to elicit increases in [Ca2+]mt in a reliable manner. The mitochondrial Ca2+ transients induced by 4-CmC were well reproducible and thus should be most dependable in studies exploring the mitochondrial Ca2+ household with regard to neuromuscular diseases (Supported by the Medical Faculty and IZKF, University of Ulm). Department of General Physiology, University of Ulm, 89069 Ulm, Germany P 43-11 COMPARISON OF TWO METHODS TO QUANTIFY THE ELECTRICAL ACTIVITY OF HUMAN SKELETAL MUSCLE ISOKINETIC CONTRACTIONS S. J. Brown Quantification of the surface recorded electromyographic signal (EMG) may be used to indicate the level of neural activation required for a given muscle force. Indices to quantify EMG include mean signal amplitude and standard deviation, integration of the rectified signal (Day et al. 1998), calculation of the signal root-mean-square, and analysis of component signal frequencies (Hagg, 1992). Measuring the slope of the cumulative integration of the rectified signal provides an alternative quantitative tool, and the present study compared this index of EMG quantification with the standard deviation of the EMG signal during isokinetic concentric and eccentric muscle actions. With ethical approval and informed consent, 14 subjects performed a maximal effort concentric and eccentric muscle action on an isokinetic dynamometer at .35 rad s-1 through a .87 rad range of motion, using their left bicep brachii. Force (KinCom, Chattecx, TN., USA) and muscle EMG (signal filtered between a 1Hz high pass and 1KHz low pass, and sampled at 2 KHz) were recorded (PowerLab, AD Istruments, Hastings, UK). The standard deviation of the mean EMG signal (EMG sd) and the slope of the cumulative integration of the rectified signal (EMG slope) were used to determine the force per unit of electrical activity (N/mV). Mean (+/- SD) concentric force was significantly lower than eccentric force (265 +/-109 N vs 452 +/-102 N; P<0.001, paired t-test), although electrical activities for concentric and eccentric were not different. Mean (SD) EMG slope for concentric was .82 +/.39mV and eccentric was .85 +/-.43mV (P>0.05, paired t-test). For both contraction types, electrical activity measured using the EMG sd technique recorded higher values than the EMG slope technique (both P<0.001). When determining the force per unit of electrical activity for concentric muscle actions, a bias of 74.6 N/mV toward higher values with the EMG slope method was recorded, and for eccentric muscle actions, a bias of 124 N/mV toward higher values with the EMG slope method was recorded. A trend towards an increasing bias (when using the EMG slope method) with higher forces was observed, particularly for eccentric muscle actions. As subjects were required to generate maximum effort actions, higher values for EMG slope with higher eccentric forces may not indicate increased levels of rate coding in stronger subjects. Eccentric muscle actions may require different activation for a given force compared to concentric muscle actions (Nardone, Romano, & Schieppati, 1989; Enoka 1996), however, the EMG slope and the EMD sd techniques could not distinguish these differences. Day, S., Donnelly, A., Brown, S. & Child, R. (1998) Muscle Nerve, 21 (7):961-963; Enoka, R. (1996) J Appl Physiol 81(6): 23392346; Hagg, G.(1992) J Appl Physiol 73, 1211-1217; Nardone, A, Romano, C. & Schieppati, M. (1989) J Physiol 409: 451-471 DeMontfort University, Bedford, MK40 2BZ P 44-1 BPDZ 154: A NOVEL ACTIVATOR OF KATP CHANNELS IN ßCELLS A.T. Lee1, K.E. Cosgrove 1, P.D. Barnes1, K.J. Lindley2, A. AynsleyGreen2, P. de Tullio3, B. Pirotte3, P. Lebrun4, M.J. Dunne1

Diazoxide is a selective agonist of the ß-cell ATP sensitive K+ (KATP) channel, which is composed of SUR1 and Kir6.2. Clinically, diazoxide is used to treat hyperinsulinism, caused by insulinomas or Hyperinsulinism in Infancy (HI). The responsiveness of patients to diazoxide is highly variable, and whilst this is in part related to the molecular heterogeneity of hyperinsulinism, diazoxide is poorly tolerated with patients suffering a number of complications including hypertension and hypertrichosis. The aim of this study was to examine the actions of a novel benzothiadiazine-derivative BPDZ 154 (6,7-dichloro-3isopropylamino-4H-1,2,4-benzothiadiazine 1,1-dioxide) on ß-cell KATP channels and insulin release. Secretion studies were carried out using isolated rat islets of Langerhans, and ion channel studies were performed using either human ß-cells from patients with HI (n=4) or adenoma (n=1) following surgery (with permission), or the rat insulinsecreting cell line BRIN-BD11. Patch-clamp studies were performed using the cell-attached patch and inside-out patch configurations under standard conditions. Results: BPDZ 154 was found to inhibit 16.7 mM glucose-induced insulin release with an EC50 value of 0.28µM. Under the same experimental conditions the EC50 for diazoxide was approximately 20µM. The mechanism of action of BPDZ 154 was due to the selective activation of K ATP channels since BPDZ 154 increased the activity of KATP channels in both intact cells (10-50µM, n=5) and in inside-out patches exposed to 500µM ATP, (10nM-50µM, n=23/23). BPDZ 154 was consistently found to be more potent than diazoxide (n=7/7) and was less readily reversible upon removal of the compound. The sulphonylurea tolbutamide, reversed the effects of BPDZ 154-induced activation of KATP channels (500µM ATP/ 25µM BPDZ 154/ 100µM tolbutamide, n=13/13). In ß-cells isolated from HI patients, BPDZ 154 was effective in those patient tissues where KATP channel function remained (Type 3 disease, 10-50µM, n=24/24), but was ineffective in patients where KATP channels were functionally absent (10-50µM, n=10/10). In adenoma ß-cells, BDPZ 154 consistently activated KATP channels (500µM ATP/ 10µM BPDZ 154, n=6/6). Summary: These data document the inhibition of insulin release and the activation of ß-cell KATP channels by BPDZ 154, a novel compound structurally based upon the KATP channel openers diazoxide and pinacidil. BPDZ 154 was more potent than diazoxide and had improved selectivity for SUR1/Kir6.2 channels over non-ß-cell KATP channels. These types of compounds may offer a therapeutic potential in the future treatment of Hyperinsulinism in Infancy. 1

Department of Biomedical Science, Sheffield University, Sheffield, UK; 2Institute of Child Health, University College London, London, UK; 3Department of Medicinal Chemistry, University of Liège, Liège, Belgium; 4Laboratory of Pharmacology, Free University of Brussels, Brussels, Belgium P 44-2 SEGREGATED EXPRESSION OF KIR6.1 AND KIR6.2 IN SUBTYPES OF ASTROCYTES IN HUMAN HIPPOCAMPUS K. Hüttmann, G. Seifert, A. Wallraff, W. Schröder, R.W. Veh, C. Steinhäuser Human astrocytes were investigated with the patch clamp technique in the CA1 area of acute hippocampal brain slices obtained from surgical specimens of patients suffering from pharmaco-resistant temporal lobe epilepsy. Astrocytes were identified by their functional properties, immunohistochemically, and by performing single-cell RTPCR subsequent to the recording to test for S100β transcripts. We have previously shown that in patients with significant neuronal cell loss (Ammon’s horn sclerosis) astrocytes display reduced K+ inward rectifier currents. Here we tested whether ATP-sensitive K + ( K ATP) channels are expressed by human astrocytes and contribute to inward rectification in these cells. K ATP channels are octameric protein complexes composed of pore-forming subunits (Kir6.1, Kir6.2) that co-assemble with sulfunylurea receptors (SUR1, SUR2). While both Kir6.1 and Kir6.2 are functionally expressed in mammalian neurons, recent data suggested that astrocytes exclusively possess Kir6.1 protein while Kir6.2 is absent. Pharmacological tools were used to test for functional expression of K ATP channels. Kir currents in the human astrocytes were sensitive to low concentrations of Ba2+ (0.1 mM) and were increased upon application of the Kir6/SUR2 channel activator, pinacidil (0.3 mM, increase by 15 %) or diazoxide (0.3 mM, increase by 12 %), another KATP channel agonist. Subsequently, the cytoplasm of the respective cell was harvested for transcript analysis with singlecell RT-PCR. All cells contained Kir6.2 transcripts while Kir6.1 mRNA was consistently absent (n = 22). This was surprising since in a previous

S 339 study, Kir6.1 protein was found in human astrocytes using a specific antiserum. These data suggest a segregated expression of KATP channels in different subpopulations of human astrocytes. Future work has to uncover whether these channels play a protective role in the hyperactive tissue. Supported by DFG (SFB/TR3, Ve 187/1-3) and BMBF Experimental Neurobiology, Neurosurgery, Bonn University, SigmundFreud-Str. 25, 53105 Bonn, Germany

weakly if at all by PNU 96293 (100 µM) whereas PNU 96179 (100 µM) activated the current. At SUR1, inhibition of [3H]glibenclamide binding was only partial. Surprisingly, PNU 96293 at 10 µM slightly activated the Kir6.2 / SUR1 channel whereas 100 µM induced inhibition by ≈ 50 %. PNU 96179 (10 or 100 µM) showed only channel inhibition. In these enantiomers, the sterical conformation determines the sign of the allosteric coupling to MgATP binding and, correspondingly, the opening or blocking action on the channel. Interestingly, the effects of the enantiomers are reversed when going from SUR2B to SUR1.

P 44-3 COMPARISON OF SIGNIFICANT PHYSIOLOGICAL PROPERTIES OF WT AND SUR1-KO MOUSE PANCREATIC BCELLS M. Düfer, G. Drews, L. Aguilar-Bryan*, J. Bryan*, P. Krippeit-Drews The regulation of ATP-dependent K+ (KATP) channels by metabolism is a key event in stimulus-secretion coupling and in glucose-induced membrane potential (MP) oscillations of pancreatic B-cells. SUR1 knockout (SUR1-KO) mice lack pancreatic KATP channels. Therefore, the B-cells of these mice are a good model for studying KATP-independent pathways (Seghers et al., JBC 275, 9270, 2000). We investigated metabolic and hormonal influences on B-cell properties of SUR1-KO mice by measuring the MP with the patch-clamp technique and cytosolic Ca 2+ concentration ([Ca2+] C) and mitochondrial membrane potential (∆Ψ) with fluorescence techniques. As expected for a B-cell lacking KATP channels the MP was depolarized (-36±1 mV) with continuous spike activity even in 0.5 mM glucose (0.5 G) (n=5). The plateau potential (from where action potentials start) did not alter by a change to 15 mM glucose (15 G) (-38±2 mV, n=5), but the action potentials were larger in amplitude without an obvious change in frequency. Poisoning the cells with NaN 3 (5 mM) stopped spike activity, but without hyperpolarizing the MP (n=4) as in WT B-cells. Surprisingly, [Ca2+]C in 0.5 G was low (68±3 nM, n=11) at the beginning of most experiments. Switching to 15 G elicited an almost identical response compared to WT cells with an initial decrease and a subsequent rapid elevation (n=3). However, in other cells [Ca2+]C spontaneously increased and oscillated in the continuous presence of 0.5 G (n=5). NaN3 (5 mM) in 15 G induced an almost identical response in [Ca2+]C as in WT B-cells, namely, an initial transient increase followed by a plateau and a transient drop to basal values after wash-out (n=5). The glucose (0.5 to 15 G) induced hyperpolarization of ∆Ψ seemed to be much lower (~50 %) compared to WT B-cells as well as the NaN3 (5 mM) induced maximal depolarization (n=4). Presumably, an overall higher basal [Ca2+]C already partly depolarizes the mitochondria of the SUR1-KO B-cells. Galanin (50 nM) transiently inhibited spike activity and hyperpolarized the MP of SUR1-KO B-cells by –7±1 mV (n=5). Accordingly, galanin transiently decreased [Ca2+]C in the presence of 15 G (n=4). We conclude that most of the observed differences in the behaviour of WT and SUR1-KO B-cells are indeed owing to the lack K ATP channels in the latter. In contrast to the current opinion deduced from tumor cell lines, galanin exerts its hyperpolarizing effect by activating a K + current different from the KATP current. Pharmazeutisches Institut, Auf der Morgenstelle 8, D-72076 Tübingen

Department of Pharmacology, Medical Faculty, University of Tübingen, Wilhelmstr. 56, D-72074 Tübingen and 1 Aventis Pharma Deutschland, D-65926 Frankfurt P 44-5 FINDING PROTEINS THAT INTERACT WITH A RENAL ATPDEPENDENT POTASSIUM CHANNEL BY YEAST TWO HYBRID SCREENING M.P. Roos*, J. Spencer, P. Pagel, U. Kerl, S. Silbernagl*, G. Giebisch, M.J. Caplan The ROMK channel‘s inwardly-rectifying K+ currents have been shown to underlie important physiological functions such as renal K+ secretion and reabsorption. Structure-function studies in heterologous expression systems have provided a great deal of information about the biophysical and regulatory properties of the channel. However, little is known about the mechanisms responsible for the selective transport of these channels to the apical membrane of epithelial cells and about the pathways through which channel function is regulated. In both cases it is likely that proteins involved in these processes interact with the ROMK channel’s cytoplasmic tail, whose last three amino acids, TQM, have similarity to a PDZ-interaction motif. Therefore, a search for possible interaction partners of ROMK, using the yeast two hybrid system in association with a human kidney cDNA library, was performed. 174 proteins were identified that interacted with the cytoplasmic tail of ROMK. Of these, 12 were chosen for further experiments based on some of their known functions and structures. Calponin, Acidic Calponin, Hef1, DRF, Posh, ArgAblBP2 are involved in organizing the actin cytoskeleton. NHERF2 , Shank3a, NUMB2 contain domains known to be involved in protein-protein interactions, such as PDZ or SH3 domains. Aquaporin 1, the K +-channel KvLQT1 and Antiquitin1, which is found in the cochleus and is thought to represent a K+-channel, also bound to ROMK. In order to confirm the interaction of the possible binding partners, GST pull-down experiments were performed. Due to technical problems, only interaction of NHERF, Antiquitin, NUMB and Calponin could be shown repeatedly. The results of the yeast two hybrid screen indicate that proteins associated with the actin cytoskeleton bind to ROMK. The relationship of the ArgAblBP2 to the signaling pathway of Abelson Tyrosine Kinases makes its putative association with ROMK especially interesting. In addition, the PDZ-domain containing protein Shank also interacts with the Abelson family and is involved in regulating the distributions of membrane proteins. Since ROMK is regulated by tyrosine kinases it is tempting to say that these two proteins have influence on its activity.

P 44-4 THE ENANTIOMERS OF A PINACIDIL ANALOGUE OPEN OR BLOCK THE KATP CHANNEL U. Lange, U. Russ, C. Löffler-Walz, A. Hambrock, H.C. Englert1 and U. Quast ATP-sensitive K+ channels (KATP-channels) are composed of pore- forming subunits (Kir6.x) and regulatory subunits, the sulfonylurea receptors (SURs). The cyanoguanidines, pinacidil and P1075, are openers of these channels. In contrast, the pinacidil-analogue PNU 96293 ((R)-N-cyano-N’-(1-phenylpropyl)- N’’-3-pyridylguanidine) inhibits pinacidil-induced relaxation of vascular smooth muscle (Khan et al. 1997, J Pharmacol Exp Ther 283:1207). We therefore examined the effect of this compound and its (S) enantiomer, PNU 96179, in binding studies and electrophysiological experiments. At SUR2B, inhibition by PNU 96293 of [3H]-P1075 binding was negatively affected by MgATP, a characteristic which has been shown similarly for other blockers as glibenclamide, whereas that of PNU 96179 was positively coupled, a typical behaviour for an opener (Ki values at 3 µM / 1 mM MgATP: 65 / 660 nM (PNU 96293) and 6.3 / 2.3 µM (PNU 96179)). In inside out patches, Kir6.2 / SUR2B channels were blocked very

Yale University Department of Cellular and Molecular Physiology School of Medicine 333 Cedar Street, P.O. Box 208026 New Haven CT 06520 8026, USA; *Physiologisches Institut Würzburg Röntgenring 9 D-97070 Würzburg, Germany P 44-6 GENE MUTATIONS IN ANDERSEN’S SYNDROME REVEAL IMPORTANT FUNCTIONS OF THE KIR2.1 POTASSIUM CHANNEL F. Döring, E. Wischmeyer, A. Karschin Andersen’s syndrome (AS) is a rare disorder characterized by cardiac arrhythmia, periodic paralysis and dysmorphic features. The disease occurs either sporadic or as an autosomal dominant hereditary defect. Recently the locus of Andersen’s syndrome was mapped to chromosome 17q23 near the KCNJ2 gene of Kir2.1. The investigation of 16 AS families discovered 9 mutations of highly conserved residues in the Kir2.1 inwardly rectifying potassium channel (Plaster et. al., 2001, Cell, 105, 511-519). In the present study six AS mutations of Kir2.1

S 340 were analyzed by heterologous channel expression in Xenopus oocytes. To evaluate possible functional consequences in both homo- and heterozygous carriers AS mutants were expressed either alone or in combination with wtKir2.1. Of all Kir2.1 AS mutants tested (∆95-98, S136F, G144S, G300V, E303K, ∆314-315) none evoked measurable Kir currents in the recombinant expression system. In contrast, oocytes coinjected with equal amounts of wtKir2.1 and mutated Kir2.1 cRNA displayed profound Kir currents although their amplitudes were reduced by 30% - 83% relative to wtKir2.1 currents (-46.8 µA; Vh= -100mV). The inhibitory effect of mutated Kir channel subunits on current amplitudes was augmented if the ratio of mutated and wtKir2.1 cRNA was increased. Strong dominant negative effects were not only observed in mutations within the highly vulnerable pore region (∆95-98, S136F, G144S), but also for three mutations (G300V, E303K, ∆314-315) within a 30 aa stretch in the cytoplasmic C-terminus. The severity of these ”hot spot” mutations uncovers a functioinal significance of cytosolic Kir2.1 domains. Previously only heterozygous Andersen’s patients have been identified. Our functional analysis may thus suggest that homozygous genotypes are lethal due the absence of Kir2.1 currents. Physiologisches Institut, Universität Würzburg, Röntgenring 9, 97070 Würzburg

Kir2.1 in living cells. PSD95 induces clustered expression of Kir2.1 by binding the channel C-terminus with its second PDZ domain. We fused cyan-fluorescent protein (CFP) and yellow-fluorescent protein (YFP) to PSD95 and Kir2.1, respectively, and co-expressed these constructs in HEK293 cells. Cells were imaged by means of a confocal laser scanning microscope (LSM510, Zeiss), a polychromator (250is Chromex), a 16 channel photomultiplier detector head (PML-16 Becker&Hickl) and a time-correlated single photon counting imaging module (SPC-730, Becker&Hickl). As excitation source a frequency doubled mode-locked Titanium:Sapphire laser was used. This setup enabled us to measure the fluorescence decay of both, CFP and YFP, in each pixel of an image. Within clusters of PSD95 and Kir2.1 fluorescence lifetime of CFP was significantly decreased compared to non-clustered regions. This finding suggests that PSD95 and Kir2.1 interact within these clusters and that CFP fluorescence is quenched due to resonance energy transfer between the CFP moiety (donor) of PSD95 and the YFP moiety (acceptor) of Kir2.1. Supported by the Deutsche Forschungsgemeinschaft (SFB 430) 1

Friedrich-Schiller-Universität Jena, Institut für Physiologie II, Teichgraben 8, D 07740 Jena, Germany; 2Becker&Hickl GmbH, Nahmitzer Damm 30, D 12277 Berlin, Germany; 3Eberhard Karls Universität Tübingen, Institut für Physiologie II, Ob dem Himmelreich 7, D 72074 Tübingen, Germany

P 44-7 P 44-9 ‘SLEEPY’ INWARD RECTIFIER K+ CHANNELS IN CARDIAC MUSCLE ARE ACTIVATED BY STRONG HYPERPOLARIZATION G.X. Liu, J. Daut Potassium channels of isolated guinea-pig cardiomyocytes were studied using the patch-clamp technique. At transmembrane potentials between -120 ands -220 mV we observed inward currents through an apparently novel channel. The novel channel was strongly rectifying; no outward currents could be recorded. Between -200 and -160 mV it had a slope conductance of 42.8 ± 3.0 pS (S .D .; n = 96). The open probability (P0) showed a sigmoid voltage dependence and reached a maximum of 0.93 at -200 mV, half-maximal activation was around 150 mV. The voltage dependence of P0 was not affected by application of 50 µM isoproterenol. The open-time distribution could be described by a single exponential function, the mean open time ranged between 73.5 ms at -220 mV and 1.4 ms at -160 mV. At least two exponential components were required to fit the closed-time distribution. Experiments with different external Na+, K+ and Cl- concentrations suggested that the novel channel is K+ selective. Extracellular Ba2+ ions gave rise to a voltage dependent reduction in Po by inducing long closed states; Cs + markedly reduced mean open time at -200 mV. In cell-attached recordings the novel channel frequently converted to a classical inward rectifier channel, and vice versa. This conversion was not voltage dependent. After excision of the patch, the novel channel always converted to a classical inward rectifier channel within 0 to 3 min. This conversion was not affected by intracellular Mg 2+ , phosphatidylinositol (4,5)-bisphosphate or spermine. Taken together, our findings suggest that the novel K + channel represents a different ‘mode’ of the classical inward rectifier channel in which opening occurs only at very negative potentials. It appears possible that the conversion of inward rectifier channels to the sleepy mode is responsible for the reduction in the density of the inward rectifier current observed in patients with severe heart failure.

THE ION SELECTIVITY SEQUENCE FOR THE BACKGROUND, POTASSIUM CURRENT IK SO IN CEREBELLAR GRANULE CELLS K.M. Ranatunga, D.F. Boyd, A. Mathie IKSO is a non-inactivating, ‘background’ K+ current expressed in dissociated rat cerebellar granule neurons (CGN) (Watkins & Mathie, 1996). It shares many of the biophysical and pharmacological properties of the two pore domain potassium channel, rTASK-1 (Millar et al., 2000). We present here the ion selectivity sequence for IK SO. CGNs were isolated from 6- to 8- day-old Sprague-Dawley rats and cultured as described previously (Watkins & Mathie, 1996). Currents were recorded using the whole-cell perforated patch-clamp configuration, with amphotericin B (320 µg ml-1) as described (Millar et al., 2000). External K + concentration was raised from 2.5 to 25 mM before substitution of K+ with the test ion. In the case of Tl+, 10 mM K/ TlC2H3O2 was used with NaSO4. All remaining K+ was replaced with Na+ to further test Na+ permeability. Cells were held at -20 mV and reversal potential measurements were made using a hyperpolarising 10 ms mV1 ramp of 800 ms duration. Permeability ratios were calculated using the Goldman-Hodgkin-Katz model. The measured permeability ratios relative to K+ were Rb+ (1.01 ± 0.05, mean ± SEM, n = 6) > Tl+ (0.83 ± 0.08, n = 6) > Cs+ (0.22 ± 0.02, n = 5) ≥ NH4+ (0.18 ± 0.02, n = 6) > Li+ (0.062 ± 0.001, n = 3) > Na+ (0.005 ± 0.001, n = 26). The sequence is similar to Eisenmann sequence III. The finite permeability of Cs+ is unusual for a K+ channel suggesting an altered filter region. TASK-1 channels show an identical permeability sequence (Lopes et al., 2000) which supports the hypothesis that the molecular identity of IKSO is TASK-1. Lopes, C.M.B. et al. (2000) J. Biol. Chem. 275, 1696916978; Millar, J.A. et al. (2000) P.N.A.S. 97, 3614-3618; Watkins, C.S. & Mathie, A. (1996) J. Physiol. 491, 401-412.

Institut für Normale und Pathologische Physiologie der Universität Marburg, Deutschhausstrasse 2, 35037 Marburg, Germany

Biophysics Section, Blackett Laboratory, Department of Biological Sciences, Imperial College of Science, Technology, and Medicine, Prince Consort Road, London, SW7 2BW, U.K.

P 44-8

P 44-10

PROTEIN-PROTEIN INTERACTIONS IN LIVING CELLS RECORDED BY SPECTRALLY RESOLVED RESONANCE ENERGY TRANSFER IMAGING C. Biskup 1 , A. Bergmann 2 , W. Becker 2 , C. Stockklausner 3 , J.P. Ruppersberg3, K. Benndorf1, N. Klöcker3. Members of the PSD95 family of membrane-associated guanylate kinases serve as multidomain protein modules that are centrally involved in synaptic organization. Advanced understanding of their role in synaptic development and plasticity requires visualization of the protein-protein interactions in vivo. Here we applied a novel time-correlated single photon counting (TCSPC) imaging technique to study the protein interaction between PSD95 and the inward rectifier potassium channel

MOLECULAR ANALYSIS OF DIVALENT CATION SENSITIVITY OF TASK CHANNELS C. Derst, G.X. Liu, B. Musset, S. Rajan, R. Preisig-Müller, J. Daut The Tandem Pore Domain Acid-Sensitive K+ channel (TASK) family comprises three members (TASK-1, -3 and -5). Despite the high sequence homology in the pore region, these channels show significant differences in their sensitivity to extracellular divalent cations. Heterologous expression in HEK293 cells showed that TASK-1 channels are insensitive to changes in extracellular Ca2+ or Mg2+ concentration, whereas TASK-3 channels showed an about three-fold increase in single channel conductance after removal of external divalent cations. The effects of Ca2+ and Mg2+ ions were virtually identical and could also be

S 341 mimicked by adding La3+ to the external solution. Analysis of TASK-1/ TASK-3 chimeras showed that it is the N-terminal half of the channel which confers Ca2+ and Mg2+ sensitivity. Point mutations of negatively charged residues in the M1-P1 loop identified E-70 of TASK-3 as an important determinant for Ca 2+ sensitivity. Introducing positively charged lysine or arginine residues at position 70 of TASK-3 completely abolished sensitivity to extracellular divalent cations. Introducing the reverse mutation into TASK-1 (K70E) made this channel Ca2+-sensitive. Neutral residues at position 70 of TASK-1 or TASK-3 also produced Ca2+ and Mg2+ sensitive channels, indicating that it is the presence of a positively charged residue in TASK-1 that abolishes sensitivity to divalent cations. Institut für Physiologie, Deutschhausstr. 2, 35037 Marburg, Germany P 44-11 EXPRESSION PATTERN AND FUNCTIONAL CHARACTERISTICS OF TWO NOVEL SPLICE VARIANTS OF THE TWOPORE-DOMAIN POTASSIUM CHANNEL TREK-2 W. Gu, G. Schlichthörl, J.R. Hirsch, H. Engels, C. Karschin, A. Karschin, C. Derst, J. Daut Two novel alternatively spliced isoforms of the human two- poredomain potassium channel TREK-2 were isolated from cDNA libraries of human kidney and fetal brain. The cDNAs of 2438 bp (TREK-2b) and 2559 bp (TREK-2c) encode proteins of 508 amino acids each. RTPCR showed that TREK-2b is strongly expressed in kidney (primarily in the proximal tubule) and pancreas, whereas TREK-2c is abundantly expressed in brain. In-situ hybridization revealed a very distinct expression pattern of TREK-2c in rat brain which partially overlapped with that of TREK-1. Expression of TREK-2b and TREK-2c in HEK293 cells showed that their single-channel characteristics were similar. The slope conductance at negative potentials was 163 ± 5 pS for TREK-2b and 179 ± 17 pS for TREK-2c. The mean open and closed times of TREK-2b at -84 mV were 133 ± 16 and 109 ± 11µs, respectively. Whole-cell currents were larger after transfection with TREK-2C than with TREK-2b, suggesting a role of the alternatively spliced N-terminus in regulating the number of functional channels in the surface membrane. Application of forskolin decreased the whole-cell current carried by TREK-2b and TREK-2c. The sensitivity to forskolin was abolished by mutating a protein-kinase-A phosphorylation site at position 364 of TREK-2C (construct S364A). Activation of protein kinase C by application of phorbol-12-myristate-13-acetate (PMA) also reduced whole-cell current. However, removal of the putative TREK-2b-specific PKC phosphorylation site (construct T7A) did not affect inhibition by PMA. Our results suggest that alternative splicing of TREK-2 contributes to the diversity of two- pore-domain K+ channels. Institut für Normale und Pathologische Physiologie der Universität Marburg, Deutschhausstrasse 2, 35037 Marburg, Germany P 44-12 TWO COMPONENTS OF ACTIVITY-DEPENDENT TRANSIENT K+-CURRENT (IKSLOW) IN MOUSE PANCREATIC B-CELLS S.O. Göpel, P. Rorsman, T. Kanno Background and aims: The amount insulin secreted correlates with the time β-cells are electrically active. At intermediate glucose concentrations β-cells within intact islets generate bursts of action potentials which are rarely seen in dispersed β-cell, the standard preparation for patch-clamp experiments. We applied the patch-clamp technique to functionally identified β-cells within intact islets in order to identify the processes involved in the generation of the bursts of action potentials. Material and methods: All experiments were performed in the perforated patch configuration on B-cells within intact pancreatic mouse islets. Results: Trains of simulated action potentials associated with the gradual development of an outward K+current (IKslow) with an average amplitude of 80±28 pA (n=10). IKslow flowed through both sulphonylurea-sensitive TEA-resistant K ATP channels and sulphonylurea-resistant TEA-blockable K+-channels. In the presence of tolbutamide (0.1 mM), the peak current amplitude averaged 21±3 pA (P<0.05 vs. control). Upon cessation of electrical activity, this current deactivated with a biphasic time course. The kinetics of deactivation could be described as the sum of two exponentials with time constants (τ) of 0.9s and 20 s. The rapid component was

blocked by TEA (10 mM) whereas the slow component was abolished by tolbutamide. Increasing the glucose concentration from 10 to 20 mM reduced the slow (KATP-channel-dependent) component by 50% but whilst not affecting the peak amplitude. The Ca2+-ATPase inhibitor thapsigargin increased electrical activity evoked by 10 mM glucose two-fold. This effect was associated with a 25% reduction of peak IKslow but there was no change in the holding current that could account for the stimulation of electrical activity. Conclusions: We propose that: 1) Ca2+-entry during electrical activity leads to an increased K+conductance (IK,slow) triggering membrane repolarisation; 2) IKslow flows through two types of K +-channel with distinct pharmacology, regulation and deactivation kinetics, one of which being identical to the KATP-channels; and 3) thapsigargin may stimulate electrical activity by exerting an ATP-sparing action by inhibition of the Ca2+ -ATPase thus preventing rapid reduction of the cytoplasmic ATP/ADP-level, opening of KATP-channels activity and membrane repolarisation. Institute for Physiological Sciences, Lund University, The Biomedical Center, F11, SE-221 84 Lund, Sweden. [email protected] P 45-1 EXPRESSION OF T-TYPE Ca2+ CHANNELS IN THE MURINE HEART DURING EMBRYONIC DEVELOPMENT Y. Duan, *A. Ehlich, J. Hescheler, B.K. Fleischmann Voltage dependent L-type Ca 2+ channels (VDCC) are known to be involved in excitation contraction coupling, a key function of cardiomyocytes. Beyond VDCC T-type Ca2+ channels were reported to be expressed in adult pacemaker cells and are therefore possibly involved in the mechanism generating the spontaneous electrical activity. Aim of the present study was to investigate expression, biophysical characteristics and molecular identity of T-type Ca 2+ channels in the murine embryonic heart, where the different cardiac subtypes are still spontaneously beating. In both atrial and ventricular cardiomyocytes a low voltage activated inward current was detected starting from E8.5 of embryonic development. This current component was identified as T-type Ca2+ channel because of its negative threshold of activation close to –60 mV with peak current amplitude at a potential of –30 mV. This current was also detected in Na+ free solution, in the presence of tetrodotoxin (10 µM) a selective blocker of Na2+ channels and nisoldipine (10 µM) a selective antagonist of VDCC, excluding atypical Na + or L-type Ca 2+ channels. At early stages of embryonic development (E10.5) all atrial and ventricular cardiomyocytes expressed T-type Ca2+ channels with a current density of 5.76±1.0 pA/ pF (n=10) and 3.6±0.7 pA/pF (n=11), respectively. During further development the percentage of cells expressing T-type Ca2+ channels and the current density declined. At the late stage (E16.5/E17.5) 89 % of atrial and 56 % of ventricular cells displayed T-type Ca2+ channels with current densities of 1.5±0.4 pA/pF (n=9) and 0.42±0.2 pA/pF (n=9), respectively. We conclude that T-type Ca 2+ channels are expressed in all subtypes of cardiomyocytes at relatively high density during early stages and downregulated at later stages of embryonic development. The molecular subtype of T-type Ca2+ channels and their possible functional role is currently under investigation. Institute of Neurophysiology, University of Cologne, D-50931 Cologne, Germany, *Axiogenesis, Cologne P 45-2 LOW-THRESHOLD Ca2+ SPIKES IN DENTATE GYRUS GRANULE CELLS OF THE RAT HIPPOCAMPUS C. Schmidt-Hieber, P. Jonas, J. Bischofberger In the dentate gyrus of the hippocampus, new granule cells are continuously generated from a population of progenitor cells located in the subgranular zone, even in adult mammals. Therefore a gradient of cell ages is established, with the oldest cells located more superficially and the youngest ones deep in the granule cell layer at the border to the hilus. On the other hand it was reported that granule cells (GCs) can be classified into three maturity groups according to their input resistance; into mature (Rin ≤ 400 MΩ), intermediate (400 ≤ Rin ≤ 1 GΩ) and immature GCs (Rin ≥ 1 GΩ, Liu et al. J Neurophysiol 76:10741088). To compare the active properties of superficial and deep GCs, we made whole-cell patch-clamp recordings in acute hippocampal brain slices of 2- to 3-week old rats at 24°C. GCs were identified by IRDIC videomicroscopy and biocytin filling during the whole-cell re-

S 342 cording. Superficial GCs had an average input resistance of 272 ± 29 MΩ (n=14, mean ± SEM), consistent with maturity. In contrast to the superficial GCs, which form a fairly homogenous population of neurons, the deep GCs fall into 3 different groups according to their Rin. The most abundant group were immature GCs (35 out of 40 cells, Rin = 3.47 ± 0.39 GΩ) with a membrane time constant of 114 ± 11ms and a resting potential of -66.0 ± 2.4 mV. A small depolarization (I = 5-10 pA, 800ms) above a threshold potential of –53.6 ± 0.8 mV evoked a low threshold action potential with a peak amplitude of 15.3 ± 1.1 mV (n=31) and a duration of 100 to 300 ms (mean 196 ± 14 ms, n=31). These potentials were never observed in intermediate or mature GCs. A stronger depolarization (I ≥ 500 pA, 1ms) evoked a fast action potential with an amplitude of 98.4 ± 4.7 mV and a half duration of 2.7 ± 0.2 ms, n=15). The amplitude of the low threshold spikes was not changed by application of 1µM TTX (106 ± 7% of control, n=7), but largely blocked by 50 - 100µM Ni2+ (6.3 ± 1.9% of control, n=7), indicating that these potentials are generated by low-threshold T-type Ca 2+-channels.

manner (n=5). Accordingly, glucose-elevated [Ca 2+]c was reversibly reduced to basal values by 10-20 µM ritonavir (n=5). Interestingly, an inhibition of K ATP channels by 100 µM tolbutamide was unable to increase [Ca2+]c in the presence of ritonavir (n=3). This points to an inhibition of Ca2+ channels by ritonavir, however, the electrical activity induced by glucose (15 mM) was not suppressed. Nevertheless, the pattern of spike activity completely changed. Single action potentials broadened to slow potential oscillations of more than 1 s duration on top of the plateau potential from which action potentials normally start (n=6). These slow potentials were inhibited by the L-type Ca2+ channel blocker D600 (50 µM; n=4). Ca2+ action potentials are normally terminated by deactivation of the Ca 2+ channels as a result of the opening of voltage-dependent K+ channels. Ritonavir (10 µM) partly inhibited the voltage-dependent K + current and induced strong inactivation of these currents (n=3). We conclude that ritonavir inhibits insulin scretion by changing the electrical activity of B-cells through an influence on ion channels. The observed effects of ritonavir thus contribute to the understanding of the occasional occurrence of diabetes during ritonavir treatment.

Physiologisches Institut der Universität Freiburg, Hermann-HerderStr. 7, D-79104 Freiburg, Germany. Supported by DFG (Bi 642/1-2)

Pharmazeutisches Institut, Auf der Morgenstelle 8, D-72076 Tübingen

P 45-3

P 45-5

DIFFERENTIAL Ca2+ CHANNEL ACTIVITY IN PRIMARY HUMAN NEUROENDOCRINE TUMOR (NET) CELLS AND PERMANENT NET CELL LINES FROM FORE- AND MIDGUT TUMORS ANALYZED BY A NEW DIAGRAM METHOD S. Mergler, K. Detjen, W. O. Bechstein, P. Neuhaus, B. Wiedenmann Neuroendocrine (NE) tumor cells express L-, N-, P-, and Q-type Ca2+ channels (Glassmeier et al., 1997). To address the question, if clinically distinct entities of NE tumors are associated with different patterns of Ca 2+ channel expression, the electrophysiological properties of primary NE tumor cultures derived from consecutive surgical resections and permanent BON, RIN and STC1 cells were determined. Using the whole-cell patch-clamp technique, electrophysiological data were obtained from permanent BON, RIN and STC1 cells as well as from cells of primary NE tumors or liver metastasis from fore- and midgut tumors. In addition, time- and potential-dependent current amplitudes of Ca 2+ channels were analyzed by a new diagram method. As previously described, Ca2+ channel characteristics of NE tumors mainly reflected Ltype Ca2+ channels. However, in midgut tumors we identified an additional transient component of Ca2+ channel activity with a high current density. While in primary NET cells from foregut tumors, the mean current density was 4.25 ± 0.52 pApF-1, the mean current density in NET cells from midgut tumors was 15.95 ± 0.97 pApF-1 (± SEM; n = 7 - 13). In addition, activation threshold potential shifted from -39 ± 2 mV to -44 ± 0.5 mV (n = 7 - 13). These characteristics were suggestive of T/R-type channels. Because midgut tumors but not foregut tumors frequently produce hypersecretion syndrome, the additional T/ R-type channels may be functionally relevant for hypersecretion. Since specific inhibitors of T/R-type channels exist and could provide a novel therapeutic approach to hypersecretion, further analysis of T/ R type Ca2+ expression and neuroendocrine secretion in NET cells is needed.

INFLUENCE OF INTEGRIN LIGANDS ON L-TYPE Ca 2+ CHANNEL ACTIVITY IN HUMAN RETINAL PIGMENT EPITHELIAL CELLS W. Kröger, R. Rosenthal, O. Strauß Purpose Choroidal neovascularization is the most severe complication in age-related macular degeneration (AMD). The majority of angiogenic factors is secreted by retinal pigment epithelial (RPE) cells expressing a subtype of L-type channels known to regulate secretion. Since changes in the extracellular matrix might be responsible to initiate this process we studied the effects of integrin ligands on the activity of L-type channels. Methods Ba 2+ currents through L-type channels were measured in cells of the human RPE cell line ARPE-19 using the perforated patch-clamp configuration. Results Under extra- and intracellular K+-free conditions and in the presence of 10 mM Ba2+ in the bath solution cells of the ARPE-19 cell line show voltage-dependent inward currents with characteristics of L-type Ca 2+ channels: the currents activate at potentials more positive than -30 mV, were reduced to 51.2 ± 5.4 % (n = 4) of control by 10 µM nifedipine and showed an inactivation time constant of 39.4 ± 7 ms (n = 9). Application of soluble fibronectin (0.1 µM) led to a reduction of the maximal L-type current amplitude to 65.7 ± 8.4 % (n = 3) of control. Application of soluble vitronectin (0.03 µM) reduced the maximal current amplitude to 52 ± 12 % (n = 3) of control. Application of GRGDNP peptide (100 µM), an inhibitor of fibronectin-dependent cell attachment, reduced the L-type current to 65.6 ± 5 % (n = 3) of control and GRGDSP peptide 100 µM), an inhibitor of the vitronectin-dependent cell attachment reduced the L-type currents to 52 ± 7 % of control (n = 3). Conclusions In summary, disruption of insoluble ligand binding to integrins by application of soluble integrin ligands or RGD-peptides changes activity of L-type channels in RPE cells. Thus, changes in the extracellular matrix occurring during development of AMD influences Ca 2+ channel activity. This effect might be of importance in the induction of choroidal neovascularization in AMD.

Universitätsklinikum Charité der Humboldt-Universität zu Berlin, Med. Klinik, Gastroenterologie, Campus Virchow-Klinikum, Augustenburger Platz 1, D-13353 Berlin, Germany

Institut f. Klinische Physiologie, Universitätsklinikum Benjamin Franklin, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany

P 45-4 THE HIV PROTEASE INHIBITOR RITONAVIR DIRECTLY INTERFERES WITH PANCREATIC B-CELL FUNCTION Y. Neye, P. Krippeit-Drews, M. Düfer, G. Drews Ritonavir is a protease inhibitor widely used in the pharmacological treatment of HIV infection. One of its severe side effects is the occurrence of glucose intolerance or overt diabetes mellitus. To date it is still unclear whether this is owing to an increased peripheral insulin resistance or to a decreased insulin secretion. Therefore, we tested ritonavir for effects on mouse pancreatic B-cell function. We investigated membrane potential and whole-cell currents with the patchclamp technique and cytosolic free Ca2+concentration ([Ca2+]c) with a fluorescence technique. Insulin release was measured in vitro in a steadystate incubation of whole islets by RIA. Glucose-induced insulin release was inhibited by ritonavir (10 and 20 µM) in a concentration-dependent

P 45-6 CHARACTERISATION OF IONIC CHANNELS IN RAT BASOPHILIC LEUKEMIA CELLS O. Kréneisz, A. Cavalié Rat basophilic leukemia (RBL) cells express K+ and Cl- channels as well as Ca 2+-permeable large-conductance channels and Ca2+-selective storeoperated channels (CRAC). Therefore, RBL cell lines are useful models in studies of antigen-induced calcium entry. We examined the ionic conductances of non-stimulated RBL-1 and RBL 2H3-hm1 cells using the patch-clamp technique in the cell-attached mode. The pipette solution contained (in mM) 115 Na+, 5 K+, 2 Ca2+ and 2 Mg2+ (pH 7.4) and single channel currents were searched in the range of pipette potentials between -100 mV and +100 mV. Under these conditions, we

S 343 regularly observed only inward currents with discrete amplitude levels, revealing the opening and closing of ionic channels. Typically, we recorded single channel currents with an amplitude of 2.09 ± 0.16 pA at –100 mV. Accordingly, the slope conductance between -100 mV and -50mV was 8.77 pS. Furthermore, voltage-clamp ramps showed that one important characteristic of the underlying ionic conductance is an strong inward-going rectification. The highest opening probability was observed at hyperpolarising potentials and the estimated mean open time was >0.5 s at -90 mV. Essentially, no outward current was detected up to +100 mV. Removal of K+ from the pipette solution had no effect on the single channel currents. Ionic currents through CRAC channels were not detected under the present experimental conditions. The lowconductance (~9 pS) channel identified in the present study is likely a new pathway for cationic entry into resting RBL cells. Pharmakologie und Toxikologie, Universität des Saarlandes, Homburg P 45-7 DISRUPTION OF THE CAV2.3 GENE OF VOLTAGE-GATED Ca2+ CHANNELS A. Pereverzev1, D. Sochivko2, R. Vajna1, C. Gissel1, M. Mikhna1, J. Hescheler 1, N. Smyt h3, H. Beck2, T. Schneider1 Calcium entry through voltage-gated Ca2+ channels is important for neurotransmitter release and secretion. Expressional and functional data have suggested a role for Cav2.3 Ca2+ channels in neuronal firing and hormone release. To verify its role in vivo, Ca v2.3 (-|-) mutant mice were generated, thus deficient in α1E / R-type Ca2+ channels. In hippocampal CA1 pyramidal cells and dentate granule neurons, as well as neocortical neurons of wild-type mice, Ca 2+ current components resistant to organic Ca 2+ channel blockers (ω-conotoxin GVIA 2 µM, ω-conotoxin MVIIC 3 µM, ω-agatoxin IVa, nifedipine 10 µM) were detected showing biophysical properties similar to cloned Ca v2.3 subunits. They were sensitive towards Ni2+ (100 µM: 71-86% block) but virtually insensitive to the spider toxin SNX-482 (100 nM: 4-6% block) in CA1 pyramidal and cortical cells. In dentate granule neurons, a small SNX-482 sensitive component was detected (16% block). Recordings in Cav2.3(-|-) cells confirmed that Cav2.3 underlies the major fraction of the blocker-resistant Ca2+ current in both neocortical (83%) and CA1 pyramidal cells (66%), and a significant portion of this current in dentate granule cells (46%). After intraperitoneal injection of D-glucose, Cav2.3(-|-) mutant mice (10 week old) showed a reduced glucose tolerance and an impaired insulin release. In isolated islets of Langerhans no insulin release was detected during 30 min challenge by 20 mM glucose. Stress-induced hyperglycemia was reduced in Cav2.3(|-) mice assuming a defect in catecholamine release from adrenal gland. The phenotype of Cav2.3-deficient mice is complex and reveals components related to neurological and endocrine disturbances. 1

3

Institute of Neurophysiology, Institute of Biochemistry, University of Cologne, Robert-Koch-Str. 39, D-50931 Köln, Germany. 2 Department of Epileptology, University of Bonn, Medical Center, SigmundFreud Str. 25, 53105 Bonn, Germany P 45-8 EFFECTS OF VOLTAGE SENSOR MUTATIONS ON L-TYPE CALCIUM CHANNEL GATING C. Hang, A. Alekov, W. Melzer, F. Lehmann-Horn, K. Jurkat-Rott Three mutations in voltage sensors of the skeletal muscle L-type calcium channel are well known for causing hypokalemic periodic paralysis (HypoPP). Previous electrophysiological studies from different research groups including ours yielded contradictory results and failed to explain the clinical features of the disease. We introduced the three mutations and two additional double mutations into rabbit cardiac L-type calcium channel. Whole-cell recordings were performed on HEK 293 cells transfected by rabbit calcium channel cDNAs. In normal extracellular solution, current-voltage relationship pulse program showed the channel opening of R650H corresponding to the R528H mutation in the voltage sensor of domain II of the skeletal muscle L-type calcium channel 3mV shift to the left whereas other mutants showed a 10-mV shift to left compared with wild type. The reduction in current density ranged from 40 to 70% in accordance to the clinical severity of the mutations.In the presence of 1 µM (+) Bay K8644 in extracellular solution, all calcium channel variants produced larger current enabling the performance of inactivation and recovery pulse programs. Varied effects

of mutants on the activation, deactivation, and inactivation kinetics provided more valuable information about voltage sensor function in calcium channel gating. Reduced current density and voltage sensitivity were the two common findings for all mutants, which might have important implications for triggering HypoPP symptoms. Department of Applied Physiology, University of Ulm, D-89069 Ulm, Germany P 46-1 DIFFERENTIAL EFFECT OF IKR AND IKS BLOCK ON ACTION POTENTIALS IN RABBIT HEART UNDER ß-ADRENERGIC STIMULATION S. Al Makdessi, R.F. Bosch, H. Sweidan, A. Bachmann*, U. Gerlach*, H. Goegelein*, L. Seipel, V. Kühlkamp Dofetilide is a potent blocker of the fast component (I Kr ) of the delayed rectifier potassium current (IK). Its effect is associated with a reverse rate dependence which is responsible for dofetilide-induced proarrhythmia. In contrast, the block of the slow component (IKs) by chromanol 293B exerts a frequency-independent effect. The aim of the present study was to investigate the effect of blocking both IKr and IKs under ß-adrenergic stimulation in isolated Langendorffperfused rabbit hearts. Epicardial monophasic action potentials were recorded by means of contact electrodes and the action potential duration (APD) was measured at 20% (APD20), 50% (APD50) and 90% (APD 90) repolarization. Under control conditions, dofetilide (10-7M) increased APD90 at a cycle length of 500 ms by 34.7 ± 15.3%. With a decrease of cycle length to 333 ms the change of the APD90 was no longer significant (17.5 ± 17.6%, p=ns). 293B had no significant effect on repolarization under control conditions (5.5 ± 10.9% at 500 ms, and 3.3 ± 20.1% at 333 ms, p=ns for both). Under ß-adrenergic stimulation with isoproterenol (10-8M) in spontaneously beating hearts, dofetilide (10 -7M) exerted only a moderate prolongation of APD90 (14.1±12.6%, p<0.05), while a much more pronounced effect was observed with 293B (3x10-5M) (34.7±17.9%, p<0.05). In contrast to the delay in later phases of repolarization, the APD20 and APD50 were not affected by dofetilide or 293B administration. Both substances exhibited no effect on heart rate in Langendorff-perfused hearts. Under ß-adrenergic stimulation, the effect of IKs block increases substantially and the prolongation of repolarization becomes stronger compared to IKr blockade. University of Tübingen, Department of Cardiology, Tübingen, Germany. *Aventis Pharma Germany GmbH, Frankfurt, Germany P 46-2 ANGIOTENSIN II IS NOT INVOLVED IN MODULATION OF THE TRANSIENT OUTWARD POTASSIUM CURRENT IN LEFT VENTRICULAR MYOCYTES OF RATS WITH ASCENDING AORTIC STENOSIS T. Volk, A.P. Schwoerer, M. Wagner, H. Ehmke A reduction in the transient outward potassium current (Ito) is thought to underlie action potential prolongation which is observed in several models of cardiac hypertrophy. The signaling pathways involved in the reduction of I to , however, are unknown. A possible candidate, which has been found to alter Ito magnitude and kinetics in vitro, is the peptide hormone angiotensin II (AngII). To test the hypothesis that AII is involved in the hypertrophy-associated downregulation of Ito in vivo, we subjected rats to ascending aortic stenosis (AS). AS animals develop pressure-induced cardiac hypertrophy without a contribution of the renin-angiotensin system to the hypertrophy. Therefore, a possible effect of AngII on Ito can be investigated independently from the hypertrophic stimulus. In AS animals, the heart weight/body weight ratio increased by 30% compared with sham-operated animals. Oral treatment of AS animals with the angiotensin-converting enzyme inhibitor captopril (40 mg/day) significantly reduced peripheral blood pressure, but did not alter the pressure gradient among the stenosis or the degree of cardiac hypertrophy. Since the magnitude of Ito is distributed heterogeneously among the left ventricular free wall, both, subepiand subendocardial myocytes were investigated. Treatment of AS animals with captopril did not restore Ito current density (at +40 mV) in subepicardial myocytes, instead there was a trend towards a further decrease (14.7 ± 1.8 pA/pF, n=7 vs. 19.3 ± 1.4 pA/pF, n=20, n.s.). Additionally, in subendocardial myocytes, captopril had no significant

S 344 1

effect on Ito current density. Inactivation and recovery from inactivation were unaffected by captopril treatment all groups of myocytes. In conclusion, pressure-induced hypertrophy secondary to AS reduces Ito current density independently from the renin-angiotensin system.

Aventis Pharma Deutschland GmbH, DG Cardiovascular Diseases, 65926 Frankfurt am Main, Germany and 2 Laboratory of Physiology, K. U. Leuven, 3000 Leuven, Belgium

Institut für Physiologie, Universität Hamburg, Martinistraße 42, 20246 Hamburg, Germany

P 46-5

P 46-3 INHIBITION BY FLUOXETINE OF THE HUMAN CARDIAC K+ CURRENT INDUCED BY EXPRESSION OF hERG IN CHO CELLS G. Trube, P. Pflimlin Inhibition of the hERG-mediated K+ current (IKr) in the heart by a large number of antiarrhythmic, antiallergic, neuroleptic and antibacterial drugs is a major source for the acquired long-QT syndrome which eventually may lead to life-threatening torsade-de-pointes. For fluoxetine, a 5-HT uptake inhibitor and common antidepressant drug, there is no warning about QT prolongation in the ”Physician’s Desk Reference”, but the drug is listed in the ”International Registry of Drug Induced Arrhythmias” on the basis of a few case reports. We examined the effects of fluoxetine on CHO cells expressing hERG by whole-cell patch-clamp recording. The compound inhibited the K+ current with an IC 50 of 1.0 µM. It’s main metabolite, norfluoxetine, was about 10fold less potent. In contrast to the use-dependent effects of many other inhibitors of IKr the onset of block by fluoxetine (3µM) did not depend on electrical stimulation of the cells, but recovery from block was slowed in absence of depolarising voltage pulses. Neither voltagedependent activation (V ½ = -17 mV) nor inactivation (V½ = -66 mV) was significantly changed by the compound (3µM). The IC 50 of fluoxetine for inhibiting IKr in vitro is at least 20-fold higher than the therapeutic plasma concentration of unbound drug which may explain the lack of QT prolongation when the compound is appropriately dosed. The shortened, rather than prolonged, action potential duration previously observed in cardiac cells is probably due to an additional inhibition of Ca++ channels (Pacher et al. 2000, Naunyn-Schmiedeberg’s Arch Pharmacol 361: 67). In conclusion, fluoxetine may be a compound like verapamil where inhibition of the hERG K+ channel is overruled by Ca++ channel blockade.

BLOCKING PROFILE OF DIFFERENT Na +/H + EXCHANGE BLOCKERS ON CARDIAC ATP-SENSITIVE K+ CHANNELS C. Bollensdorff, T. Zimmer, K. Benndorf In cardiac myocytes, sarcolemmal ATP-sensitive K + channels (KATP channels) are expressed to a high degree. These channels open massively when the cytosolic [ATP] decreases from several millimolar to several tens of micromolar. It is a matter of debate whether a minor opening of these channels plays a role also under physiological conditions, possibly in the sense of a metabolic feedback of the cells. Under ischemic or anoxic conditions, however, KATP channels open massively. When studying ionic fluxes in the myocardium it is often desired to block various ion transport mechanisms, among those the Na+/H+ exchange. Appropriate blockers should be maximally specific. We tested the potency of HOE694 and HOE642 (Cariporide) to block KATP channels in excised patches of mouse cardiac myocytes and of Xenopus oocytes expressing SUR2A/Kir6.2 channels and compared these results with earlier data upon the respective action of amiloride and its derivatives EIPA and DCB. At pH=7.4 HOE642 caused half maximum block (IC50) at 138 and 269 µM whereas the IC50 of HOE694 was 62 and 120 µM in myocytes and SUR2A/Kir6.2, respectively. Since it is known that the HOE694 block of the Na+/H+ exchange is more potent at lowered pH, a relevant condition upon ischemia, we also tested the block at pH=6.6. The respective IC50 values were 29 and 32 µM. These values are by two orders of magnitude higher than typical values for the Na+/H+ exchange, quite in contrast to the IC50 values of amiloride and its derivatives which are lower for KATP channels than for the Na+/H+ exchange. It is concluded that both HOE642 and HOE694 are appropriate to block the Na +/H+ exchange without affecting KATP channels. Institute of Physiology II, Friedrich Schiller University Jena, Teichgraben 8, D-07740 Jena, Germany P 46-6

F. Hoffmann – La Roche Ltd., Pharmaceuticals Division, Preclinical Research, Bldg. 70/410, CH-4070 Basle, Switzerland P 46-4 SHORTENING OF ATRIAL ACTION POTENTIAL BY CELL SWELLING IS PRIMARILY CAUSED BY I CL,SWELL : IDENTIFICATION OF DCPIB AS A NOVEL SELECTIVE BLOCKER 1 N. Decher, 1H.J. Lang, 2B. Nilius, 1A. Brüggemann, 1A.E. Busch, 1K. Steinmeyer DCPIB is a novel potent inhibitor of ICl,swell, which blocks native ICl,swell of calf bovine pulmonary artery endothelial (CPAE) cells with an IC50 of 4.1 µM. Similarly, 10 µM DCPIB almost completely inhibited the swelling-induced chloride conductance in Xenopus oocytes and in guinea pig atrial cardiomyocytes. Block of ICl,swell by DCPIB was fully reversible and voltage independent. DCPIB (10 µM) was selective for ICl,swell and did not inhibit ICl,Ca in CPAE cells, chloride currents of the CLCchloride channel family or hCFTR after expression in Xenopus oocytes. DCPIB (10 µM) also showed no significant inhibition of several native anion and cation currents of guinea pig heart like ICl,PKA, IKr, IKs, IK1, INa and ICa. The influence of ICl,swell on the action potential duration (APD) in atrial guinea pig myocytes was examined using DCPIB. In all atrial cells (n = 7) osmotic swelling produced an increase in chloride current and a strong shortening of APD. Both swelling-induced chloride conductance and AP shortening were inhibited by treatment of swollen cells with DCPIB (10 µM). In agreement with the selectivity for ICl,swell, DCPIB did not affect atrial APD under isoosmotic conditions. Preincubation of atrial cardiomyocytes with DCPIB (10 µM) completely prevented both the swelling-induced chloride currents and the AP shortening during subsequent hypotonic cell swelling. We conclude that swelling-induced atrial AP shortening is mainly caused by activation of ICl,swell. DCPIB therefore is a valuable pharmacological tool to study the role of I Cl,swell in cardiac excitability under pathophysiological conditions leading to cell swelling.

INTRAUTERINE GROWTH RESTRICTION-INDUCED INHIBITION OF L-ARGININE/NITRIC OXIDE PATHWAY IS INSENSITIVE TO HYPOXIA P. Casanello, C. Aguayo, R. Rojas, L. Sobrevia Intrauterine growth restriction (IUGR) impairs L-arginine transport and nitric oxide (NO) synthesis in human umbilical vein endothelium (HUVECs), and hypoxia reduces fetal growth and development. We studied the effect of chronic hypoxia on L-arginine/NO signaling pathway in HUVECs from IUGR. Cells from normal pregnancies or pregnancies with IUGR (birth weight <3th percentil)(Ethics committee approval obtained) were exposed to hypoxia (24 h, 2% O2). The Vmax for L-arginine transport (5.0 ± 0.3 pmol (µg protein)-1 min-1, means ± S.E.M.) was reduced (P<0.05, unpaired Student’s t test) by IUGR (3.6 ± 0.2) and hypoxia (2.8 ± 0.2). However, hypoxia did not alter IUGRreduced transport (3.8 ± 0.2). Influx of tetraphenylphosphonium was reduced in IUGR (31 ± 4 %) and hypoxia (17 ± 3 %), but not in IUGR cells exposed to hypoxia. hCAT-1 and hCAT-2B mRNA levels were reduced by IUGR (69 ± 7 and 70 ± 6 %, respectively) and hypoxia (50 ± 6 and 44 ± 5 %, respectively), but IUGR-induced reduction in hCATs mRNA was not further altered by hypoxia. eNOS mRNA and protein levels were increased by IUGR (88 ± 5 and 71 ± 8 %, respectively) and hypoxia (40 ± 4 and 25 ± 2 %, respectively), while L-[3H]citrulline synthesis from L-[3H]arginine was reduced by IUGR (67 ± 5 %) and hypoxia (80 ± 5 %). Hypoxia did not alter IUGR-induced reduction of NO synthesis. In summary, IUGR and chronic hypoxia reduce L-arginine/NO pathway activity due to lower expression of hCAT-1 and hCAT-2B transporters, and eNOS activity in HUVECs. Reduced NO synthesis and L-arginine transport could stimulate eNOS expression. Hypoxia could be a key regulatory factor at a gene level to induce IUGR. Supported by FONDECYT 1000354 & 7000354, DIUC 201.084.003-1.0 (Chile), and The Wellcome Trust (U.K.). P. Casanello and C. Aguayo hold Beca Docente University of Concepción- and CONICYT-PhD fellowships (Chile). CMPL, Department of Physiology, Faculty of Biological Sciences,

S 345 and Department of Obstetrics & Gynecology, Faculty of Medicine, University of Concepción, P.O. Box 160-C, Concepción, Chile

norepinephrine-induced hypertrophy. It may be more attributed to the increased mass of the ventricles with the accompanying remodelling of the ECM than to increased fibrosis.

P 46-7

Carl-Ludwig Institut für Physiologie, Universität Leipzig, Liebigstr. 27, 04301 Leipzig

CARDIAC SPECIFIC OVEREXPRESSION OF INOS IN TRANSGENIC MICE ATTENUATES β-ADRENERGIC STIMULATION IN VIVO A. Molojavyi, J. Heger, J. Schrader, A. Gödecke Clinical and experimental studies associate overexpression of inducible NO-synthase (iNOS) with the development of heart failure. Interestingly transgenic mice tolerate the chronic cardiac-specific overexpression of iNOS (TGiNOS) and show only minimal functional consequences of cardiac function ex vivo (Langendorff-preparation) (Circ. Res., in press). For further analysis we studied the hemodynamic and contractile response to β-adrenergic stimulation in vivo with a Millar-pressurevolume-catheter (1.4F). Mean arterial blood pressure (MAD) and heart rate (HR) measured in mice anaesthetized with urethane were reduced in TGiNOS mice: (MAD: WT: 78,6±14,1; TGiNOS: 69,9±12,6 mmHg; n=20; P<0.05; HR: WT: 581±51; TGiNOS: 536±74 min -1 ; n=20; P<0.05). Basal parameters such as developed pressure (WT: 92±8; TGiNOS: 82±9 mmHg; P<0.05), dP/dtmin (WT: -7192±1172; TGiNOS: -5919±1031 mmHg/s2, P<0.05), end-diastolic pressure (WT: 3.2±0.9; TGiNOS: 6.1±1.4 mmHg, P<0.001) and time constant of relaxation (τ) (WT: 3±0.9; TGiNOS: 5.7±1.9 ms; P<0.05) were slightly altered. β-adrenergic stimulation (Dobutamin 4 mg/kg*min i.v. 15 min) increased dP/dtmax (basal: 12000±2060; Dob. 14600±1900 mmHg/s2, P<0.05), HR (basal: 607±48; Dob.: 699±32 min -1 ;P<0.05), stroke volume (SV) (basal: 18±9; Dob.: 22±9 ml; P<0.005) and stroke work (SW) (basal: 1440±758; Dob: 1826±860 mmHg*ml, P<0.05) in control hearts. In TGiNOS mice the Dobutamine-induced increase was either fully blunted (dP/dtmax, SV, SW) or the increase was significantly attenuated (HR). These data show, that cardiac-specific overexpression of iNOS only minimally affects basal cardiac contractility in vivo. However, under β-adrenergic inotropic and chronotropic response is considerably reduced.

P 46-9

Institut für Herz- und Kreislaufphysiologie, Heinrich-Heine-Universität, Universitätsstr. 1, 40225 Düsseldorf

INCREASE IN SERUM BUT NOT MYOCARDIAL TNFα-CONCENTRATION DURING PACING-INDUCED HEART FAILURE IN RABBITS S. Aker, T. Neumann, S. Belosjorow, I. Konietzka, G. Heusch, R. Schulz An increase in the myocardial TNFα-concentration (mTNFα) during intracoronary TNFα-infusion is associated with the development of contractile dysfunction in dogs. In heart failure (HF) patients with left ventricular (LV) systolic dysfunction the serum TNFα-concentration (sTNFα) is increased. To study whether this increase in sTNFα is a cause or consequence of HF, HF was induced by rapid LV pacing (400 bpm) in ten chronically instrumented rabbits. After 3 weeks, rabbits displayed clinical signs of HF; echocardiography revealed an increase in LV end-diastolic diameter from 15±2 to 18±2 mm and a reduction in shortening fraction from 35±5 to 11±4% (Mean±SD; both p<0.05). Nine sham-operated rabbits served as controls. sTNFα, mTNFα and liver TNFα-concentrations (lTNFα) were measured using the WEHIcell assay. sTNFα was increased in HF rabbits compared to controls (267±27 vs. 171±24 U/ml, p<0.05). Also lTNFα was increased in HF rabbits compared to controls (2215±295 vs. 1495±125 U/g, p<0.05). In contrast, mTNFα did not differ between HF and controls (140±11 vs. 125±14 U/ml, ns). In the liver (HE-staining) the number of leukocytes was higher in HF rabbits compared to controls (112±40 vs. 48±14 cells/mm2, p<0.05). Close, linear correlations between the number of leukocytes and lTNFα (r=0.78) as well as between lTNFα and sTNFα (r=0.68) were demonstrated. In this HF model, with reduced LV systolic function and clinical signs of HF, mTNFα remains unchanged. The increase in sTNFα appears to be a consequence rather than a cause of heart failure, and is possibly a result of gastrointestinal congestion and subclinical inflammation.

P 46-8

Abteilung für Pathophysiologie, Universitätsklinikum Essen, Hufelandstr. 55, 45122 Essen, Germany

EFFECT OF TRIIODOTHYRONINE TREATMENT ON THE EXPRESSION OF mRNA OF CYTOKINES AND OF EXTRACELLULAR MATRIX PROTEINS IN RAT HEARTS B. Ziegelhöffer-Mihalovièová, W. Briest, B. Raßler, H-G. Zimmer Increased expression of cytokines as well as of extracellular matrix proteins (ECMP) was reported in several models of cardiac hypertrophy e.g. by treatment of rats with norepinephrine infusion. In this study the effect of triiodothyronine (T3) on the expression of mRNA for cytokines and ECMP was investigated. Methods: Female SpragueDawley rats were treated daily with T3 in a dose of 0,2 mg/kg of body weight s.c. Changes in the left (LV) and right (RV) ventricular function were measured 6, 12, 24, 48, 72 hours and 7 and 14 days after the first T3-injection using Millar ultraminiature pressure catheters. RNA was isolated from left (LV) and right (RV) ventricular tissue and the expression of cytokines and ECMP was investigated using the ribonuclease protection assay. Results: T3-treatment induced a significant increase in heart rate (p<0,001), as well as LV dP/dtmax, RV dP/dtmax, LV dP/dtmin and RV dP/dtmin (p<0,05) as soon as 24 hours after the first injection of T3. The RV systolic pressure increased 72h after the first T3 injection, whereas the LV systolic pressure remained unchanged. After 72h, the heart weight/body weight ratio was increased (p<0,001). Hypertrophy of the RV was more prominent than that of the LV (155,9% vs. 137,7% respectively). In all groups the expression of mRNA for IL-6, TNF-α, IL-1α or IL-1β in both ventricles did not change (p>0,05). There was a significant increase in the mRNA for colligin 24h after the T3 injection in both, LV (p<0,01) and RV (p<0,05). This was followed by an increase in the mRNA for collagen I and III 72h after the first T3-dose (p<0,05 in RV; p<0,01 in LV). At this point, the mRNA for TIMP-2 increased (p<0,01) in the LV only. Moreover, after 7 days also the mRNA for MMP-2 increased (p<001). Neither TIMP-2 nor MMP-2 were affected in RV. Conclusions: T3-treatment of rats induced significant cardiac hypertrophy without any changes in the mRNA for cytokines. The increase in collagen RNA, though significant, was moderate as compared to that found in the model of

P 46-10 THE EFFECT OF SOMATOTYPE ON THERMOGRAPHIC IMAGING OF THE BACK A.I. Heusch, P.W. McCarthy We are currently compiling reference bank of standard thermal images from humans. Although controversial. one factor that may affect surface heat production and the resulting thermal image, is subcutaneous fat. Therefore this has been subjected to investigation. 18 male volunteers (age range 19 - 45 years old) were imaged. The subjects disrobed and equilibrated for twenty minutes within a stable environment (23±0.5°C). The dorsal surface of the lower torso was imaged and the average temperature calculated. We found the skin temperature of endomorphs, as recorded by infrared thermography appears lower than that of mesomorphs and ectomorphs. A further difference relates to the pattern of infrared emission from the lower torso. Non-obese subjects tend to have V or Y patterns (Woodrough, 1985), in contrast, obese subjects tended to have various patterns, although still appearing symmetrical. Therefore, consideration of somatotype is important in determining standard images. Woodrough RE., (1976) Thermographic screening for scoliosis in adolescents. Acta Thermographica 1: 63-66. This work was funded by the EPSRC. Welsh Institute of Chiropractic and School of Computing, University of Glamorgan, Pontypridd CF37 1DL P 46-11 THE RELATIONSHIP BETWEEN HOT SPOTS AND THE POTENTIAL FOR BLISTERING IN EPIDERMOLYSIS BULLOSA SIMPLEX P.W. McCarthy, B. Bath, A.I. Heusch

S 346 Epidermolysis Bullosa is a collection of inheritable skin disorders, where minor mechanical stress can lead to blister formation. The least severe is the simplex form (EBS). Although a genetic keratin anomaly has been identified, a number of other factors may be involved. 10 volunteers (age range 25-50 yr. old) were imaged following an equilibration period (20 min at 23+0.5 oC) having not exercised, showered or imbibed vasoactive agents in the previous 12hrs. Dorsal and ventral surfaces of the feet were imaged with an infra-red camera (SC500, Flir) and the average temperature calculated (CTHERM). The feet were then placed in a loose plastic cover and immersed in water at 20 oC for sixty seconds. Plantar surface was then imaged every minute for a 15-minute period. EBS subjects had increased temperature in regions associated with new blisters but an unchanged response to the thermal challenge. Welsh Institute of Chiropractic and School of Computing, University of Glamorgan, Pontypridd CF37 1DL P 46-12 NEONATE MICE HAVE ALTERED AUTONOMIC CONTROL OF HEART RHYTHM V. Chu, J. Otero, I. Amende, J. Morgan, R. Volk, T. Hampton Objective: Characterize electrocardiograms (ECGs) in conscious neonatal mice to see whether they correspond to observations in larger mammals, including humans. Methods: The AnonyMOUSE ECG Screening system for non-invasively recording ECGs in conscious mice (no anesthesia, no implants) and e-MOUSE analyses software in male FVB mice (n=6) at 6-days of age (n=5) and again at 4 weeks of age. Results: When the mice were 6-days of age, heart rate was significantly depressed compared to observations made when the mice had aged to 4-weeks (338 ±26 bpm vs. 755 ±16 bpm, P<0.05). Corresponding ECG interval durations were significantly prolonged in neonates compared to mice at 4-weeks. Heart rate variability was significantly reduced in neonates compared to 4-weeks later (0.9 ± 0.1 bpm vs. 7.0 ± 1.6 bpm, P<0.05). FVB 6-days old

100 ms

FVB 4-weeks old

100 ms

Conclusions: Neither anesthetic nor surgery was required to record beautiful ECG signals in very young mice. The results in neonates may be attributable to reduced sympathetic and parasympathetic signaling. The developmental changes we observed in mice parallel those in human neonates, lambs, and newborn rats. Our approach, adaptable to the study of small mice that might otherwise die from anesthesia or surgery, might be valuable in examining developmental changes and abnormalities. Mouse Specifics, Inc. and The CuraVita Corporation, P.O. Box 200821, Boston MA USA 02120, USA

response from non-junctional ryanodine receptors, while the junctional RyR response displayed a modest increase (~180%). The inwardlydirected Ca2+ gradient observed during control stimulation was reversed during the agonist application. Similar observations were made in cells that were challenged with 100 nM ET-l for 5 min. We found a positive inotropic response (585 ± 52%, n=8), whereby the non-junctional RyRs displayed an increase in the Ca2+ signal of (449 ± 75%, n=8) while the periphery exhibited a smaller increase (38 ± 15%; n=8). We recently described that in atrial myocytes peripheral, junctional RyRs co-localise with inositol 1,4,5 trisphosphate receptors (InsP 3Rs). We thus investigated how specific stimulation of the InsP3Rs by a membrane permeant InsP3 ester (InsP3BM) alters the spatio-temporal properties of atrial myocyte ec-coupling. Incubation of atrial cells with 10 µM InsP3BM for up to 6 min resulted in a positive inotropic response that was again mediated by the pronounced increase in the Ca2+ signal recorded at central locations. The response remained constant at peripheral release sites, while the non-junctional regions responded with a substantial increase of the Ca2+ transient amplitude (487 ± 65% n=5). From these data we conclude that the inotropic state of the rat atrial myocyte is predominantly determined by the release of Ca 2+ from central, non junctional RyRs. Thus, the sole change of the spatiotemporal properties of subcellular Ca 2+ signals underlies agonist controlled changes in the contractile response of rat atrial myocytes. Laboratory of Molecular Signalling, The Babraham Institute, Cambridge, CB2 4AT, UK P 47-2 EFFECTS OF MEMBRANE POTENTIAL ON STEADY-STATE CARDIAC [Ca2+]i IN THE ABSENCE OF Na-Ca EXCHANGE P. Kupittayanant, A.W. Trafford, M.E. Díaz, S.C. O’Neill, D.A. Eisner The aim of this work was to investigate the control of resting Ca concentration ([Ca2+]i ) and, in particular, the contribution of various voltage-sensitive sarcolemmal pathways. Experiments were performed on single rat ventricular myocytes using the perforated patch technique. The sarcoplasmic reticulum was disabled with thapsigargin and/or ryanodine. [Ca2+]i was measured with Indo-1 or fluo-3. In control (145 mM Na) solution with 12 mM Na in the pipette, a maintained depolarisation from –40 mV to 0 mV produced an increase of [Ca2+]i which was maintained for 10-20 sec before decaying slowly. Depolarisation to more positive potentials produced smaller increases of [Ca2+]i. The increase of [Ca2+]i was abolished by nifedipine (10 µM) showing a contribution from L-type Ca channels. With 30 mM pipette Na, nifedipine had less effect suggesting a role for Na-Ca exchange (NCX) mediated Ca entry under conditions when Nai is elevated. To eliminate contribution from NCX, other experiments were performed in Nafree (both bath and pipette). Under these conditions, the increase of [Ca2+]i on depolarisation was abolished by nifedipine. In addition, as depolarisation was increased from 0 to +100 mV, the increase of [Ca 2+] i first became smaller and, eventually, further depolarisation produced a decrease of [Ca2+]i. Furthermore, hyperpolarisation increased [Ca2+]i . This increase of [Ca2+]i as well as the decrease on depolarisation were still observed in the presence of nifedipine. This suggests an additional pathway for Ca entry in addition to the L-type channel and NCX. Although the identity of this pathway is unclear at present. Unit of Cardiac Physiology, Faculty of Medicine, University of Manchester, Oxford Road, Manchester, M13 9PT, UK

P 47-1 RECRUITMENT OF NON-JUNCTIONAL RYANODINE RECEPTORS CONTROLS THE INOTROPIC RESPONSE OF A TRIAL MYOCYTES TO HORMONAL STIMULATION L. Mackenzie, M.D. Bootman, M.J. Berridge, P. Lipp Electrical stimulation of rat atrial myocytes results in the release of Ca2+ from the peripheral sarcoplasmic reticulum where the junctional ryanodine receptors (RyRs) reside. Under control conditions, the Ca2+ response does not propagate into deeper layers of the cell. This study we investigated how the spatio-temporal properties of Ca2+ signals are modulated during agonist induced inotropic responses. We used isolated fluo4-1oaded rat atrial myocytes and subjected them to electrical field stimulation while simultaneously recording subcellular Ca2+ signals and cell length changes with high speed (120 frames/s) confocal imaging. Upon stimulation of atrial cells with 1 µM isoproterenol for 3 min, the inotropic response of the cells increased by 480% (n=8). We found that this increase was mainly due to a ~750% increase of the Ca 2+

P 47-3 AGEING RETARDS THE DIASTOLIC Ca2+ TRANSIENTS AND CHANGES THE DISTRIBUTION OF SERCA2 ISOFORMS (MOUSE VENTRICULAR MYOCYTES) G. Isenberg*, U. Borschke The hypothesis that the failing heart were a model of the ageing heart predicts that ageing would impair the SERCA mediated diastolic Ca2+decay (Swyngedauw, Physiol Rev 79, 215, 1999). We tested this idea by analysing the cytosolic Ca2+-transients and the subcellular distribution of the isoforms SERCA2a and SERCA2b (colocolization with αactinin). Ventricular myocytes were enzymatically isolated from mice in the age of either 3 months or 19-24 months. Cells were superfused with Tyrode solution (1.8 mM CaCl2, 37 oC) and dialyzed with an KClelectrode solution containing 50 µM K 5Indo1. Ca2+ transients were measured with a pair of photomultipliers. After averaging 10 traces,

S 347 the time course of diastolic Ca2+-decay was fitted with a single exponential function. Voltage-clamp pulses of 40 ms duration depolarised from –45 to 0 mV at a rate of 0.2, 1, 2, 4 or 8 Hz. In myocytes from both young and old mice, the rate of diastolic Ca2+ decay increased with the frequency of stimulation. The rate constants were 10, 12, 15, 18 and 20/s in the young and 8, 9, 12, 14 and 14/s in the old cells. SERCA2 distribution was evaluated with conventional antibody staining. 1) In cells from young mice, SERCA2b localized in stripes of low intensity at z- and m bands. 2) In both old and young cells, SERCA2b contributed small prickles all over the cyotosol and appeared in the nuclear envelope. 3) In cells from old but not from young mice, SERCA2b formed a network adjacent to the surface sarcolemma. In old cells, this SERCA2b network entered the sarcomeres and replaced the regular stripes of α-actinin. 4) In old cells, SERCA2a was expressed to smaller extent (no quantitative data), however, with similar distribution than in the young cells. SERCA2a co-localized with α-actinin, and it did not build a networklike structures as SERCA2b. Conclusion: In myocytes form aged mice, the rate of diastolic Ca2+ decay is only modestly reduced. We discuss that the hypothetical loss of function of SERCA2a in the old animals is partially compensated by augmented expression SERCA2b. In addition, intrusion of SERCA2b networks into the sarcomer may impair the mechanical function of the contractile myofilaments.

ing events. In contrast, fully propagating events (Ca2+ waves) were evident in FKBP12.6 over-expressing cells at an identical intracellular [Ca 2+]. Localized Ca2+ release (Ca 2+ sparks) were evident when cells were perfused with ~170nM Ca2+. Analysis of Ca2+ spark characteristics indicated that peak fluorescence value (F/Fo) decreased from 1.9+0.1 (n=464 events, 4 cells) to 1.6±0.1 (n=303 events, 4 cells) on FKBP12.6 transfection (P< 0.01). Only small (nonsignificant) decreases in width and duration of the Ca2+ spark occurred after FKBP12.6 transfection, but spark frequency decreased from 75.8+6.3 to 53.5+4.2 µs-1 µm-1 (P< 0.01). Preliminary results using rapid application of caffeine (10mM) to assess SR Ca 2+ content, suggest that FKBP12.6 over-expressing cells had increased SR content when compared to the LacZ control cells. Thus FKBP12.6 over-expression in cultured rabbit myocytes has pronounced effects on Ca2+ spark characteristics and the ability of spontaneous SR Ca2+ release to generate propagating Ca2+ waves.

Department of Physiology, Martin-Luther-University Halle., D-06097 Halle

ADENOSINE AFFECTS THE SARCOPLASMIC RETICULUM CA2+ RELEASE BY ACTING ON A2A RECPTORS IN SAPONINSKINNED FERRET CARDIAC FIBRES W. Hleihel, C. Huchet-Cadiou, C. Léoty In mammalian skeletal muscle, it was demonstrated that adenosine inhibits caffeine-induced Ca2+ release in Saponin-skinned fibres by acting on the RyR1 receptors and/or via A 1 receptors present at the sarcoplasmic reticulum level (Hleihel et al., Naunyn Shmiedebergs Arch pharmacol 2001;364(3):259-268). Then, it was of interest to see whether a similar mechanism was present in cardiac muscle. Experiments were performed on saponin (50 µg/ml) or Triton (1 % v/v) skinned fibres (200 µm of diameter and 5 mm length) isolated from papillary muscles of adult ferret heart, and animals were humanely killed in accordance with French guidelines. In saponin-skinned fibres, adenosine (1-100 nM) and the specific A2A receptor agonist CGS 21680 (50 nM) increased the caffeine (2.5 mM) contracture amplitude. The data plotted relative to adenosine concentrations showed a sigmoïd relationship with a Hill coefficient (nH) of 2.1 ± 0.3, a C50 of 35.9 ± 3.6 nM and a maximal potentiation of 36.1 ± 4.9 % (n=6). In addition, The potentiation by adenosine (50 nM: 23.1 ± 2.1 %; n=6) or CGS 21680 (50 nM: 17.4 ± 0.9 %; n=6) of caffeine responses was reduced by the specific A 2A receptor antagonist ZM 241385 (50 nM) to 12.3 ± 2.5 % (n=4) and to 8.3 ± 1.6 % (n=4), respectively. By contrast, the specific A1 receptor agonist (CCPA: 1-50 nM) or antagonist (DPCPX: 25-50 nM) had no significant effect on caffeine-induced Ca2+ release. In Triton-skinned fibres, in presence of caffeine (2.5 mM), the maximal Ca 2+-activated tension of contractile proteins (41.3 ± 4.1 mN/ mm2; n=8), the Hill coefficient (nH=2.2 ± 0.1; n=8) and the pCa50 (6.2 ± 0.1; n=8) were not significantly modified by adenosine (100 nM) or by CGS 21680 (50 nM). We proposed that adenosine potentiates caffeine responses by acting on A2A receptors present at the sarcoplasmic reticulum level and/or by activating directly RyR2 Ca2+ release channel.

P 47-4 DYSSYNCHRONOUS SUBCELLULAR Ca 2+ RELEASE IN PIG VENTRICULAR MYOCYTES F.R. Heinzel2, V. Bito1, K.R. Sipido1 It has been shown mainly in rodents that rapid Ca2+ transients in ventricular muscle cells are based on the simultaneous release of Ca2+ from densely arranged release sites in the sarcoplasmic reticulum (SR). However, little is known about the time course of Ca 2+-release in larger animals with slower heart rates. We examined the time course of spatially resolved Ca2+ release in voltage clamped pig ventricular myocytes (-70 to 0 mV step for 225 ms, K+-based pipette solutions, 37°C) using confocal microscopy (line-scans at 300 Hz, Ca2+-indicator Fluo-3). In 67% of cells (n=47) we observed areas of delayed Ca2+-release along the scanned line. Compared to the early sites, maximal [Ca2+]i was delayed by 84±8 ms (mean±SEM) and tended to be smaller (2.37±0.12 vs 2.73±0.15 F/F0, P= 0.08)(ncells= 13, nregions= 24), whereas [Ca2+]i decay times were comparable. The delayed release areas extended to a width of up to 37 µm. Their spatial distribution remained constant during steady state stimulation. Even with large amplitudes of global Ca2+ release as obtained after a 30 s rest or after stimulation with isoproterenol (3 µM), the release pattern along the line did not become homogeneous. We visualized T-tubules with Di-8-ANEPPS and found irregular areas without T-tubular structures throughout the cell. Simultaneously recorded [Ca2+]i showed early Ca2+-release preferentially in the regions with a dense and intense T-tubular signal. Conclusion: In porcine ventricular myocytes, regions of delayed Ca2+-release can be identified which seem to be related to areas of lower T-tubular density.

Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK. *Department of Cardiology and Pneumology, Georg-August-University Goettingen, D-37075 Goettingen, Germany. P 47-6

1

Labo Exp Cardiology. K.U. Leuven, Campus Gasthuisberg. Herestraat 49, 3000 Leuven, Belgium; 2Dept. of Pathophysiology, University of Essen, Hufelandstr.55, 45122 Essen, Germany

Faculté des Sciences et des Techniques, Laboratoire de Physiologie Générale, CNRS UMR 6018, 2, rue de la Houssinière 44322 Nantes cedex 03-France

P 47-5

P 47-7

Ca 2+ SPARK AND Ca 2+ WAVE CHARACTERISTICS IN PERMEABILSED ADULT RABBIT CARDIOMYOCYTES ARE ALTERED BY FKBP12.6 OVEREXPRESSION Loughrey CL, Rankin A, Reynolds DF, Hasenfuss G *, Prestle J * & Smith GL FKBP12.6 is known to be tightly associated with the cardiac SR Ca2+release channel, and to regulate channel function. We have examined the characteristics of the SR in cultured rabbit myocytes over-expressing FKBP12.6 (via adenoviral transfection) compared to a control adenovirus (LacZ). Myocytes were permeabilised by brief exposure to 0.01 mg β-escin and subsequently superfused with a mock intracellular solution with a [Ca2+] of ~350nM. SR Ca2+ release was imaged using a confocal microscope with l0µM Fluo-3 in the perfusion solution. At this [Ca2+], both untransfected and LacZ transfected cultured myocytes showed local, non-propagating SR Ca2+ release and partially propagat-

VARIABILITY IN MECHANICALLY INDUCED CHANGES IN CARDIAC ACTIONPOTENTIALS: EFFECTS OF PROBE ATTACHMENT M. Lei, P.J. Cooper, P. Kohl Mechanically induced action potential (AP) shortening, cross-over and lengthening have been observed in cardiac cells and may contribute to rhythm disturbances [1]. Here we study the effects of probe attachment for subsequent application of axial stretch. Isolated Guinea pig ventricular myocytes [2] were studied in perforated patch mode and subjected to local deformation by multiple carbon fibres (12 µm each) or glass filaments (50-150 µm) followed, in the case of carbon fibres, by axial stretch. Effects on APs and whole cell currents during AP-clamp were studied in control and during block of cation non-selective stretch-activated channels (SAC) by 40 µm Streptomycin. Moderate stretch (5-7%) generally causes AP shortening,

S 348 on average by 11% (n=13). This is explained by a weakly inwardrectified current (-0.594 ± 0.188 pA/pF at -40 mV, Erev ~ -11 mV, n = 5). Probe attachment, in contrast, may cause AP shortening, crossover or lengthening, depending on contact area and extent of compression. AP shortening is usually accompanied by resting membrane depolarisation and is inhibited by SAC block. AP lengthening often coincides with resting membrane hyperpolarisation and is insensitive to SAC block. This suggests i) that SAC contribute to AP shortening, and ii) that additional mechano-sensitive mechanisms can be activated by the attachment of the very probes needed to study stretch-induced AP effects. 1) Kohl et al. Cardiovasc Res 2001/50:280289; 2) Linz & Meyer. Cardiovasc Res 1997/33:110-122. Supported by BHF, MRC & Royal Society, London, UK. University Lab of Physiology, Parks Road, Oxford OX1 3PT, UK P 47-8 STRETCH INDUCED CHANGES IN Na AND Ca IN MICE VENTRICULAR MYOCYTES D. Kondratev, M. F. Gallitelli We follow the idea that the biomechanical phenotype modulation in cardiac muscle is regulated by the spatial/temporal pattern of mechanically induced increase of Ca-concentration ([Ca]). If Ca 2+influx induced by mechanical deformation of the cell membrane triggers intracellular signal cascades, an increase in i.c. [Ca] should be expected as a first consequence of stretch. Here we ask whether during stretch calcium merely accumulates in the narrow space underneath the surface membrane or also diffuses deeply into the cytosol to central cell compartments. Single enzymatically isolated ventricular myocytes of mice were stretched for 1-2 minutes with a pipette, while a second patch pipette allowed to record the stretch-induced membrane currents in the silver holder in which the cell eventually was quick frozen and cryocut for electronprobe microanalysis (EPMA), as described previously (Gallitelli et al. J. Physiol. 1999). The [Na] and [Ca] in subsarcolemmal and central cytosol, mitochondria as well as in the nuclear envelope were analysed at the spatial resolution of 16 nm. Elemental digital maps of [Na] and [Ca] were also collected over regions of the cell including the above compartments. The results indicate that 1-2 minutes stretch increased [Ca] in subsarcolemmal cytosol from 1.86±0.66 mmol/kg dw (or 422±106 µM) to 2.76±0.75 mmol/kg dw (or 630± 170 µM). Also [Na] increased in subsarcolemmal cytosol from 104.53±18.78 mmol/kg dw (or 23.7±4.25 mM) to 234.49±30.86 mmol/kg dw (or 53.29±7.01 mM). [Ca] increased also in central cytosol from 2.11±0.49 mmol/kg dw in control (or 480 µM) to 2.93±0.51 mmol/kg dw in stretched myocytes (or 666 µM). In the nuclear envelope [Ca] increased from 1.93±0.47 in control to 4.17±1.99 mmol/ kg dw in stretched myocytes. The data indicate that stretch locally increases [Na] and [Ca] in subsarcolemmal cytosol as expected for stretch-dependent activation of nonselective cation channels. Additionally, the [Ca] increase in central cytosol and in the nuclear envelope indicate that the stretch-induced changes of [Ca] involve also compartments located in more central regions of the cell. Supported by the Deutsche Forschungsgemeinschaft, SFB/TR 02, to M.F. Gallitelli Julius-Bernstein-Institute of Physiology, University of Halle, Magdeburger Str. 6, D-06097 Halle/Saale, Germany P 47-9 GRADIENTS IN CYTOSOLIC SODIUM CONCENTRATION [Na + ] C CAUSED BY STRETCH ACTIVATED CURRENTS (GUINEA-PIG VENTRICULAR MYOCYTES) V.Kazanski*, G. Isenberg Axial stretch between two glass tools has been shown to activate nonselective cation channels (SACs, Kamkin et al., Cardiovasc Res 2000). Here, we studied whether the current through SACs (ISAC) can increase [Na +]c. Myocytes were superfused with warm (37oC) Tyrode solution composed of (in mM) 150 NaCl, 5.4 KCl, 1.8 CaCl2, 1 MgCl2, 10 glucose, 5 HEPES /NaOH (pH 7.4) and dialyzed with 150 KCl, 0.5 MgCl 2, 10 HEPES/KOH (pH 7.2), 0.005 EGTA. Whole cells were voltage-clamped with 70 ms pulses from –70 to 0 mV (1 Hz) and axially stretched by increasing the distance between a glass stylus and the patch electrode (by approx. 15%) with a stepping motor (Sutter MP 285). Cells were loaded with the Na-indicator sodium green

tetraacetate (20 µM) for 60 min, the membrane stained with ANEPPS (2 µM) for 20 min. Images were taken by a 3-D camera system (12 bit), straylight was removed off-line by a deconvolution algorithm. 1 Hz stimulation increased global Na-fluorescence by 6.2%. x-y, x-z and y-z planes (8 sections à 0.5 µm in z) revealed a striped fluorescence pattern, fluorescence was 1.7 times larger in the middle of the sarcomere than close to the t-tubules. From the fluorescence, a 2.5-fold [Na+]c gradient can be calculated. Axial stretch induced ISAC of –0.45 ±0.09 nA (evidence: reversal potential at –5 mV, modest outward rectification and block by 8 µM GdCl3). During continuous stimulation, 60 s stretch increased global sodium green fluorescence by 8.5 ±3% which would translate in a global 25% increase in [Na+]c. Axial stretch increased the fluorescence heterogeneity, i.e. the local maxima were 1.25-fold higher during stretch than before stretch. Assuming a resting [Na+]c of 7 mM, stimulation plus stretch would increase [Na+]c globally to 10 mM, and it would induce local peaks of subsarcolemmal [Na+]c as high as 25 mM. Conclusion: Na+ influx through SACs can increase [Na+]c similar as Na+ influx through Na+ channels (see WendtGallitelli et al., 1993). The increased [Na+]c shows spatial heterogeneities. Resulting increments in peak [Na+]c are high enough for modulating Na+,Ca2+-exchange and thereby contraction. Department of Physiology, Martin-Luther-University, D-06097 Halle

P 47-10 ION SELECTIVITY OF STRETCH-ACTIVATED CATION CURRENTS IN MOUSE VENTRICULAR MYOCYTES A. Kamkin, I. Kiseleva, G. Isenberg Recently we have described stretch activated whole-cell currents (ISAC) in rat, guinea-pig (GP) and human isolated ventricular myocytes (IVM). Here, we analysed ISAC in IVM from adult mice, cells that have action potentials (APs) of approx. 20 ms, short in comparison with those of GP or human IVM. Axial stretch was applied between the patchpipette as fixpoint and a motor-driven glass-stylus. During current clamp (1 Hz stimulation), stretch depolarised the diastolic potential and induced extra-AP, however, it did not increase the AP-duration. In comparison to rat IVM, cell from mice were less stretch sensitive, significant effects needed stretches of 6 µm or larger. Cells were voltage-clamped with 140 ms pulses from –45 to 0 mV. ISAC was defined as difference of the currents before and after stretch. Using Cs+ ions in both bath and pipette solution, ISAC increased with the amplitude of stretch, it was -0.10±0.01 nA (6 µm), -0.13±0.04 nA (8 µm), -0.22±0.05 nA (10 µm) or -0.92±0.02 nA (12 µm). ISAC was abolished after depolymerization of F-actin with 100 µM cytochalasin D (dialysis via the patch pipette, n=12). I SAC was insensitive to cell dialysis of 5 mM BAPTA., e.g. was not a Ca2+-activated current. 5 µM Gd3+ abolished ISAC. ISAC followed a linear voltage-dependence with a Erev close to 0 mV. ISAC was insensitive to the substitution of Cl- ions by aspartate- ions. These results suggest that ISAC flows through non-selective cation channels. The permeability ratio for monovalent cations was estimated by replacing extracellular (e.c.) Na+ by Cs+ or Li+ ions, from the currents at –45 mV we evaluated PCs: PNa : PLi = 1.77 :1.0 :0.60. Substitution of e.c. Na + with TEA + or NMDG + reduced I SAC from 0.88±0.07 to 0.42±0.009 and to 0.06±0.001, respectively. Substitution of 150 mM e.c. Na + with 75 mM Ca 2+ abolished I SAC. Removal of e.c. Ca2+ ions increased in ISAC approx. 2.5-fold. These results support the idea that Ca 2+ ions permeating through SACs exert a fast flickering channel block. Our data suggest that stretch induced Na+ influx through SACs is responsible for the diastolic depolarisation that can induce extra APs when reaching the threshold. The missing effect on the duration of the AP is attributed to the result that repolarizing K+ currents are large in comparison to ISAC in mouse IVM however not in IVM from GP. Dept. of Physiology, Martin-Luther-University, D-06097 Halle, Germany P 47-11 COMPARISON OF Ca 2+ WAVE CHARACTERISTICS IN PERMEABILSED CARDIAC AND SKELETAL MUSCLE FIBRES D. Borstlap, N. MacQuaide, C.M. Loughrey, R.H.A. Fink*, G.L. Smith Ca2+ release from SR of both cardiac and skeletal muscle cells is regulated by clusters of ryanodine receptors arrange in regular arrays at the junction between the T-tubules and SR. Under conditions of increased

S 349 SR Ca2+ load, a local SR Ca2+ release can initiate an intracellular Ca2+ wave that propagates from one release site to the next throughout the length of the cardiac cell. Comparable waves cannot be normally initiated in skeletal muscle cells. In this study, skeletal muscle fibres were isolated from rat extensor digitorum brevis by enzymatic dissociation of isolated whole muscle. Isolated rat cardiac muscle cells were prepared by perfusion of whole hearts with enzyme solution. Both cell types were permeabilised with beta-escin (0.lmg/ml) and exposed to a mock intracellular solution of the following composition (mM): KCl, 100; MgATP, 4.5; CrP, 10; HEPES, 25; Mg2+, 1; EGTA 0.05; pH 7.0 (20-22°C). pH 7.0 Intracellular Ca2+ was monitored using confocal/ linescan imaging (Biorad Radiance 2000; Ar laser) using 0.01mM Fluo3. In isolated permeabilised rat ventricular myocytes, perfusion with 240-270 nM Ca2+ caused regular Ca2+ waves with an average velocity of 157+ 17 µms-1 (SEM n=3) and wave frequency of 0.231±0.03s (SEM n=3). Perfusion of permeabilised skeletal muscle fibres with an identical solution caused no detectable Ca2+ waves. On addition of 2mM caffeine regular low amplitude waves with an average velocity of 17.2 ±1.3 µms-1 (SEM, n=3) and an average wave frequency of 0.096±0.006Hz (SEM, n=3) were observed. Perfusion of permeabilised rat cardiac cells with 2mM caffeine (in 240-270 nM Ca2+) abolished Ca2+ waves. This data is consistent with the view that the Ca2+ sensitivity of the skeletal muscle ryanodine receptor is normally insufficient to initiate and sustain Ca2+ waves, but addition of caffeine increases the sensitivity sufficiently to allow wave initiation and propagation. Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK; *Universität Heidelberg, Institut für Physiologie und Pathophysiologie, Im Neuenheimer Feld 326, 69120 Heidelberg

P 48-1 THE ANTIOXIDATIVE EFFECTS OF THE MANNITOL AND GLUTATHIONE ON GLOBAL MYOCARDIAL FUNCTION AND MHC-ISOENZYME DISTRIBUTION S.C. Garcia Pomblum, V.J. Pomblum, E. Gams, J.D. Schipke Background: Free oxygen derived radicals can initiate or amplify postischemic ventricular dysfunction. The oxidative stress could induce protein oxidation, e.g. of myosin, an important constituent of cardiac muscle. Objective: Test the effects of two antioxidants (mannitol: man and glutathione: GSH) on functional recovery after ischemia/ reperfusion. To investigate whether these antioxidants can prevent myocardial protein from oxidation, the distribution of myosinisoenzymes in cardiac tissue and protein levels of troponin I (TnI) and creatine kinase (CK) in the perfusate were evaluated. Method: Experiments were performed on 33 isolated, blood-perfused rabbit hearts. In series I (n=23; pla: 6; man: 7; GSH: 10) global ventricular function was assessed. In series II (n=5; pla; function and distribution of myosin isoenzymes (minigel PAGE) were assessed. In series III (n=5; pla: 1; man: 2; GSH: 2), function, isoenzyme distribution and proteins in the perfusate were assessed. Protocols: 30 min stabilisation, 15 min control, 20 min ischemia, and 60 min reperfusion. Man (8.6 mM) and pla (0.9 % NaCl) were administered at the onset and GSH (0.3 mM) 20 min after the onset of stabilisation. Results: Results for ventricular systolic function (cardiac output: CO and dP/dtmax) are given for 30 min (rep30) and 60 min (rep60) after onset of reperfusion relative to control (series I) series I rep 30 rep 60

CO (%) pla man 12 44 14 62

GSH 87 91

dP/dtmax (%) pla Man 53 64 50 79

GSH 89 93

Series II: In normoxic myocardium, the ratio between V1 and V3 was 41 to 46 %. In the postischemic hearts, the ratio was inverted to 70 to 30 %. Series III provided no consistent results except for pla, where function was low and TnI and CK levels were high. Conclusion: The data suggest that both antioxidants exert cardioprotection. The increase of V 1 isoenzyme in the postischemic hearts – which is associated with an inefficient contraction- might contribute to the dysfunction in the postischemic hearts. Zentrum Operative Medizin I, Forschungsgruppe Experimentelle Chirurgie, Universitätsklinikum Düsseldorf, Moorenstr. 5, D-40225 Düsseldorf

.

P 48-2 EFFECT OF CARDIAC HYPERTROPHY UPON THE EXPRESSION AND ACTIVITY OF THE ASPARTATE TRANSPORTER, EAAC-1 N. King., J.D. McGivan., M.-S. Suleiman. The non-essential amino acid aspartate has been implicated in myocardial protection [1], where its effectiveness is likely to be influenced by the characteristics of aspartate transport across the cardiac sarcolemma. In rat hearts, a high-affinity sodium dependent transporter mediates aspartate transport and is associated with the expression of EAAC-1 [2]. However all the work so far has concentrated upon normal hearts with little known about whether aspartate transport is affected by pathological conditions. Therefore, in this study we have investigated the effects of one such condition, namely hypertrophy, upon the expression and activity of the aspartate transporter, EAAC-1. Cardiac sarcolemmal vesicles were prepared from hypertrophic hearts of spontaneously hypertensive (SHR) and control Wistar Kyoto (WKY) rats by differential centrifugation [2]. L-[14C]aspartate transport was measured by rapid filtration. Western blotting followed by scanning densitometry was used to assess the level of EAAC-1 expression [2]. Aspartate uptake was of a similar high affinity (Km) in SHR vesicles (20.06 ± 4.63µM) compared to WKY vesicles (15.63 ± 3.22µM). The maximal velocity (Vmax) of aspartate uptake into the SHR vesicles was 27.4 ± 2.17pmol/mg/s, which was significantly greater than the WKY vesicles at 9.16 ± 0.78pmol/mg/s (n=6 ± sem, p<0.05, ttest). This increase in the rate of aspartate uptake was associated with a 2.6 fold increase in the expression of EAAC-1 in SHR vesicles. These results suggest that whilst the same transporter mediates aspartate transport in the normotrophic and hypertrophic heart, the level of expression and rate of transport is upregulated in the hypertrophic heart. 1. Pisarenko OI. (1996). Clin Exp Pharmacol Physiol. 23, 627-633 2. King N et al. (2001). Card Res. 52, 84-94 Bristol Heart Institute, Bristol Royal Infirmary, Bristol. BS2 8HW. England

P 48-3 THE COMITOGENIC EFFECT OF NOREPINEPHRINE IN CARDIAC FIBROBLASTS IS MEDIATED BY IL-6 AND MAPK M. Leicht, W. Briest, H.-G. Zimmer Norepinephrine (NE) is involved in many cardiovascular diseases such as congestive heart failure. We have recently reported that NE had a comitogenic effect in isolated cardiac fibroblasts, and that it activated p42/p44MAPK (Leicht et al., Cardiovasc. Res., 2000;48:274-84). This study was designed to characterize a possible mechanism involved in the proliferative effect of NE. Isolated rat cardiac fibroblasts were exposed to NE (10 µM) for up to eight hours, and interleukin-6 (IL-6) expression was measured by Ribonuclease Protection Assay and Western Blotting. IL-6/GAPDH mRNA was increased by NE in a timedependent manner reaching 23-fold stimulation after one hour compared to untreated samples. Immunoreactivity to IL-6 was not found in controls. After 16 h of exposure to NE, IL-6 protein was detected. It further increased up to 48 h. The effect of NE on IL-6 mRNA was abolished by the â-adrenoceptor blockers propranolol, metoprolol (α 1) and ICI 118.551 (α 2), but not by the á-adrenoceptor blockers prazosin (α 1) and yohimbine (α 2). The MAPK-inhibitor PD98059 suppressed the NE-induced MAPK activation in a concentration-dependent fashion after five minutes, attenuated the NE-induced IL-6 expression after two hours, and suppressed the proliferative effect of NE from 53 % to 18 % after 48 h. Recombinant IL-6 caused an increase in proliferation by 31 % after 48 h. Simultaneous application of the IL-6 antibody reduced the NE-induced proliferation to 34 %, and completely prevented the IL-6-induced effect. These results suggest that NE induces proliferation of rat cardiac fibroblasts in part by increasing the expression of IL-6 through regulation of MAPK. Carl-Ludwig-lnstitut für Physiologie, Universität Leipzig, Liebigstrasse 27, 04299 Leipzig

S 350 P 48-4 TEMPERATURE DEPENDENT EFFECT OF INSULIN ON L-TYPE Ca-CURRENT RECORDED FROM GUINEA-PIG ISOLATED VENTRICULAR MYOCYTES V.K. Pabbathi, M.-S. Suleiman, J. Hancox Insulin has long been known to play a central role in modulating metabolic processes in many tissues, including the heart. It has also been implicated in the regulation of cardiac function as it stimulates Ltype calcium current (ICa,L) in isolated rat ventricular and human atrial myocytes (Aulbach F, et al. 1999; Maier S, et al. 1999). However this observation was made at room temperature. The present investigation was undertaken to determine whether or not the effect of insulin on ICa,L also occurs at a normal physiological temperature. Adult male guinea pigs were humanely killed by cervical dislocation and myocytes were isolated by a combination of mechanical and enzymatic dispersion. Whole cell recordings of ICa,L were made at both room temperature and 37oC in a K+-free external solution, using a Cs+-based pipette solution. ICa,L was elicited by a two-step protocol applied from a holding potential of –80mV: an initial step to –40mV for 100 msec to activate and then inactivate fast sodium current; a second step to +10mV for 600 msec to elicit ICa,L. ICa,L magnitude was measured as the difference between peak and end-pulse current. At room temperature, 1 µM insulin stimulated I Ca,L (127.16 ± 18% of basal level, mean ± S.E.M; n = 6 cells), in agreement with previous reports (Aulbach F, et al. 1999; Maier S, et al. 1999). However, at 37°C, insulin decreased the ICa,L amplitude (50.08 ± 6% of basal levels , mean ± S.E.M; n = 10 cells). These initial data suggest that the effect of insulin on ICa,L from guinea-pig ventricular myocytes may be temperature dependent and that at physiological temperature, the effect is an inhibitory one. Aulbach F, et al. (1999) Cardiovasc Res. 42(1): 113-20; Maier S, et al. (1999) Cardiovasc Res. 44:390-397. Bristol Heart Institute, Bristol Royal Infirmary, Marlborough Street, Bristol, BS2 8HW. UK and Department of Physiology, University of Bristol, Bristol, Bristol BS8 1TD, UK. P 48-5 THE BEHAVIOUR OF PHYSIOLOGICAL PARAMETERS IN ISOLATED PERFUSED RAT HEART C.R. Revnic, V. Voiculescu, M. Sozanski, D. Cergan, S. Botea, F. Revnic Heart is a complex organ system with a buil in capacity for self regulation and adaptation. In particular, regulation of contractile proteins in response to mechanical overload, alteration in autocrine/paracrine regulation in metabolical stressed hearts. Learned tolerance to deprivation of muscle blood perfusion are examples of extent of the ”cardiac plasticity”. The aim of our study was related with the investigation of the behaviour of water-contractile proteins interaction in isolated perfused rat heart muscle as well as the investigation of physiological parameters and the impact of 60 minutes perfusion upon dna stability. Materials and method: Isolated perfused rat hearts from 6 and 24 months old female rats have been mounted and perfused with Krebs-Hanseleit buffer at 37 C in Langerdoff retrograde perfusionsystem at a constant pressure. Physiological parameters such as: coronary flow (CF), heart rate (HR) and left ventricle pressure developed (LVPD) have been recorded at the following intervals of time: 10‘,20‘,30‘40‘,50‘ and 60‘. Apoptotic DNA ETDBr laddering kit has been used to assay heart cells for apoptosis after 60 minutes perfusion. In old rats the values of HR were higher than in controls; CF is vardiable in time of experiment. LVPD values were net elevated in old rats, but with fluctuation in time. There were no signs for apoptosis in heart cells concerning internucleosomal fragmentation and DNA laddering for this experimental model of 60 minutes reperfusion. National Institute of Gerontology and Geriatrics” Ana Aslan”, 9 Caldarusani Str. Sector 1, 78178, Bucharest, Romania P 48-6 1H NMR STUDIES OF CARDIAC CONTRACTILITY IN GEROVITAL H3 TREATED RATS C.R. Revnic, S. Botea, F. Revnic In order to obtain new data concerning the molecular mechanisms of Gerovital H3 action in rat heart, we have investigated the effect of Gerovital H3 incubated in vitro upon heart proton transverse relaxation

times and upon 3H ATP uptake. Our data concerning the ionic behaviour of contractile apparatus in contraction and relaxation by 1H NMR technique have revealed the existence of two proton transverse relaxation times: T2sspin-spin proton transverse relaxation time and T2l spin-lattice proton transverse relaxation time. The existence of two proton transverse relaxation times accounts for the existence of two water compartments: Heart fragments incubated in vitro in Contraction state with Gerovital H3 (200mg/l) have pointed out a decrease inT2s and T2l accaunting for an intensification in H+ binding to the polar groups of contractile proteins. An increase in 3H ATP uptake has been recorded in heart muscle incubated with Gerovital H3 versus controls. Gerovital H3 may have a stimulatory effect upon contraction mechanisms in increasing the affinity of ATP ase for its substrate. N.I.G.G.”ANA ASLAN”, 9 Caldarusani Str. Sector 1,78178, Bucharest, Romania P 48-7 INHIBITION OF THE NADH-UBIQUINON-OXIDOREDUCTASE (COMPLEX I) REDUCES THE REPERFUSIONINDUCED CONTRACTURE OF MYOCARDIUM B. Rother, C. Schäfer, Y. Abdallah, C. Meuter, S. Walther, A. Gkatzoflia, K. Heinz, H.M. Piper Reperfused myocardium can develop Ca2+-independent contracture (recontracture), caused by a slow ATP recovery. The aim of the present study was to investigate whether mitochondrial substrates can influence the reperfusion-induced recontracture and if complex I plays a crucial role in this mechanism. Isolated cardiomyocytes were perfused anoxically under Ca2+-free conditions (100 min, substrate-free) and than reoxygenated (15 min, 2.5mM glucose). During reoxygenation pyruvate or succinate were given with the onset of reoxygenation. Additionally DPI (10µM) for inhibition of complex I was added. Pyruvate increased dose-dependently cell shortening (cell length in % of endischemic cell-length: 1mM 87.0±0.2%; 5mM 85.9±1.1%; 10mM: 79.8±1.1%; 20mM: 74.7±2.5%; p<0.05 vs. Ctr). However, succinate reduced reperfusion-induced recontracture (200µM: 90.3+0.5%; 1mM: 91.8+0.2%; Control: 87.8±0.4%; p<0.05 vs Ctr.). This result shows that complex I substrates (pyruvate) increase, complex II substrates (succinate) reduce recontracture. In presence of DPI the impaired effect of pyruvate was abolished (20mM pyruvate+DPI: 85.2+2.0 p<0.05 vs. 20mM pyruvate). The data indicate that acceleration of recontracture can be due to complex-I activation during reperfusion. This effect might be mediated by generation of oxygen radicals at mitochondrial complex I. Institute of Physiology, Justus -Liebig -University Giessen Aulweg 129, D-35392 Giessen, Germany P 48-8 THE SARCOLEMMAL Na+/Ca2+ EXCHANGER IMPORTANTLY CONTRIBUTES TO THE HYDROXYL RADICAL-INDUCED DAMAGE OF RABBIT VENTRICULAR CARDIOMYOCYTES H. Kögler, S. Wagner, G. Hasenfuss Isolated cardiomyocytes when exposed to oxygen-derived free radicals suffer cytoplasmic Ca2+ overload, hypercontracture, and finally cell death. We examined the contribution of the cellular Ca 2+ gain mediated by the reverse mode of the sarcolemmal Na+/Ca2+ exchanger (NCX) to the Ca2+ overload that occurs after treatment with hydroxyl radical (*OH) in rabbit ventricular cardiomyocytes. Isolated cardiac myocytes over a period of 4 min were exposed to 0.75 mM hydrogen peroxide (H2O2), which in the presence of the catalyst Fe3+/NTA (10µM/ 20µM) generates *OH via a Fenton reaction. This treatment after a latency of several minutes caused the cells to hypercontract, such that during the observation period of 30 min the fraction of intact, rodshaped cells among the total number of cells per view area (FR) continuously decreased. After incubation with 5 µM KB-R7943, a specific inhibitor of the NCX reverse mode, this process was significantly prolonged: FR at 10, 15, and 20 min after onset of *OH treatment was 0,87 ± 0.02, 0,50 ± 0,06, and 0,29 ± 0,08, respectively, under control conditions, whereas in the presence of KB-R7943 the corresponding values were 0,94 ±0,02, 0,72 ± 0,05, and 0,48 ± 0,08 (p=0,036, 0,003, and 0,045, respectively, n=564 and 583 from N=8 myocyte isolation batches). On the contrary, preliminary data from cells in which we

S 351 overexpressed NCX after adenovirus-mediated gene transfer indicate an acceleration of the radical-induced development of hypercontracture. Thus, NCX reverse mode-mediated influx of Ca2+ makes an important contribution to the radical-induced myocyte Ca2+ overload. In the face of the enhanced expression of NCX that has been reported in human heart failure and in various animal heart failure models this mechanism would be suitable to explain the known enhanced susceptibility of failing myocardium toward radical-induced damage.

was not different. In vsmc, augmentation of ICa,L by ECMP was reported for both RGD-motives of insoluble FN interacting with α5β1 and the LDV-motives of soluble FN binding to α4β1 (Wu et al., 2001). Our results suggest for I Ca,L of cardiac ventricular myocytes a similar responsiveness to ECMP, the underlying mechanism, however, remains to be elucidated. Department of Physiology, Martin-Luther-University Halle, D-06097 Halle, Germany

Georg-August-University Göttingen, Department of Cardiology and Pneumology, Robert-Koch-Str. 40, D-37075 Göttingen. P 49-2 P 48-9 THE ROLE OF REACTIVE OXYGEN SPECIES IN VASCULAR STUNNING V.J. Pomblum, S.C. Garcia Pomblum, E Gams*, J.D. Schipke Background: The injury in the coronary vasculature in the postischemic/ reperfused myocardium (=vascular stunning) can in part be attributed to reactive oxygen species. We sought to further investigate the protective effect of an •OH-radical scavenger. Methods: The experiments were performed on 21 isolated, blood-perfused rabbit hearts. An endothelium-dependent dilator (substance P; SP; 5 nM) was used to test the responsiveness of the coronary arterial system. The hearts were randomly assigned to one of three groups: (1) Control; administration of SP after 120 min normoxic perfusion. (2) Stunning; 40 min normoxic perfusion → 20 min no-flow ischemia → 60 min reperfusion → SP. (3) Stunning + mannitol; protocol as in (2) but 4.15 mM mannitol in the perfusate. The ventricular systolic function (left ventricular peak pressure and aortic flow) and and the coronary blood flow were assessed during baseline and 5 min after SP administration. Results: The systolic function before SP administration was not different from baseline in all groups. In the control hearts, SP increased coronary flow transiently by 39 %. In the stunned hearts, SP increased CBF by only 9 %. In the presence of mannitol, the SP-induced vasodilation was improved to 23 %. In turn, SP did affect the systolic function in neither group. Conclusion: Vascular stunning can exist in the absence of myocardial stunning. The severely reduced response towards the endotheliumdependent agent suggests an endothelial dysfunction that seems owing to the cytotoxicity of •OH-radicals that can be scavenged using mannitol. Zentrum Operative Medizin I, Forschungsgruppe Experimentelle Chirurgie, *Klinik für Thorax- und kardiovaskuläre Chirurgie, Universitätsklinikum Düsseldorf

THE NEGATIVE INOTROPIC EFFECT OF NATRIURETIC PEPTIDES IN RAT ATRIAL CARDIAC MYOCYTES IS ASSOCIATED WITH AN INCREASE IN INTRACELLULAR cGMP LEVELS N.S. Freestone, A. Feldheiser, M. Scheuermann-Freestone Introduction: In heart failure there is a well documented elevation in the levels of circulating atrial natriuretic peptide (ANP). However, the inotropic consequences of this increase in natriuretic peptide concentration on mammalian atrial myocytes has not been extensively investigated. The effects of ANP and C-type natriuretic peptide (CNP) are thought to be mediated via intracellular mobilization of the second messenger, cGMP. This study seeks to determine the effect of acute addition of ANP and CNP on the contractility of rat atrial myocytes and cGMP mobilization in suspensions of the same cells. Method: Rat atrial myocytes were acutely isolated, viewed on a microscope and displayed on a monitor whilst being electrically stimulated at 0.5 Hz to contract. Movement of cells was monitored by computer and chart recorder via a photodiode. For measurement of cGMP, atrial cells were preincubated with 1 mM IBMX and then exposed to 10–6 M ANP and CNP. Cellular cGMP levels were then measured by radioimunnoassay. Results: Both natriuretic peptides caused a significant diminution in contraction amplitude of beating atrial myocytes. At 10 –7 M ANP there was a 33% reduction in contraction amplitude compared to control (p=0.009) from 12.8±1.9 µM to 8.6±1.5 µM (n=10). CNP at the same concentration reduced contraction amplitude by 22% from 13.1±2.6 µM to 10.21±2.78 µM (p=0.0004; n=7). In comparison, isoprenaline increased contraction amplitude by 24% (p=0.0004; n=14). ANP and CNP also increased intracellular cGMP levels by 40% (p=0.001) and 65% (p=0.004) respectively in atrial cells. Conclusion: ANP and CNP exert direct negative inotropic effects on rat atrial myocytes. This decrease in contractility is associated with an increase in intracellular cGMP levels. This action of the natriuretic peptides may play a role in the impaired contractile performance of the diseased heart. School of Chemical and Pharmaceutical Sciences, Kingston University, Kingston, Surrey, KT1 2EE, UK

P 49-1 2+

MODULATION OF CARDIAC Ca CURRENTS BY LIGATION OF EXTRACELLULAR MATRIX PROTEINS TO INTEGRINS U. Rueckschloss*, G. Isenberg Binding of extracellular matrix proteins (ECMPs) to integrins is known to modulate ICa,L of vascular smooth muscle cells (vsmc) via tyrosine phosphorylation (Wu et al., 2001). Since cardiac hypertrophy upregulates the expression of ECMP and integrins, we investigated whether ECMPs have a similar effect on ICa,L of cardiac ventricular myocytes (cvm). This extrapolation is not trivial because the corresponding tyrosin residues are not present in the cardiac isoform of the α subunit. Guinea-pig cvm were superfused with Tyrode solution (22 °C) containing (in mM) 150 NaCl, 5.4 CsCl, 1.8 CaCl2, 1.2 MgCl2, 10 glucose, 10 HEPES (pH 7.4). Patch pipettes were filled with 150 mM CsCl. Cvm were voltage clamped with 280 ms pulses from –45 to 0 mV at 0.5 Hz. Zero, 5, 10 and 15 min after intervention, i-v curves were measured. At control, ICa,L run down during the above periods to 91 ±3, 83 ±4 and 74 ±4% (n=8, mean ±S.E.M.). Bath application of soluble fibronectin (FN, 80 nM) significantly attenuated ICa,L run down (93 ±5, 94 ±4, 95 ±4%; n=6, p <0.05 at 15 min). Run down was not modified (93 ±5, 89 ±6, 87 ±6%, n=8) by a synthetic peptide containing the RGD in multiple copies motive, suggesting that augmentation of I Ca,L does not depend on the FN RGD-motive. Immobilized FN augmented ICa,L, i.e. cvm adhered to FN-coated coverslips responded with an enhanced ICa,L (102 ±3, 104 ±8, 105 ±4%; n=5; p <0.05 vs control at 10 and 15 min). Immobilized FN increased ICa,L even more when patch electrodes contained additional 4 mM ATP (118 ±8, 123 ±6 and 121 ±4%; n=5). Since ATP also attenuated run down of ICa,L the effect of immobilized FN on ICa,L in the absence and presence of ATP

P 49-3 DIFFERENTIAL EFFECTS OF ANP AND CNP ON RAT VENTRICULAR MYOCYTE CONTRACTILITY IN HEART FAILURE N.S. Freestone Introduction: Atrial natriuretic peptide (ANP) has been shown to depress the contractility of normal cardiac cells whilst having no effect on the contractility of hypertrophied myocytes derived from pressure-overloaded hearts (Tajima et al, 1998). However, little information is available on the action of ANP and C-type natriuretic peptide (CNP) on the contractility of myocytes derived from animals in overt heart failure. Method: Rat ventricular myocytes were isolated from rat hearts 4 weeks after aortocaval shunt and their age-matched sham-operated controls. The myocytes were viewed on a monitor via a camera linked to a microscope whilst being electrically stimulated at 0.5 Hz to contract. Movement of cells under control conditions and after addition of ANP and CNP was monitored by computer and chart recorder via a photodiode. Results: ANP had no significant effect on contractility of myocytes from shunted hearts whilst having a moderate negative inotropic effect on myocytes from sham-operated animals at 10-6M (p<0.002; n=17). In contrast CNP was negatively inotropic in myocytes from both shunt and sham-operated animals. At 10-7M, myocyte contraction amplitude was reduced from 13.7 ± 1.5 µM to 8.1 ± 1.2 µM in the sham-operated group (p<0.0001; n=14) whilst there was a reduction from 16.0 ± 2.1 µM to 12.5 ± 1.4 µM in the

S 352 shunt-operated group (p=0.018; n = 13). Conclusion: CNP exerts more potent effects on myocytes from both shunt and sham-operated animals than ANP. These results indicate that CNP may be a more effective regulator of cardiac contractility than ANP under both normal and pathophysiological conditions. School of Chemical and Pharmaceutical Sciences, Kingston University, Kingston, Surrey, KT1 2EE, UK

analysis. Preliminary data obtained from permeabilized muscle fibers indicate an enhanced myofilament responsiveness to [Ca 2+]. These results demonstrate a substantially impaired Ca2+ homeostasis in MCT myocardium. A likely cause for this behavior is the reduction in SERCA2a expression levels. Enhanced myofilament Ca2+ responsiveness could represent a compensatory mechanism explaining the baseline hypercontractility. Georg-August-University Göttingen, Department of Cardiology and Pneumology, Robert-Koch-Str. 40, D-37075 Göttingen

P 49-4 DEVELOPMENT OF VULNERABILITY OF THE IMMATURE HUMAN HEART TO ISCHEMIA-REPERFUSION INJURY P. Modi, M. Caputo, H. Imura, A. Pawade, A. Parry, G.D. Angelini, M.S. Suleiman Background: The vulnerability of the immature heart to ischemiareperfusion injury in experimental models is low during early development, increases in mid-term and decreases again in late development. Whether this pattern also exists in human hearts is not known. Methods: Twenty-eight children (>12 months), 25 infants (1-12 months) and 30 neonates (<1 month) undergoing open-heart surgery for correction of congential cardiac defects were recruited. All hearts underwent ischemic cardioplegic arrest using cold St Thomas’s I crystalloid cardioplegia. Myocardial injury was assessed by measuring post-operative release of cardiac troponin I (TnI) over 24 hours. Clinical outcomes were prospectively collected. Results: There was significantly (p<0.05) more myocardial injury in infants (total TnI 80.3±9.2 ng/ml) than neonates (47.5±4.3 ng/ml) than children (35.2±3.7 ng/ml) despite neonates having longer durations of ischemia than infants. When the 3 age groups were matched for ischemic times, infants had significantly more injury (total TnI 96.0±12.9 ng/ml) than either neonates or children (47.5±4.3 ng/ml and 42.7±9.1 ng/ml respectively). The durations of inotropic support and ICU stay were dependent on ischemic duration and, for matched ischemic times, were significantly less in children than neonates (25.0±2.4 vs 47.4±3.5 hours, p=0.01; 1.6±0.2 vs 3.9±0.4 days p=0.01 respectively) and both were less than infants (83.1±18.9 hours, p=0.01; 6.5±1.5 days, p=0.03 respectively). Conclusions: The tolerance of the immature human heart to ischemic cardioplegic arrest is high during the neonatal period, decreases during infancy and increases again in childhood. Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, Bristol, BS2 8HW. UK.

P 49-5 BASAL HYPERCONTRACTILITY AND DISTURBED CA2+ HOMEOSTASIS IN MONOCROTALINE-INDUCED RIGHT VENTRICULAR HYPERTROPHY OF THE RAT H. Kögler, O. Hartmann, R. Roncon-Albuquerque, P. Schott, P. Nguyen Van, K. Leineweber, A. Leite-Moreira, O.-E. Brodde, G. Hasenfuss Treatment of rats with the Crotalaria alkaloid monocrotaline causes obliterative vasculitis of the lung arterioles, resulting in increased vascular resistance within the lung circulation and right ventricular (RV) hypertrophy. In some of the animals heart failure develops. The consequences of this hypertrophy for RV myocardial contractility have not been studied in detail. We examined at 37oC isometric twitches of intact RV trabeculae from rats in the compensated phase of hypertrophy, 20 to 24 days after a single s.c. injection of 50mg/kg monocrotaline. Control animals were treated with saline. Baseline contractility at [Ca 2+]o 1.25mM and 2Hz stimulation rate in MCT preparations was enhanced by 174% compared to controls (p=0.006; n=8 each). In control preparations, raising the stimulation rate resulted in a monotonous increase in developed force, while in MCT preparations the force-frequency relationship was negative, such that at a physiological rate of 5Hz no significant difference in developed force was observed. Increasing [Ca2+]o to 2.5mM enhanced developed force in controls by 128±21% while in MCT preparations the increase amounted to only 61.5±3.5% (p<0.001; n=9 each). Examination of the post-rest behavior in MCT preparations compared to controls revealed a dramatic reduction in the ability of the sarcoplasmic reticulum (SR) to accumulate Ca2+ during rest intervals exceeding a duration of 2s (p=0.01; n=8 controls, n=9 MCT). SR Ca2+-ATPase (SERCA2a) mRNA expression in MCT myocardium was reduced by 53% (p<0.001; n=8 each), which was also confirmed at the protein expression level using Western blot

P 49-6 INCREASE IN EXTRACELLULAR K+-CONCENTRATION PROTECTS CARDIOMYOCYTES AGAINST REPERFUSION INJURY Y. Abdallah, K. Heinz, C. Schäfer, B. Rother, C. Meuter, S. Walther, A. Gkatzoflia, H.M. Piper It could be shown in earlier studies that reoxygenation-induced Ca2+oscillations play an important role in the development of reoxygenation induced hypercontracture and therefore reperfusion injury. The aim of the present study was to investigate whether increase in extracellular K+-concentrations can reduce the frequency of spontaneous Ca2+oscillations and therefore protect cardiomyocytes against reoxygenation-induced hypercontracture. Isolated cardiomyocytes of adult rats were perfused anoxically (pH0 6.4, 60 min) with following reoxygenation (pH0 7.4, 15min) in presence of different extracellular K + -concentrations (+5mM: protocol 1; +9mM: protocol 2). The time course of intracellular Ca2+ (Fura-2) and cell length (% of endanoxic length) were estimated. During anoxia cardiomyocytes developed cytosolic Ca2+-overload (Fura-2 ratio (a.u): end-anoxic 4.4±0.1; pre-anoxic 1.1 ± 0.1; *; p<0.05). During reoxygenation spontaneous Ca2+-oscillations occurred. The increase in extracellular K+-concentration reduced dose-dependently the frequency of Ca2+-oscillations (min1 ) (2min Reox. Ktr:19.7±3; P1:11.5±1,9; P2:11.5±2,3; p<0.05 vs Ktr.) and protected against hypercontracture (cell-length in % of endischemic length: Ctr.: 64.3+2.4%; P1: 83+2.7%; P2: 82.9±2.8%; p<0.05 vs Ctr.). The recovery of cytosolic Ca2+ was accelerated (Fura-2 ratio (a.u.) after 2min of reoxygenation: Ctr: 3.0+0.1; P1: 2.6±0.1; P2: 2.7±0.2; p<0.05). Increased cytosolic K +-concentration reduces reoxygenation-induced Ca 2+-oscillations and protects against hypercontracture. These results can be explained by the acceleration of the cytosolic Ca 2+-recovery and could be due to an increase of activity of the Na+/K+-ATPase in the early phase of reperfusion. Institute of Physiology, Justus -Liebig -University Giessen Aulweg 129, D-35392 Giessen, Germany P 49-7 α IN CLASSIC ISCHEMIC NO IMPORTANCE OF TNFα PRECONDITIONING S. Belosjorow, I. Bolle, A. Duschin, G.Heusch, R. Schulz The cytokine TNFα plays an important role for the signal transduction in the late phase of ischemic preconditioning (IP), since pretreatment with TNFα-antibodies (TNFα-Ab) abolished the reduction of infarct size by IP (Br. J. Pharmacol. 131, 415-422, 2000). Whether TNFα is also important for the signal transduction in classic preconditioning, is unknown. Anesthetized rabbits were either untreated (group 1, n=7) or preconditioned by one cycle of 5 min ischemia (I) and 10 min reperfusion (R) (group 2, n=6) and then subjected to 30 min coronary artery occlusion and 180 min R. Pretreatment with TNFαAb was performed without (group 3, n=6) or with one cycle of IP (group 4, n=6) starting 1 h prior to the 30 min coronary artery occlusion. Heart rate, mean and maximal aortic blood pressures were similar throughout the protocol among all groups. With a comparable area at risk, infarct size in group 1 was 44±11 %. Infarct size was reduced to 13±7 % (p< 0.05), 23±8 % (p< 0.05) und 19±12 % (p< 0.05) in groups 2, 3 und 4, respectively. Serum TNFα-concentration was comparable at baseline in all groups. Serum-TNFα-concentration was increased during I/R in group 1 from 752±403 to 1542±482 U/ml (p< 0.05), but it remained unchanged in groups 2, 3 and 4. Conclusion: Pretreatment with TNFα-antibodies reduces infarct size per se, but does not interfere with the infarct size reduction by ischemic preconditioning during myocardial ischemia/ reperfusion in rabbits.

S 353 Abteilung für Pathophysiologie, Universitätsklinikum Essen, Hufelandstrasse 55, 45122 Essen

study of arrays of individual cardiac cells. This system can be interfaced with optical detection systems to allow the use of ion/voltage sensitive dyes.

P 49-8

Dept. Electronics, Oakfield Avenue, University of Glasgow, Glasgow G12 8LT; *Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ

ABSENCE OF GELSOLIN INFLUENCES BLOOD PRESSURE IN VIVO AND CONTRACTION OF CARDIOMYOCYTES IN VITRO R. Meyer1, J. Weisser-Thomas3, H. Kempelmann1, P.C. Djoufack2, C. Grohé3, K.B. Fink2 A comparison of the L-type calcium current in cardiac myocytes of neonatal gelsolin deficient mice (gsn-/-) with that of cells from gsn+/+ mice exhibited a highly increased current in myocytes from gsn-/animals (Lader et al. 1999, Am J Physiol 277: C1277-1283). Gelsolin depolymerises actin, gsn-/- should thus have a higher polymerised cytoskeleton. Aim of this study was to test whether the elevated L-type current is preserved in adult animals and whether it has consequences in the living animal. Ventricular cardiac myocytes from adult gsn-/- and gsn+/+ mice were isolated. In these cells the L-type calcium current was recorded in whole cell mode after inactivation and blocking of overlapping currents. β-Stimulation of the L-type current was performed by orciprenaline. Sarcomere shortening was monitored in myocytes stimulated externally. In living animals blood pressure and heart rate were recorded intraarterially by a catheter during enflurane anaesthesia. Control measurements were performed by telemetric recordings. The L-type current density in gsn-/- cells was 12.8 pA/pF during clamp steps to 10 mV. β-Stimulation increased the current density significantly by 36%. In gsn+/+ cardiac myocytes L-type current density was 10.5 pA/pF and β-stimulation increased it significantly by 13.4%. Although, current density in gsn-/- compared to that of gsn+/+ animals appeared higher it did not reach the level of significance. However, cell shortening was significantly higher in gsn-/- cells at frequencies =4 Hz. At 8 Hz stimulation frequency extent of shortening was higher in gsn+/ + cells. Mean arterial blood pressure was 109 mmHg in gsn+/+ and 122 mmHg in gsn-/- mice (p<0.05). Heart rate did not differ between both animal groups. The remarkably increased L-type current in neonatal gsn-/- myocytes as compared to gsn+/+ reported in literature could not be shown in adult cells. An increased shortening in gsn-/- cells was found at low frequencies, which may contribute to the differences in blood pressure.

P 50-1

1 Physiologisches Institut II, 2Institut für Pharmakologie und Toxikologie, 3Medizinische Klinik und Poliklinik II, Universitätsklinikum Bonn

NEURONAL NITRIC OXIDE SYNTHASE INHIBITION RESETS THE TUBULOGLOMERULAR FEEDBACK IN VOLUMEEXPANDED RATS A. Ollerstam, R. D. Brown, A. E. G. Persson In the kidney the neuronal isoform of nitric oxide synthase is located in the macula densa cells. These cells are known to be the sensor in the tubuloglomerular feedback (TGF), which influences the tonus of the afferent arteriole. The TGF response is attenuated during volumeexpansion (VE) allowing for increased water and salt excretion. This study was designed to investigate whether the inhibition of neuronal nitric oxide synthase re-establishes the TGF response caused by acute extracellular volume expansion. Experiments were performed on inactin-anesthetized Sprague-Dawley rats. VE was achieved by an infusion of isotonic saline solution at 5ml/hr/100g bw. When urine flow was stabilised the TGF response was evaluated by measuring changes in proximal tubular stop-flow pressure (Psf) in response to various loop of Henle perfusion rates. The loop of Henle was perfused with artificial ultrafiltrate under control conditions and under addition of 1mM non-specific nitric oxide synthase inhibitor, Nω-nitro-L-arginine methyle ester. In additional rats the neuronal nitric oxide synthase inhibitor, 7-nitro indazole, was given i.p. (25 mg/kg bw). Maximal decrease in Psf was used to indicate TGF reactivity. To further evaluate the TGF responsiveness single nephron glomerular filtration rate (SNGFR) was measured while perfusing the loop of Henle at zero and 40 nl/min. After addition of Nω-nitro-L-arginine methyle ester to the artificial ultrafiltrate the maximal Psf response and ∆SNGFR increased, indicating an increased TGF. 7-Nitro indazole also elicited an increase in TGF responsiveness seen both in the ∆Psf and ∆SNGFR. These results suggest that a functioning nitric oxide system, especially through the neuronal nitric oxide synthase, may be important in mediating normal renal responses and that an increased production of and/or sensitivity to nitric oxide during sustained VE plays an important roll in the adaptive mechanism of the TGF.

P 49-9

Dept of Medical Cell Biology, Division of Integrative Physiology, Uppsala University, Box 571, S-751 23, Uppsala, Sweden

AN INVESTIGATION OF THE CELLULAR CHARACTERISTICS OF STIMULATED ISOLATED CARDIOMYOCYTES WITHIN MICROSYSTEM TECHNOLOGIES N. Klauke, X. Cia, J. Cooper, GL Smith* Increased demands for systems to allow higher-throughput-screening of compounds based on single cell assays has led to the development of arrays of microtitre chambers to allow arrays of isolated cells to be examined individually. Microchambers of 50 (w) x 200 (l) x 400 (h) µm dimension were microfabricated in PDMS and aligned against an array of stimulating gold electrodes (10 µm wide and 10 µm high). Hearts were removed from terminally anaesthetized rabbits (1mg/kg euthatol), and myocytes were isolated from the left ventricle by perfusion with collagenase solution. Individual cells were placed parallel to the electrical field in the chambers. Chambers were designed such that they either enabled washout of byproducts of cellular metabolism/electrolysis or accumulation and further amperometric analysis of metabolites. Stimulating conditions were investigated by monitoring cellular contraction and simultaneously measuring intracellular calcium, extracellular pH and the current on the electrodes. Threshold for contraction in the limited volume (∼400 pl) was generally 40 V/cm passing ∼50 nA through the titre chamber. Use of a bipolar stimulus (8 ms duration) prevented the accumulation of electrolysis products. Under these circumstances isolated cardiac cells could be continuously paced for up to 60 mins, following frequency changes between 0.2 Hz and 1 Hz. Mitochondrial uncoupling (1 µM FCCP) prevented twitching after ∼30 mins. Lactate content of quiescent mycocytes rose from 1.6 mM (control cells) to 5.0 mM (mitochondrial uncoupled cells) as detected amperometrically after saponin permeabilization of individual cells placed in the microchamber. These studies indicate that pl chambers with associated stimulus electrodes can be synthesized to allow the

P 50-2 ANGIOTENSIN II STIMULATE CALCIUM AND NITRIC OXIDE RELEASE FROM MACULA DENSA CELLS THROUGH AT 1 RECEPTORS R. Liu, A. E. G. Persson In present study, a fluorescent nitric oxide (NO) indicator, 4,5diaminofluorescein diacetate and calcium indicator indo-1 were used to detect NO and calcium concentration in macula densa (MD) cells challenged by angiotensin II (Ang II), using 488 nm and UV confocal laser scanning microscopy. Glomeruli with attached cortical thick ascending lambs and containing the MD plaque were isolated, and then transferred to a chamber fixed to the stage attached to a laser confocal system. Ang II (10-6 M) caused the delta cytosolic calcium concentration ([Ca2+]i) increase by 125.8 ± 16.3 nM (n = 17) from bath and 52.3 ± 11.5 nM (n = 18) from lumen. AT 1 antagonist CV-11974 (10 -6 M) blocked the calcium responses both from bath and luminal side caused by Ang II, but AT2 antagonist PD-123319 (10-6 M) could not in both sides. AT 2 agonist CGP-42112A (10-6 M) had no effect on [Ca 2+]i in MD cells both in bath and lumen. In calcium free experiments, there were no significant different compared with that in normal Ca 2+ solutions. Ang II (10-6 M) increased the delta NO production by 16.12% ± 3.4% (n = 26) from bath and 17.86 ± 3.1 nM (n = 24) from luminal. AT 1 antagonist CV-11974 (10 -6M) blocked the NO responses both from bath and luminal side, but AT2 antagonist PD-123319 (10-6M) could not in both sides. AT 2 agonist CGP-42112A (10 -6M) had no effect on NO in MD cells. In calcium free experiments, there were no significant difference compared with that in normal Ca2+ solutions. In conclusion, we found that Ang II could elevate [Ca2+]i and stimulate

S 354 NO productions in MD cells both from basolaterial and luminal sides through AT 1 receptors. Department of Medical Cell Biology, Uppsala University, BMC Box 571, 75123 Uppsala, Sweden. P 50-3 NITRIC OXIDE IS A MODULATOR AND NOT A MEDIATOR OF MACULA DENSA CONTROLLED RENIN SECRETION F Schweda, A. Kurtz Nitric oxide has a critical role in the regulation of the renin system. However there is still uncertainty if NO acts as a mediator or as a modulator of renin secretion and to what extend the eNOS or nNOS isoform, both constitutively expressed in the kidney, are involved in the regulation of renin secretion. To address these questions we adapted the technique of the isolated perfused kidney model to the anatomical conditions of mice and investigated the macula densa control of renin secretion (RS) in C57BL/6 mice as well as in eNOS and nNOS knockout mice and their wildtype controls. After the confirmation of a physiological involvement of the cAMP- and NO-pathway as well as of the renal baroreceptor in the regulation of the renin secretion in our new experimental approach we tested the macula densa control of renin secretion. We found that a blockade of macula densa salt transport by bumetanide 100µM significantly stimulated renin secretion. This stimulation was completely prevented or reversed by a blockade of the NOsystem (L-NAME or ethyl-ITU hydrobromide). Although basal renin secretion rates were significantly lower in eNOS-/- and nNOS-/- compared to their controls, bumetanide stimulated renin secretion in the kidneys of both knockout animals to the 2-fold of control. As these findings argue against a regulatory role of one of these isoforms in this stimulation and for a more permissive effect of NO, we inhibited the endogenous NO-synthases and added exogenous NO (SNAP) in a fixed concentration to the perfusate. In this setting inhibition of the macula densa salt transport stimulated RS significantly while it did not increase RS when SNAP was not added. Our data clearly demonstrate a critical involvement of the NO system in the control of the renin secretion. However NO does not act as the mediator of the macula densa control of renin secretion.

P 50-5 CARBON MONOXIDE (CO) STIMULATED ERYTHROPOIETIN (EPO) SECRETION IN CONSCIOUS RATS: INFLUENCE OF RENAL NERVES F. Petroktistis, A. Grenz, D. Kloor, K. Klingel1, C.H. Gleiter, H. Osswald Eckardt et al have reported (1992) that renal denervation (DNX) does not affect EPO-secretion (EPOS) stimulated by hypoxia, haemorrhage and CO in conscious rats. Since in this report no attempts were made to determine a dose-response relationship between the degree of hypoxia and EPOS and since DNX was induced only in one kidney, we reexamined the role of renal nerves in CO-induced EPOS by constructing a dose-response relationship between EPO serum levels and CO concentrations in inspired air in conscious rats with innervated (controls) and bilateral denervated kidneys. Methods: Rats were placed into a cage with an air inflow between 6-10 l/min. CO was mixed with room air to generate concentrations of CO between 100 -1400 ppm. The actual CO concentration was monitored by a CO sensor (Testo, Reutlingen). After 4 hours of CO exposure the rats were sacrificed, blood samples were collected and stored at - 80°C until analysis. Bilateral denervation was carried out by cutting all visible nerves and painting kidney hilus with 10 % alcoholic phenol solution. Completeness of DNX was assessed by measuring renal norepinephrine content with HPLC (Mühlbauer and Osswald, 1994). For constructing a dose-response relationship we used for each CO concentration 4-6 rats (total 64). Serum EPO levels was measured with ELISA (Medac). Data are expressed as mean ± SEM. Results: In control rats EPO serum rose from 9.5±1.5 up to 2280±210 mU/ml at CO concentration of 1200 ppm within 4 hours. However, in DNX rats the maximum response of EPOS to CO exposure (1200 ppm) was 1180±130 mU/ml. Analysis of the dose-response curves revealed a left-shift with EC50 values of 870±65 ppm in controls and 602±32 ppm in DNX rats indicating a higher sensitivity of DNX rats to CO intoxication compared to control rats. Conclusion: Our data support the concept, that renal nerves exert a significant influence on CO-induced EPO secretion in conscious rats. Supported by BMBF, 0lEC000l. Departments of Pharmacology and Pathology1, Faculty of Medicine, University of Tubingen, Wilhelmstr.56, 72074 Tubingen

Institut für Physiologie, Universität Regensburg, 93040 Regensburg P 50-6

AGE-DEPENDENCY OF RENAL FUNCTION IN CD-1 MICE G. Luippold, B. Pech, S. Schneider, H. Osswald, B. Mühlbauer It is well known that kidney function declines with age in rats and humans [Epstein, JASN, 1996]. A sensitive indicator for the functional integrity of the kidney is the ability to increase glomerular filtration rate (GFR) in response to intravenous infusion of amino acids (AA). Since mice are increasingly used to explore pathophysiological and therapeutical concepts we investigated whether the age-dependent decline in renal response to AA infusion holds true also for these animals. Methods: In young (YNG: 5 weeks, n=8) and adult (ADT: 20 weeks n=8) pentobarbital anesthetized CD-1(ICR)BR mice GFR was determined by renal clearance of 3[H]-inulin during infusion of Ringer solution at baseline (BAS) and during infusion of a body weight-related mixed solution of AA (10 or 12.5%). Body weight of YNG and ADT mice was 30.5±0.4 and 37.4±0.6 g, respectively. Results: AA infusion significantly increased GFR from 0.39±0.03 at baseline to 0.66±0.12 ml⋅min-1⋅g-1 kidney weight (kw) in YNG but not in ADT (0.37±0.02 to 0.38±0.07 ml⋅min-1⋅g-1 kw). YNG and ADT mice at BAS, showed similar values for urinary flow rate (3.13±0.76 vs. 2.69±0.51 µl⋅min-1⋅g-1 kw, respectively) and renal sodium excretion (0.20±0.06 vs. 0.25±0.05 µmol⋅min-1⋅g-1 kw, respectively). In both groups, AA infusion increased urinary flow rate and sodium excretion to a similar extent (approximately 1.8- and 2-fold, respectively). Arterial blood pressure and heart rate slightly increased due to AA infusion. Conclusions: The data show that the first signs of age-related changes in kidney function concern alterations in renal hemodynamics while tubular function appears to be preserved.

ADENOSINE - A LINK BETWEEN TRANSMETHYLATION REACTIONS AND ENERGY METABOLISM D. Kloor, G. Luippold, U.Delabar, B. Mühlbauer, H. Osswald Adenosine (Ado) plays an important role in the regulation of renal function (control of glomerular filtration rate, electrolyte excretion, release of renin). Ado formation in the kidney is enhanced when ATP hydrolysis rate prevails over ATP synthesis rate. Since Ado inhibits the hydrolytic activity of S-adenosylhomocysteine (SAH) hydrolase and since synthesis of SAH is thermodynamically favoured, it can be predicted that SAH levels in the kidney may increase in parallel of those of Ado. Therefore, we analyzed in the present study in vivo the effect of Ado and homocysteine (Hcy) on SAH tissue levels in the ischemic rat kidney. Methods: The tissue content of SAH, Sadenosylmethionine (SAM) and Ado as well as the tissue contents of the adenine nucleotides in the rat kidney was measured under normoxic and hypoxic conditions after freeze clamp, tissue extraction and subsequent HPLC analysis. Results: Under normoxic conditions tissue contents (mean ± SEM, n=4-8) for SAH, SAM and Ado were 0.68±0.05, 44.1±1.0 and 3.8±0.1 nmoles/g wet weight, respectively. Within 5 min of renal artery occlusion tissue levels of SAH and Ado increased 510 fold, whereas SAM tissue levels remained unchanged. Increases of Ado tissue content up to 57.1±4.5 nmoles/g wet weight induced by intrarenal administration of Ado did not result in further SAH accumulation. However, when Hcy was administered to the ischemic kidney, SAH levels rose up to 35.3±3.2 nmoles/g wet weight, indicating that Hcy is rate limiting for SAH synthesis in the rat kidney. Conclusion: The increase of SAH tissue content in the hypoxic kidney as a result of an inhibition of SAH hydrolysis by Ado leads to a marked reduction of methyltransferases activity. Whether this 50-fold increase of SAH during ischemia may contribute to altered gene expression, e.g. by inhibition of mRNA capping, remains to be determined.

Department of Pharmacology, Faculty of Medicine, Eberhard-KarlsUniversity, D-72074 Tübingen, Germany.

Department of Pharmacology, Faculty of Medicine, Eberhard-Karls University of Tübingen, Germany

P 50-4

S 355 P 50-7

P 50-9

ALDOSTERONE INHIBITS DEPOLARISATION-INDUCED VASOCONSTRICTION IN RENAL AFFERENT ARTERIOLES J. Schjerning, P.B. Hansen, R. Nørregaard, T.R. Uhrenholt, B.L. Jensen, O. Skøtt Aldosterone has been suggested to elicit vascular contraction by increasing intracellular calcium concentration via a non-genomic mechanism. We tested this proposal in small resistance vessels and large arteries. In microdissected, perfused rabbit afferent arterioles aldosterone did not elicit contraction in concentrations from 10-10-10-5 M, but after exposure to aldosterone, the ability of 100 mM KCl to induce contraction was abolished. In a concentration-response study we observed this inhibitory effect from an aldosterone concentration of 10-12 M. The effect of aldosterone on the KCl-induced contraction was maximal after 20 minutes. This effect was fully reversible and it was blocked by 10-7 M spironolactone. Norepinephrine-induced contraction was unchanged. Aldosterone (10-9 M) did not affect the increase in intracellular calcium concentration induced by 100 mM KCl as measured with flourescence-microscopy with Fura-2. An inhibitory effect of aldosterone on depolarisation-mediated contraction was also observed in rat aorta rings, but only at very high aldosterone concentrations (10-4 M at 20 min, and 10-5 M at 1h). Mineralocorticoid receptor (MCR) and 11β-hydroxysteroid dehydrogenase 2 (11β-HSD2) mRNA expression were demonstrated in microdissected rat preglomerular renal vasculature and cultured rat preglomerular vascular smooth muscle cells by RT-PCR. Immunolabelings with MCR and 11β-HSD2 antibodies were positive in acid-macerated, microdissected rat renal preglomerular vessels. Expression of MCR and 11β-HSD2 mRNA were demonstrated in rat aorta by ribonuclease protection assay. We conclude that aldosterone in physiological concentrations inhibits depolarisation-induced vasoconstriction in small renal resistance vessels, but not in large arteries, through the classical mineralocorticoid receptor, and independently of changes in intracellular calcium concentration.

A MODEL OF ISOLATED AUTOLOGOUSLY HEMOPERFUSED PORCINE SLAUGHTERHOUSE KIDNEYS D.A. Groneberg, C. Grosse-Siestrup, C. Fehrenberg, A. Fischer, C. Peiser Background: The rapidly evolving field of transplantation research with a focus on ischemic and reperfusion injuries has gained importance since the methodology of organ preservation significant limits graft survival. Numerous models of isolated perfused kidneys have been established in the past years but limitations such as organ size, perfusate and ethical standards have restricted a widespread research in this area. Methods: A model of hemoperfused isolated porcine slaughterhouse kidneys was established which encompasses the advantages of autologous blood as optimal perfusate and a reduction of animal experiments. Results: The size and geometry of the porcine kidney is more comparable to human conditions and various renal functions, blood parameters and morphology can easily be accessed in the present model. Stable organ function can be maintained over 2 hours with an amount of 500 to 1000 ml of autologous blood which is metabolically controlled via a dialysis system. Conclusion: In summary, the present model describes a new and economic approach for targeting renal function in transplantation models by combining autologous blood as optimal perfusate with a well defined organ geometry and function and slaughterhouse animals as a source.

Jeppe Schjerning, Physiology and Pharmacology, University of Southern Denmark, Winsløwparken 21, DK-5000 Denmark

P 50-8 IS THERE A SIGNIFICANT TUBULAR CREATININE SECRETION IN THE DOG? J. Höchel, D. Lehmann, C. Große-Siestrup, H. Hartmann In healthy humans up to 20% of the endogenous creatinine are not excreted via glomerular filtration, but tubular secretion. It was suggested that organic cation transporters (OCT) are responsible for this route of excretion. In contrast to humans, there is still a controversy, whether tubular creatinine secretion plays a significant role in the dog, the model animal for many renal studies. If tubular secretion takes place via OCT, this process should be saturable. Hence, an increasing plasma concentration of creatinine should lead to changes in the creatinine/inulin clearance ratio (CLcrea/CLin). This question was addressed with experiments in isolated haemoperfused canine kidneys (n = 20). In 8 kidneys plasma creatinine concentration was increased from 0.18 to 1.3, and finally 12.7 mmol/l in 3 successive sampling periods. 12 kidneys served as controls with a constant plasma creatinine concentration of 0.7 mmol/l. Sinistrin (= inulin) and creatinine clearance were measured in parallel. In addition, electrolyte excretion and oxygen consumption were determined. CLcrea/CLin values did not change significantly in the experimental or control groups during the successive sampling periods. Its mean ± SD was 0.88 ± 0.11. There was not a significant correlation of CLcrea/CLin with any of the determined parameters as renal blood flow, glomerular filtration rate, reabsorbed sodium, diuresis, and oxygen consumption. The functions investigated did not differ between the kidneys of male or female animals. Our results suggest a tubular net reabsorption of creatinine, which has been described only occasionally in the literature. Although alterations in the isolated perfused kidney model compared to the situation in vivo cannot be excluded, we conclude that a tubular net secretion as in humans, does not play a significant role in the dog. The unchanged CL crea /CL in values, despite marked changes in the plasma creatinine concentration, do not support an OCT mediated tubular transepithelial transport pattern in the dog. Freie Universität Berlin, Institut für Veterinär-Physiologie, Oertzenweg 19b, D-14163 Berlin, Germany

Dept. of Pediatric Pneumology and Immunology, Charite Campus Virchow, BMFZ Forum 4, Augustenburger Plate 1, D-13353 Berlin, Germany P 50-10 SLOW VOLUME EXPANSION IN CONSCIOUS DOGS: THE EFFECTS OF LOW AND HIGH SALT INTAKE N.C.F. Sandgaard, S. Wamberg, P. Bie The aim of this study was to investigate the effects of sodium diets on the regulatory responses to slow volume expansion. After 8 days of low salt (LS, 0.4±0.1 mmol Na+/kg BW/day) or high salt diet (HS, 7.2±0.2 mmol Na+/kg BW/day) 4 female beagles were studied by slow infusions of isotonic saline (0.13 ml/kg/min for 120 min ∼1.56% of BW). Plasma volume was measured by Evans Blue, and glomerular filtration rate (GFR) by the clearance of exogenous creatinine. During HS, plasma volume was 5% higher than during LS, while BW or mean arterial blood pressure (MABP) did not change. Sodium excretion was 0.4 µmol/min in LS and about 28 µmol/min in HS. During LS, plasma angiotensin II (ANGII) and aldosterone (ALDO) were markedly higher than during HS (22±4 vs. 7±1 pg/ml and 814±325 vs. 31±13 pg/ml, respectively). In contrast plasma atrial natriuretic peptide (ANP) was higher in HS than in LS (101±15 vs. 64±8 pg/ml). Slow volume expansion during LS decreased plasma ANGII (to 12±3 pg/ml) and ALDO concentration (to 265±50 pg/ml), while MABP, GFR, or ANP levels did not change; sodium excretion increased only slightly from 0.4±0.1 to 4.0±1.6 µmol/min. During HS, slow volume expansion tended to suppress ANGII (4.7±0.7 pg/ml) and ALDO (17±6 pg/ml). Sodium excretion, however, was augmented to 78±10 µmol/min, without significant changes in MABP, GFR, or ANP. In conclusion, in absolute terms LS diet almost abolished the natriuretic effect of slow volume expansion, despite large suppressions of ANGII and ALDO. In HS dogs volume expansion produced natriuresis without significant acute suppression of ANGII and ALDO. The data indicate that the acute natriuretic response to slow volume expansion is very dependent on sodium status but independent of MABP, GFR and ANP, and may point to a threshold value for renin system activity below which other factors become prominent modulators of sodium balance. Department of Physiology and Pharmacology, University of Southern Denmark, 21 Winsloewparken, DK-5000 Odense C, Denmark P 51-1 POSSIBLE CONTRIBUTION OF IMPAIRED SODIUM EXCRETION TO THE HYPERTENSION IN ADULT PRAGUE HYPERTENSIVE RAT I. Vanìèková The objective of the present study was to examine the role of the kidney in the development and maintanance of hypertension in the

S 356 Prague hypertensive rat. The function of isolated perfused kidneys from Prague hypertensive rat (PHR) and Prague normotensive rat (PNR) was evaluated in young and adult animals either at low (110 mm Hg) or hypertensive (150 mm Hg) perfusion pressures. No significant difference in heamodynamic parameters were found in young (6-weekold) animals either at low or high perfusion pressures. Also water and sodium excretion were the same in both substrains at both perfusion pressures. The only different value was higher fractional sodium reabsorption found at 110 mm Hg in PHR, reflecting thus a change in excretory function of these kidneys. In adult (12-week-old) PHR, both perfusate flow and glomerular filtration rate were significantly reduced at 110 mm Hg. At hypertensive perfusion pressure, these values were similar in both PHR and PNR. Perfusate flow and glomerular filtration rate were well autoregulated in adult PNR over the perfusion scale. Sodium excretion was lower, and its tubular reabsorption higher, in PHR than in PNR at both levels of perfusion pressures. Higher perfusion pressure was needed in PHR to achieve the same rate of sodium excretion resembling that of PNR perfused at lower pressure. These results indicate that kidneys of PHR retain more sodium, which may contribute to the development of hypertension in this strain. Institute for Clinical and Experimental Medicine, Prague, Czech Republic

domain of R-R wave intervals in electrocardiograms, we calculated a cardiac index of parasympathetic activity (CIPA) in chronically instrumented dogs at different stages of pacing-induced HF at rest and during treadmill exercise. Continuous pacing at 200 - 240 bpm for 5 weeks caused progressive sympathovagal imbalance. While CIPA fell most rapidly already during the first week of pacing and further declined to less than 10% of control in the second week, relevant increases overall sympathetic tone as assessed by plasma noradrenaline concentrations occurred only during weeks 3-5. These changes in parasympathetic and sympathetic activity were closely reflected by concomitant decreases in cardiac contractility and increases in left ventricular end-diastolic pressure, respectively. Similarly different time courses for parasympathetic withdrawal and sympathetic activation were observed during treadmill exercise. Our data demonstrate that the new method accurately describes the early depression of parasympathetic cardiovascular control in pacing induced HF, which occurs well before the onset of general sympathetic activation together with decline in cardiac contractile function. We conclude that real-time measurements of cardiac vagal tone at early stages of cardiac diseases could become a useful diagnostic and prognostic tool for patients at risk of developing heart failure. Pfizer Global Research & Development, Ramsgate Road, Sandwich, Kent, CT13 9NJ, United Kingdom

P 51-2 P 51-4 ACTIVATION OF THE ENDOTHELIN SYSTEM IN UREMIC CARDIOMYOPATHY – INFLUENCE OF ENDOTHELIN RECEPTOR ANTAGONISTS S.C. Wolf, B.R. Brehm, F. Gaschler, F. Lang*, T. Risler Objective of the study: Patients with renal failure die due to cardiac complications. Uremic cardiomyopathy is characterized by reduced contractility and myocardial hypertrophy. Aim was to analyze the endothelin (ET) system in uremic cardiomyopathy. The effects of ET receptor antagonists were investigated. Methods: Male rats were subtotally nephrectomized and treated with an endothelin A-receptor antagonist (30mg/kg/d, LU302146) or an endothelin AB -receptor antagonist (30mg/kg/d, LU302872) for 12 weeks. One group was left untreated (SNX) and one was sham operated (sham). Determined were ET-1 serum-, ET-1 urin- and ET-1 tissue-concentrations. mRNA levels of prepro-ET-1, ETA and ETB receptors were analyzed. Left ventricular contractility (LC) and relaxation (LR) was measured. Results: ET-1 levels increased in serum, urine, myocardium and in kidney. PreproET-1 mRNA rose in the kindney of SNX (Co 30±2,7, SNX 104±11,5 amol/µg). ETA reduced prepro ET-1 to 75±5,6 and ETAB to 57±4,7 amol/µg. Myocardial prepro-ET-1 increased to 7,0±0,4 amol/µg in SNX (Co 2,0±0,2). Both ET antagonists reduced prepro-ET-1 in myocardium. Kidney ETA mRNA was downregulated in SNX to 38±0,2 amol/µg (Co 72±0,4) and not influenced by ET antagonists. In the heart ETA mRNA was unchanged in SNX but decreased by both antagonists. Kidney ETB mRNA increased from 204±19 to 296±23 amol/µg and was reduced by both antagonists. Myocardial ETB mRNA increased from 410±19 to 453±18 amol/µg and reduced by both ET antagonists. Contractility was decreased in SNX by 40%. ETAB partly and ETA antagonist completely prevented the reduction in left ventricular function. Conclusion: The ET-system is activated during uremia in myocardium and kidney. ET reduced cardiac contractility in SNX. Inhibition of the ET-system improves cardiac function. ET receptor antagonists downregulate ETA and ETB receptors in kidney and heart. Medizinische Klinik III, Otfried-Müllerstr. 10, *Institute of Physiology, Gemlinstrasse 5, University of Tübingen, Germany

pH-DEPENDENT APOPTOSIS CAUSED BY NEPHROTOXINS IN TWO RENAL CELL LINES (IHKE AND MDCK-C7) G. Schwerdt, R. Freudinger, C. Schuster, S. Silbernagl, M. Gekle Nephrotoxins exert their action in part via induction of apoptosis. We have shown recently that nanomolar concentrations of the fungal metabolite ochratoxin A (OTA) causes caspase-3 activation with subsequent apoptosis in human proximal tubule-derived cells (IHKE cells) and collecting duct cells (MDCK-C7). Cisplatin, a known nephrotoxin, also induces apoptosis in both cell lines. We determined apoptosis induction by measuring caspase-3 activity and DNA ladder formation. First, to investigate the mitochondrial role in OTA or cisplatin induced apoptosis, we inhibited mitochondrial function at the level of the respiration chain (using rotenone), ATP synthesis (by oligomycin), or ATP export (by bongkrekic acid) and by uncoupling (using CCCP). Both cell clones responded differently to these manoeuvres: in IHKE cells, inhibition of mitochondrial function led to an inhibition of OTA or cisplatin induced apoptosis. In MDCK-C7, inhibition of mitochondrial function led to an increase in apoptosis. Concommittant with mitochondrial inhibition, an increase in lactic acid production and export into the media was observed. This exported lactic acid caused an extracellular acidification. Therefore, we investigated the role of extracellular pH on OTA or cisplatin induced apoptosis in the two renal cell lines. Again, both cell clones responded differently on extracellular pH changes: in IHKE cells an acidification of the media (by adding HC1) led to inhibition of apoptosis induction and after alkalinisation of the media (by adding NaOH) an increase in OTA or cisplatin induced apoptosis was noticeable. Again, MDCK-C7 cells responded contrary: acidification led to increased apoptosis and alkalinisation to decreased apoptosis induced by OTA or cisplatin. Thus, we conclude that the extracellular pH plays an important but cell type dependent role in the apoptotic process of renal cells. (This sudy was supported by DFG Grant Ge 905/3-4) Physiolog. Institut, Universitat Würzburg, Röntgenring 9, 97070 Würzburg P 51-5

P 51-3 REAL-TIME MEASUREMENT OF CARDIAC VAGAL TONE IN DOGS WITH PACING INDUCED HEART FAILURE M. Dewhurst, C. Connaughton, H. West, M. Wiltshire, A. Popovic and J. Zanzinger Heart failure (HF) is associated with and driven by chronic increases in sympathetic and decreases in parasympathetic activity, which aggravate and reinforce the disease process. However, currently used methods (spectral analysis) to detect to autonomic dysfunction in early stages of heart failure suffer from limited specificity and high variability of results. Using phase demodulation of a high-resolution time

KV CHANNELS AND BKCA CHANNELS IN SINGLE RAT JUXTAGLOMERULAR CELLS U.G. Friis, F. Jørgensen, B.L. Jensen, O. Skott The contribution of different ion channels to the membrane conductance of juxtaglomerular cells freshly isolated from rat kidneys were studied by the whole cell patch-clamp technique. Using standard intra- and extracellular solutions, the voltage dependency of the membrane conductance was examined by a voltage step protocol where the voltage was changed from –110 to 130 mV in 30 mV steps for 60 ms from a holding potential of –30 mV. The steady state currentvoltage relationship was typical for both isolated and in-situ JG-cells1,2.

S 357 The cells displayed a marked outward current rectification at positive membrane potentials, and very limited net currents in the votage range –50 mV to 10 mV. At 130 mV, the outward current amounted to 2.35 ± 0.33 nA (n=35). This outward current was partially inhibited by 1 mM 4-aminopyridine (4-AP) and also by 5 mM TEA, both applied from the outside via a superfusion system. 4-AP inhibited 20.6 ± 9.5% of the outward current at 130 mV (P<0.05, n=12), whereas TEA inhibited 83.3 ± 7.5% of the outward current at 130 mV (P<0.05, n=4). The blocking effects of 4-AP and TEA were additive and the outward current was inhibited by 91.5 ± 2.9% at 130 mV (P<0.05, n=4). With 2 mM EGTA in the internal solution, the outward current at 130 mV was reduced to 0.11 ± 0.02 nA (P<0.05, n=7). To identify the current components further, the effect of charybdotoxin applied via an application pipette was tested. In the presence of charybdotoxin (30 nM), the inhibition of the outward current at 130 mV amounted to 50.6 ± 6% (P<0.05, n=3) and at 100 nM charybdotoxin the outward current was inhibited by 88.7 ± 1.2% (P<0.05, n=4). The blocking effects of 4-AP suggest that under resting conditions part of the outward current is through voltage-gated KV channels, whereas the Ca2+ dependence of the outward current together with the blocking by charybdotoxin suggest that the major current in single juxtaglomerular cells was carried through calcium-activated voltage-gated BKCa channels. 1. Kurtz A, Penner R. Proc. Natl. Acad. Sci. USA 1989; 86: 3423-7. 2. Friis UG, Jensen BL, Aas JK, Skott O. Circ. Res. 1999, 84: 929-36.

was investigated using clearance technique in non-diuretic ? Wistar rats. Plasma and urine concentrations of Na+, K+, Mg2+, Ca2+, Cl- and PO42- were measured by ionic chromatography. GFR was determined with 3H inulin. Second, microinjections of 45CaCl2 and 109CdCl2 (with tracer doses of 3H inulin) into the lumen of early and late superficial proximal tubules (PCT) and early and late distal tubules (DCT) were performed in mild-diuretic rats. Results: The acute infusion of Cd2+ did not significantly modify the urinary flow rate and the GFR. Fractional excretions of Na+, K+, Mg2+, Ca2+, Cl- and PO42- were increased while plasma concentrations and filtered amounts were not modified. Expressed in % of the amount injected, 45Ca reabsorption fluxes were 3.3 ± 0.4 (n=8) in the PCT, 66.1 ± 2.6 (n=15) in the loop of Henle and 22.2 ± 0.9 (n=15) in the DCT. Lanthanum (1mM) and cadmium (20µM) strongly decreased the reabsorption of 45Ca in the DCT by 43% and 57% respectively. Nifedipine (20µM) did not modify this reabsorption. 109Cd microinjections showed that 9.8 ± 1.7% (n=18) of injected Cd2+ were taken up by PCT, 46.8 ± 3.8% (n=20) by the loop of Henle and 38.3 ± 3.1% (n=20) by DCT. Gadolinium blocked distal Cd2+ reabsorption by 78%. Conclusion: Acute Cd 2+ intoxication decreases overall renal ion transport, Cd2+ and Ca2+ are mainly reabsorbed in the loop of Henle and the DCT. In the DCT , Cd2+ and Ca2+ probably share identical reabsorption pathways. The Cd2+ blockade of Ca2+ reabsorption in DCT could explain the hypercalciuria during Cd2+ intoxication.

Physiology & Pharmacology, University of Southern Denmark, Winsloewparken 21,3, DK-5000 Odense C, Denmark

UMR CNRS 6548 Université de Nice-Sophia Antipolis, 06108 Nice cedex2 France

P 51-6

P 51-8

ANGIOTENSIN II AND ADENOSIN A1 RECEPTOR AGONISTS INCREASE CYTOPLASMIC FREE CALCIUM IN RENIN SECRETING CELLS OF RAT KIDNEY J. Laske, U. Russ, U. Quast The release of renin from the juxtaglomerular cells (JGC), located in the media of the afferent arteriole close to the entrance into the glomeruli of the kidney, determines the activity of the renin-angiotensin-aldosteron system. In contrast to other exocytotic processes, the rate of renin release is inversely related to the cytoplasmic free calcium concentration ([Ca2+]i), the ”calcium-paradox”. Renin secretion is inhibited by angiotensin II (ANGII) and adenosine (ADO). ANG II increases [Ca2+]i in JGC; however, the effects of ADO on [Ca2+]i are unknown. To treat this question, glomeruli with the afferent arteriole attached were prepared from rats kept on a low NaCl diet. Arterioles were loaded with FURA 2 and [Ca2+]i was measured from 5-10 cells in the distal region of the arteriole (mostly JGC). Whereas 3 nM ANGII increased [Ca2+]i in all arterioles ADO (0.01-10 µM) was only weakly active. The A1 receptor agonist N6-cyclohexyladenosine (CHA, 0.01-10 µM) affected the number of responding arterioles and [Ca2+]i with a bell-shaped concentration dependence. Best effects were observed at 1 µM CHA where 11 of 22 vessels responded and the [Ca2+]i increase in the responding arterioles was 40 ± 11 % of the ANGII (3 nM) effect. Application of 1µM CHA in combination with 3 nM ANG II increased [Ca2+]i more than either stimulus alone. No evidence for an effect of endogenous ADO on [Ca2+]i was found in this preparation. The experiments show that ADO A 1 receptor stimulation increases [Ca2+]i in afferent arterioles in some but not all cases and that the effect is small. The reason for the variability of the response is not yet known. Despite this shortcoming, evidence for a synergistic action of ANGII and CHA on [Ca2+]i in afferent arterioles was obtained.

PKC-REGULATION OF HOAT1 LEADS TO INTERNALIZATION OF THE CARRIER AND IS INDEPENDENT OF SEVERAL CONSERVED OAT1 PKC CONSENSUS SITES N.A. Wolff, K. Thies, B. Friedrich, F. Lang, G. Reid, G. Burckhardt The renal basolateral organic anion transporter (OAT1), which exchanges substrates like PAH for intracellular dicarboxylate, is downregulated by agents known to stimulate PKC activity. In the present study, PAH uptake in Xenopus laevis oocytes mediated by the cloned h(uman)OAT1 was almost completely extinguished by a 1-hr preincubation with 5 µM of the PKC-activating agent 1,2-Dioctanoylsn-glycerol (DOG). This downregulation could be partially suppressed by the simultaneous presence of 10 µM staurosporine. Immunocytochemistry showed that DOG treatment led to an internalization of the C-terminally FLAG-tagged carrier. Of the 6 potential PKC-phosphorylation sites in hOAT1, S129, S271, T284 and S521 were selected for site-directed mutagenesis based on conservation with OATs1 cloned from species known to possess a PKC-sensitive PAH transport pathway in their renal proximal tubules. Of these, mutation of S129, T284, or S521 to A did not significantly attenuate PKC-mediated downregulation of hOAT1-dependent PAH uptake. However, mutation of the best conserved site S271 led to an almost complete loss of hOAT1 function. Immunocytochemistry indicated that this was due to a retention of mutant S271A in the oocyte cytoplasm; only very little staining was visible at the oocyte surface. Mutation of S271 to E did not preserve function, suggesting that the reason for the altered localization of mutant S271A is not only the requirement for a negative charge in position 271, i.e. a phosphorylated serine, for incorporation and/or stabilization of the carrier in the oocyte membrane. In contrast to mutation of S129, T284 or S521, deletion of the last exon (E10) in hOAT1 led to a significant decrease in DOG sensitivity, which might be taken as an indication that the very C-terminus of hOAT1 contains sequence motifs used in the PKC-activated internalization pathway. In summary, PKC activation in oocytes led to downregulation of hOAT1-mediated transport by carrier retrieval from the plasma membrane, which does not require phosphorylation in positions S129, T284, or S521.

Department of Pharmacology, Medical Faculty, University of Tübingen, Wilhelmstr. 56, D-72074 Tübingen, Germany P 51-7 CALCIUM AND CADMIUM REABSORPTION ALONG THE RAT NEPHRON: A MICROINJECTION STUDY O. Barbier, G. Jacquillet, M. Tauc, P. Poujeol Objective: Cadmium (Cd2+) exposure induces nephropathies and bone pathologies resulting from a defect of calcium (Ca2+) homeostasis. Notably, Cd2+ administration increases the urinary Ca2+ excretion leading to kidney stone formation. The aim of this study was therefore to investigate Ca2+ and Cd2+ transport along the nephron in order to localise the sites in which Cd2+could interact with Ca2+ reabsorption. Methods: First, the effect of an acute load of Cd2+ (10µM) on renal function

P 51-9 CALCIUM SIGNALING AT DIFFERENT SITES ALONG THE RENAL INTERLOBULAR ARTERIOLE AND IN CREMASTER MUSCLE ARTERIOLES A. Just, W.J. Arendshorst The response to sympathetic and other stimuli varies substantially among different vascular beds. The myogenic response has been shown to be faster in the kidney than in the skeletal muscle vasculature and

S 358 other data suggest that this is also true for sympathetic vasoconstriction. This prompted us to compare the responses of intracellular Ca 2+concentration ([Ca 2+]i) to increased external KCl (50 mM vs 5 mM) and to norepinephrine (NE 10 µM) in microdissected vessels from rat renal interlobular arterioles (ILA, n=6 rats) and from similarly sized cremaster muscle arterioles (CMA, n=6). In ILA [Ca2+]i increased in response to NE from baseline (51 ± 6 nM) to a transient maximum peak after 5-15 sec (104 ± 4 nM), after which it declined to an elevated plateau level at 20 - 60 sec (78 ± 4 nM). High KCl evoked a nearly square-shaped response (peak 81 ± 4 nM, plateau 74 ± 5 nM). The responses were similar among the proximal (n=2), middle (n=6) and distal (n=4) segment of the ILA. The peak response of [Ca2+]i to high KCl was considerably larger in CMA than in ILA (+66 ± 10 vs +32 ± 3 nM, p<0.01), whereas the plateau and the response to NE were similar in both vessels. In dose response curves for KCl (10 – 100 nM) and NE (10 nM – 30 µM) CMA showed a higher peak response to KCl, whereas sensitivity (ED50) and plateau response were similar to ILA. The peak response to KCl was unaffected by α-adrenergic blockade (n=2) indicating it was not mediated by perivascular nerves. Responses to NE were the same in both vessel types and completely abolished by α-adrenergic blockade. The time to peak was not significantly faster in ILA (9 ± 2 s) than CMA (13 ± 2 s), and was the same in both vessel types after high KCl (10 ± 2 vs 11 ± 1 s). The time course for complete contraction after NE was also only slightly faster in ILA than CMA (12 vs 15 s). In conclusion, the response to NE was surprisingly similar between both vessel types and along the renal vascular tree, whereas the response to KCl showed an exaggerated peak response in CMA. Dept. of Cell and Molecular Physiology, University of North Carolina at Chapel Hill, Med. Sci. Res. Bldg. CB#7545, Chapel Hill, NC 27599, USA P 51-10 REGULATION OF BASOLATERAL ORGANIC ANION UPTAKE IN PROXIMAL TUBULAR OK CELLS: EPIDERMAL GROWTH FACTOR AFFECTS OAT BY A CALCIUM INDEPENDENT BUT ARACHIDONIC ACID DEPENDENT MECHANISM C. Sauvant, H. Holzinger, M. Gekle In case of damage, proximal tubular cells produce epidermal growth factor (EGF) and EGF-receptor. EGF is supposed to play a key role in repair processes of the tubular epithelium. In former studies we could show that EGF stimulates basolateral uptake of the organic anion paraaminohippuric acid (PAH) acutely by activation of mitogen-activated protein kinases. EGF action was prevented by inhibition of phospholipase A2 (JBC 2001; lUPS 2001). To gain more detailed insight into the mentioned signalling cascade, OK-cells were cultivated on permeable supports. [14C]-PAH was used as organic anion for initial basolateral uptake studies. AA and its metabolites were detected using by HPLC separation. Cytosolic free Ca2+ was determined using the Ca2+ -sensitive dye fura-2 AM and fluorescence microscopy. EGF (10 ng/ml for 10 min) led to release of AA in OK cells. In addition, a release of PGs was detectable, while no other AA metabolites were detected. EGF also led to an increase of cytosolic Ca2+ OK cells. Ca2+ -reduction in the extracellular medium to µM concentrations reduced cytoplasmic Ca 2+ and led to a decrease of the PAH uptake rate in OK cells. However, EGF stimulation still took place during extracellular Ca 2+ -reduction. Increasing concentrations of the specific cyclooxygenase 2 (COX2) inhibitors indomethacin amid or indomethacin n-heptyl ester did not influence basolateral PAH uptake rate, while the COX1 specific substance SC-560 inhibited PAH uptake in a dose dependent manner. In summary, we showed that EGF stimulation of organic anion uptake in OK cells is Ca2+ independent, whereas basal activity depends on Ca2+. Additionally EGF leads to release of AA which is converted to PGs via COX1. These data confirm and complement the above mentioned data on short term EGF action on basolateral organic anion uptake rate in OK cells. (supported by DFG-Grant Ge 905/3-4) Physiol. Inst., Univ. Würzburg; Röntgenring 9, 97070 Würzburg, Germany P 52-1 LUMINAL P2Y 2 RECEPTOR-MEDIATED INHIBITION OF ELECTROGENIC Na + ABSORPTION IN MOUSE DISTAL COLON H. Lehrmann, O. Wittekind, J. Leipziger

Luminal P2 receptors are expressed in nearly all epithelia. Previously we demonstrated that luminal ATP/UTP stimulated a P2Y 2 receptor located in the luminal membrane of mouse CCD principal cells and inhibited electrogenic Na+ absorption (Lehrmann et al. JASN 13: [2002] in press). Similarly, also in rat distal colonic mucosa, a luminal P2Y 2 was identified (Kerstan et al. Pflügers Arch, 436:712-16 [1998]). Here we investigate the effect of luminal ATP/UTP on electrogenic Na+ reabsorption in distal colonic mucosa of mice treated on a low NaCl diet for 35±3 days. Transepithelial electrical parameters were recorded in an Ussing chamber. Baseline parameters: transepithelial voltage (Vte): –7.9±0.85 mV (lumen-negative), transepithelial resistance (Rte): 24.3±1.3 Ω • cm2, equivalent short circuit current (Isc): -334±46 µA/cm 2 (n=27). Amiloride (100 µM) completely inhibited I sc to – 6.1±9 µA/cm 2 (n=27). Luminal ATP had two distinct effects: 1. A rapid and transient (1 min) reduction of Vte from –5.9 to –1.1 with a parallel drop of Rte by 1.55±0.4 Ω • cm 2 (n=23).This effect was completely blocked by 5 mM luminal Ba 2+ (n=5). 2. A slowly onsetting and persistent inhibition of amiloride-sensitive I sc . After 4.4±0.4 min a maximal effect was reached. Luminal ATP decreased Vte to 4.3±0.5 resulting in a reduction of amiloride sensitive I sc by 109.8±16.4 µA/cm2 (37.9 % inhibition, IC50: 2.4 µM). Also, luminal UTP inhibited 27.9 % of amiloride-sensitive Isc (n=7, IC50: 2.5 µM). In the presence of amiloride luminal ATP induced the first effect only. Incubation with BAPTA-AM or cyclopiazonic acid attenuated the first effect, without effecting nucleotide-induced inhibition of Na + transport. RT-PCR from isolated colonic crypt mRNA identified specific transcripts for a P2Y2, P2Y4 and P2Y6 receptor. Immuno-histochemistry using an anti-rat-P2Y2 antibody showed significant colonic crypt cell staining. In conclusion, this data indicate that mouse distal colonic mucosa expresses a luminal P2Y2 receptor. Stimulation of this receptor leads to activation of a transient Ba2+-sensitive K+ secretion and a sustained inhibition of electrogenic Na + absorption. Thus, in all steroide-sensitive Na+ absorptive epithelia (lung, kidney, colon) this luminal P2Y2 receptor inhibits Na+ transport. Institute of Physiology, Århus University, Ole Worms Allé 160, 8000 Århus C, DK

P 52-2 THE HIGH-AFFINITY SODIUM-DEPENDENT DICARBOXYLATE TRANSPORTER CAN MEDIATE LITHIUM-DEPENDENT SUCCINATE CURRENTS C. Menzel, G. Burckhardt, J. Steffgen, B.C. Burckhardt The active transport of Krebs cycle intermediates, such as succinate, α-ketoglutarate and citrate, is mediated by sodium-coupled, lithiuminhibitable transporters found on the plasma membrane of many epithelial cells. At least two classes of sodium-coupled dicarboxylate transporters have been identified, distinguished by differences in their relative affinity for succinate and related dicarboxylates. The cDNAs coding for several high-affinity transporter orthologs, called NaDC-3, have been cloned from kidney, placenta or brain of rat, human, mouse, and flounder. In this study, we used the two-electrode voltage clamp technique to study currents associated with the influx of 3 sodium ions and one divalent dicarboxylate into oocytes expressing the highaffinity sodium-dicarboxylate transporter from winter flounder kidney, fNaDC-3. The succinate-induced k0.5 was approx. 30 µM independent of membrane potential in the range between –90 and 0 mV. In contrast to the low affinity sodium-dicarboxylate transporters, fNaDC-3 accepted both 2,2- and 2,3-dimethylsuccinate (DMS) with 2,2-DMS preferred over 2,3-DMS. Succinate-induced currents could be inhibited by lithium with half maximal inhibition at 1.2 mM lithium. Albeit lithium inhibited the succinate-induced currents, it did not significantly change the affinity: At –60 mV, k0.5 for succinate was 52 ± 9 µM and 39 ± 7 µM in the absence and presence of lithium (1 mM, n=5), respectively. In addition, lithium supported succinate-dependent currents itself. When all sodium was replaced by lithium, succinate (1 mM) induced concentration-dependent outward currents at negative, and inward currents at more positive clamp potentials. Such currents were not observed in uninjected oocytes. The nature of these unexspected succinate-dependent currents in lithium remains to be determined. Zentrum Physiologie und Pathophysiologie, Abt. Vegetative Physiologie und Pathophysiologie, Universität Göttingen, Humboldtallee 23, 37073 Göttingen

S 359 P 52-3

P 52-5

GUANYLYL CYCLASE C (GC-C) IS NOT THE RECEPTOR FOR GUANYLIN (GN), UROGUANYLIN (UGN) AND STA EFFECTS IN MOUSE AND HUMAN CORTICAL COLLECTING DUCTS (CCD) A. Sindic, C. Basoglu, J.R. Hirsch, R. Potthast, M. Kuhn, E. Schlatter GC-C is the receptor for GN, UGN, and STa in intestine, but in GC-C deficient mice saliuresis and diuresis caused by UGN and STa are not disturbed. Effects of these peptides on membrane voltage (V m) were measured by the slow-whole-cell patch-clamp technique in mouse and, for the first time, in human CCD. mRNA for GC-C was examined in isolated mouse and human tubules. Basal Vm of principal cells of human and mouse CCD were –62.0±2.8 mV, n=20 and -67.9±1.3 mV, n=95, respectively. Amiloride (1 µM) hyperpolarized (human: -4.5±0.6, n=14; mouse: -8.0±1.0 mV, n=38) and Ba2+ (1 mM) depolarized cells (human: 18.5±5.0, n=4, mouse: 28.4±1.5 mV, n=33). GN, UGN and STa (all 10 nM) depolarized cells (3.6±1.4 mV, n=7; 2.1±0.5 mV, n=5; 3.1±1.3, n=7 respectively) in human CCD. In mouse CCD, GN, UGN and STa (all 10 nM) either hyperpolarized or depolarized cells (GN: -3.9±0.7 mV, n=8; 3.4±0.4 mV, n=17; UGN: -3.7±0.5 mV, n=27; 3.6±1.5 mV, n=9; STa: -4.1±0.7 mV, n=17, 2.8±0.5 mV, n=10). All effects in mouse were mediated by changes in K+ conductances as they were abolished by 1 mM Ba 2+. 8Br-cGMP (0.1 mM) in human and mouse CCD hyperpolarized cells (-3.0±0.2 mV, n=4, -4.4±1.2 mV, n=8, respectively). There were no significant differences between effects in CCD isolated from GC-C deficient compared to wild type mice. mRNA for GC-C was detected in mouse and human proximal tubules but not in CCD. These data suggest that the signaling pathway of these peptides in human CCD is cGMP-independent. In mouse CCD we propose the existance of two signalling pathways, one possibly involving cGMP but distinct from GC-C.

EXPRESSION PATTERN AND REGULATION OF THE ACCESSORY Atp6n1b (a4) VACUOLAR H +-ATPase SUBUNIT, DEFECTIVE IN AN INHERITED FORM OF DISTAL RENAL TUBULAR ACIDOSIS C.A. Wagner1, K.E. Finberg2, P.A. Stehberger1, F.E. Karet3, G. Giebisch1, R.P. Lifton2, J.P. Geibel1,4 Vacuolar-type H+-ATPases (V-H +-ATPases) are the major H +-secreting protein in the distal portion of the renal nephron and are involved in net H +-secretion (bicarbonate generation) or H+-reabsorption (net bicarbonate secretion). In addition, V-H +-ATPases are involved in HCO3- -reabsorption in the proximal tubule and loop of Henle. V-H+ATPases are made of at least 13 subunits the functions of which have not all been elucidated to date. Mutations in the accessory Atp6n1b (a4 isoform) vacuolar H+-ATPase subunit have recently been shown to cause an inherited form of distal renal tubular acidosis in man. Here we studied the localization and regulation of expression of this subunit in mouse kidney in response to acid-base and electrolyte intake. Atp6n1b was localized by immunofluorescence to the apical compartment of the proximal tubule (S1/S2 segment) and was highly abundant in both type A and B cells in the collecting duct. NH4Cl or NaHCO3-loading as well as K+-depletion for 24 hrs, 48 hrs, or 7 days had no influence on a4 protein expression levels in either cortex or medulla as determined by Western blotting. Immunofluorescence, however, demonstrated a fast redistribution of a4 in response to the different stimuli. NH4Cl and K+-depletion led to a pronounced apical staining in type A cells. Taken together, these results suggest that under conditions where increased V-H+-ATPase activity is required a4 is regulated by trafficking but not protein expression. This may allow for the rapid adaptation of V-H+ATPase activity to altered acid-base intake and homeostasis.

Medizinische Klinik und Poliklinik D, Experimentelle Nephrologie, Universitätsklinikum Münster, Domagkstr. 3a, 48149 Münster Germany

Departments of 1Cellular and Molecular Physiology, 2Genetics, and 4 Surgery, Yale University, School of Medicine, 333 Cedar Street, New Haven, CT 06520 USA, 3 Departments of Medical Genetics and Nephrology, Cambridge Institute for Medical Research, Addenbrooke’s Hospital, Cambridge, UK

P 52-4 REGULATION OF THE EXPRESSION OF THE Cl -/ANION EXCHANGER PENDRIN IN MOUSE KIDNEY BY ACID-BASE STATUS C.A. Wagner1, K.E. Finberg2, P.A. Stehberger1, R.P. Lifton2, G. Giebisch1, P.S. Aronson3, J.P. Geibel1,4 Pendrin belongs to a superfamily of Cl-/anion exchangers and is expressed in the inner ear, the thyroid gland, and the kidney. Mutations in pendrin cause Pendred syndrome characterized by sensorineural deafness and goiter. Recently pendrin has been localized to the apical pole of nontype A intercalated cells of the cortical collecting duct, and reduced bicarbonate secretion was demonstrated in a pendrin knock-out mouse model. To investigate a possible role of pendrin in acid-base transport in the cortical collecting duct we examined the regulation of expression of pendrin by acid-base status in mouse kidney. Challenging mice with NH4Cl reduced pendrin expression levels after only 1 day and decreased levels to 30 % of control after 1 week. Concomitantly, pendrin protein was shifted from the apical membrane to the cytosol and the relative abundance of pendrin positive cells declined. Similarly, in chronic K+depletion causing a metabolic alkalosis, pendrin expression decreased and was shifted to a more intracellular pool with the relative number of pendrin positive cells being reduced. In contrast, after NaHCO3-loading pendrin was found exclusively in the apical membrane and the relative number of pendrin positive cells was increased. Protein levels, however, were only increased after 24 h of treatment and fell thereafter to control levels. In summary, three different mechanisms may regulate pendrin activity in the cortical collecting duct: i) protein expression levels, ii) trafficking, and iii) relative abundance of pendrin expressing cells ie. plasticity of intercalated cells. These results are in agreement with a potential role of pendrin in bicarbonate secretion and regulation of acid-base transport in the cortical collecting duct. Departments of 1Cellular and Molecular Physiology, 2Genetics, 3Internal Medicine, and 4Surgery, Yale University, School of Medicine, 333 Cedar Street, New Haven, CT, 06520 USA

P 52-6 cAMP INDUCED RENIN GENE TRANSCRIPTION IN JUXTAGLOMERULAR CELLS IS MEDIATED VIA THE CRE BINDING PROTEIN J. Klar, P. Sandner, M.W.H. Müller, A. Kurtz Although the cyclic AMP signalling cascade is considered as the main activator of renin gene expression in renal juxtaglomerular cells, the molecular pathways along which cAMP exerts this effect are still a matter of controversy. The aim of this study was to investigate the role of the CRE binding protein in the cAMP-signaltransduction pathway for renin gene transcription. We used the well established mouse JG-cell line As4.1, which shares a number of functional similarities with native JG cells. We found that forskolin in the presence of IBMX time dependently increased renin mRNA levels and prorenin secretion. The stimulation of renin gene expression by forskolin/ IBMX was markedly attenuated by a PKA inhibitor (H-89, 10µM). Inhibition of the general gene transcription by actinomycin D had no effect on the decline of renin mRNA in forskolin/IBMX treated cells. Conversely forskolin/IBMX increased the activity of a 2.8kb fragment of the renin promoter 3-fold. We narrowed down the responsible region for the stimulatory effect of forskolin/IBMX to three 4bp of the mouse ren1c gene, which are known as putative CRE consensus sites. We found that the CRE binding protein is activated by phosphorylation in forskolin/IBMX treated cells. To further investigate the role of CRE binding protein we performed a gel shift assay with an As4.1 cell nuclear extract and found that at least one of the three putative CRE-consensus sites shows a DNA-protein interaction with a member of the CREB/ ATF family. These data suggest that cAMP stimulates renin gene transcription in JG cells by activating protein kinase A and subsequent phosphorylation of the CRE binding protein. Institut für Physiologie I, Universitätsstr. 31, Universität Regensburg, 93040 Regensburg

S 360 P 52-7

Rikke Norregaard, Physiology and Pharmacology, University of Southern Denmark, Winsloewparken 21,3, DK-5000 Odense, Denmark

GENE EXPRESSION OF PROSTANOID RECEPTORS ALONG THE MOUSE NEPHRON H.Vitzthum, J. Sailer, V. Schubert, A. Kurtz Prostaglandins and thromboxane have various intrarenals effects, e.g. play a major role in the regulation of glomerular filtration rate, renal hemodynamic, and are involved in the regulation of salt and water excretion. Since prostanoids exert their physiological effects via membran receptors linked to inhibitory or excitatory G-proteins, we investigated the distribution of prostanoid receptors in microdissected nephron segments from the mouse by RT-PCR to characterize the target structures of PGs and TXA within the kidney. We determined the mRNA expression of the TXA 2-receptor TP, PGI2 -receptor IP, PGF 2α-receptor FP, and the four subtypes of PGE2-receptors EP1, EP2, EP3, and EP4. The results are summarized in the table: Glom TP IP FP EP1 EP2 EP3 EP4

PCT DTL / TL

mTAL cTAL

DCT

CT/ CCD

17 %

20 %

100 % 73 % 50 %

90 % 21 %

100 % 25 % 18 % 9 %

OM CD

IM CD

100 % 100 %

45 %

100 %

100 %

22 %

30 % 24 %

46 % 100 % 35 % 18 % 19 %

27 %

Table 1: Semiquantification of PCR products standardized to ß-actin. Relative mRNA levels of TXA and PG receptors in different nephron segments as % of the segment with the highest expression level. Taken together TP and IP mRNA is highly abundant in the glomeruli and apparently absent in all other nephron segments. FP is strongly expressed in the distal convoluted tubule (DCT) and the cortical collecting duct (CT/CCD), a weaker expression is found in the outer medullary collecting duct (OMCD) and the thick ascending limb of the Henle’s loop (TAL). The EP1 is found in the whole collecting duct system and EP2 in Glomeruli and thin descending limb of Henle’s loop (dTL). The EP3 mRNA is mainly localized in TAL and a weaker expression is found in DCT and the collecting duct. Finally the EP4 mRNA is mainly expressed in glomeruli, a weak expression of EP4 was detected in dTL, CT/CCD, TAL, OMCD, and DCT.

P 52-9 CIRCULATING GLUCOCORTICOIDS CONTROL CYCLOOXYGENASE 2 EXPRESSION IN THE RAT KIDNEY DURING POSTNATAL DEVELOPMENT K. Madsen, J. Stubbe, O. Skøtt, S. Bachmann, B.L. Jensen Cyclooxygenase-2 (COX-2) is strongly induced in the early postnatal period from postnatal day 5 (P5) in the rat kidney. COX-2 activity is necessary for late stages of kidney development. In the adult rat, adrenal gluco- and mineralocorticoids suppress renal COX-2 expression. In the present study we tested the hypothesis that endogenous changes in circulating adrenal gluco- and mineralocorticoids control COX-2 expression during rat kidney development. Plasma corticosterone and aldosterone concentrations were determined in the postnatal period in rats. Corticosterone changed markedly during postnatal days with high levels at the time of birth, P0 (69.4 ± 15.0 ng/mL), low levels at P3 (16.9 ± 3.9 ng/mL) to P9 (12.3 ± 0.5 ng/mL) and then high levels again from P21 (63.5 ± 14.9 ng/mL) and onwards. Aldosterone concentration was high at birth, P0 (414.3 ± 160.1 pg/mL) and remained elevated through to P28 (180.9 ± 98.0 pg/mL) and then decreased to very low levels in adult rats, P57 (< 11 pg/mL) Two litters of rats were treated by subcutaneous injection of corticosterone (20 mg/kg) and with the synthetic glucocorticoid receptor agonist dexamethasone (2 mg/kg) in the ”window” with low endogenous corticosterone levels (P5-P10). Dexamethasone and corticosterone potently suppressed (4.1-fold and 1.8-fold, respectively) renal COX-2 mRNA levels as determined by ribonuclease protection assay. A second litter was given the mineralocorticoid receptor antagonist canrenoate (20 mg/kg) from P1-P5. Canrenoate had no effect on renal COX-2 expression. Gluco- and mineralocorticoid receptors were stably expressed at birth and in the early postnatal period in control kidneys. We conclude that renal COX-2 expression is regulated by glucocorticoids and not by mineralocorticoids in the early postnatal period. Regulation is likely to be due to changes in circulating levels of glucocorticoids and not to changes in receptor expression in kidney tissue. Endogenous changes in circulating glucocorticoids may impact on late renal development through this mechanism.

Institut für Physiology, University of Regensburg, 93040 Regensburg

Kirsten Madsen, Physiology and Pharmacology, University of Southern Denmark, Winsloewparken 21, 3, Dk-5000 Odense, Denmark

P 52-8

P 52-10

STIMULATION OF 11β-HSD2 mRNA AND PROTEIN IN RAT COLON BY A LOW-SALT INTAKE Norregaard R., Jensen BL., Skott O We investigated whether the level of dietary salt intake affects sensitivity to adrenal corticosteroids in primary epithelial target organs. Rats were given high (Na+ 3%, n=8) or low salt (Na+ 0.01%, n=8) diet. After 10 days, they were killed by decapitation and plasma and tissue were sampled. Plasma renin and aldosterone concentrations were measured by radioimmunoassay. Ribonuclease protection and western immunoblotting assays were used to assess mRNA and protein expression of mineralocorticoid and glucocorticoid receptors (MCR and GCR, respectively) and of 11β-hydroxysteroid dehydrogenase-type 2 (11βHSD2) in kidney regions and in terminal colon. Immunohistochemistry was used to localize expression of 11β-HSD2. Plasma renin and aldosterone concentrations were significantly and markedly increased by dietary sodium restriction (renin: 73.3 ± 12.5x10-5 GU/ml vs 9.4 ± 2.2x10-5GU/ml and aldosterone: 7100 ± 432.9 pg/ml vs 26.7 ± 10.8 pg/ ml). 11β-HSD2 mRNA and protein levels were significantly (3.2-fold) elevated in rat terminal colon in response to low salt diet. In contrast, MCR mRNA levels were small but significantly lower in rat colon in the low salt group compared to high salt, whereas GCR levels did not change. In rat kidney cortex and medulla there were no changes in 11β-HSD2, MCR or GCR mRNA levels in response to dietary salt intake. Also 11β-HSD2 protein levels were not regulated in rat kidney cortex by changes in dietary salt intake. Immunohistochemistry showed strong labeling of colon epithelial cells in the mid-crypt region by the anti-11β-HSD2 antibody in low-salt rats, whereas labeling of colon in the high salt group was hardly detectable.Terminal colon seems to be a site where mineralocorticoid sensitivity and selectivity is regulated by changes in dietary salt intake.

ACUTE AND CHRONIC REJECTION OF RAT KIDNEY TRANSPLANTS MODIFY EXPRESSION OF RECEPTORS AND TRANSPORTERS INVOLVED IN Na+-AND H2O-REABSORPTION J.R. Hirsch, S. Schonebeck, R. Thomas, R. Schröter, H. Stegemann, E. Schlatter, G. Gabriëls The kidney plays an important role in regulation of Na+- and H2Ohomeostasis. Key regulators are a.o. the renin-angiotensin-system (RAS) and natriuretic peptides. After kidney transplantation patients often show increased Na+- and H2O-retention, but not much is known about the underlying mechanisms. We used two different rat rejection models to investigate changes in the expression of regulators and transport systems in the kidney. Kidneys from Lewis-Brown-NorwayF1 (LBN) rats transplanted into Lewis rats after uninephrectomy were removed after 1, 2, 4 and 5 days (acute rejection) and kidneys from Fisher rats transplanted into Lewis rats with bilateral nephrectomy were removed after 7-10 weeks (chronic rejection). Semiquantitative RT-PCR was used to investigate expression of the natriuretic peptide receptor type A (GC-A), the angiotensin II receptor (AIIR), the Na+/ H+-exchanger type 3 (NHE-3), the secretory K+ channel (ROMK), the epithelial Na+-channel (ENaC), the mineralocorticoid receptor (MiR) and the Na+/K+-ATPase (NKA). GAPDH served as the reference signal. The expression of GC-A increased by 60% from 0.82±0.05 in control kidneys to 1.25±0.18 in transplanted kidneys (n=5, 1 day after transplantation; d.a.t). NHE-3 decreased by 85% from 0.47±0.18 to 0.07±0.06 (n=5, 1 d.a.t.). No significant changes could be seen in the AIIR or NKA signals (0.78±0.06 to 0.64±0.07; 0.11±0.05 to 0.16±0.11, respectively, n=5). While ROMK and MiR showed no significant regulation of expression, ENaC was strongly inhibited by 62% (n=5, 1 d.a.t.). In the chronic rejection model only NHE-3 showed a significant inhibition (63%). An increase in GC-A and a decrease in

S 361 NHE-3 expression could be expected as counterregulation to the known initially reduced Na +-excretion. Surprisingly, neither AIIR nor MiR seem to be regulated due to acute or chronic rejection. Therefore, factors responsible for renal sodium retention after transplantation still need to be identified. Universitätsklinikum Münster, Med. Klinik und Poliklinik D, Exp. Nephrologie, Domagkstr. 3A, 48149 Münster, Germany P 52-11 KIDNEY COLLECTING DUCT VASOPRESSIN ESCAPE IN RATS WITH LIVER CIRRHOSIS IS ASSOCIATED WITH DECREASED cAMP ACCUMULATION AND AQP2 EXPRESSION L. Brønd1, N. Salling1, S. Nielsen2, S. Christensen1, T. E.N. Jonassen1 In addition to sodium and water retention the pathophysiological condition of liver cirrhosis is associated with elevated levels of plasma vasopressin (AVP) as a result of non-osmotic probably baroreceptor mediated AVP secretion. AVP stimulates the V 2 receptor in the basolateral membrane of the kidney collecting duct principal cell inducing cAMP accumulation and thereby regulating the aquaporin 2 (AQP2) expression. Previously, we and others have demonstrated that liver cirrhosis induced by common bile duct ligation (CBL) is associated with decreased AQP2 expression despite elevated plasma vasopressin levels indicating the presence of an escape form increased plasma AVP levels. The aim of the present study was to elucidate the mechanism behind this AVP escape phenomenon in CBL rats. As previously shown the plasma AVP levels were elevated in the CBL rats (CBL 3.3 ± 0.2 pg/ml vs. control 1.4±0.4 pg/ml, n=14 and n=11 in the two groups, p < 0.0005) and the AQP 2 water channel expression was decreased (CBL 62±12% vs. control 100±0%, n=13 in both groups, p<0.005). In addition to this a significantly decreased expression of the active water carrying AQP2 water channel was demonstrated by measuring the expression of the phosphorylated AQP2 using a specific antibody recognizing the AQP2 water channel phosphory-lated at serine 256 (CBL 50±6% vs. control 100±0%, n=13 in both groups, p<5⋅10 -8). Dose response studies of cAMP accumulation of isolated micro-dissected cortical collecting ducts upon AVP stimulation demonstrated a significantly (p<0.01) decreased cAMP accumulation within the CBL rats (see figure). Cortical collecting ducts fmol cAMP / mm tubule

15

10

CBL Sham CBL

5

0 -15.0

-12.5

-10.0

-7.5

-5.0

-2.5

0.0

log (AVP conc, M)

Together these results indicate that decreased cAMP accumulation is involved within the mechanism behind the AVP escape phenomenon in rats with liver cirrhosis induced by CBL. 1

Dept. of Pharmacology, University of Copenhagen, Blegdamsvej 3, Blg. 18.6, 2200 Copenhagen N, Denmark, 2 University of Aarhus, Denmark

P 53-1 TRUNK MUSCLE ACTIVITY DURING TREADMILL WALKING – DEPENDENCY ON BACK PAIN, DEFINED LOADING AND PROFESSIONAL EXPERIENCE IN CARRYING OBJECTS Ch. Anders, Ch. Puta, R. Graßme, H. Wagner The world over Low Back Pain (LBP) is a major problem. In developed countries chronic LBP causes costs of billions of dollars annually. For individuals LBP is associated with certain psychological and physical characteristics. Especially functional abnormalities in the co-ordination of their trunk muscles are worthy of continued attention. Trunk muscles can be subdivided into two functionally different groups: the local

system and the global system. Muscles of the local system which insert at the spinal column act at low force levels independently from movement direction are thought to generate stability. Muscles of the global system are activated depending on movement direction and are therefore responsible for movements. Nine subjects without professional experience in carrying objects and four waiters were investigated at different speeds on a treadmill. Treadmill speed was varied between 2 km/h and 6 km/h in steps of 0,5 km/h. Using surface EMG (SEMG) M. rectus abdominis, M. obliquus externus and internus and M. multifidus were investigated. The investigation was carried out both before and after a loading situation which consisted of wearing a waistcoat with an additional weight of 25% of the upper body mass for two hours. The amplitude of all investigated muscles increased with increasing speed. During locomotion in general M. obliquus internus showed the highest SEMG amplitude, followed by the multifidus muscle. External oblique and rectus abdominis muscles only showed negligible amplitudes. This experimental setup as is did not enable consistent observation of pattern changes in relation to chronic LBP, profession or load. However, intra-individual comparison shows activation pattern between sides in particular in internal oblique and multifidus muscles, low amplitude levels of the internal oblique and amplitude changes due to load application. Therefore, evidence has still to be substantiated to assign the found pattern to pain history, profession or premorbid conditions. Supported by Kompetenzzentrum Interdisziplinäre Prävention, University of Jena and BGN. Institute for Pathophysiology, Motor Research Group, Jena, Germany P 53-2 TRUNK MUSCLE ACTIVITY DURING ACYCLIC EXTERNAL SENSOMOTOR PROVOCATION OF THE POSTURAL SYSTEM C. Puta, C. Anders, H. Wagner, H.H.W. Gabriel The function and co-ordination of the small inter segmental muscles of the spinal column in Low Back Pain (LBP) patients came into scientific focus recently. A basis for the classification of the trunk muscles was given by Bergmark (Bergmark 1989). He subdivided trunk muscles into local and global muscles. The aim of this study was to prove the significance of the trunk muscles in the context of postural activity. By means of external acyclic sensomotor provocations (Posturomed®) postural stability (bipedal standing) was investigated. The provocations were carried out both, latero-lateral and anteriorposterior directions, each with eyes open and closed, respectively. Nine subjects were investigated before and after a static-dynamic loading of the thorax (wearing of a waistcoat for two hours). Surface EMG (SEMG) of trunk muscles (M. obliquus internus, M. obliquus externus, M. rectus abdominis, M. multifidus lumbalis) and movement signals of the Posturomed were measured simultaneously . The mean SEG amplitude during a reference situation (bipedal standing on solid ground) showed higher SEMG amplitudes compared to the situation while standing on the Posturomed. Therefore, the Posturomed does not facilitate the trunk muscles. Side differences occurred were observable, but unsystematically between the subjects. External provocation in latero-lateral direction predominantly activated the multifidus and internal oblique muscles (local system). Furthermore, the muscles of the local system show the highest amplitudes independently from provocation direction and the two hours loading of the thorax. The M. obliquus internus, as an important stabiliser of the front trunk wall during bipedal standing and during external provocation, plays a prominent role in the postural program of humans. It’s main function can be described as facilitator of local and ultimately global stability. This underlines it’s significance for the diagnostic and therapeutic entrance towards LBP. Bergmark A. Stability of the lumbar spine. A study in mechanical engineering. Acta Orthop Scand Suppl 1989; 230:1-54. Supported by Kompetenzzentrum Interdisziplinäre Prävention, University of Jena and BGN Institute of Sport Science, Dept. Sport Medicine, University of Jena, D-07749 Jena, Germany P 53-3 COMPARISON OF TREADMILL AND SHUTTLE WALK TESTS FOR ASSESSING WALKING PERFORMANCE IN PATIENTS WITH INTERMITTENT CLAUDICATION I. Zwierska, J.M. Saxton, S. Nawaz, C.H. Wilkinson, R.D. Walker,

S 362 A.G. Pockley, R.F.M. Wood Treadmill walking is used to assess the clinical status of patients with intermittent claudication. However, patients can experience concomitant-cardiovascular problems (e.g. hypertension) during exercise (Gardner et al., 1995), which may be exacerbated by incline walking. This study compared the cardiovascular responses to an incline treadmill walking test (TWT) with those of an incremental shuttlewalking test (SWT; Singh et al., 1992), that involves walking at a prescribed pace between two points on level ground. Following ethical approval from the North Sheffield Research Ethics Committee, fiftyfive patients with stable intermittent claudication performed the TWT (3.2 km.h-1, 12% gradient) and SWT (initial speed of 3 km.h-1, increasing 0.5 km.h -1 every min), each on three occasions in random order on separate days. The initial test served to familiarise patients with the testing protocols. Claudication distance (CD), maximum walking distance (MWD), heart rate, blood pressure and the ankle-brachial pressure index (ABPI) were assessed for each test. Patients also completed a test preference questionnaire. TWT induced greater increases in systolic blood pressure and heart rate (P<0.01; paired ttest). The ABPI decreased by 10% after TWT compared to 6% following SWT (P<0.05; paired t-test). Walking distances for SWT were more than double (P<0.01; paired t-test) those achieved in TWT, with the reproducibility of MWD for both tests being equivalent. The proportion of patients preferring the SWT was 43%, in comparison to 24% for the TWT. We therefore conclude that SWT induces lower cardiovascular stress and should be considered as a safe, alternative means of assessing walking performance in patients with intermittent claudication. Gardner, A.W., Skinner, J.S., Bryant, C.X., Smith, L.K., (1995). J. Cardiopulm. Rehabil. 15(2): 134-42. Singh, S.J., Morgan, M.D.L., Scott, S., Walters, D., Hardman, A.E., (1992). Thorax, 47: 1019-24.

INTRAMUSCULAR INTERLEUKIN-6 PRODUCTION AND RELEASE IN YOUNG AND ELDERLY HUMANS DURING EXERCISE M. Pedersen, A. Steensberg, C. Keller, H. Bruunsgaard, T. Osada, M. Zacho, H.W. Hendel, B. Saltin, M.A. Febbraio, B.K. Pedersen It has been shown that contracting skeletal muscles produces and releases the cytokine interleukin (IL)-6. IL-6 expression in contracting skeletal muscles is augmented in the presence of low intramuscular glycogen stores. Since elderly subjects may rely on glycogen more than younger subjects, it is possible that they may produce more IL-6 during exercise. To test this hypothesis seven healthy elderly males (mean age 70 yr) and 6 healthy young males (mean age 26 yr) performed 3 hours of dynamic knee-extensor exercise at 50% of Wmax. IL-6 mRNA and glycogen in muscles were analysed and the IL-6 release were estimated before, during and after. The absolute net-IL-6 release and the decline in muscle glycogen concentration during exercise did not differ between age groups. However, when corrected for workload the glycogen utilization and the change in net IL-6 release were higher in the old subjects. Our results suggest that elderly subjects release a greater amount of IL-6 from contracting muscle and this may be related to a higher intramuscular glycogen utilization.

Centre for Sport & Exercise Science, Sheffield Hallam University, S10 2BP, UK

The Copenhagen Muscle Research Centre, Rigshospitalet, University of Copenhagen, Blegdamsvej 9, 2100 Copenhagen, Denmark

sion of human PBMC on trans-criptional level in response to an appropriate signal. Further research is necessary to explore the functional significance of this adaptional effect of ET on immunity. Medical Clinic, Dept. of Sports Medicine; *Dept. of Transfusion Medicine, University of Tubingen, 72074 Tübingen, Germany P 53-5

P 53-4 P 53-6 IMPACT OF ENDURANCE TRAINING ON THE REGULATION OF THE INDUCIBLE NITRIC OXIDE SYNTHASE (iNOS) IN HUMAN BLOOD MONONUCLEAR CELLS Niess, A.M., E. Fehrenbach*, M. Vogel*, I. Waldvogel*, H. Northoff*, H.-H. Dickhuth Objective: Nitric oxide as generated by iNOS plays an important role in host defence. Lu et al. (Am JPhysiol, 1999, 276: R482) documented an augmented responsiveness of iNOS expression to LPS and IFN-γ in peritoneal macro-phages of trained mice when compared to control animals. The goal of the present study was to determine whether regular endurance training (ET) exert a priming effect on iNOS expression in human peripheral blood mononuclear cells (PBMC). Methods: PBMC were obtained from 22 trained (TR: 31.4, 28.5-34.6 yrs. (mean, 95% Cl); 10.2, 8.7-11.9 h ET/week) and 20 sedentary male controls (UT: 28.5, 25.4-31.6 yrs.) at rest. After in-vitro stimulation of the cells with heat stress (HS: 41.6°C, 2 h), PMA or a combination of LPS, IFN-γ and TNF-α (CYT), expression of iNOS-mRNA and protein was determined by RT-PCR and FACS, respectively. Untreated cells from the corresponding samples were analyzed as controls in parallel. Results: Stimulation of PBMC by HS or PMA did not affect expression of iNOS-mRNA. Treatment of PBMC of TR with CYT induced a slight but significant rise of iNOS-mRNA, while PBMC of UT did not respond to stimulation (see figure).

INFLUENCE OF EXERCISE, HEAT AND HYPOXIA ON PLASMA-HSP72 E. Fehrenbach, A.M. Niess, M. Jesse, H.-H. Dickhuth, H. Northoff Heavy endurance exercise is known to activate the immune system. Environmental conditions like elevated temperature and hypoxia boost the exercise stress response. Soluble heat shock protein HSP70 is suggested to play a role as signaling molecule in the immune response. Therefore we asked if endurance exercise and additional environmental heat stress or hypoxia affect the level of inducible HSP72 in the plasma of endurance athletes. Seven subjects performed continuous treadmill runs (CR, 1 h, 75% VO2max) at 18°C (N) and at 28°C (H) ambient temperature. Another group of 10 subjects performed extensive interval training (IT, 10x1000 m) at 350 m above sea level (88% VO2max) and at moderate altitude of 1800 m (84% VO2max) yielding similar lactate levels at both altitudes. HSP72- and interleukin-6 (IL6)-levels in plasma were determined by ELISA before and 0, 0.5, 3, 24 h after exercise. Plasma-HSP72 and IL-6 were significantly increased immediately and 0.5 h after CR with a prolonged high HSP72-level in H. IT at both altitudes induced a significant increase of soluble HSP72 and IL-6. High ambient temperature but not hypoxia had an additional impact on the extracellular HSP72 response. The parallel increases of HSP72 and IL-6 in plasma after exercise may suggest a role for soluble HSP72 in the regulation of the exercise-induced stress response. Abteilung Transfusionsmedizin, Eberhard-Karls-Universitaet Tuebingen, Otfried-Mueller-Str. 4/1, D-72076 Tuebingen, Germany, Email: [email protected] P 53-7

In both groups, cytoplasmic expression of iNOS protein was not affected by stimulation. Conclusions: ET seems to prime iNOS expres-

DO GENDER DIFFERENCES EXIST IN RESPONSE TO ADDED EXTERNAL DEADSAPCE DURING CONSTANT WORKLOAD EXERCISE IN HUMANS? # E. Kilbride, *C.G. Gallagher, #H.R. Harty, #†P. McLoughlin Recent studies have suggested that smaller lung volumes and lower maximal flow rates in women may predispose them to ventilatory limitation during high level exercise (Harms et al.,1998). External deadspace provides an additional ventilatory stimulus during exercise

S 363 and we hypothesised that females may exhibit more limitation during lower levels of exercise in the presence of additional deadspace compared to males. In order to make suitable physiological comparisons we chose to study subjects at both a set VCO2 of 1l/min and at a workload halfway between 20w and the gas exchange threshold (MID WL) with an added external deadspace that corresponded to 12% of each individual’s forced vital capacity (FVC). Ten males and ten females (taking oral contraceptives) performed two 12min constant workload cycle tests (randomised order) in the presence (DS) and absence (CT) of an additional deadspace (males: 638 ± 18.8ml, females: 480 ± 9.6ml). Subjects exercised for 6mins at a workload targeting 1l/min VCO2 followed by 6mins at MID WL.

∆VE (l/min) ∆VT (l) ∆FR (bpm) ∆PetCO2 (mmHg)

FEMALES 1L VCO2 7.41 (0.93) 0.19 (0.07) 2.90 (1.76) 0.13 (1.06)

MID WL 5.94 (1.53) 0.18 (0.07) 2.24 (1.96) 0.51 (0.98)

MALES 1L VCO2 7.74 (0.57) 0.35 (0.04) 0.22 (0.31) 0.51 (0.98)

MID WL 9.42 (1.81) 0.37 (0.09) 0.51 (0.55) 0.45 (0.38)

Table 1: Mean (SEM) delta values for the last 2mins of exercise at 1l VCO2 and at MID WL. While deadspace increased VE in both males and females by significantly increasing VT, we found no significant gender differences during constant workload exercise in response to an added external deadspace in this group of normal adults. Harms, C.A., McClaran, S.R., Nickele, G.A., Pegelow, D.F., Nelson, W.B. & Dempsey, J.A. (1998). J. Physiol. 507:619-628. This work was funded by a UCD Presidential Scholarship and Enterprise Ireland and was approved by St Vincent’s University Hospital Ethics committee. # Department of Human Anatomy and Physiology and † Conway Institute of Biomolecular and Biomedical Sciences, University College Dublin; * Department of Respiratory Medicine, St Vincent’s University Hospital, Dublin, Ireland.

P 53-8 CO2 INFLUENCES PUPIL FLUCTUATIONS M. Möller, T. Schäfer, M.E. Schläfke In the dark spontaneus fluctuations in the size of the pupil can be measured. This behaviour of the pupil is seen as a decreasing influence of the central adrenergic system on the Edinger-Westphal nucleus. Does CO2 influence this mechanism? In 20 healthy subjects we have performed two pupillographic measurements (AmTech, Weinheim). The subjects breathed through a nose-mask allowing the variation of air composition. On each subject one pupillographic measurement has been performed in normocapnia, the other in hypercapnia (FiCO 2 0.05). The sequence has been randomized. We compared the mean of the pupillary-unrest-index (PUI) of the normocapnic and hypercapnic tests using the Wilcoxon-test for matched pairs. Results: The subjects with a low PUI in normocapnic conditions already (below the median of 6.06 mm/min) did not show any changes in the PUI in hypercapnia (normocapnia 4.19 ± 1.23 mm/min versus hypercapnia 4.53 ± 1.68 mm/min, ns). In contrast the PUI was significantly decreased during hypercapnia in those subjects showing a higher baseline PUI (normocapnia 11.08 ± 4.62 mm/min versus hypercapnia 6.51 ± 2.24 mm/min, p<0.01). In the group of subjects with a low normocapnic PUI we did not find any influence of CO2 on the heart rate (normocapnia 73.02 ± 11.01 min-1 versus hypercapnia 76.50 ± 9.07 min-1, ns). In the group of subjects with a high normocapnic PUI however the heart rate increased significantly under CO 2 (normocapnia 74.67 ± 7.94 min -1 versus hypercapnia 82.20 ± 10.39 min-1, p<0.05). The subjects of both groups did not show any differences in the ventilary response to CO2. Conclusions: The data indicate that CO2 is able to decrease the PUI in that group in which the heart rate became increased at the same time. The question arises whether this reaction indicates the CO2 sensitivity of the central adrenergic system. Dept. of Applied Physiology, Ruhr-Universität Bochum, D-44780 Bochum

P 53-9 SIMULTANEOUS MODULATION OF FREQUENCY AND AMPLITUDE OF MIDDLE FREQUENCY CURRENTS BETWEEN HIGH TONE AND LOW ULTRASOUND RANGE - APPLIED PHYSIOLOGY FOR DIFFERENTIATED ELECTROTHERAPY CONCERNING THE DISTINCTION AND SPECIFIC PROPORTIONAL MIXTURE OF STIMULATORY AND NONSTIMULATORY (METABOLIC AND/OR BLOCKING) EFFECTS H.U. May Apart from DC numerous kinds of currents are offered in electrotherapy. It is usual to classify them according to their frequencies (f) as low, middle and high f currents (LF, MF, HF). This mode of classification goes back to physiological investigations. The term „MF“ was introduced 1944 by GILDEMEISTER. 2-100 kHz is the central MF range, 1 - 2 kHz and 100 - 300 kHz are transition ranges to LF and HF. MF currents have special properties, a) in general a lack of correlation between the polarity of the phase of the current and the generated AP, at higher intensities (i) the generation of a reversible sustaining depolarisation, in nerves accompanied with a block, in muscles with a physiological contracture, b) f-dependent a decrease of the impedance and of the systemic electrotoxicity, an increase of the thresholds and an acceleration of fading of sensation. Furthermore non-stimulatory electro-chemical shaking effects exist, increasing with f and i, beneficial for a facilitation of diffusion. Nevertheless until now the lower MF range around 4 kHz is clearly preferred for therapeutical purposes because of the lower thresholds for stimulatory effects. A new equipment was developed to enhance distinctly the non-stimulatory efficacy up to the legal power limit of 5000 mW. New is the possibility of simultaneous modulation of amplidude and f between 4096 and 32768 Hz (212 and 213 Hz), a) slowly in 72 quarter tone steps (1 step/s) along the individually calculated threshold curve to maximize the non-stimulatory effects, b) fast with adjustable modulation f in the LF range to generate stimulatory effects simultaneously with adjustable portions of non-stimulatory effects, realized by adjustable variations of the gradient of the line crossing the threshold curve and representing the simultaneous modulations of amplitude and f (parallel stimulation and paradoxical stimulation). All stimulations are well tolerable or even comfortable. For example, 10 Hz can be used for stimulations of peripheral sympathetic nerves with vasoconstrictory and antiedematous effects, 20 Hz for strong asynchroneous tetanizing muscle stimulations, 100 Hz for pain relief. Dr. med. Hans Ulrich May, Nervenarzt, Albert-Einstein-Str. 1, 75015 Bretten P 53-10 EFFECTIVENESS OF A LIGHT-WEIGHT ICE-VEST FOR BODY COOLING IN FIRE-FIGHTER’S WORK J. Smolander, K. Kuklane, D. Gavhed, H. Nilsson, E. Karlsson, I. Holmer The aim of the present study was to examine the effects of wearing a light-weight ice-vest (1kg, water) on physiological and subjective responses in fire fighter’s work. The experiments were carried out in a climatic chamber, in a container under extreme radiant heat, and during simulated smoke-diving. In addition, the physical cooling effect of the ice-vest was measured with a thermal mannequin. Four experienced fire fighters participated in the experiments. In all tests, the subjects wore the standard clothing for fire fighters with the selfcontained breathing apparatus. The total extra weight carried was 2123 kg. The ice-vest was worn over the underwear. The laboratory tests consisted of 30 minutes of treadmill walking at a moderate (4km/h, no inclination), and a heavy (4 km/h, inclination of 4 degrees) work intensity in the heat (45oC, 30%) without and with the ice-vest. The results showed that the ice-vest effectively reduced skin temperatures under the vest, especially on the back under the breathing apparatus. Wearing the ice-vest did not affect the metabolic rate, skin temperatures outside the vest or the rise in rectal temperature. On average, heart rate was ca. 10 beats/min lower, amount of sweating was reduced by 13%, and subjective sensation of effort and warmth were lower during work with the ice-vest compared to work without it. The results from tests in the container and in the smoke-diving house largely supported the laboratory results. According to the thermal mannequin tests, the useful energy available from the vest for body cooling was rather high (58%). In conclusion, the light-weight ice-vest clearly reduces circulatory, thermal, and subjective strain during demanding smoke-

S 364 diving tasks. It was estimated that the added benefit is 10-15%.

521, 2P, 2-3P).

ORTON Research Institute, Tenholantie 10, FIN-00280, Helsinki, Finland

Institute of Biomedical & Life Sciences, Glasgow University, Glasgow G12 8QQ, Scotland, United Kingdom

P 54-1

P 54-3

ECHOCARDIOGRAPHY IN THE PHYSIOLOGICAL LAB COURSE DEMONSTRATION OF THE HEART CYCLE IN 2DMODE AND QUANTIFICATION OF STROKE VOLUME IN MMODE W. Moll, S. Holmer Scope.The classical warm blood animal experiment, abandonned because of animal rights, provided the opportunity to the students to measure cardiac pumping function and related essential data on cardiophysiology. In the present paper it shall be shown that echocardiography, performed in collaboration with a cardiology department, provides the opportunity to obtain similar quantitative data on cardiac output in human. Protocoll. Two-dimensional (2D) echocardiograms (4chamber view) and parasternal time-motion (Mmode) echocardiograms with ECG were taken on students in a plenary session and documented on video film. Echos demonstrating heart failure were also taken. The video films were digitized, put in the university net and made available using the student (zip) pool of computers. In addition, hard copies on the M-mode echocardiogram and the 2D enddiastolic and endsystolic left ventricle were printed in true dimensions (scale 1cm:1cm). In the lab course, the students replayed the 2D movies demonstating the heart cycle. Single frames demonstrate the ventricular diameters in 40 ms intervals. On the hard copies of the parasternal M-Mode echocardiogram, the students measured the (minor) ventricular diameter in endsystole and enddiastole. Since the major diameter is approximately twice the minor diameter, the ventricular volume is easily derived as cube of the minor diameter (Dodge et al. 1960). From enddiastolic and endsystolic volume stroke volume, and, using heart rate from ECG, cardiac output is fairly easily and conceptually derived. Conclusions. 2D-Echocardiograms as videos allow to study the time sequence of left ventricular movements. Hard copies of M-mode – echocardiograms with ECG allow the quantification of stroke volume, ejection fraction and cardiac output.

PROBLEM BASED LEARNING - FIRST EXPERIENCES IN SEMINARS OF PATHOPHYSIOLOGY U. Zwiener, M. Eiselt, R. Bauer The efficiency of physiological education for the thinking of the future physician fundamentally depends on the continuation of physiological thinking by connecting disciplines to clinical medicine such as pathophysiology and pathobiochemistry. Because of the reduction or the cessation of these disciplines at the moment and the structure of the planned „Approbationsordnung“ one must try to demonstrate the essential attractiveness and the methodical efficiency up to and during the clinical education. Otherwise the morphological part of pathology will determine alone the clinical thinking. Problem based learning (PBL) could be a possible way. This is in progress worldwide. The study of medicine as a whole becomes very probably an essential part in the PBL-form, at least in a long-term perspective. This is also a reason to deal with this new didactical possibility, although the staff capacity isn’t guaranteed for the education in small groups today. However, this could be temporary compensated by teaching of further tutors from engaged clinical disciplines. PBL is mainly interdisciplinary. Therefore in the seminar of Harvard-PBL ,,pathomechanisms“, physiologists have to represent at least partly pathomorphology, pathobiochemistry, and clinical chemistry and clinical propaedeutics, too. It is an advantage that we can demonstrate the efficiency of pathophysiological thinking in simple examples of clinical routine. In Jena, we teach PBL components in seminars by this way: The student trains his pathophysiologic and clinical-prodaedeutical skill by fictitious patient. Only when mistakes occur the student is corrected. The seminar ,,patient“ (e.g. coronar heart diseases with Mobitz II) demonstrates a main pathophysiological and clinical-epidemiological issue. If students ask, the tutor gives additional functional clinical-physiological or paraclinical data which must be evaluated by the students. There are introducing lectures and student’s studies for (1) clarification of basal problems of understanding, (2) the definition of the problems, and there is a seminar (3) for the collection of ideas for problem solving, (4) a systematic classification, (5) the formulation of aims of further learning, in a final step (6) the covering of the whole content, and (7) the synthesis and discussion. The students reacted very positively to this mixed form of lectures and PBL. Students of weak performances will have difficulties with PBL, and PBL in this second part of medical studies must be combined with introducing lectures.

Physiologisches Institut und Klinik und Poliklinik für Innere Medizin II der Universität Regensburg, Universitätsstr. 31, D-93040 Regensburg P 54-2 TEACHING APPARATUS FOR INVESTIGATING POISEUILLE’ S RELATION O. Holmes, M.L. Lucas Physiology is an experimental discipline and there is a need to promote the inclusion of experimental work in the undergraduate curriculum. We have developed apparatus, for student use, to produce experimental data for investigating Poiseuille’s relation. The associated teaching material incorporates the principles of problem-based learning. The laboratory schedules guide the student to make experimental observations, to interpret them and to relate them to the functioning of the body. Where appropriate, the student designs and performs further experiments to confirm or refute the ideas that have arisen from the observations.This demonstration is of the apparatus for investigating Poiseuille’s relation. Water flows from a reservoir that provides an adjustable pressure head. The water flows through nylon catheters of various radii and the flow is measured by collection of the water and measurement of volume per unit time. The student plots graphs of flow as a function of pressure and investigates the range over which the relationship is linear . This leads to a consideration of the division of energy into the proportion expended in overcoming viscous forces and the proportion imparting kinetic energy to the liquid. Relevance to the flow of blood in the circulation is considered.For low velocities of flow, the experimental data, plotted on log log graph paper, falls convincingly on a straight line with a slope around 4. As the velocity of flow increases, the fourth power relationship breaks down because a significant amount of energy is required to impart kinetic energy to the liquid. The students are guided to appreciate the physiological and clinical importance of the fourth power relationship between flow and radius. The apparatus is part of a modular design that includes other apparatus previously demonstrated to the Society (J. Physiol. 1999,

Institute of Pathophysiology, Friedrich-Schiller-University Jena, 07740 Jena P 54-4 PROBLEM-BASED LEARNING ON A SYSTEMATIC BASIS: THE BOCHUM APPROACH TO REFORMED PRECLINICAL STUDIES OF MEDICINE T. Schäfer, H. Lieverscheidt, P. Jansen, P. Scheid The classical division of medical studies into a preclinical and a clinical part has often been criticized because of a lack of clinical expertise, skills, and decision making competence even among advanced medical students due to the delay in confronting clinical issues. A new concept to integrate clinical aspects into the preclinical courses has been developed: The preclinical studies consisting of lectures, seminars and practical courses within the main fields of anatomy, biochemistry, physiology and medical psychology have been shortened by half a semester by introducing block courses and by rearrangement and selection of the most important topics. The remaining time is then devoted to the mandatory course of ”Introduction to Clinical Medicine”. This course has been rearranged in the sense of problem-based learning (PBL) in small student groups of 7 members with a professional tutor. The students are trained to work in a self-directed way according to the classical steps of PBL with definition of learning goals, elaboration, and discussion of the results after a one week interval. Afterwards we offer an expert lecture on the clinical features and newest developments in the fields addressed by the paper cases. At this moment the cases

S 365 have been constructed with respect to relevant preclinical aspects of (patho)physiology, (patho)biochemistry, anatomy and medical psychology. The main topics include: blood, heart, lung and airways, muscle, gastro-intestinal tract, liver, kidney, thyroid, the endocrine system, always containing psycho-social and ethical aspects. We now test the cases in a peer group of students in order to see, whether appropriate learning goals are defined. The official start of the programme is in the summer 2002 with a cohort of about 325 students in their 4th semester. The new idea is that PBL is used to motivate the students to re-activate their knowledge of preclinical studies so as to analyse clinical cases and to explain observations using the basis of their systematic learning. Finally, each student is asked to give structured feedback in order to evaluate the process, together with an objective evaluation procedure, namely the preclinical examination, called ”Physikum”. Medical Faculty, Ruhr-Universität Bochum, D-44780 Bochum P 54-5 NON-INVASIVE HUMAN EXPERIMENTS FOR STUDENT’S LAB COURSE IN PHYSIOLOGY H.-V. Ulmer Most activities in student’s lab courses can be divided into I. animal experiments II. human experiments III. simulated ”experiments”. Animal experiments are disapproved in Germany; often handling a pcmouse is regarded more positive than handling a lab mouse. Sim-”experiments” may help to learn about physiology, but not to transmit the feeling for the handicraft aspect and its cussedness referring to the future duties of physicians. Lab courses (Praktika) should include these handicraft aspects, they should not only be used as a bit-transfer for physiological knowledge. – For more than 20 years the author has developed non-invasive human experiments with small groups, at most 4 students have to perform a duty (subject, measurements, recording – e.g.: http://www.uni-mainz.de/FB/Sport/physio/ pdffiles/325wAAZ.pdf). Simple methods reduce the time, that is needed to introduce the applied methods, on a minimum, e. g. for the experiment: Perrotatory and postrotatory nystagmus with subgroups of 4 students: 1) subject, 2) guide 3) recorder 4) timer. No. 1 closes the eyes, feels the movements of the eyes with his fingers and indicates each movement by saying ”now” during and after vertical rotation, guided by No. 2. No 3 marks every ”now” on a protocol sheet with 5”sections each. No 4 indicates the 5”intervals acoustically.

n/5'' perrotatorical 3

postrotatorical

2 1 0

0

10

20

30

0,

40

50

60

1,

70

80

90

Zeit (min) 1,5

Figure 1: Typical results. Group of 24 students, 6 subgroups, 19 complete experiments in 30 min. Experim. leader: Ch. Girtsios, 2000. The results enable discussions about physiology of nystagmus and orientation in space, especially in combination with an additional experiment (motor act of walking with opened and closed eyes after rotation). Altogether this type of human experiments mediates physiological knowledge and how to handle experiments, and this by doing and not under the reduced conditions of sim-”experiments”. Dept. Sportphysiology , FB 26, Johannes Gutenberg-Universität Mainz, D 55099 MAINZ, [email protected] P 54-6 A MODULAR SYSTEM OF COMPUTER BASED EXAMINATION TECHNIQUES IN PHYSIOLOGY: PERSONALISED TESTS AND COURSE LOGISTICS M. Baumann, P. Blum, C. Jaeger, H.-J. Foerster and H. Schmid-Schönbein A novel four semester programme developed in our institution is based

on intensive student self-evaluation and objective monitoring of cognitive progress during first 3 semesters of the programme (PSIP, see accompanying abstract SCHMID-SCHÖNBEIN et al.). It’s principle aim lies in the preparation of the students for an essay test (free formulation) at the end of the 4 th semester and is preceded by an evolving series of paper-based and electronic palmtop based examinations. At the end of the 1. semester „free questions“ concerning the course in cell biology must be answered, results being evaluated by instructors from separate institutions and stored into a database. The following lectures on „physiological propedeutics“ and the physiological laboratory courses are being accompanied by a so called „credit examination“ (machine readable, repeatable at will). During each session of the 10 subsequent laboratory courses, the students are given written examinations with gradual progression from simple multiple choice to „open“ questions, short statements of 10 lines of free answer and in numerical calculations. Depending on the question type, student answers are evaluated automatically, semi-automatically or manually. A formal multiple choice test (MC-technique) is required at the half time and the end of the compulsatory laboratory course, thereby preparing our candidates for the same technique when applied by the state licensing authorities (60 % passing level). The written essay test has to be passed before success in „Physiological Seminars“ is attested (at the end of the 4th semester). These comprehensive tests are systematically prepared during the entire academic year, in that piece by piece a total of about 100 questions are published in advance. Each student has to answer a subset of questions chosen from the entire question pool, which is stored in the palmtop’s memory and can be adapted at any time. In addition to these didactic tasks, the PSIPorganiser algorithm (written in TeX, public domain) is automatically implementing the organisation of the many types of instruction to the various subsections of the large (2 x 260) student cohorts. Furthermore, the programme automatically performs sophisticated test evaluations for each MC-question; finally, it automatically prints out official forms required for submission to the licensing agencies (Landesprüfungs-Amt). [email protected]. Department of Physiology, RWTH Aachen, Pauwelsstrasse 30, D 52074 Aachen

P 54-7 USE OF INTERACTIVE COMPUTER SIMULATIONS IN THE PHYSIOLOGY LECTURE A. Deussen, V. Zürich Lectures may be expected to maintain an important role in future curricula as they allow to instruct a large audience. The lecturing platforms, however, may change as multimedia techniques become more widely available. We like to suggest 3 examples of commercially available software that were found helpful in teaching 2nd year medical and dental students. Background: Students at our faculty have 4 lectures of physiology, 1 seminar and one practical course per week. The attempt was made to better integrate the contents of these different teaching/learning modalities. Materials: Simulations were employed in main lectures using a laptop computer in connection with a beamer. The lecturer had full control of the programs used. Simulation I dealt with “Regulation of contractile force in skeletal muscle” using “SimMuscle” (Thieme). Simulations addressed the stimulus-effect relationship, superposition of twin stimuli, and tetanic force development in an isolated frog muscle preparation. Simulation II dealt with “Regulation of heart rate by neuro-humoral effectors” using “Physiology Labs” (Critical Concepts, Inc.). Simulations addressed the effects of block of adrenergic and cholinergic transmission in intact humans. The same software was used for simulation III which dealt with “Circulatory effects of sympathomimetics”. Effects of αand β-adrenergic stimuli on cardiac and circulatory parameters were evaluated including endogenous reflex effects. Results: The students have very positively responded to these simulation exercises. They have found it helpful in assisting their thinking and have been stimulated by the directness of the simulation rather than being exposed to diagrams. The students have used “data” obtained during these simulations to calculate the “experimental” results and discuss them on their own. However, only in very limited extent the students have been challenged to use simulation programs from our computer pool. Conclusions: Using simulations in the lecture theatre may be helpful in facilitating the mediation of teaching contents. It may better prepare for practical courses and allow to address teaching contents that cannot be addressed

S 366 otherwise with a similar directness. Institut für Physiologie, Medizinische Fakultät Carl Gustav Carus, TU Dresden, Fetscherstr. 74, 01307 Dresden

pathophysiology by summer 2003, and to make them available to interested faculties thereafter. Institut für Herz- und Kreislaufphysiologie, Heinrich-Heine-Universität, Postfach 101007, 40001 Düsseldorf, Germany

P 54-8 P 54-10 NEW MEDIA IN MEDICINE – CURRICULAR AND EXTRACURRICULAR STUDIES M. Illert, G. Karsten, C. Pfeiffer1, P. Scheid2, A. Schill3 The aim of this joint project of the Physiological Departments of the Universities Kiel, Rostock and Bochum and the Department of System Architecture of the University of Dresden are development, implementation and evaluation of a teleteaching/telelearning platform in medicine. The platform will be integrated in the curricular studies of the pre-clinical stage of medicine and dentistry. Furthermore it will support physicians during lifelong learning. The object oriented database of the platform will contain reusable multimedia based high quality knowledge objects. Emphasis is put on the field of physiology and content related clinical and pre-clinical fields. However, the concept allows the integration of objects from the complete field of medicine and beyond. The platform enables attendance teaching, synchronous and asynchronous distance teaching as well as tutor lead and selfdetermined learning. Tools like chat, black board and white board function are integrated means for communication between students and student and tutor. Evaluation of the platform by the Department of Science Education (University of Kiel) will not only ensure its quality with regard to didactics but will give deeper insight into cognitive processes during multimedia based learning. Further information on this project is given on the homepage: http://mailserv.physiologie.unikiel.de/teleteaching/ This project is supported by the Federal Ministry of Education and Research (BMBF) within the scope of the “Zukunftsinvestitionsprogramm”. Department of Physiology, Christian-Albrechts-University of Kiel, Olshausenstr. 40, D-24098 Kiel, Germany; 1 Department of Physiology, University of Rostock, Gertrudenstraße 9, D-18057 Rostock, Germany; 2 Department of Physiology, Ruhr-University of Bochum, Universitätsstr. 150, D-44780 Bochum, Germany; 3 Department of System Architecture, TU Dresden, Hans-Grundig-Straße 25, D-01062 Dresden, Germany. P 54-9 DEVELOPING AN INTRANET-BASED LEARNING SYSTEM FOR CARDIOVASCULAR PHYSIOLOGY AND MEDICINE V. Laudahn, M. Heydthausen, A. Deussen, J. Schrader, U.K.M. Decking In recent years, isolated internet-based teaching applications were developed at several sites. This was facilitated by technical advances, such as increasing speed of data transmission and better access to true multimedia production sites within universities. To overcome the rather fragmented structure of current applications and to bridge the gap between preclinical and clinical teaching, the Multimedia Centre of the Medical Faculty of Düsseldorf initiated the collaboration of 11 Departments of Physiology, Pharmacology, Cardiology and Cardiothoracic Surgery from 4 German universities (Düsseldorf, Dresden, Hamburg, Köln). Focusing on coronary artery disease (CAD) a teaching and learning system is being developed that will present current and precise information on cardiovascular physiology and pathophysiology of ischemic heart disease, as a basis to the diagnosis and treatment of CAD. This system will provide user-specific information. It is intended to accompany the current faculty curriculum, but will also provide knowledge required for further specialization in cardiovascular medicine. To achieve this goal, so far 400 individual modules (sections) have been defined that are accessible to the individual user according to his current knowledge and his learning objectives. Within cardiovascular physiology and pathophysiology, we have identified 104 modules, each requiring about 15 min to study. Each module is written by a system author in XML, a recently developed internet language, which enables the integration of the different modules in a common database and facilitates the embedding of media (3D animations, audio, video). This may prove especially valuable in the teaching of dynamic processes, e.g. ionic currents, flow, cardiac motion, ECG. A first set of modules will be evaluated by medical students in spring 2002, and will accompany the lab course in physiology. The current goal is to complete the units on cardiovascular physiology and

APPLYING A FACE-RECOGNITION ALGORITHM TO THE SERIAL RECONSTRUCTION OF ELECTRON-MICROSCOPIC IMAGES C. Kayser, V. Bonin, R. P. Würtz, P. König Evaluation of electron microscopic images of serial sections is a timeconsuming process requiring a high level of expertise. Here we present an algorithm to ease and accelerate this process. It is a modification of an algorithm successfully used in computer vision for face recognition. However, rather than recognizing individual structures we estimate the spatial mapping of a whole section onto the consecutive one. This mapping is used to transfer labeled information of the very first section, e.g. a classification by a human expert of different visible structures, onto structures visible in the next section. We investigate its performance on an artificially constructed benchmark as well as on real electron microscopic samples taken in primary visual cortex and demonstrate its potential for dramatically facilitating the evaluation process of serial sections. Supported by ZNZ & SNF (31-65415.01). Institute of Neuroinformatics, Uni/ETH Zürich, Winterthurerstr. 190, 8057 Zürich, Switzerland P 54-11 AUTUMN ACADEMY OFFERING SCIENTIFIC KNOWLEDGE AND SKILLS FOR STUDIES OF INTEGRATED SYSTEMS IN MAN M.E. Schläfke, T. Schäfer Medical research involving human beings is subjected to special requirements with respect to ethical aspects on one side and to special methodological approaches, measures and skills on the other side. Meanwhile recent developments of non-invasive or minimally invasive techniques, as well as computational features allow for quantitative analyses of integrated systems and system interactions in man. On behalf of the Special Interest Group ”Applied, Clinical, and PathoPhysiology” (AKP) in the German Physiological Society, we are preparing a four days workshop entitled ”How to study integrative systems in man”, taking place in Bochum in October, 2002. The academy will feature both, basic and advanced education of the theoretical basis including ethical guidelines and requests to Ethics committees, different designs of studies in man including German and EU guidelines for clinical studies, the issues of ”Good Clinical Practice” and ”Good Scientific Practice”, statistical, epidemiological and other mathematical tools including linear and non-linear algebra and synergetics, and new methods of data acquisition. These issues of knowledge acquisition are completed by the training of special skills including noninvasive measurement techniques such as cardio-vascular and neuroelectrophysiology, respiratory and blood gas measurements, quantification of blood pressure and microcirculation, assessment of vigilance and reactivity. The academy addresses at doctoral fellows as well as at basic scientists and clinicians involved in physiological studies in man offering advanced education and professional development. Finally, a structured evaluation process will be applied to analyze the effectiveness of this approach, which is considered as a new, regular institution of the AKP in the German Physiological Society. Dept. of Applied Physiology, Ruhr-Universität Bochum, D-44780 Bochum; [email protected] P 55-1 THE ENDOGENOUS PENTAPEPTIDE QYNAD INDUCES ACUTE NERVE CONDUCTION BLOCK IN VITRO F. Weber a,b, R. Rüdel b, P. Aulkemeyerb, H. Brinkmeier b Reversible block of sodium channels by endogenous substances was claimed to account for the fast relapses and remissions seen in demyelinating autoimmune disorders. The pentapeptide QYNAD, isolated from the CSF of MS or GBS patients, blocked Na+ channels in various

S 367 types of cultured cells. To explain the fast symptom changes in MS and GBS, we extended our studies of QYNAD effects from those on sodium channels functionally expressed in test cells to those on the compound nerve action potential (CNAP) of isolated rat sciatic nerve, a model that is closer to the in vivo situation. 100 µM QYNAD bath-applied to isolated rat sciatic nerve causes a decrease of amplitude and area as well as an increase in CNAP latency. After 60 min the decline of amplitude of the action potential caused by QYNAD was more than 50 %. On average, it took about 20 min of incubation with 100 µM QYNAD to see an inhibitory effect on amplitude and area of CNAPs and an increase of latency. After 60 min, the amplitude and the area were decreased to 61 and 68 %, respectively, of their initial values, whereas the latency was increased to 130 %. Washout reversed the changes in part. The endogenous pentapeptide QYNAD may have an anti-excitatory effect on intact myelinated axons and nerves in vivo and may contribute to the fast symptom changes in MS and GBS. a

German Air Force Institute of Aviation Medicine, Department of Neurology, PO Box 1264 KFL, D-82242 Fürstenfeldbruck, Germany, b Department of General Physiology, University of Ulm, D-89069 Ulm, Germany P 55-2 CONCENTRATION OF THE NA+ CHANNEL BLOCKER QYNAD DETERMINED IN CSF AND SERUM OF GBS PATIENTS USING ION TRAP MASS SPECTROSCOPY H. Brinkmeier1, P. Aulkemeyer1, K. Bechter2, K.H. Wollinsky3, R. Rüdel1 Ion trap mass spectroscopy allows one to determine with high specificity and at picomolar resolution mass, amino acid sequence and concentration of small peptides in body fluids. An enormous advantage of the method is that no extensive purification of the tested fluids is required. We have used this method to determine more exactly the concentration of QYNAD, a pentapeptide with Na+ channel blocking properties that we had discovered in samples of CSF from patients with GuillainBarré syndrome (Brinkmeier et al., Nature Med. 6:808-811, 2000). For one CSF sample, in which, using HPLC methods, we had previously determined 44 µM QYNAD, the ion trap value was 59 µM. In five newly obtained samples, the concentration varied between 25 and 323 µM. In all eight control samples we found no QYNAD at all, i.e. our previously determined HPLC value of 3.2 µM for the same set of samples had been exaggerated by background. The importance of this finding is that we now know that QYNAD is much more specific than previously thought. Sera from the same six GBS patients, drawn at the same time as the CSF, all contained QYNAD although at about half the concentration (10-164 µM). This suggests that QYNAD readily passes the blood-CSF barrier, but nothing can be said about the real gradient because QYNAD binds to serum albumin (Aulkemeyer et al. Neurosci Lett 216:37-40, 1996). The sera and CSFs also contained to a lesser extent pyroQYNAD, a decay product of the pentapeptide. The CSFs from six psychiatric patients suffering from schizophrenia or affective psychoses were also tested none of them contained QYNAD. Supported by DFG, Br-1139/4. 1 Dept.of General Physiology, University of Ulm, D-89069 Ulm, 2Dept of Psychiatry II, Regional Hospital Günzburg, D89301 Günzburg, 3Rehabilitation Hospital Ulm, Albert-Einstein-Alle, D-89081 Ulm

P 55-3 TESTING REFLEX PATHWAYS IN THE CHRONIC LOW-DOSE MPTP-MODEL OF PARKINSON’S DISEASE IN THE CAT F. Steigerwald, H. Kümmel, M. Gerlach, G. Deuschl, M. Illert H-reflex studies in human Parkinson Disease patients suggested an altered excitability of Ia- and I b-Interneurons in the spinal cord reflex pathways. These changes correlated with the intensity of rigidity and were therefore hypothesised to represent an electrophysiological basic mechanism of the parkinsonian symptom rigor (Delwaide et al. 2000). To test the hypothesis of altered descending control of spinal reflex pathways in Parkinson Disease, we investigated spinal reflexes from Ia, Ib- and flexor reflex - afferents in a new-developed model of Parkinson’s Disease in the cat by a conditioning-test paradigm. Application of lowdose MPTP (2.0mg/kg bodyweight) over 3 weeks was used to gradually induce an akinetic state. In these cats, anaesthetised by chloralose, monosynaptic test reflexes from the nerves to the M. gastrocnemius or the M. biceps femoris and M. semitendinous were recorded in the

ventral roots. Conditioning stimuli were applied to the nerve of the M. flexor digitorum longus at 1.5-, 2-, 5- and 10- times stimulation threshold. The effects of stimulation at different conditioning-test time intervals were compared in acinetic and normal control cats. First results did not reproduce the lowered excitability of I b-Interneurons seen in reflex conditioning studies of human Parkinson Disease patients. Furthermore, the study of reflex conditioning via the flexor reflex afferents showed no significant changes between acinetic and normal control cats. Physiologisches Institut der Christian-Albrechts-Universität zu Kiel, Ohlshausenstraße 40, D-24098 Kiel P 55-4 MOTOR RECOVERY IN CHILDREN AFTER TRAUMATIC BRAIN INJURY: HAND FUNCTION AND CLINICAL ASSESSMENT M. Gölge 1, B. Hoppe 1, J.P. Kuhtz-Buschbeck1, M. Dreesmann 1, M. Müller1, B. Benz2, A. Ritz2, M. Illert1 In a longitudinal study in children we investigated restitution of motor functions after traumatic brain injury. 19 children (age: 5-15 years, mean: 10 years), who were admitted to a rehabilitation center, were selected by the grade of brain injury (≥ II°). Patients were examined four times; the first date (T0) was defined by the Barthel Index, further examinations followed after one (T1), two (T2) and five months (T3). The Glasgow Coma Scale and the Injury Severity Score were chosen to evaluate the trauma severity. Neurological outcome was measured by the clinical examinations, Rappaport Disability Rating Scale, Functional Independence Measure and Gross Motor Function Measure (GMFM). To evaluate restitution of hand function, standardised tests were used: the Jebsen-Taylor Test of hand function and the Purdue Pegboard. The aim of this study was to examine, if these methods are sensitive enough to assess motor recovery, and whether different motor functions restitute in parallel. All three tested clinical scores were appropriate to describe the restitution of motor performance in children with TBI. Hand motor skills and co-ordination improved in children with traumatic brain injury. The total duration of the Jebsen-Taylor Test of hand function decreased significantly. Further the median score for the Purdue Pegboard increased significantly. Taken together, parallel and significant improvements of both, the activities of daily living and the more specific hand functions could be observed throughout the rehabilitation period. 1

Department of Physiology, Christian-Albrechts-University, Hermann Rodewald Str. 5, 24098 Kiel, Germany. 2Neurological Rehabilitation Centre Friedehorst, Rotdornallee 64, 28717 Bremen, Germany P 55-5 SPREADING DEPRESSIONS IN DYSPLASTIC NEOCORTICAL TISSUE ORIGINATE FROM PARA-MICROGYRAL AREAS R. Köhling1, U.-R. Koch1, C. Redecker2, G. Hagemann2, U. Mußhoff1, E.-J. Speckmann 1 Freeze-lesioning of the neocortex of neonatal rats constitutes a model for cortical dysplasia. It results in the formation of a sulcus 1-2 mm in diameter consisting of only the upper three layers of neocortex, which persists during the animal´s natural life span. In this sulcus, but also in adjacent cortex, distinct changes in e.g. network excitability and receptor composition occur. To test whether these alterations make dysplastic cortex vulnerable also to the loss of extracellular ion homeostasis, the generation, and origins of spreading depression reactions were monitored in dysplastic neocortical slices of rats, using tissue sham animals as controls. Experiments were carried out on 28 neocortical slices of adult rats (9 dysplastic, 7 sham). Spreading depression (SD) reactions, if not occurring spontaneously, were elicited by superfusion of 5 µM bicuculline. Using 4 extracellular electrodes (12 MΩ) placed in variable intervals alongside layers II/III, the origin of of SD was calculated assuming constant spreading velocities. In some experiments, neuronal and glial reactions were recorded using sharp microelectrodes, and extracellular K+ was monitored with ion-selective electrodes. In non-dysplastic tissue, SD never appeared spontaneously, but could be elicited in 64% (total n=14) slices. SD originated in all cortical areas (frontal, hindlimb, Par1 and 2) with no significant preference for any region. By constrast, in dysplastic tissue, SD appeared in 86% (total n=14) slices, either spontaneously (29%) or

S 368 elicited by bicuculline (57%). They preferentially were generated in para-dysplastic areas, corresponding to frontal or medial Par1 regions. Changes in K+ were larger in dysplastic cortex and areas where SD were initiated as compared to areas merely invaded by SD. In conclusion, dysplastic tissue appears to be predisposed also to disturbances of extracellular ion homeostasis as reflected by SD. Supported by IZKF grant G7. 1

Institut für Physiologie, Westfälische Wilhelms Universität Münster, and 2Klinik für Neurologie, Universität Jena, Germany

P 55-6 IMPACT OF LEUCINE AND α-KETOISOCAPROATE ON NEURONAL NETWORK ACTIVITY DETECTABLE WITH MULTI ELECTRODE ARRAYS: IMPLICATIONS FOR ACUTE ENCEPHALOPATHY IN MAPLE SYRUP URINE DISEASE P. Görtz, B. Schwahn, U. Wendel, M. Siebler Increased concentrations of leucine and its ketoacid α-ketoisocaproate (KIC) in plasma and cerebrospinal fluid (CSF) are related to an acute encephalopathy in patients with Maple Syrup Urine Disease (MSUD). In order to explore the neurophysiological mode of action, we investigated the influence of leucine and KIC on electrophysiological properties of primary dissociated rat neurons in the concentration range [1–10 mmol/L] found in patients suffering from acute encephalopathy. Primary dissociated embryonal neurons of Wistar rats growing on multi electrode arrays (MEA) (60 channels, NMI Reutlingen) display network-driven activity patterns characterized by synchronized network bursts and correlations between different electrodes, detectable from day 4 in vitro and lasting for more than 60 days on MEAs. This neuronal network activity, which includes presynaptic and postsynaptic mechanisms, was reversibly reduced in a dose-dependent manner by leucine and KIC. Furthermore application of both agents resulted in a partial compensation of their effects (see table). By means of patch-clamp technique we could show that primary postsynaptic mechanisms were unaffected: The neuronal resting potential, the voltage dependent Na +- and K +-currents, the GABA, glutamate, and the glycine membrane currents were constant during incubation with leucine and KIC [10 mmol/L]. These data suggest that the impact of leucine and KIC on the neuronal network activity may be explained by an imbalance in the presynaptic neurotransmitter release. mmol/l ∆DF

KIC 1

KIC 5

KIC 10

leucine 1

leucine 10

28% ± 4 % (n=4)

50% ± 13 % (n=3)

total suppression (n=3)

51%±11% (n=3)

total suppression (n=3)

leucine + KIC 5 + 5 26 %± 14 % (n=3)

Table: Elevated levels of leucine or KIC decreased the depolarization frequency (DF) of a neuronal network in a dose dependent manner. Application of both resulted in a partial compensation of their effect. Neurology Department, Heinrich-Heine-University of Düsseldorf, Moorenstr. 5, 40225 Düsseldorf

P 55-7 RELEASE OF HISTAMINE FROM MAST CELLS OF CRANIAL DURA MATER – INVOLVEMENT IN PATHOPHYSIOLOGY OF MIGRAINE K. Meßlinger, N. Bernardini, M. Dux In patients suffering from migraine, histamine (HA) can elicit migrainelike headache attacks. Increased plasma levels of HA have been found in these patients, most likely derived from mast cells (MCs). Experimental stimulation of trigeminal afferents have shown signs of MC degranulation in rat dura mater encephali. We asked therefore which stimuli are able to release HA by degranulation of dural MCs in an in vitro preparation of rat meninges. To demonstrate HA and degranulation of dural MCs we employed immunohistochemical and electron microscopical (EM) methods. For measurements of HA release we used a competitive ELISA. Hemisected rat skulls devoid of brain but with intact dura mater were rinsed in oxygenated artificial interstitial fluid, mounted in a humid atmosphere of 37°C and filled with fluid for periods of 10 min. After two control periods, solutions of MC degranulators, excitatory substances or neuropeptides were applied. The HA concentration was assessed in the superfusates, and samples

of dura mater were processed for histological examination. Without stimulation MCs in the dura mater appeared intact. The basal HA concentration in the superfusate was around 12 ng/ml. Compound 48/ 80 (2.5 µg/ml) increased HA release by 18.5, codeine (0.3 mg/ml) by 0.8, low pH (5.7-4.7) by 2.3-2.8 times of the basal value. MCs in the dura treated with these substances showed low HA immunoreactivity and appeared degranulated in the EM. Capsaicin (10-5 M) and donators of nitric oxide (10-4 M) did not elevate HA release. CGRP (10-5 M) increased HA release by 2.7 times, while substance P and endothelin (both 10 -5 M) were ineffective. The results show that low pH and CGRP are effective MC degranulators in the dura mater. Plasma CGRP (but not substance P) levels have been shown to be increased in attacks of migraine, while pH lowering may be common in meningeal disorders. HA release may thus accompany these conditions aggravating nociceptive processes that lead to severe headaches such as migraine pain. Supported by the DFG (SFB 353) and OTKA T031894. Department of Physiology & Experimental Pathophysiology, University of Erlangen-Nürnberg, Universitätsstraße 17, D-91054 Erlangen P 55-8 COMPARATIVE INVESTIGATION OF CENTRAL AND PERIPHERAL NEUROTOXIC CHANGES CAUSED BY SINGLE OR COMBINED TREATMENT BY ENVIRONMENTAL XENOBIOTICS IN RATS L. Nagymajtényi, T. Vezér, A. Papp Heavy metals (Pb, Hg, Cd, Zn) in inorganic forms are known environmental pollutants causing biochemical, functional and morphological alterations of the central and peripheral nervous system. The aim of this study was a comparative analysis of the changes of some neurophysiological parameters of rats subchronically administered by relatively low doses of lead, mercury, cadmium and zinc. 12 week old male Wistar rats were treated, per os by gavage, for 4, 8 or 12 weeks with 80.0 and 320 mg/kg Pb, 0.4 and 16 mg/kg mercury, 3.5 and 14.0 mg/ kg of cadmium, 1.87 and 6.48 mg/kg zinc (in form of C4H6O4Pb .3H2O, HgCl2, CdCl2, ZnCl2, respectively). The analyzed neurophysiological parameters were the ECoG, cortical evoked potentials, conduction velocity of peripheral nerve. The administered metals caused dose-, and treatment duration-dependently alterations of the investigated functions without any clinical signs of the chronic metal intoxication. The index (ratio of the slow and fast frequency wave bands) of EcoG in the somatosensory, visual and auditory areas was decreased, the latency and duration of the somatosensory, visual and auditory evoked potentials was lengthened, the conduction velocity of the peripheral nerve was lowered. Although, compared to the control, the changes were singificant mostly in the groups given the higher dose for 12 weeks only, the alterations showed a clear biological trends in every case. On the basis of this fact it can be supposed that these inorganic metals, especially in case of long lasting exposure, can also represent a real functional risk for human beings. Univ. of Szeged, Dept. Public Health, H-6720 Szeged, Dóm tér 10. Hungary P 55-9 MARKED RESIDUAL EFFECT OF SUBCHRONIC ORGANOPHOSPHATE TREATMENT SEEN ON SPONTANEOUS AND EVOKED CORTICAL ACTIVITY IN RATS A. Papp, L. Pecze, L. Nagymajtényi Organophosphates (OPs) are neurotoxic substances widely used as insecticide agents. Their principal way of action is irreversible block of acetylcholinesterase (AChE). OPs are known to induce characteristic changes in spontaneous and evoked cortical activity in humans and animals. It is still an open question, however, if inhibition of AChE is alone responsible for the electrophysiological effects seen after OP treatment. In our previous work it was found that different OPs, administered per os for up to 12 weeks to young adult Wistar rats, induce reproducible and often significant changes in the spontaneous activity (electrocorticogram, ECoG) and sensory evoked potentials (EPs) recorded from the primary sensory areas (somatosensory, visual, auditory) of rat cortex. It was supposed that by determining how long the changes last after cessation of the treatment, a valuable information can be obtained for the future identification of

S 369 the underlying biochemical mechanisms. The rats (10/group) were treated from their 12th to 20th or 24th week of life with 1/25 and 1/100 LD50 (7.0 and 28.0 mg/kg b.w.) of dimethoate, an OP. Control groups received distilled water. Following the 8 or 12 weeks treatment, half of the groups was kept for another 4 weeks. Electrophysiological recording was done from the exposed cortical surface in the animals anesthetized with urethane. ECoG was recorded for 5 min. The relative power of the standard frequency bands was calculated and plotted. EPs were evaluated by their latency. ECoGs showed a shift to higher frequencies which was slight after 8 but significant after 12 weeks and was dose dependent. Records taken after 4 weeks without treatment were between the controls and the corresponding treated group. The same was seen on the latency (and duration) increase of the sensory EPs after 12 and 12+4 weeks. Those after 8 and 8+4 weeks were inconclusive. Subchronic dimethoate treatment had a lasting effect on cortical electrophysiology, possibly by modifying a long-lived biomolecule. Univ. of Szeged, Dept. Public Health. H-6720 Szeged, Dóm tér 10. Hungary

P 55-10 SUBCHRONIC EXPOSURE BY METHYL MERCURY IN MALE WISTAR RATS: EFFECTS ON NEUROBEHAVIORAL PERFORMANCE T. Vezér, A. Papp, L. Nagymajtényi Human mercury exposure arises from industry, from dental amalgam, from the use of mercurials as fungicides and bactericides and also from food via mercury compounds accumulating in the food-chain. Being lipid soluble, methyl mercury permeates the placenta, the blood-brain barrier and cell membranes. Being neurotoxic, mercury alters the number and structure of synapses, resulting in altered behavioural, cognitive, coordinative and emotional functions of humans and animals. In the present experiment, spatial learning ability (in an 8-arm radial maze) locomotor activity and psychomotor performance was studied in young adult male Wistar rats (10/group) treated subchronically with low-level methylmercury (II) chloride (0.5 and 2.0 mg/kg b.w.) per os by gavage. At the end of the treatment, it was also investigated what heavy metalspecific functional neurotoxicological changes are seen in the central nervous system generally (spontaneous cortical activity), and in centres with special role in learning (hippocampus CA1 region). During acquisition in the maze test, the treated and control animals showed dissimilar memory performance. On the 2nd and 4th days of acquisition and from the 13 th day of behavioural investigations (short-term retention tests) a significant memory deficit developed in the treated groups. In the long term retention test the treated groups’ performance decreased by further ca. 15 %. In the open field tests performed in 10 min sessions in the 5 th and 10th week of treatment, decreased locomotor activity and diminished spontaneous exploration was seen. MEM-treatment caused a moderate shift in the distribution of spontaneous cortical and hippocampal activity to higher frequencies, most markedly in the primary visual field and in the hippocampal CA1. In the latter, there was some increase also in the slowest, delta, waves. The changes seen in the methyl mercury-treated rats were similar to those found in exposed humans which stresses the importance of monitoring and control of direct and indirect environmental exposure. Univ. of Szeged, Dept. Public Health. H-6720 Szeged, Dóm tér 10. Hungary

P 55-11 GENETIC ALTERATIONS IN THE SDH GENES LEAD TO ONCOGENESIS OF PARAGANGLIOMAS S. Braun, K. Riemann, M. Pfister, K. Sotlar, N. Blin, C. Pusch, S. Kupka Paragangliomas of the head and neck region are usually beninge tumors developing from chemoreceptors of paraganglionic origin in the majority of patients. These receptors play an important role in sensing and regulation of the blood CO 2 -level. Genetic alterations in the mitochondrial enzyme complex II (SDH), which is involved in respiratory chain and citrat cycle reactions, have been shown to lead to sporadic as well as familiar cases of these tumors. Therefore we analyzed our collective containing sporadic cases of patients with paragangliomas for genetic changes in the SDH-genes SDHD, SDHC and SDHB. We

detected several new DNA mutations in samples derived from tumor patients. Furthermore we demonstrated loss of heterozygosity (LOH) in these tumor samples, which was connected with oncogenesis of paragangliomas. Elucidation of the genetic regions involved in tumor development is a basis for understanding their contribution to normal and pathogenic cellphysiology. Department of Anthropology and Human Genetics, University of Tübingen, Germany

S 01-1 FUSION PORES IN THE CONTROL OF SURFACTANT RELEASE FROM TYPE II PNEUMOCYTES T. Haller, P. Dietl, K. Pfaller, K. Maly In alveolar type II cells, the release of surfactant is considerably delayed after the formation of exocytotic fusion pores, suggesting that content dispersal may be limited by fusion pore diameter and subject to regulation at a postfusion level. To address this issue, we used confocal FRAP and FM 1–43, a dye yielding intense localized fluorescence of surfactant when entering the vesicle lumen through the fusion pore (Haller et al., Proc. Natl. Acad. Sci. USA, 1998, 95, 1579– 1584). Thus, we have been able to monitor the dynamics of individual fusion pores up to hours in intact cells, and to calculate pore diameters using a diffusion model derived from Fick’s law. After formation, fusion pores were arrested in a state impeding the release of vesicle contents, and expanded at irregular times thereafter. The expansion rate of initial pores and the probability of late expansions were increased by elevation of the cytoplasmic Ca 2+ concentration. Consistently, content release correlated with the occurrence of Ca2+ oscillations in ATP–treated cells, and expanded fusion pores were detectable by EM. This study supports a new concept in exocytosis, implicating fusion pores in the regulation of content release for extended periods after initial formation. Department of Physiology, University of Innsbruck, Fritz-Pregl-Str. 3, A-6020 Innsbruck

S 02-1 ASTROCYTE-NEURONE SIGNALING AND SPONTANEOUS CALCIUM OSCILLATIONS V. Crunelli, H.R. Parri Changes in astrocytic [Ca 2+]i are a form of excitability that can be propagated to other astrocytes via gap junctional or chemically mediated mechanisms. In culture studies, chemical and mechanical stimulation of astrocytes has been shown to lead to transient [Ca2+]i increases and glutamate release via a probable exocytotic mechanism, while astrocytic stimulation in in situ preparations has been shown to modulate neuronal activity. In acute slices of neonatal rat ventrobasal thalamus (VB) we find that astrocytes respond to a variety of thalamic neurotransmitter agonists, including glutamate, GABA and noradrenaline, with [Ca 2+] i increases. Trains of electrical stimuli delivered to either the sensory or cortical afferents also cause astrocytic [Ca 2+ ] i increases, indicating activation of astrocytic receptors by synaptically released transmitters. In addition, we observed thalamic astrocytes that could display spontaneous [Ca 2+ ] i oscillations. Spontaneous oscillations are due to [Ca2+]i release from intracellular stores and are independent of neuronal activity since they are not blocked by TTX. Groups of astrocytes can display correlated [Ca2+]i increases, and oscillations can propagate to adjacent astrocytes, via a probable chemical mechanism. Recordings from thalamocortical (TC) neurones in the VB reveal large long-lasting (τdecay~2.7s) inward currents which are correlated to propagating [Ca2+]i waves in adjacent astrocytes. These inward currents are blocked by APV, and so are due to NMDA receptor activation. The incidence of the spontaneous astrocytic oscillations decreases after P9, so that the main expression of spontaneous activity corresponds to a time of TC neurone dendritic arborisation and thalamic synaptogenesis. These findings therefore suggest a developmental role for spontaneous astrocytic [Ca 2+ ] i oscillations and the neuronal currents that they induce. School of Biosciences, Cardiff University, Museum Avenue, Cardiff, CF10 3US, Wales, UK

S 370 S 02-2 RAPID SIGNALING BETWEEN AXONS AND SCHWANN CELLS VIA P2 NUCLEOTIDE RECEPTORS? P. Grafe In a peripheral nerve trunk, several unmyelinated axons are in close contact with a single ensheathing Schwann cell. It is likely, therefore, that rapid signaling exists between these two cell types such as between glia and neurons in the central nervous system. However, at present, knowledge about such interactions is fragmentary mainly due to the fact that unmyelinated axons and Schwann cells in an intact nerve preparation are not accessible by conventional electrophysiological techniques. In the last few years, we have used confocal Ca2+ imaging, threshold tracking, and K + - sensitive micro-electrodes in order to study receptors for neuroligands on axons and Schwann cells suitable for rapid transcellular signaling. Thereby, the presence of at least four different ionotropic and metabotropic P2 nucleotide was observed. Activation of ATP receptors on Schwann cells results in intracellular Ca 2+ transients (P2Y 1, P2Y2) or Schwann cell depolarisation (P2X7). On unmyelinated axons, an ATP-induced increase in excitability was found and attributed to P2X and/or adenosine A2B receptors. It is interesting that activation of axonal P2X receptors results in an increase of extracellular K+ activity. Recent observations indicate a function of these receptors in signaling from axons to Schwann cells or in the opposite direction. Strong evidence for the first mechanism was found in co-cultures: ATP released during trains of action potentials from axons was found to activate P2Y receptors on Schwann cells (Fields & Stevens; TINS 23, 2000). Conversely, ATP-induced Ca2+ transients in Schwann cells seem to release compounds able to modify axonal excitability. Possible mediators are NO, prostaglandins, cytokines, and/ or neuropeptides. Such signaling via P2 nucleotide receptors may contribute to transduction of nociceptive stimuli, mechanosensitivity, inflammation, gene expression and/or cell differentiation. Supported by the Deutsche Forschungsgemeinschaft. Department of Physiology, University of Munich, Pettenkoferstr. 12, 80336 Munich, Germany S 02-3 NEURON-GLIA SIGNALING VIA α 1-ADRENOCEPTORMEDIATED [Ca2+]i RELEASE IN BERGMANN GLIA IN SITU K. Ballanyi Simultaneous recording of [Ca2+]i and membrane potential or current was done on superficial Bergmann glial cells of cerebellar slices from juvenile mice. [Ca2+]i was microfluorometrically determined with 100 µM Fura-2, added to the patch pipette solution. Repetitive electrical stimulation with bipolar electrodes (20-100 V, to-50 Hz, 2 s) in the granular cell layer caused an inward current of up to -400 pA or a depolarization by up to 25 mV and a concomitant [Ca2+]i rise by -100 nM. No difference in the amplitude of the [Ca2+]i response was observed under current-clamp or voltage-clamp. Block of synaptic transmission with TTX (1 µM) or Ca 2+-free superfusate suppressed the stimulusevoked inward current and abolished the [Ca2+]i transient. A mixture of the ionotropic glutamate receptor blockers CNQX (20 µM) and APV (100 µM) had no effect on the evoked [Ca2+]i rise, but attenuated the current response. Bath-application of norepinephrine (10 µM) mimicked the stimulus-evoked elevation of [Ca 2+]i. Depletion of intracellular Ca 2+ stores with the endoplasmatic Ca 2+ pump blocker cyclopiazonic acid (30 µM) diminished the stimulus- and noradrenalineinduced [Ca2+]i rises while the current was not affected. The Ca2+ transients were blocked by the α 1-adrenoceptor antagonist prazosine (2 µM), whereas antagonists of α2- (yohimbine, 5 µM) and β-adrenoceptors (propranolol, 5 µM) had no effect. The results show that synaptic activity increases glial [Ca2+]i due to α1-adrenoceptor mediated Ca2+ release from intracellular stores. Perinatal Research Centre & Departments of Physiology and Pediatrics, University of Alberta, Edmonton, CanadaT6G, 2S2

triggers neuronal death. In simulated ischaemia of hippocampal slices, using receptors in CA1 pyramidal cells to sense released glutamate, we have previously shown that most glutamate release is by a reversal of glutamate uptake, caused by the rundown of ion gradients which occurs when [ATP] falls and the Na/K pump is inhibited (Rossi et al. (2000) Nature 403, 316). Here we investigate whether neuronal or glial transporters release glutamate, and whether Na/K pump inhibition alone determines glutamate dynamics in early ischaemia. Blocking the glial transporter GLT-1 with dihydrokainate, or knocking it out in transgenic mice, had no significant effect on the rise of [glutamate]o in ischaemia, suggesting that glutamate release is by a neuronal transporter (possibly an as yet uncloned presynaptic transporter) and that glia do not take up significant glutamate in ischaemia. Reversal of neuronal transporters may occur more easily than reversal of glial transporters because [glutamate] i is normally lower in glia due to glutamate conversion to glutamine by glutamine synthetase. In non-ischaemic solution, blocking the Na/K pump with ouabain produced a large transient inward current like the glutamate-mediated current underlying the start of the anoxic depolarization in ischaemia, but produced less sustained [glutamate]o rise than did ischaemia. However, in the presence of methionine sulfoximine to block glial glutamine synthetase, ouabain produced a maintained [glutamate]o rise more similar to ischaemia. This suggests that inhibition of glutamine synthetase by the fall of ATP occurring in ischaemia may prevent a protective uptake of glutamate into glia in early ischaemia: with glutamine synthetase inhibited, uptake of only a little glutamate released from neurons may be sufficient to raise [glutamate]i in glia sufficiently to inhibit further uptake. Dept Physiology, University College London, Gower St., London WC1E 6BT S 02-5 Ca2+-DEPENDENT GLUTAMATE RELEASE FROM ASTROCYTES: MOLECULAR ASPECTS, PHYSIOLOGICAL AND PATHOLOGICAL RELEVANCE A. Volterra1,2, M. Domercq1, F. Grohovaz3, J. Meldolesi3, P. Bezzi1,2 Astrocytes often ensheath brain synapses with fine processes expressing receptors for neurotransmitters and other mediators. Such astrocyte receptors are in the position to sense neuronal activity and translate it into intracellular calcium ([Ca 2+]i) elevations which, in turn, start local or long-range glial communication, notably by glutamate release. We find that stimulation of astrocyte receptors for the neurotransmitter glutamate (mGluR5 and AMPAR) or the chemokine SDF1α (CXCR4) triggers Ca2+-dependent glutamate release via a process distinct from reversed transport or swelling-induced release. Such process is selectively inhibited by blockers of exocytosis (tetanus neurotoxin and bafilomycin A1) and peculiarly implicates TNFα (in the case of CXCR4) and prostaglandin E2 (PGE2) as necessary, sequential intermediates. TNFα stimulates PGE2 formation acting as an autocrine/ factor, whereas the role of PGE2 is apparently to amplify the [Ca2+]i rise responsible for glutamate release. Indeed, blockade of PGE2 synthesis results in a dramatic reduction of both GluR-dependent [Ca2+]i elevation and glutamate release. In hippocampal slices, PGE2-evoked glutamate release from astrocytes elicits [Ca 2+]i responses in neighbouring neurons, thereby providing a modulatory input to the ongoing activity. In pathological conditions this pathway may become neurotoxic. Thus, when glial cells are ”reactive” and microglia migrates in apposition to astrocytes, CXCR4 stimulation is followed by a significantly higher TNFα production and, as a consequence, by potentiated astrocyte glutamate release, which triggers neuronal apoptosis. This CXCR4-dependent death cascade can be activated by the HIV-1 coat glycoprotein, gp120 IIIB and play a role in the pathogenesis of AIDS dementia. Agents interfering with it provide neuroprotection both in vitro and in vivo. 1 Dept. Pharmacological Sciences, 2 Center of Excellence on Neurodegenerative Pathologies, University of Milan, Via Balzaretti 9, 20133 Milan and 3Department of Neurosciences , HSR, Milan, Italy

S 02-4 S 03-1 THE ROLE OF NEURONAL & GLIAL GLUTAMATE TRANSPORTERS IN EARLY BRAIN ISCHAEMIA D. Attwell, D.J. Rossi, M. Hamann During brain ischaemia, a rise of extracellular glutamate concentration

α-ADRENERGIC CORONARY VASOMOTION – AN EXAMPLE OF COOPERATIVE STUDIES BY BASIC CARDIOVASCULAR PHYSIOLOGISTS AND CLINICAL CARDIOLOGISTS

S 371 R. Schulz, D. Baumgart, R. Erbel, G. Heusch In the early eighties, detailed analyses of regional myocardial blood flow and function were not yet possible in man, but only in animals. We therefore studied the effects of acute cardiac sympathetic nerve activation in anesthetized dogs and found profound α2-adrenergic coronary vasoconstriction resulting in net lactate production and contractile dysfunction distal to severe stenoses (Circ Res 53, 1983,8-15). Subsequently, these findings were confirmed in chronically instrumented conscious dogs during treadmill exercise (Circ Res 62, 1988, 436-442). In the early nineties, methods for detailed analyses of regional myocardial blood flow (Doppler+angiography) and function (transesophageal Echo) in man became available. Studies in patients with significant coronary artery disease confirmed profound α2-adrenergic vasoconstriction resulting in EKG changes and net lactate production (Circulation 99, 1999, 2090-2097). The susceptibility to such α2-adrenergic coronary vasoconstriction has a genetic component (Circ Res 85, 1999, 965-969). α-Adrenergic coronary vasoconstriction also acts to limit myocardial perfusion and contractile function following interventional reperfusion of myocardial infarction; conversely αblockade improves the reperfusion result (Circulation 99, 1999, 482490). Thus, data obtained originally in anesthetized dogs have been successfully translated into patients undergoing coronary catheterization (Circulation 101, 2000, 689-694). Conversely, the primarily clinical observation of coronary microembolization (J Am Coll Cardiol 36, 2000, 22-24) has been successfully translated into an experimental model in anesthetized dogs. Mechanistic studies in this model revealed a marked perfusion-contraction mismatch, significant inflammation (Am J Physiol 279, 2000, H2587-2592) and a causal involvement of TNFα (J Mol Cell Cardiol, in press). In our experience, close communication and active cooperation promotes scientific achievements both in the direction from basic to clinical and back from clinical to basic science. Abteilungen für Pathophysiologie und Kardiologie des Universitätsklinikums Essen, Germany

S 03-3 MODULATION OF POSTERIOR PITUITARY FUNCTION IN HEALTH AND DISEASE M.L. Forsling The stimuli for the release of the posterior pituitary hormones, oxytocin and vasopressin are well established. More recently it has become clear that their release may be modulated by circulating hormones including the gonadal steroids and melatonin. Studies in the rat have shown that ovariectomy reduces basal plasma concentrations of oxytocin and vasopressin, as well as attenuating the response to increased plasma osmolality and decreased plasma volume. The attenuated responses are reflected by altered neuronal activity, expressed in terms of immediate early gene expression in the hypothalamic magnocellular neurones, and reduced hormone synthesis, as indicated by the SON levels of mRNA for the two hormones. The responses can be restored by administration of oestrogen. Changes in plasma concentrations of ovarian steroids have similar effects in the human. Pinealectomy alters the daily rhythm of posterior pituitary hormone release in the rat, hormone concentrations normally peaking at the end of the light phase of the 24h cycle. It also attenuates stimulated hormone release and the accompanying increase in neuronal activity, an effect reversed by melatonin administration. Altered nocturnal patterns of melatonin release in the human also affect the normal nocturnal increase in oxytocin and vasopressin concentrations. Thus posterior pituitary function is affected both by reproductive status and time of day and it has additionally been demonstrated that these two factors interact. Recent work has shown that the effects of these two factors are not confined to release of vasopressin, but also renal responsiveness to the hormone in both the rat and man. Uterine responsiveness to oxytocin is similarly affected. Alterations in the normal patterns of hormone release over reproductive cycles and the day night cycle can lead to a number of disturbances including altered fluid balance. Neuroendocrine Laboratories, New Hunt’s House, GKT School of Medicine, Guy’s Campus, London Bridge London SE1 1UL, UK

S 03-2 IMPACT OF ALTERED FETAL GROWTH ON ORGAN DEVELOPMENT - PATHOPHYSIOLOGICAL IMPLICATIONS ON ARTERIAL HYPERTENSION AND NEURODEVELOPMENTAL DISTURBANCES R. Bauer, B. Walter, U. Zwiener Asymmetrical intrauterine growth restriction (IUGR) results mainly from inadequate nutritional supply due to uteroplacental insufficiency in the late gestation period. Reduced fetal growth can be viewed as a compensatory mechanism in regard to the reduced supply as long as the fetal demand is not critically restricted so that a decompensation with asphyxia and even death can result. The period of well-compensation, characterized by reduced growth due to restricted glucose and amino acid availability but widely compensated placental respiratory function reflects obviously a functional state in the fetus. This is able to induce contradictory effects of acceleration or delay in organ maturation. Therefore, we studied key functions of kidneys, skeletal muscles, and cerebral dopamine metabolism in newborn naturally occurring IUGR piglets in order to clarify whether or not accelerated or disturbed organ development appears. (i) We found an increase in specific force of calf muscles in IUGR piglets compared with normal weight (NW) piglets (P<0.05) and an increased proportion of type I fibers in flexor digitalis superficialis and gastrocnemius medialis in IUGR piglets (P < 0.05). These data clearly indicate that contractile function is accelerated in newborn IUGR piglets. (ii) Furthermore, (18F)-labelled FDOPA was used together with positron –emission tomography (PET) to estimate the activity of the aromatic amino acid decarboxylase (AADC) a key enzyme for cerebral dopamine synthesis in the brain of newborn IUGR piglets. We found that IUGR induces an upregulation of AADC activity (P<0.05), which is not related to alterations in brain oxidative metabolism. (iii) We studied the effects of IUGR on renal structure and function in newborn piglets. We found a reduced nephron number (P<0.01) and a reduced excretory function (GFR ↓, osmotic clearance ↓) (P<0.05) in IUGR piglets, which indicates a markedly disturbed kidney development. Thus, evidence is provided that IUGR induces organ-specific effects of both acceleration or disturbance of structural and functional development.

MUSCLE ACTIVITY AND RADICALS O. Hänninen, M. Atalay, D. Laaksonen, C.K Sen During intensive exercise oxygen consumption increases by around 810 folds, and oxygen flux through the muscle may increase by 90-100 folds. Even moderate exercise increases free radical production and overwhelms antioxidant defenses, resulting in oxidative challenge. Regular exercise, on the other hand, appears to protect against oxidative stress. Our studies indicate that oxidative stress is elevated in both healthy and diabetic subjects both in human and animal models. We have shown increased lipid, protein and DNA markers of oxidative stress in response to exercise in plasma, blood, skeletal muscle, heart muscle and liver. Increased glutathione disulphide or disulphide-total glutathione (GSH) ratio has also been shown to be a sensitive index to exercise induced oxidative stress. A variety of endo- and exogenous antioxidants act in concert to protect against oxidative stress. GSH plays a central role in the synergism. Apart from enzymatic and nonenzymatic decomposition of reactive oxygen species (ROS), GSH in its reduced form is suggested to be implicated in maintaining a favourable redox milieu of crucial antioxidants, vitamins E and C. Vitamin E and ubiquinol are lipid phases antioxidants. Alpha-Lipoic acid (LA) is a natural thiol antioxidant, implicated with the regeneration of major physiological antioxidants such as vitamin E, ubiquinol, ascorbate and glutathione. Our studies have shown that supplementation of vitamin E, LA combined with endurance training or alone, enhanced antioxidant protection and decreased lipid and protein indices of oxidative stress during exhaustive exercise. Regular exercise has potentially enormous therapeutic value in the normal population, although its long term benefit is poorly defined in disease states such as diabetes. The relative benefits and risks of exercise may be markedly different in diabetic patients, having increased oxidative stress due to the disease. Our data showed a protective effect of fitness in young, otherwise healthy IDDM men.

Institut für Pathophysiologie, Klinikum der Friedrich-Schiller-Universität, D-07740 Jena, Germany, ([email protected])

Department of Physiology, University of Kuopio, 70211 Kuopio, Finland

S 03-4

S 372

Institute of Pathophysiology, Martin Luther University HalleWittenberg, Magdeburger Str. 18, D-06097 Halle, Germany

and medullary reticular formation. We have found that approximately 80% of lamina I projection neurons labelled from each of these sites express the NK1 receptor. We have also identified a population of dorsal horn projection neurons with the NK1 receptor which have cell bodies in lamina III or IV and dendrites that enter the superficial laminae. By using combined confocal and electron microscopy in conjunction with retrograde tracing and immunocytochemistry we have found that substance P-containing afferents selectively innervate both lamina I and lamina III/IV projection neurons which express the NK1 receptor. Asymmetrical synapses were found at most of the contacts between the substance P afferents and the cell bodies or dendrites of these projection neurons. Substance P is thought to act through volume transmission, and these synapses are likely to underlie glutamatergic transmission, since glutamate is believed to be the principal neurotransmitter for all primary afferents in the spinal cord. Many C fibres do not contain neuropeptides. These end mainly in lamina II and it is thought that most are nociceptors. These axons can be identified because they bind the lectin Bandeiraea simplicifolia isolectin B4. Even though the dendrites of the lamina III/IV NK1 receptor-immunoreactive neurons pass through lamina II, they appeared to receive few contacts from non-peptidergic C afferents. These findings indicate that there is a powerful and selective monosynaptic connection between substance P-containing nociceptive afferents and projection neuron in the dorsal horn which express the NK1 receptor.

S 03-6

Spinal Cord Group, IBLS, West Medical Building, University of Glasgow, Glasgow G12 8QQ UK

S 03-5 INDUCTION OF OXIDATIVE STRESS IN VASCULAR CELLS: FROM BENCH TO BEDSIDE H. Morawietz Objective of study. Oxidative stress is considered as an important risk factor in the development and progression of endothelial dysfunction and atherosclerosis. A main source of vascular reactive oxygen species is the superoxide anion generating NAD(P)H oxidase. Results We show an upregulation of vascular NAD(P)H expression and superoxide anion formation by proatherosclerotic stimuli like oxidatively modified low density lipoprotein (oxLDL), angiotensin II and endothelin-1 in vitro and in animal models. Furthermore, increased NAD(P)H oxidase activity correlates with cardiovascular risk factors in clinical studies. In addition, we find a downregulation of vascular NAD(P)H oxidase expression by angiotensin II receptor type 1 (AT1) blockade or HMG-CoA reductase inhibitor therapy before coronary bypass surgery in internal mammary arteries of patients with coronary artery disease. Conclusions These data support the importance of experimental studies in the development of new pathophysiological concepts and therapeutic strategies in the treatment of cardiovascular diseases.

INTERCELLULAR COMMUNICATION BETWEEN BONE MARROW STROMAL CELLS AND CD34+ HEMATOPOIETIC PROGENITOR CELLS IS MEDIATED BY CX43 GAP JUNCTIONS J. Dürig*, A. Gellhaus, K. Halfmeyer, M. Wiemann, K. Passarge, K. Schirrmacher, E. Winterhager, D. Bingmann The existence of functional gap junctions between hematopoietic progenitor cells (HPC) and stromal cells of the hematopoietic microenvironment in the human system is still controversial. To address this issue primary CD34+ HPC isolated from leukapheresis products were coincubated with the human fibroblastoid bone marrow stromal cell line L87/4 in short-term liquid culture. Using the highly sensitive double whole cell patch clamp technique we found that the majority (91%) of CD34+ HPC were electrically coupled to L87/4 cells. Importantly, efficient coupling was observed within one hour after the attachment of CD34+ HPC to plastic adherent L87/4 cells. By comparison, homologous cell pairs formed by L87/4 cells exhibited a significantly higher electric coupling. Analysis of single channel conductances revealed an electric profile characteristic of Cx43-gap junctions for both homologous and heterologous cell pairs. The Cxphenotype was confirmed using Cx43 specific monoclonal antibodies in a flow cytometric assay and RT-PCR for the detection of Cx43 mRNA. The electrophysiological studies were further complemented by dye transfer experiments using the recently described „parachute“ technique which allows for the monitoring of dye diffusion without disruption of the plasma membrane. Taken together, our data indicate that functional Cx43 gap junctions exist between stromal cells and immature HPC and thus form a functional syncytium which may contribute to the regulation of blood cell formation in the bone marrow. Furthermore, the functional role of Cx43 in hematopoiesis was recently validated by studies employing a Cx43-deficient mouse model in which defects in embryogenic hematopoiesis and impaired hematopoietic recovery from chemotherapy occurred.

S 04-2 WIND-UP IN THE NOCICEPTIVE SYSTEM AND THE SPINAL MEMORY OF PAIN J. A. Lopez-Garcia Wind-up was first described as a frequency-dependent facilitation of spinal cord neuronal responses mediated by afferent C-fibres. This phenomenon is likely to represent the physiological correlate of the sensitisation to sustained stimulation which characterises pain sensation in contrast to the habituation observed in other sensory systems. A clear and complete mechanistic explanation as to how this phenomenon occurs is not yet available. In this talk I will discuss how network, presynaptic and post-synaptic factors may all contribute to the generation of wind-up in nociceptive neurones. I will describe a simple procedure to record sub- and supra-threshold events related to wind-up in an in vitro model of the rat spinal cord. This model is particularly well suited to pharmacological experiments. I will present results showing how specific ion channel modulators and endogenous neuromodulators associated with descending antinociceptive systems modify wind-up. Finally I will discuss data showing that induction of an experimental inflammation prior to the extraction of the spinal cord leads to an increased capability to produce wind-up of spinal neurones in isolation from descending inputs and from the peripheral injured area. This indicates that isolated spinal circuits retain a memory of the peripheral injury. Time course experiments demonstrate that consolidation of this spinal memory is a slow process that develops gradually giving some clues about the underlying mechanisms. Preliminary experiments indicating the involvement of ERK signaling in the first steps of this process will also be discussed. Supported by the Spanish Ministry of Science (SAF-2000-0199) and the Madrid Regional Government (Contrato Programa) to Prof. Cervero. Department of Physiology, University of Alcala. 28871 Madrid, Spain

Department of Haematology, University of Essen, D-45122, Germany S 04-3 S 04-1 ORGANIZATION OF PAIN PATHWAYS IN SPINAL DORSAL HORN A.J. Todd Nociceptive primary afferents terminate mainly in the superficial dorsal horn of the spinal cord (laminae I and II) and many contain the peptide substance P. The neurokinin 1 (NK1) receptor (on which substance P acts) is expressed by certain dorsal horn neurons, including some of those in lamina I as well as scattered cells in deeper laminae (III-VI). Many lamina I neurons have axons that project to brain sites, including the thalamus, periaqueductal grey matter, parabrachial area

LEARNING AND MEMORY IN PAIN PATHWAYS J. Sandkühler The neurons of the central nervous system not only have the capacity to transmit, inhibit and weigh information, but they may also store information for prolonged periods of time. Synaptic plasticity in hippocampus is a cellular model of learning and memory and recent studies suggest that similar mechanisms also apply to pain pathways and may account for some forms of hyperalgesia, allodynia and analgesia. The discovery of synaptic long-term plasticity in nociceptive systems provides a relatively simple and straight forward concept for a number of clinically relevant phenomena. Hyperalgesia and allodynia

S 373 following trauma, inflammation or acute nerve injury are, at least in part, caused by an enhanced sensitivity of nociceptive spinal dorsal horn neurons to sensory stimulation. This phenomenon has been termed central sensitization. Well known “wind-up” and central sensitization are, however, not equivalent. Synaptic plasticity may instead underlie some forms of central sensitization. Repetitive electrical stimulation of fine primary afferents or strong natural noxious stimuli such as skin burns, contusions, inflammation and acute nerve injury induce longterm potentiation of synaptic strength in Aδ-and in C-fibers in vitro and in vivo. The conditioning stimuli that induce synaptic LTP in fine primary afferents are similar to those that trigger hyperalgesia. Spinal LTP and injury-induced hyperalgesia share signal transduction pathways, time course and pharmacological profile, which makes use-dependent LTP at Aδ- and C-fiber synapses an attractive cellular model of injury-induced central sensitization and hyperalgesia. LTP induction requires a steep rise in postsynaptic [Ca 2+]i . One can therefore predict that LTP induction is sensitive to pre- and/or postsynaptic inhibition as this would reduce any synaptically evoked rise in [Ca2+]i. Descending antinociceptive pathways are known to exert a pre- and postsynaptic inhibition on nociceptive spinal dorsal horn neurons. When this inhibition was abolished (e.g. by spinalization) LTP was not only induced by strong noxious stimuli but also by less severe stimuli that were previously ineffective. Thus, endogenous analgesia may not only reduce acute pain but may also protect the spinal cord from sensitization. Thus, long-lasting changes in synaptic strength in superficial spinal dorsal horn may underlie some clinically relevant changes in pain sensitivity.

respiratory induced nasal cooling decreases the brain temperature during heavy physical exercise or during exposure to high environmental temperatures. The cooled nasal vein blood cools the brain arterial blood in the cavernous sinus – carotid artery complex. Tracheal intubation wills shortcut the mechanism, since the airflow through the nasal cavities is stopped. It was investigated, whether an artificial nasal flow could cool the brain. Thermometers were inserted into the brain and rectum of anaesthetized, intubated pigs and rats. The temperatures were measured every two seconds. Ten min periods of nasal oxygen flushing (different flows) induced rapid and reversible decreases (0.5 – 2.0 oC) in the brain temperatures. The decrease was dependent of the airflow. A similar result was obtained in one patient (Einer-Jensen and De Tommasi, pers. com.). Extubation of the pigs induced an 0.6 oC decrease of the brain temperature. Cooling was thus found in both the pig, which has a Rete Mirabile, and in the rat, which like man, does not have a Rete.The results indicate that intubation of patients may induce brain damage in patients with high brain temperature due to lack of cooling. The nasal cavities should be flushed with oxygen as prevention. N Einer-Jensen, KE Baptiste, F Madsen, MH Khorooshi. Can intubation harm the brain in critical care situations? Medical Hypotheses, 2001, in print. Institute of Medical Biology, University of Southern Denmark, Winsloewparken 21, DK-5000 Odense M, Denmark, e-mail: [email protected] S 05-2

Brain Research Institute, Vienna University, Spitalgasse 4, A-1090 Vienna S 04-4 LONG-TERM POTENTIATION AS A MECHANISM OF NEUROGENIC HYPERALGESIA IN HUMANS R.-D. Treede Enhanced pain sensitivity surrounding an injury site (secondary hyperalgesia) is due to sensitisation of nociceptive pathways within the central nervous system, and not due to sensitisation of the peripheral terminals of primary nociceptive afferents. Using intradermal capsaicin injections to mimick the nociceptive afferent barrage of an injury, central sensitisation has been documented for nociceptive neurons in the spinal cord (Simone et al. 1991, J Neurophysiol 66: 228-246). The observation that central sensitisation can last for many hours following one brief conditioning injury suggests that its mechanisms may be related to long-term potentiation (LTP). LTP has been found to occur at the first synaptic relay of the nociceptive pathways in the spinal cord (Randic et al. 1993, J Neurosci 13: 5228-5241). Using the same tetanic electrical stimulation patterns that elicit LTP in spinal cord slices, we found a long-lasting 2-3fold enhancement of pain sensitivity in human subjects. This neurogenic hyperalgesia consisted of a leftward shift of the stimulus-response function for pain elicited by punctate mechanical stimuli. The leftward shift was comparable to that elicited by intradermal capsaicin injections or observed in some patients with neuropathic pain. These fndings show that neurogenic hyperalgesia is a psychophysical correlate of LTP in the human nociceptive system. The mechanisms of central sensitisation differ from homosynaptic LTP, because previous studies have shown that the conditioning stimuli activate another input (capsaicin-sensitive C-fibres) than the test stimuli (capsaicin-insensitive A-fibres; Magerl et al. 2001, Brain 124: 17541764). In summary, neurogenic hyperalgesia in humans can be induced by LTP of central nociceptive pathways, but the mechanisms of neurogenic hyperalgesia include heterosynaptic effects that have not yet been characterised neurobiologically. Supported by NATO collaborative research grant CRG 95032540495. Institute of Physiology and Pathophysiology, Johannes GutenbergUniversity Mainz, Saarstr. 21, D-55099 Mainz S 05-1 COOLING OF THE BRAIN THROUGH LOCAL TRANSFER OF HEAT/COLD BETWEEN THE BLOOD VESSELS TO THE BRAIN N. Einer-Jensen, M.H. Khorooshi A high body temperature may damage organs. The most heat sensitive organ is the brain. Many mammals have developed a system, where the

LOCAL TRANSFER OF HORMONES BETWEEN THE BLOOD VESSELS TO THE BRAIN J. Skipor, N. Einer-Jensen Local transfer of substances in the head has recently been found in the sheep, pig, rabbit, and rat under both in vivo and in vitro conditions. The transfer is based on transfer from the brain venous blood to brain arterial blood in areas with close connections between the vessels e.g. where the carotid artery penetrates the cavernous sinus. Since the transfer probably is passive, a shared area of the blood vessels with a short distance of diffusion will facilitate transfer. Several species (pig, sheep) form a Rete Mirabile (a set of thin parallel arteries) within the cavernous sinus; thus creating a huge shared area. Transfer was, however, also found in specie (rat) without a Rete Mirabile. Both small (tritiated water, tyrosine) and larger substances (steroid and peptide hormones) were found to be present in the carotid blood in higher concentrations than in ”peripheral” arterial blood after nasal application or after infusion into the inferior nasal vein. Higher than expected concentrations in the brain tissue was also found. Even drugs may be transferred locally to the carotid blood after nasal administration. The local transfer system between the brain vessels probably has physiological importance. A substance produced in one part of the brain and transferred from the vein blood to the brain arterial blood may have a regulatory influence on other parts of the brain. References: Skipor, W. Grzegorzewski, T. Krzymowski, N. Einer-Jensen: Local transport of testosterone from the nasal mucosa to the carotid blood and the brain in the pig. Polish Journal of Veterinary Sciences 2000, 3, 19-22. N. Einer-Jensen, L. Larsen: Local transfer of diazepam, but not of cocaine, from the nasal cavities to the brain arterial blood in rats. Pharmacology and Toxicology 2000, 87, 276-78 Reproductive Endocrinology and Pathophysiology, Polish Academy of Sciences, Prawochenskiego 5, 10-718 Olsztyn, Poland E-mail: [email protected] S 05-3 OVARIAN TEMPERATURE GRADIENTS AND A CONTRIBUTION OF COUNTER-CURRENT MECHANISMS R.H.F. Hunter, N. Einer-Jensen After recalling male gonadal physiology in respect of tissue temperatures within the scrotal sac, and raising questions concerning abdominal testes, attention turned to mature Graafian follicles and ovarian stroma. Temperature gradients between such tissues were summarised for human, rabbit, pig and cow, and generally fell in the range of 1.3–1.7oC: follicles were always cooler than stroma. Measurements were made principally by means of a thermo-sensing camera at mid-ventral laparotomy, but also using microelectrodes or thermistor probes sited in the follicular antrum of rabbits and pigs, respectively. When thermo-

S 374 imaged under the fimbriated extremity of the Fallopian tube, mature pig follicles and stroma could still be distinguished. Such follicles cooled slightly more rapidly during the first 10 seconds of a 60 second recording interval, after which curves for the two tissues remained parallel. Arresting ovarian blood supply for 5 minutes had a negligible influence on the temperature differentials. Endoscopy in three models recorded mean differentials between follicles and stroma of 0.6 ± 0.1oC to 1.1 ± 0.1oC, but such follicles had not attained mature diameter. Temperature gradients were thought to be generated at least in part by endothermic reactions within mature follicles, reflecting hydration of large extracellular matrix molecules such as proteoglycans. Temperature gradients would be maintained locally by counter-current heat exchange mechanisms and, in this context, the microvasculature and lymphatic flow of individual follicles were found to be appropriate. Observations on the temperature of pre-ovulatory follicles appear relevant to procedures of in vitro maturation and in vitro fertilisation.

There are also instances where the several components of a complex endocrine pathway occur within a single organ and likely exert local homeostatic control there by paracrine signalling. For example, all components of the renin-angiotensin system are expressed in the kidney. Recent work, with transgenic mice that express human angiotensinogen in the proximal tubules, strongly supports the hypothesis that a tissue-specific renin-angiotensin cascade exists in the kidney. Countercurrent exchange has hitherto been seen as a mechanism for maintaining high local concentrations of hormones in target tissues – for example, within the ovary and testis. This concept can be extended to encompass improved efficiency of paracrine mechanisms, since countercurrent exchange would limit the loss of signal molecules from an organ. Additionally, by restricting entry of these molecules into the circulation, countercurrent exchange might reduce the risk of local control mechanisms spilling over into systemic effects.

Faculty of Veterinary Medicine, University of Cambridge, Madingley Road, Cambridge, CB3 0ES, England

Department of Physiology and Pharmacology, University of Southern Denmark, Winsloewparken 21, Third Floor, DK-5000 Odense, Denmark

S 05-4 LOCAL TRANSFER OF HEAT AND HORMONES FROM VAGINA TO UTERUS AND URETHRA E. Cicinelli, N. Einer-Jensen Anatomically and functionally, the vagina is closely connected to the uterus, bladder and urethra. Clinical trials in women have demonstrated an unexpected high efficiency of hormones administered vaginally compared to the effect of oral administration. The phenomenon is supposed to be based on local counter current transfer from lymph vessels and veins to the arteries. The hypotheses was based on the finding that after vaginal administration of progesterone, serum levels of progesterone in the uterine artery was twice as high as levels in the radial artery. Transfer of heat has been used as a tool for investigating the extent of the transfer system. A thin 4-point temperature catheter was inserted into the urethra in women and vagina cooled with a cold plastic device. The temperature dropped throughout urethra independent of the physical distance to vagina. The results were taken to indicate a local transfer system and may explain the positive impact of vaginal administration of estrogens on urethral atrophy without endometrial stimulation in menopausal women. Temperature catheters were also inserted into the uterine lumen in women, followed by vaginal cooling. The uterine temperature decreased. The decrease was significantly smaller in the tubal part (10-15 mm from the ostium) than in corpus. This was taken to show that the corpus was supplied with cooled blood from the uterine artery, while the non-cooled tubal part of uterus was supplied from the ovarian artery via the tubal arterial communicants. From a physiological point of view, the results indicate that the tubal part of uterus is locally influenced by the ipsilateral ovarian hormones being secreted to the ovarian veins, locally transferred to the ovarian arterial blood, and reaching the ovary-tubeuterine tubal corner. N Einer-Jensen, E Cicinelli, P Galantino, V Pinto and B Barba. Preferential vascular-based transfer from vagina to the corpus but not to the tubal part of the uterus in postmenopausal women. Human Reproduction, 2001, 16, 1329-1333. Department of Obstetric and Gynaecology, University of Bari, I-70124 Bari, Italy. E-mail: [email protected] S 05-5 COUNTERCURRENT TRANSFER AND PARACRINE SIGNALLING A.M. Carter Autocrine and paracrine signalling, in which molecules released from a cell act on the same cell or on target cells in the same organ, are recognised mechanisms of homeostatic control. In several instances, it has become apparent that conventional views on endocrine regulation need to be revised to take account of paracrine signalling. It had, for example, long been thought that growth hormone regulated bone growth by stimulating secretion of insulin-like growth factor-I (IGF-I) from the liver. However, when the Cre/LoxP system was used to generate mice with liver-specific and inducible IGF-I gene knockout, serum IGFI levels decreased by more than 80%, but postnatal growth was normal. The implication is that extrahepatic IGF-I, produced locally in the tissues, is sufficient for maintenance of postnatal body growth in mice.

S 06-1 ROLE OF CIRCULATING ATP ON THE REGULATION OF SKELETAL MUSCLE BLOOD FLOW DURING EXERCISE J. González-Alonso, B. Saltin Blood flow to skeletal muscle (MBF) increases in proportion to the level of metabolic activity of the contracting muscle fibres. This process therefore requires a precise mechanism matching the O2 demand by the tissue and the O2 supply by the blood. A recent model proposes that the erythrocyte acts as a sensor of the local regulation of MBF and O2 delivery, by either releasing ATP and/or NO in direct proportion to the offloading of O2 from haemoglobin (Hb) (Ellsworth et al. 1995; Stamler et al. 1997). We presently examined whether the progressive increase in MBF during incremental exercise is associated with parallel elevations in plasma ATP. Furthermore, because the release of ATP from the erythrocyte depends on the red blood cell deoxyhaemoglobin concentration, we determined if pronounced alterations in O 2 Hb produced by systemic hypoxia, systemic hyperoxia and CO inhalation were met by reciprocal changes in MBF and circulating ATP. Seven young healthy subjects completed three to four 4 min knee-extensor exercise bouts at 27±1, 46±1, 64±2 and 85±2% (S.E.M.) of their peak power output (78±5 W; ~60 rpm), separated by ~1 h of rest, under the following conditions that drastically altered arterial O2 content (Ca,O2): 1) normoxia (~185 ml l -1); 2) hypoxia (~143 ml l -1); 3) hyperoxia (~202 ml l-1), and 4) CO+normoxia (~144 ml l-1). In normoxia, the increases in TBF and thigh vascular conductance (TVC) during incremental exercise were closely associated with increases in both arterial plasma [ATP] and femoral venous [ATP]. With hypoxia compared to normoxia, TBF, TVC and plasma [ATP] were elevated, whereas they tended to be reduced with systemic hyperoxia. In CO+normoxia, TBF and TVC were elevated as in hypoxia, yet the increase in plasma [ATP] was attenuated, suggesting that binding of CO to Hb inhibited the release of ATP from the red blood cell. In conclusion, our findings suggest that ATP release from the erythrocyte is an important signal in the regulation of blood flow and O2 delivery to skeletal muscle. Supported by the Danish National Research Foundation (504-14). The Copenhagen Muscle Research Centre, Rigshospitalet, University of Copenhagen, DK 2100 Copenhagen, Denmark S 06-2 THE ROLE OF NITRIC OXIDE AND ADENOSINE IN THE REGULATION OF SKELETAL MUSCLE BLOOD FLOW Y. Hellsten Although much focused on, the effect of NO on exercise hyperemia remains somewhat controversial. Nevertheless, several human studies in which either ultrasound doppler or thermodilution techniques have been used to monitor blood flow demonstrate that NO synthase inhibition does not reduce blood flow during exercise. Exercise hyperemia is, however, likely to be the product of several vasoactive compounds interacting and a compromised synthesis of one vasoactive compound can probably be compensated for by an enhanced formation of other

S 375 vasodilators. With the use of the microdialysis technique it has been shown that NO synthase inhibition in humans does not increase the interstitial concentration of the vasodilators adenosine, prostacyclin or potassium, suggesting that these compounds are not responsible for maintaining flow during a diminished formation of NO. However, in a recent study it was shown that the combined inhibition of Cytochrome P450 2C9 (CYP 2C9) and NO synthase caused a significant reduction in flow whereas inhibition of CYP 2C9 alone did not affect exercise blood flow. Based on this observation and findings from in vitro studies there appears to be a close interaction between NO and the product of CYP 2C9, 11,12 epoxyeicosatrienoic acid (11, 12 EET), where either compound can compensate for a reduced synthesis of the other. Another potent vasoactive compound formed extracellularly in skeletal muscle during contraction is adenosine. Although human studies on the role of adenosine in exercise hyperemia have been difficult to perform due to a lack of non-toxic adenosine receptor inhibitors some evidence exist for an involvement of adenosine in exercise hyperemia. Adenosine also appears to interact with prostanoids, as evidenced by an increase in the muscle interstitial concentration of adenosine during inhibition of cyclooxygenase in humans, possibly indicating a redundancy also between these two compounds. In conclusion, we propose that exercise induced skeletal muscle blood flow is the result of an integrated response by locally formed vasodilators where NO and 11, 12 EET are critical players and where compounds such as adenosine and prostanoids also are likely to be involved. Ylva Hellsten, Copenhagen Muscle Research Centre, August Krogh Institute, AHF, Universitetsparken 13, DK- 2100 Copenhagen 0 S 06-3 ADENOSINE, NITRIC OXIDE AND THE REGULATION OF OXYGEN SUPPLY IN HYPOXIA J. M. Marshall In the rat we have shown that the vasodilatation that occurs in skeletal muscle during systemic hypoxia is largely mediated by adenosine which is released by the endothelium and acts on adenosine A1 receptors to cause dilatation in a nitric oxide (NO)-dependent manner. The adenosine component of the hypoxia-induced dilatation is also dependent on prostaglandin (PG) synthesis. By using an NO electrode and ex vivo preparations of rat artery, we confirmed that A1 -stimulated release of NO is attenuated by PG synthesis inhibition, that PGI2 stimulates NO synthesis and that A1-stimulated NO release is attenuated by adenyl cyclase inhibition. Thus, the NO release evoked by A1 receptors is at least partly dependent on PGI 2-cAMP -mediated stimulation of NO synthase (NOS). However, our recordings of muscle vascular conductance and intravital microscopy indicate that although hypoxia-induced dilatation of proximal arterioles is attenuated by NOS inhibition, their dilatation can be restored by infusion of an NO donor. This ”restored” dilatation is also mediated by adenosine, suggesting that when NOS is inhibited, adenosine is still released by hypoxia and then induces dilatation by acting on vascular smooth muscle, providing there is a background level of NO. By contrast, the hypoxia-induced dilatation of terminal arterioles is attenuated by NOS inhibition, but not restored by infusion of a NO donor, suggesting their dilatation requires increased synthesis of NO stimulated by adenosine. Studies on the relationship between O2 delivery (DO2) and muscle O2 consumption (VO2) suggest that adenosine- and NO-mediated dilatation of terminal arterioles is crucially important in systemic hypoxia allowing VO 2 to remain constant until DO2 is severely compromised. Our most recent studies indicate that the release of adenosine from the endothelial cells in hypoxia is also NO-dependent, probably because NO competes with O2 at the binding site on mitochondrial cytochrome oxidase so increasing its sensitivity to falls in O 2 tension. Indeed, in shear-stimulated, endothelial cells, adenosine release was stimulated by an NO donor or bradykinin which is known to stimulate NOS. Department of Physiology, The Medical School, Birmingham, B15 2TT, UK S 06-4 REGIONAL COORDINATION OF VASOMOTOR RESPONSES IN THE MICROCIRCULATION: ROLE OF EDHF AND NO U. Pohl, C. de Wit In order to achieve a high conductivity of a skeletal muscle vascular

bed not only terminal arterioles but also the vessels located upstream (larger arterioles and even small arteries) have to dilate in a well coordinated manner. The mechanisms underlying this co-ordination are still not fully understood. This is not only true for the signals that are generated but also for their way of transmission to upstream vessels. Intravital microscopic studies suggest that such a co-ordination could be achieved by the ”ubiquitous” effects of an increase in blood flow on all arterial vessels. The resulting increases of endothelial wall shear stress augment release of NO and perhaps the endothelium-derived hyperpolarising factor (EDHF) which in effect leads to compensatory and synchronised vasodilatation in all regions exposed to elevated shear stress. Thus, local conductivities can principally be adapted to elevated flow even in upstream arteries which are probably not exposed to metabolic signals of these cells which are finally supplied by these vessels. It has also been shown that venules and small veins draining blood from active tissue can affect the diameter of adjacent arterioles and small arteries by diffusion of NO and other endothelial mediators. This may represent a second, indirect way of co-ordinating arteriolar metabolic dilation with dilation of upstream located supplying arteries. The most direct way, however, would be the transmission of dilator signals from capillaries and terminal arterioles to the vessels located upstream. In fact ”conducted dilations” can be elicited by strictly local application dilators or electrical stimulation of muscle fibres. These dilations are based on the electrotonic propagation of changes in membrane potential of endothelial cells or vascular smooth muscle cells that can be elicited e.g. by EDHF. Experimental evidence suggests that these membrane potential changes are transmitted via gap junctions over distances up to several mm along the vascular tree. Recent experiments indicate that especially one of the gap junction forming proteins, connexin (CX) 40, has an important role in transmitting these signals, since in CX40 -/- animals the vascular distances over which signals are conducted are significantly reduced. It seems to be a direct consequence of this effect that these animals are also hypertensive. Surprisingly little is known about the regulation of the conductivity of gap junctions for these signals. Gap junctional communication between endothelial cells can be, however, modified by NO and prostaglandins. Recent experiments support a dual role of NO in this setting: It reduces conductivity of existing gap junctions thereby reducing potential feed back mechanisms in peripheral vasodilatation but also supports de novo formation of gap junctions which may be important in angiogenesis as well as adhesion of blood cells. It remains to be established whether this dual effect can be attributed to different connexins in the vascular tissue and whether it is also important in redirecting signal propagation from smooth muscle to the endothelium and vice versa. Institute of Physiology, Ludwig-Maximilians-University Munich S 06-5 INTERACTIONS BETWEEN LOCAL METABOLIC EVENTS AND SYMPATHETIC NERVE FIBRES IN CONTROLLING BLOOD FLOW AND OXYGENATION IN EXERCISING MUSCLE J. Hansen The tight coupling of oxygen supply and utilization in exercising skeletal muscle is the result of complex interactions between local mechanisms that control muscle blood flow and substrate utilization and systemic mechanisms that control cardiac output and arterial pressure. The role of the sympathetic nervous system in the integration of these responses, specifically the interaction between sympathetic activation and local vasodilator mechanisms in exercising muscle remains incompletely understood. The functional consequence of sympathetic activation in exercising skeletal muscle has been the subject of considerable debate. Previous studies in animals and humans have suggested that sympathetic vasoconstricton in active muscle is a) well maintained and serves to limit active hyperemia, thereby preventing muscle blood flow from outstripping cardiac output in order to preserve blood pressure and vital organ perfusion, or b) greatly attenuated in order to optimize muscle perfusion, a concept that has been termed ”functional sympatholysis”. Studies performed over the past 70 years have provided conflicting evidence regarding the relative importance of sympathetic vasoconstriction versus functional sympatholysis in exercising skeletal muscle. The focus of this talk is mainly on recent studies in anesthetized animal preparations and in conscious humans that have provided evidence for the metabolic modulation of sympathetic vasoconstriction in contracting skeletal muscle and have identified a number of key underlying mechanisms

S 376 that extend the initial concept of sympatholysis. Copenhagen Muscle Research Centre, Rigshospitalet Section 7652, Blegdamsvej 9, DK-2100 Ø Copenhagen, Denmark S 07-1 LOCAL CALCIUM GRADIENTS DURING EXCITATIONCONTRACTION COUPLING AND ALTERNANS IN ATRIAL MYOCYTES. L. A. Blatter, K. A. Sheehan, J. Kockskämper Subcellular Ca2+ signalling during normal excitation-contraction coupling (ecc) and during Ca2+ alternans was studied in atrial myocytes with fast confocal microscopy and Ca2+ current (ICa) measurements. Ca2+ alternans, through modulation of Ca2+-dependent ionic conductances, causes electromechanical alternans which has been implicated in the generation of ventricular fibrillation and sudden cardiac death. Cat atrial myocytes lack transverse tubules and contain junctional (j-SR) and non-junctional SR (nj-SR) which both have ryanodine receptor Ca2+ release channels. During ecc Ca2+ entering through voltage-gated membrane Ca2+ channels (I Ca) triggered Ca 2+ release at discrete peripheral j-SR release sites. Discrete Ca2+ spark-like increases of [Ca2+]i then fused into a peripheral ‘ring’ of elevated [Ca2+]i, followed by propagation (via Ca2+-induced Ca2+ release, CICR) to the cell center resulting in contraction. j-SR Ca2+ release could be terminated instantaneously by interrupting ICa, whereas nj-SR Ca2+ release continued, once initiated, even after ICa and j-SR Ca2+ release were terminated. In atrial myocytes Ca2+ alternans, a beat-tobeat alternation of the amplitude of the action potential-induced [Ca2+]i transient, could be elicited by increasing stimulation frequency or inhibition of glycolysis. Marked subcellular variations in the time of onset, the magnitude, and the phase of alternans were observed. Longitudinal and transverse gradients of Ca2+ alternans were found as well as neighboring subcellular regions alternating out-of-phase. Moreover, focal inhibition of glycolysis resulted in spatially restricted Ca2+ alternans. When two adjacent regions within a myocyte alternated out-ofphase, steep [Ca2+]i gradients developed at their border giving rise to delayed propagating Ca2+ waves. The results demonstrate that Ca2+ alternans is a subcellular phenomenon caused by modulation of SR Ca2+ release which is mediated, at least in part, by local inhibition of energy metabolism. The generation of arrhythmogenic Ca 2+ waves by subcellular variations in the phase of Ca2+ alternans represents a novel mechanism for the development of cardiac disrhythmias. Dept. Physiology, Loyola Univ. Chicago, Maywood, IL 60153, USA

S 07-2 TARGETING CALCIUM CYCLING PROTEINS BY GENE TRANSFER IN VITRO AND IN VIVO EXPERIMENTS R.J. Hajjar, F. del Monte Cardiomyocytes isolated from failing human hearts are characterized by contractile dysfunction including prolonged relaxation, reduced systolic force and elevated diastolic force. These contractile abnormalities are paralleled by abnormal Ca2+ homeostasis such as reduced sarcoplasmic reticulum (SR) Ca2+ release, elevated diastolic Ca2+ and reduced rate of Ca2+ removal. In addition, failing human myocardium is characterized by a frequency-dependent decrease in systolic force and Ca 2+ as opposed to normal myocardium where an increase in pacing rate results in potentiation of contractility and an increase in SR Ca2+ release. In the failing heart, the decrease in SR Ca2+ load has been linked to a decrease in SR Ca2+ ATPase (SERCA2a) function. We have recently shown that overexpression of SERCA2a by adenoviral gene transfer restores contractile function in cardiac myocytes from failing human hearts. In addition, we have shown that overexpression of SERCA2a in a model of pressure-overload hypertrophy in transition to failure improves contractile function and reserve in these animals. We have explored the effect of long-term expression of SERCA2a in failing animals (rodents and swine) along with the energy cost of SERCA2a expression using NMR methods. We are also using a different strategy to improve SR Ca 2+ ATPase activity which involves decreasing the expression of phospholamban by antisense strategies to enhance SR Ca 2+ ATPase activity. The Na/Ca exchanger was also targeted to enhance calcium removal in failing hearts. Action potential prolongation is attributed to reductions in transient outward current (Ito) density in human heart failure. This prolongation can

improve contractility but can also cause afterdepolarization. Using gene transfer of various K channels responsible for Ito, we are investigating the molecular and the ionic basis of action potential prolongation in cardiac hypertrophy and failure and we have examined how intracellular calcium handling changes in response to alterations in action potential duration. Gene transfer, which serves initially as an experimental tool, may provide a novel therapeutic approach. Cardiovascular Research Center, Massachusetts General Hospital, Charlestown, MA 02129, USA S 07-3 REMODELING OF THE Na + -Ca 2+ EXCHANGER IN HYPERTROPHIED RAT CARDIAC MYOCYTES A.M. Gomez1, G. Vassort1, E. Niggli, M. Egger* In cardiac myocytes the Na + -Ca 2+ exchanger (NCX) is the most important transporter for Ca2+ extrusion and it may also be involved in excitation-contraction coupling. In the present study we determined the activity of the NCX in myocytes isolated from post myocardial infarction (PMI) rats. We examined whether remodeling during cardiac hyperthrophy had altered the ion transport characteristics of the NCX. Intracellular Ca2+ concentration jumps were generated in isolated PMI rat ventricular myocytes by flash photolysis of caged Ca2+ (DMnitrophen). The resulting NCX current (I NaCa ) was recorded in the whole-cell configuration of the patch-clamp technique while the intracellular Ca2+ concentration changes were imaged with a confocal microscope using fluo-3. Myocytes from PMI hearts had a significantly increased I NaCa density (approx. 2.6 fold) and the expression of the exchanger was upregulated. Surprisingly, the Ca 2+ transport rates remained constant, despite the fact that INaCa was increased. It is assumed that the Ca 2+ transport rate and the corresponding I NaCa change in parallel, unless the stoichiometry of the exchanger is modified. However, there was no reason to assume stoichiometry changes, because no changes in the I NaCa /voltage relationship were found. The development of hypertrophy is a gradual process that also includes changes of cell size and shape. INaCa is proportional to membrane surface. Conversely, the Ca 2+ transport rate (d[Ca 2+]i/dt) depends on the myocyte volume. Since the cell surface/volume ratio changes in hypertrophy, the cells express higher I NaCa densities, but exhibit unchanged Ca2+ transport rates. We conclude that the Ca2+ transport rate via the exchanger is maintained at a constant level during hypertrophy. Therefore, myocytes need to increase NCX density. This makes it likely that the remodeling of the Na+-Ca2+ exchange is an adaptive and not a causal process. Supported by the Swiss National Science Foundation. * Department of Physiology, University of Bern, Bühlplatz 5, CH3012 Bern, Switzerland,1INSERM U390, Montpellier, France

S 07-4 CHANGES IN INTRACELLULAR CALCIUM HANDLING IN A FERRET MODEL OF LEFT VENTRICULAR HYPERTROPHY AND HEART FAILURE M.E. Díaz, D.A. Eisner, A.W. Trafford In clinical practice cardiac hypertrophy is frequently associated with the development of contractile dysfunction. Given the strong dependence of the contractile state of the heart on SR Ca content (Trafford et al, 2001) we have sought to investigate whether SR Ca loading is altered in cardiac myocytes isolated from hypertrophied hearts and the mechanisms by which such changes may occur. Left ventricular hypertrophy was induced in adult ferrets by ascending aortic coarctation. Single cardiac myocytes were isolated from the hypertrophied hearts before clinical symptoms of heart failure were present. Cells were voltage clamped and changes in intracellular Ca concentration ([Ca2+]i) were measured using Fluo-3. SR Ca content was assessed by application of 10mM caffeine and measuring the integral of the resulting Na-Ca exchange current. The amplitude and rate of decay of the systolic Ca transient were significantly reduced in cells isolated from hypertrophied hearts. At all frequencies of stimulation the SR Ca content was also significantly lower. Inhibition of SR (10mM caffeine) and Na-Ca exchange (10mM Ni) function was then used to investigate the relative contributions made to the decay of the systolic Ca transient by the SR and Na-Ca exchanger. In hypertrophied hearts

S 377 the contribution from the SR Ca-ATPase was reduced. These findings indicate that the contractile dysfunction observed in cardiac hypertrophy may be due to a reduced availability of Ca within the cardiac myocytes. The altered balance between SR Ca uptake and Na-Ca exchange mediated Ca removal from the cytosol may underlie this reduction in SR Ca content. Trafford,AW, Díaz,ME, Eisner,DA. Coordinated control of cell Ca2+ loading and triggered release from the sarcoplasmic reticulum underlies the rapid inotropic response to increased L-type Ca2+ current. Circ Res 2001;88:195-201. Cardiac Physiology, University of Manchester, Manchester M13 9PT UK

S 07-5 THE INFLUENCE OF MITOCHONDRIAL MEMBRANE POTENTIAL ON CARDIAC PRECONDITIONING AND ISCHEMIAREPERFUSION-INJURY K. Meuter, C. Schäfer, Y. Abdallah, B. Rother, H.-M. Piper Opening of mitochondrial ATP-sensitive K+ (KATP) channnels plays a pivotal role in ischemic preconditioning that provides protection against ischemia-reperfusion-injury. As one possible mechanism a moderate decrease of mitochondrial membrane potential has been discussed, which may lead to a reduced ATP-depletion during ischemia and a better energy recovery during reoxygenation. We investigated the extent and progress of mitochondrial membrane depolarisation after incubation with either the KATP-opener diazoxide (l00µM) or the mitochondrial uncoupler FCCP (5nM). As experimental model isolated cardiomyocytes of adult rats were used. Changes in mitochondrial membrane potential were measured with fluorescent indicator JC-1 (as percentage of control fluorescence). The cells were exposed to 100 min anoxia and 15 min reperfusion. Two different experimental protocols were carried out: 1. Incubation for 10 min with diazoxide or FCCP followed by 10minute washout (preconditioning model) 2. Incubation with diazoxide or FCCP immediately before anoxia Treatment of cardiomyocytes with either diazoxide (D) or FCCP caused a comparable, moderate depolarisation which persisted l0min after washout (Ctr.: 100%, D: 87±0.11%, FCCP: 85±0.04%; p<0.05 vs Ctr.). Pretreatment with diazoxide followed by 10-minute washout showed a significant protection against reoxygenation-induced injury. The same protocol with FCCP had no protective effect (extent of cell hypercontracture in per cent, related to end-anoxic cell length: D: 9.22±1.05%, Ctr: 12.93±1.49%, p<0.05 vs Ctr; FCCP: 13.01±1.94%, n.s. vs Ctr.). Pretreatment with diazoxide or FCCP immediate before anoxia had no protective effect (D: 17.33±3.09%, FCCP: 14.51±2.97%, Ctr: 12.23±2.41%; n.s. vs Ctr.). The K ATP-opener diazoxide and the mitochondrial uncoupler FCCP both induce a comparable, moderate decrease of mitochondrial membrane potential. In a preconditioning model diazoxide provides protection against ischemia-reperfusion-injury. The results of this study demonstrate that the cardioprotective effects of the K ATP-openers are not induced by a mitochondrial depolarisation. Physiologisches Institut, Justus-Liebig-Universität, Aulweg 129, 35392 Giessen

S 07-6 MYOGLOBIN ACCELERATES CARDIAC POSTISCHEMIC RECOVERY U. Flögel, A. Gödecke, J. Schrader Myoglobin (Mb), an intracellular oxygen binding hemoprotein in skeletal and cardiac muscle, is known to be a key player in intracellular oxygen supply. Previously, we have shown that Mb also significantly contributes to NO breakdown in the heart. Furthermore, in vitro studies suggested that Mb may react with reactive oxygen species (ROS, e.g. superoxide radicals). Since both NO and ROS are considered to be critically involved in ischemia-reperfusion (IR) injury, the present study explored the role of Mb during IR making use of the Mb knockout (myo-/-) mouse recently generated in our laboratory. Isolated hearts of wild-type (WT) and myo-/- mice were subjected to 12 min ischemia followed by 60 min reperfusion. 31P NMR was used to monitor the energy state of the hearts, and lucigenin-enhanced chemiluminescence was employed to assess the release of superoxide radicals. During is-

chemia cardiac function and energetic parameters declined similarly in both groups. In reperfusion, however, WT hearts showed a significantly faster recovery of postischemic function compared to myo -/hearts: left ventricular delevoped pressure (LVDP) was 35.6±7.5 vs. 22.4±5.3 % of control, resp., after 10 min reperf, P<0.05, n=8). 31P NMR revealed that, concomitantly, a significantly larger overshoot of the phosphocreatine (PCr) signal occured in myo -/- hearts (125±5 vs. 110±7 % of control, resp., after 10 min reperf, P<0.05, n=8). The retarded restoration of functional and metabolic parameters in myo-/hearts was accompanied by an increased release of superoxide radicals (465±87 vs. 287±73 cts, resp., after 5 min reperf., P<0.05, n=8). At the end of reperfusion, baseline function was almost completely restored in both groups. In an additional series of experiments, the NO synthase inhibitor L-NMMA accelerated the recovery of LVDP in WT and myo-/- hearts. Again, the recovery of cardiac contractility and energetics was markedly retarded in myo -/- hearts, which was also paralled by an increased release of superoxide in myo-/- hearts. The larger overshoot of the PCr signal in myo -/- hearts indicates a more pronounced inhibition of cytoplasmic creatine kinase (CK) during reperfusion which is most likely caused by free ROS. This impaired CK activity impedes the efficient delivery of energy to utilization sites with resulting mechanical dysfunction. Mb can contribute to the breakdown of ROS via H2O2- and superoxide-consuming reactions. Therefore, it is most likely that both pathways are involved in the cardioprotective effect of Mb during reperfusion. Our results demonstrate that Mb significantly attenuates myocardial ischemia-reperfusion injury and substantially contributes to the breakdown of ROS in the heart. Institut fur Herz- und Kreislaufphysiologie, Heinrich-Heine-Universität Düsseldorf Postfach 101007,40001 Düsseldorf, Germany

S 08-1 ELECTRONEUTRAL SALT ABSORPTION - MOLECULAR MECHANISMS G. Lamprecht Electroneutral NaCl absorption in the intestine is mediated by parallel Na +/H+- and Cl -/HCO3--exchange, which are functionally coupled by the intracellular pH. The ion transporters involved are the Na +/H+exchanger isoform 3 (NHE3) and the newly discovered anion exchanger dra (down regulated in adenoma). Both transporters have a common structure with a N-terminal transmembrane domain and a Cterminal cytoplasmic domain, which mediates regulatory signals. Both inhibition and stimulation of NHE3 activity exist and involve: 1) Inhibition through phosphorylation of S552 and S605 (rat NHE3) by protein kinase A (PKA). This mechanism requires the presence of the PDZ adapter proteins NHERF or E3KARP, which are thought to couple NHE3 to Ezrin, which acts as an A kinase anchoring protein and a linker to the actin cytoskeleton. 2) Clathrin mediated endocytosis, which may be down stream and time delayed of the PKA induced phosphorylation of NHE3. 3) cAMP induced remodeling of the actin cytoskeleton and its direct/indirect physical interaction with NHE3, which is independent of the cAMP induced phosphorylation of the transporter. 4) Modulation of the constitutive recycling of NHE3 between the plasma membrane and recycling endosomes: Insertion into the plasma membrane (activation of NHE3) is enhanced by Epidermal Growth Factor acting via PI3-kinase and the down stream tyrosin kinase Akt. Retrieval from the plasma membrane (inhibition) is mediated by protein kinase C. 5) In addition to their transcriptional activation of NHE3 glucocorticoids stimulate NHE3 through the activation of SGK1 (serum/glucocorticoid stimulated kinase), which interacts with one of the PDZ domains of E3KARP but not NHERF. These mechanisms are probably related to each other, but the details have not been worked out. Much less is known about the anion exchanger dra: 1) It is a member of a new gene family (SLC26), which differs from the conventional AEs (SLC4). 2) Naturally occurring mutations are causative for congenital disease (this has not been shown for NHE3). 3) dra has a PDZ interaction motif, that facilitates binding to the same PDZ domain proteins as NHE3, which may be a mechanism for common regulation or targeting of the two transporters. 1 st Medical Dept. Univ. Tübingen, Otfried-Müller-Str. 10, 72076 Tübingen

S 378 S 08-2 PERIPHERAL MELATONIN MEDIATES NEURAL STIMULATION OF INTESTINAL SECRETION G. Flemström, M. Sjöblom. Some common diseases in humans, including gastroduodenal ulcer and irritable bowel syndrome, show circadian rhythms in pain and discomfort. The neurohormone melatonin is released from enterochromaffin cells in the intestinal mucosa and from the pineal gland but its role in gastrointestinal function is largely unknown. We have studied the involvement of melatonin in stimulation of the mucosa-protective alkaline secretion in duodenum in anesthetized rats. Close intra-arterial infusion of melatonin significantly increased the secretion. Pretreatment with the melatonin (predominantly MT 2 receptor specific) antagonist luzindole (600 nmol/kg, i.v.) almost abolished the stimulation induced by melatonin. In contrast, luzindole did not affect secretion induced by intra-arterial infusion of bethanechol, guanylin, CCK-8 or VIP. Intracerebroventricular (i.c.v.) infusion of the alpha1-adrenoceptor agonist phenylephrine caused an up to fivefold increase in the alkaline secretion. This neurally-induced response was markedly inhibited by the melatonin antagonist luzindole, by the nicotinic antagonist hexamethonium, or by cutting all peri-carotid nerves at the sub-laryngeal level, including both vagal trunks and sympathetic chains. Removal of the pineal or pituitary glands did no affect the secretory response to i.c.v. phenylephrine. Effects on enterocyte [Ca2+]i were studied in aggregates of rat or human duodenal enterocytes, isolated by mild digestion and loaded with Fura-2. The duodenal enterocytes responded to the secretagogues CCK-8 (acting at CCK1-receptors) and carbachol (acting at M3-receptors) with a transient rise in [Ca 2+] i . Melatonin (10-500 nM) caused a more long-lasting increase in [Ca 2+]i and, furthermore, potentiated the [Ca 2+]i responses to CCK-8 and carbachol. Peripheral melatonin thus stimulates duodenal mucosal HCO 3 ¯ secretion and endogenous melatonin, very likely released from mucosal enterochromaffin cells, seems involved in mediating neural stimulation of the secretion. Stimulation probably reflects an action on duodenal enterocyte melatonin MT2 receptors and [Ca2+]i is one likely mediator of the response. Department of Physiology, Uppsala University Biomedical Center, POB 572, SE-75123 Uppsala, Sweden

Physiologisches Institut, Winterthurerstr. 190, CH-8057 Zürich S 08-4 ENTERIC SENSORY NEURONS W. Kunze The intestine contains sensory neurons that lie wholly within the organ. In the guinea-pig, intrinsic sensory neurons (ISNs) are multipolar, have an oval soma in both myenteric and submucous plexuses and have processes that run circumferentially. They innervate their own and other ganglia and have collaterals supplying the mucosa. ISNs transmit to other ISNs in a self-reinforcing network, to inter- and motor neurons and probably to vagal afferents. ISNs have a unique electrophysiology when disconnected from sensory input. They exhibit large action potentials with TTX-sensitive Na+ and N-type Ca2+ inward conductances, and with a pronounced inflection on the repolarizing phase. The action potential is followed by a slow, delayed afterhyperpolarization caused by the opening of intermediate conductance (∼70 pS) Ca2+ modulated K+ channels. A variety of transmembrane ionic currents have been distinguished using sharp electrode or whole cell1 intracellular recording. These include transient and delayed outwardly rectifying K+ currents, several Ca2+- modulated K+ (IK(Ca) ) currents, an inward rectifier (IKir), a hyperpolarizing activated cationic current (Ih), a low threshold, slowly inactivating Na+ current (INa,LTSI), and perhaps a low threshold Ca2+ ICa,T current. The presence of ICa,T, INa,LTSI, and Ih currents is suggestive of a pacemaker capability in ISNs. These currents are modulated by synaptic input and interact to mediate transitions between the 4 clearly distinguishable firing states that ISNs may enter: 1. hyperpolarized, leaky, 2. rapidly accommodating, 3. slowly accommodating and 4. depolarized, inactivated. Switching to states 3 and 4 is facilitated by conditioning low frequency stimulation of presynaptic inputs to ISNs. This can give rise to long term excitation in a manner dependent on the previous stimulation history. Such contingent state switching represents a form of use-dependent memory which may underlie normal and pathological sensory functions of the intestine. 1Rugiero et al. 2002 J Physiol. 538, 447-463. Zentrum Medizinische Forschung (ZMF), Allgemeine Chirurgie, Waldhörnlestr. 22, 72072 Tübingen, Germany

S 08-3 S 08-5 FUNCTION OF APICAL K+ CHANNELS R. Warth, D. Heitzmann, F. Grahammer, M. Garcia Alzamora, A. Schmitt-Gräff, R. Nitschke, M. Bleich, P. Meneton, F. Verrey, J. Barhanin In the intestinal tract luminal K+ channels determine K+ content of the luminal fluid and hyperpolarize the luminal membrane thereby regulating the driving force for electrogenic transport. Little is known about the molecular identity of these luminal K+ channels. We examined the role of heteromeric KCNE1/KCNQ1 K+ channels for gastro-intestinal function using the KCNE1 knockout mouse (E1 -/-) as a model. Metabolic cage experiments revealed an increased loss of Na+ and K+ in E1 -/- mice indicating a possible role of this channel complex during electrolyte reabsorption. However, in the distal colon Na+ reabsorption via epithelial sodium channels was not diminished but it was enhanced in E1 -/- mice suggesting a deficient Na+ reabsorption in proximal colon and/or small intestine. In renal proximal tubules, luminal KCNE1/KCNQ1 channels are involved in nutrient reabsorption. Therefore, it is tempting to speculate that this K+ channel complex also plays a role for Na+-coupled nutrient reabsorption in small intestine. In rodent stomach both KCNE1 and KCNQ1 are expressed. Gastric acid secretion is almost completely abolished in KCNQ1 knockout mice and strongly reduced by pharmacological inhibition of KCNQ1. The pore-forming K+ channel subunit KCNQ1 is co-localized with H +/K+ ATPase in the luminal membrane of parietal cells. These data indicate that KCNQ1 secretes K+ into the luminal compartment which then can be pumped into the cell by H +/K+ ATPase. Interestingly, gastric acid secretion was not affected by KCNE1 gene knockout; in contrast, peak acid output after hormonal stimulation was even higher in E1 -/ - mice. Recently it was shown, that KCNE2, a KCNE1-related protein, co-assembles with KCNQ1 in the luminal membrane compartment of parietal cells. In conclusion, KCNQ1 associated to KCNE1 or KCNE2 forms luminal K+ channels which are important for gastro-intestinal electrolyte transport. Further studies are required to elucidate the identity and function of other luminal K+ channels in the gastro-intestinal tract.

EPITHELIAL BARRIER FUNCTION BY TIGHT JUNCTION PROTEINS M. Fromm, S. Amasheh, A.H. Gitter, T. Schöneberg, J. Mankertz, J.D. Schulzke Primary properties of epithelia are transport and barrier functions. The paracellular barrier in epithelial tissues is maintained by the zonula occludens - the tight junction (TJ). This most apical intercellular junction is arranged in strands associated with the cytoskeleton. Proteins located in this multiprotein complex have been identified previously: ZO-1, ZO-2, ZO-3, cingulin, 7H6 antigen, symplekin, as well as the transmembrane proteins occludin, junction adhesion molecule, and the claudin family with 24 members at present. Although the number of proteins known to contribute to the TJ structure is growing, only little data addressing their physiological relevance is available. In particular, functional information of how they assemble and how they participate in barrier function is still largely unknown. Occludin was the first transmembrane protein discovered in TJ. Because of its localization, it was reasonable to assume that occludin plays a physiological role in maintaining the TJ structure. However, its functional relevance has been questioned by recent knock-out studies. Occludin-deficient epithelial cells still possess TJ strands. Moreover, the knock out mouse model did not display a perturbation of epithelial barrier function, although a complex pathophysiological phenotype was observed with growth retardation, chronic inflammation and hyperplasia of the gastric epithelium, and calcification in the brain. The claudin family shows an organ and tissue specific expression of individual members. Deficiency of distinct claudins has been reported to be associated with severe pathophysiological consequences, e.g. autosomal recessive deafness in case of claudin-14, hypomagnesaemia linked to mutations of claudin-16 and the fact that CNS myelin and sertoli cell TJ strands are absent in Osp/claudin-11 null mice. To study the function of occludin and claudin family members, two cell strains, C7 and C11, of Madin-Darby canine kidney (MDCK) cells were

S 379 electrophysiologically and biochemically analyzed. Although claudin2 was present in low resistance C11 cells, we found an almost complete lack of claudin-2 expression in C7 cell which are characterized by high transepithelial resistance. Detailed electrophysiological characterization and transfection experiments demonstrate the responsibility of claudin2 for tight junctional cation permeability. Dept. of Clinical Physiology, Dept. of Gastroenterology, and Dept. of Pharmacology, Benjamin Franklin University Hospital, Freie Universität Berlin, Hindenburgdamm 30, 12200 Berlin, Germany

in specimens from the groups receiving saline only, or CGP 42112A in presence PD123319. Mesenteric blood flow increased moderately upon administration of CGP42112A. CONCLUSION: Activation of AT 2receptors elicits an elevation of jejunal mucosal NO output. Part of this response may be due to an increase in the expression of mucosal. Dr L. Fändriks, Dept of GastroResearch, Sahlgrenska Academy, Göteborg University, POBox 75038, SE40036 Göteborg, Sweden. S 09-1

S 08-6 NOS EXPRESSION IN THE GUT: IMPLICATIONS FOR INTESTINAL MOTOR FUNCTION D. Saur, H.-D. Allescher Nitric oxide (NO) is an important non-adrenergic non-cholinergic mediator within the enteric nervous system. It is generated by NADPH dependent conversion of L-citrulline to L-arginine by nitric oxide synthase (NOS). In addition to its action as neurotransmitter NO act also as a neuromodulator facilitating or attenuating the release of other neurotransmitters like VIP or ACh. The primary mode of action of NO in the enteric nervous/smooth muscle cell system is the activation of sGC/cGMP/PGK dependent signal transduction pathways regulating free intracellular Ca2+ levels and Ca2+ sensitivity and thus the contractile activity of enteric smooth muscle cells. Neuronal NOS (nNOS) is the predominant isoform of NOS in the enteric nervous system besides endothelial NOS and inducible NOS. An altered expression or biological activity of nNOS in the gut has been linked to several physiological conditions like aging and pregnancy, as well as different diseases, such as achalasia, diabetic gastroparesis and hypertrophic pyloric stenosis. nNOSα mutant mice showed a gastrointestinal phenotype resembling hypertrophic pyloric stenosis with delayed gastric emptying. In addition these mice have a hypertensive lower esophageal sphincter with impaired relaxation, similar to the clinical condition of achalasia. A recent study comprising families with inherited infantile pyloric stenosis identified nNOS as a susceptibility gene for this disorder and expression of nNOS exon 1c mRNA is significantly reduced in the pyloric sphincter of such patients. Diabetic rats and mice with defects in gastric emptying and pyloric NANC relaxation show reduced nNOS mRNA levels in the pyloric sphincter, but not in the central nervous system. Interestingly nNOS expression and NANC relaxation are restored to normal levels by insulin treatment. Therefore it is of physiological and pathophysiological interest to investigate the molecular basis of nNOS gene expression and regulation in the gut. Technische Universität München, 2. Medizinische Klinik, Ismaningerstr. 22, 81675 München S 08-7 ANGIOTENSIN AT2-RECEPTORS AND INTESTINAL NO PRODUCTION S Ewert, M Laesser, B Johansson, A Pettersson, A Åneman, L. Fändriks Effects of the renin-angiotensin system and its key mediator angiotensin II (AngII) on intestinal epithelial functions are only sparsely investigated. It has recently become evident that AngII can act on at least two receptors, the AT1 and AT2 receptors, situated in the intestinal mucosa. In the present paper we investigated the possibility that selective AT2 receptor stimulation influences jejunal mucosal nitric oxide (NO) formation, the latter known to regulate a number of epithelial processes. METHODS: Experiments were performed on αchloralose anaesthetized pigs. Systemic and mesenteric hemodynamics were recorded. Jejunal luminal NO output was assessed by means of intraluminal tonometry and analysed by a chemiluminescense analyser. Western blotting was used to detect and quantify mucosal iNOS. RESULTS: Regression analysis showed that baseline luminal NO output correlated to the baseline mucosal iNOS content. The NO output was increased in the group treated i.v. with the AT 2-receptor agonist CGP 42112A at the dose-intervals 0.1 (p<0.05) and 1.0 µg kg-1 min.-1 ( p<0.01). In the presence of the AT2-receptor antagonist PD 123319 (0.3 mg kg-1 iv. bolus and 0.03 mg kg-1 h-1 iv. infusion) CGP 42112A failed to stimulate the NO output. Western blotting showed a significant ( p<0.05) quantitative increase of iNOS protein in the specimens from the CGP 42112A treated group. The iNOS expression did not change

INTERACTIONS OF MYELOID CELL LINES, NEUTROPHILS AND LYMPHOCYTES WITH S128R E-SELECTIN UNDER FLOW D.Haskard, R. Rao, R.C. Landis Selectins mediate the rolling of circulating leukocytes on activated vascular endothelial cells (EC). A polymorphism of E-selectin has been described in which serine is substituted for arginine at position 128 (S128R) in the EGF domain, and this has been associated with both premature ischaemic heart disease and SLE. We investigated whether the substitution alters the ligand-binding properties of Eselectin under shear flow by studying the capacity of CHO cell transfectants expressing Wild Type (WT) or S128R E-selectin to support interactions of either K562 cells or HL60 cells in a parallel plate flow chamber. Unlike transfectants expressing WT E-selectin, S128R clones supported tethering of K562 cells, and this was calciumand shear-dependent but heparin- and neuraminidase-resistant. Furthermore, K562 cells transfected with the β2-integrin LFA-1 arrested on S128R but not on WT E-selectin CHO monolayers. HL60 cells exhibited more rolling (27.2 ± 3.5 vs. 10.8 ± 1.3 cells/field, p<0.05) and arrest (24.6 ± 2.8 vs. 10.5 ± 1.2 cells/field, p<0.05) on CHO cells expressing S128R than those expressing an equivalent density of WT, without affecting rolling velocity (33.8 ± 2.0 µm/s vs. 41.3 ± 2.0µm/ s). Neuraminidase inhibited all interactions of HL60 cells with WT Eselectin, but resulted in a residual tethering interaction on S128R. The measured dissociation constants, koff, for tethering on S128R by K562 and by neuraminidase-treated HL60 are the same (-3.9s-1), suggesting a common molecular mechanism. Experiments have confirmed that increased numbers of neutrophils arrest on S128R compared to WT monolayers (48.8 ± 6.5 cells/field vs. 25.2 ± 2.3 cells/field, p<0.05), although both rolling number and velocity are unchanged. We have also demonstrated increased numbers of lymphocytes arresting on S128R compared to WT monolayers (15.9 ± 3.9 vs. 8.6 ± 1.7 cells/ field). Significantly, a population of CLA-negative T-lymphocytes rolled and arrested on S128R but not on WT monolayers, suggesting that S128R may influence lymphocyte subset recruitment. Overall, these results indicate that that the S128R polymorphism alters the ligand-binding properties of E-selectin. This may enhance leukocyteEC interactions in vivo and influence the predisposition to inflammatory disease. BHF Cardiovascular Medicine Unit, National Heart and Lung Institute, Imperial College, Hammersmith Hospital, Du Cane Road, London W12 ONN S 09-2 EARLY ACTIVATION OF THE P42/P44 MAPK PATHWAY MEDIATES ADENOSINE-INDUCED NITRIC OXIDE PRODUCTION IN HUMAN ENDOTHELIAL CELLS: A NOVEL CALCIUM INSENSITIVE MECHANISM A.W.Wyatt, J.R.Steinert, J. D. Pearson, L. Sobrevia*, G.E.Mann Adenosine is released from the myocardium, endothelial cells and skeletal muscle in ischemia and is an important regulator of coronary blood flow. We have previously shown that acute activation of A2a purinoceptors stimulates NO production in human fetal umbilical vein endothelial cells (1), and have recently established a key role for p42/ p44MAPK in the regulation of the L-arginine-NO signaling pathway. Expression of mRNA for the A 2a-, A2b- and A 3-adenosine receptor subtypes was abundant whereas A1-adenosine receptor mRNA levels were negligible. Activation of A2a purinoceptors by adenosine (10 µM) or the A2a receptor agonist CGS21680, (100 nM) resulted in an increase in L-arginine transport and NO release, which was not mediated by changes in intracellular [Ca2+]i, pH or cAMP. Stimulation of endothelial cells with adenosine is paralleled by a membrane hyperpolarization and phosphorylation of p42/p44 MAPK . L-NAME abolished the

S 380 adenosine-induced hyperpolarization and stimulation of L-arginine transport, whilst SNP activated outward K + currents. Genistein and PD98059 and U0126 (inhibitors of MAPK kinase ½) prevented adenosine-stimulated L-arginine transport, NO production and p42/ p44MAPK phosphorylation. We found no evidence for the activation of eNOS via the serine/threonine kinase Akt/PKB (protein kinase B) in adenosine-stimulated cells. Thus, the endothelial cell L-arginine-NO pathway is acutely (2 min) activated by A2a-purinoceptor agonists in a Ca 2+-insenstive manner, involving p42/p44MAPK, with release of NO leading to a membrane hyperpolarization and activation of L-arginine transport. Supported by MRC, BHF, FONDECYT (19711321-1997), Universidad de Concepcion (DIUC9733871D) Centre for Cardiovascular Biology & Medicine, GKT School of Biomedical Sciences, King’s College London, Guy’s Campus, London SE1 1UL, UK, Cellular & Molecular Physiology Laboratory (CMPL), Faculty of Biological Sciences, University of Concepción, PO Box 160-C, Concepción, Chile S 09-3 THE DYNAMIC ENDOTHELIUM: GENES, HORMONES AND CYTOKINES A.D. Hingorani The healthy vascular endothelium exerts vasodilator, anti-thrombotic and atheroprotective actions through the release of mediators such as nitric oxide synthesised from L-arginine by the action of endothelial nitric oxide synthase (eNOS). Originally described as a constitutive enzyme, it is becoming increasingly clear that the expression and activity of eNOS is subject to complex regulatory mechanisms that mediate physiological and pathological changes in endothelial vasodilator function. The time-course, magnitude and mechanisms underlying these changes in endothelial function can be modelled in healthy volunteers and patients by assessing the responses of the endothelium in the dorsal hand vein, the brachial artery and forearm resistance vessels using the Aeillig technique, brachial ultrasound and venous occlusion plethysmography respectively. Using these models, physiological changes in endothelial function have been identified following exercise, within the normal menstrual cycle and during pregnancy. Levels of bioavailable NO from the human vascular endothelium are also altered as a result of pathology, being up-regulated during sepsis and down-regulated following exposure to classical cardiovascular risk factors, following ischaemia-reperfusion, after lowgrade inflammatory stimuli and after mental stress that all cause endothelial dysfunction. These changes in bioavailable NO stem, in part, from the concomitant up- or down-regulation of endothelial tetrahydrobiopterin (BH4), a co-factor for eNOS that stabilises the active form of the enzyme, lowers the Km for the substrate L-arginine and attenuates eNOS-derived superoxide production. Regulation of BH4 availability in the endothelium occurs through the expression of GTP cyclohydrolase I, the rate-limiting synthetic enzyme. The absolute change in the functional response of the endothelium to physiological or patholical stimuli differs from subject to subject, and part of the within subject variability may be attributable to genetic factors. A common Glu298Asp polymorphism in the eNOS gene that has been associated with susceptibility to ischaemic heart disease (IHD), hypertension and pre-eclampsia appears to account for part of this between subject variability. Since endothelial dysfunction appears predictive of future cardiovascular events in patients with established IHD, an understanding of the mechanisms that underlie this phenomenon and an appreciation of the genetic contribution to this disorder may allow enhanced prediction of future disease risk and targeted, mechanism-based therapies. Centre for Clinical Pharmacology, BHF Laboratories, Rayne Institute, University College London, London WC1E 6JJ, UK S 09-4 REGULATION OF ENDOTHELIAL CELL NITRIC OXIDE SYNTHASE BY PHOSPHORYLATION I. Fleming The endothelial NO synthase (eNOS) was initially assumed to be regulated solely by changes in the intracellular Ca 2+ concentration ([Ca2+]i). It is now clear that eNOS can be phosphorylated on serine, threonine and tyrosine residues and that a battery of protein kinases is

involved in the regulation of eNOS activity in both, the presence and the absence of an increase in [Ca2+]i. Of the potential phosphorylation sites within the eNOS sequence, most is known about the consequences of the phosphorylation of Ser 1177 in the reductase domain by Akt/ protein kinase B, protein kinase, AMP-activated kinase and CaM kinase II. Phosphorylation of Ser 1177 enhances the Ca2+ sensitivity of eNOS and enhances the electron flux from the reductase to the oxygenase domain, thus increasing enzyme activity by approximately 2-fold. Depending on which kinase phosphorylates Ser1177 the activation of eNOS can occur in a Ca2+-dependent (e.g. bradykinin-induced activation of CaMKII) or -independent manner (e.g., shear stress-induced activation of Akt). While the phosphorylation of Ser 1177 enhances eNOS activity, phosphorylation of Thr495 in the calmodulin-binding domain decreases eNOS activity by interfering with the binding of Ca2+/CaM. In endothelial cells, Thr495 is constitutively phosphorylated by PKC and dephosphorylation of this residue by the phosphatase PP1 is a prerequisite for calmodulin binding and the burst in enzyme activity normally observed following agonist-stimulation of endothelial cells. The tyrosine phosphorylation of eNOS is marked only in native and primary cultures of endothelial cells. Fluid shear stress enhances eNOS tyrosine phosphorylation, probably via Src. At present no hint for a specific role of tyrosine phosphorylation in the acute regulation of enzyme activity but it seems to be involved in the formation of protein-protein complexes and in the subcellular translocation of eNOS. Institut für Kardiovaskuläre Physiologie, Klinikum der J.W.G.-Universität, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main S 09-5 GAP JUNCTIONS IN VASCULAR ENDOTHELIUM : ROLE IN VASCULAR CONTROL U. Pohl Vascular cells contain four different kinds of connexins (Cx 37, 40, 43 and 45) of which gap junctions are formed. These constitute direct pathways for the exchange of small molecules and local changes of membrane potential between adjacent cells. The functional role of gap junctions in intact vessels is just emerging at present. EDHFdependent dilations may involve direct transfer of one or several signal molecules between endothelium and smooth muscle cells, in certain vessels. Moreover, dilation in response to several endotheliumdependent agonists is attenuated when gap junctional coupling is reduced. Gap junctional coupling might also be important for the exchange of feedback signals between endothelial cells and vascular smooth muscle as shown by the increase in endothelial calcium upon smooth muscle contraction. At present, the best defined function of gap junctions is the transmission of signals from capillaries and terminal arterioles to the vessels located upstream. These rapidly propagating vasomotor responses are based on the electrotronic spreading of changes in membrane potential of endothelial or smooth muscle cells that can be elicited by EDHF and other reagents. Recent experiments indicate that especially one of the gap junction forming proteins, Cx 40, has an important role in transmitting these signals, since in Cx 40 -/animals the vascular distances over which acetylcholine induced dilation is conducted, are significantly reduced. Surprisingly little is known about the regulation of the conductivity of gap junctions for these signals. Gap junctional communication between endothelial cells can be, however, modified by nitric oxide and prostaglandins. Recent experiments support a dual role of NO in this setting. It reduces conductivity of existing gap junctions, thereby reducing potential steal mechanisms in peripheral vasodilation, but also supports the novo formation of gap junctions which may be important in angiogenesis as well as for the adhesion of blood cells to the endothelium. It remains to be established whether this dual effect can be attributed to different connexins in vascular tissue and whether it is also important in redirecting signal propagation from the smooth muscle to the endothelium and vice versa. Institute of Physiology, LMU, Schillerstraße 44, 80336 Munich S 10-1 EPILEPSY AND ION CHANNEL DYSFUNCTION F Lehmann-Horn, K Jurkat-Rott, H Lerche Ion channels provide the basis for the regulation of excitability in the central nervous system and in other excitable tissues such as skeletal

S 381 and heart muscle. Consequently, mutations in ion channel encoding genes are found in a variety of inherited diseases associated with hyperor hypoexcitability of the affected tissue, the so-called ‘channelopathies’. An increasing number of epileptic syndromes belongs to this group of rare disorders: autosomal dominant nocturnal frontal lobe epilepsy is caused by mutations in a neuronal nicotinic acetylcholine receptor (affected genes: CHRNA4, CHRNB2), benign familial neonatal convulsions by mutations in potassium channels constituting the M-current (KCNQ2, KCNQ3), generalized epilepsy with febrile seizures plus by mutations in subunits of the voltage-gated sodium channel or the GABAA receptor (SCN1B, SCN1A, GABRG2) and episodic ataxia type 1 – which is associated with epilepsy in a few patients – by mutations within another voltage-gated potassium channel (KCNA1). These rare disorders provide interesting models to study the etiology and pathophysiology of disturbed excitability in molecular detail. On the basis of genetic and electrophysiologic studies of the channelopathies, novel therapeutic strategies can be developed as has been shown recently for the antiepileptic drug retigabine activating neuronal KCNQ potassium channels. Department of Applied Physiology, University of Ulm, Albert-EinsteinAllee 11, D-89069 Ulm

S 10-2 FUNCTIONAL STUDIES OF DUCKY MICE THAT HAVE A MUTATION IN THE CALCIUM CHANNEL Α2 ∆ ∆-2 SUBUNIT J. Brodbeck, A. Davies, J.-M. Courtney, A. Meir , N. Balaguero, C. Canti, F.J. Moss, K. M. Page, W. S. Pratt, S. P. Hunt✝, J. Barclay*%, M. Rees* and A. C. Dolphin The mouse mutant ducky, a model for absence epilepsy, is characterized by spike-wave seizures, and cerebellar ataxia. A mutation in Cacna2d2, the gene encoding the α2 δ-2 voltage-dependent calcium channel accessory subunit has been found to underlie the ducky phenotype. The α2 δ-2 mRNA is strongly expressed in cerebellar Purkinje cells. Here we show that du/du mice have abnormalities in their Purkinje cell dendritic tree. The mutation in α2 δ-2 results in the introduction of a premature stop codon and predicts the expression of a truncated protein encoded by the first three exons of Cacna2d2, followed by 8 novel amino acids. We show that both mRNA and protein corresponding to this predicted transcript are expressed in du/du cerebellum, and present in Purkinje cells. Whereas the α2 δ-2 subunit increased the peak current density of the Ca V2.1/β4 channel combination when co-expressed in vitro, coexpression with the truncated mutant α2 δ-2 protein reduced current density, indicating that it may contribute to the du phenotype. Departments of Pharmacology, Paediatrics and Child Health and Anatomy and Developmental Biology, University College London, Gower Street, London, WC1E 6BT, UK

S 10-3 INTERICTAL ACTIVITY LEADS TO DOWN-REGULATION OF THE K-Cl COTRANSPORTER, KCC2, AND TO IMPAIRED Cl EXTRUSION IN RAT CA1 PYRAMIDAL NEURONS K. Kaila Fast hyperpolarizing postsynaptic inhibition depends on extrusion of chloride by the neuron specific K-Cl cotransporter, KCC2. In rat hippocampal slices, interictal-like activity generated in the absence of Mg 2+ down-regulates KCC2 in CA1 pyramidal neurons at both the mRNA and protein level. The plasmalemmal KCC2 has a very fast rate of degradation, and the degradation is further accelerated in an activity-dependent manner. The loss of functional KCC2 in response to pathophysiological activity leads to a reduced capacity for neuronal chloride extrusion. Such a mechanism may contribute to the progressive increase in neuronal hypersynchrony characteristic of temporal lobe epilepsies. A decrease in the efficacy of chloride extrusion may be a general early response involved in various kinds of neuronal trauma. Department of Biosciences, University of Helsinki, 00014 Helsinki, Finland

S 11-1 NEUROENDOCRINE FUNCTION AND CHRONIC INFLAMMATORY STRESS M.S. Harbuz The development of inflammation in the rat model of adjuvant-induced arthritis is associated with alterations in hypothalamic regulation of the hypothalamo-pituitary-adrenal (HPA) axis. These alterations have a profound impact on the ability to respond to acute stress. In contrast, the response to an acute immune challenge remains intact in these animals suggesting differential activation of the HPA axis dependent on the nature of the challenge. Impaired HPA axis responsiveness to stress has been suggested to be predictive of susceptibility to disease. We have utilised behavioural models to correlate HPA axis activation and severity of disease. These data suggest that HPA axis responsivity to stress is not a good predictor of subsequent disease activity. Dysregulation of the HPA axis has been implicated in Rheumatoid Arthritis (RA) although the mechanisms underlying this remain obscure. The dexamethasone-corticotrophin-releasing factor (DexCRF) test is a sensitive test of HPA axis activity. This test has been used to determine alterations in HPA activity in psychiatric conditions such as depression. We have recently used this test in patients with RA and compared their responses with age and sex-matched controls. These studies have identified a sub-population of RA patients who were able to escape from Dex suppression and mount a cortisol response to CRF. The reasons for these differences in response to the Dex-CRF test in RA patients remain to be elucidated. URC Neuroendocrinology, University of Bristol, BRI, Marlborough Street, Bristol, UK. BS2 8HW S 11-2 NEW INSIGHTS INTO BRAIN SEXUAL DIMORPHISM: FUNCTIONAL IMPLICATIONS R. Grossmann, A. Jurkevich, A. Köhler The nonapeptide hormone arginine vasotocin (AVT) is the main regulator of fluid balance in avian species and, in addition, is involved in oviposition in these species. The neuroendocrine AVT system is represented by the hypothalamo-neurohypophysial (HN) system and consists of magnocellular neurons located within the hypothalamic nuclei supraopticus and paraventricularis. Axons travel via the HN-tract to the posterior pituitary and from there the nonapeptide is released into the blood stream. In addition to this ‘classical’ neuroendocrine system, the avian AVT system includes sexually dimorphic brain structures located mainly in the areas of the bed nucleus of the stria terminalis (BnST) and the lateral septum. It is known from various mammalian and bird species that these brain areas are involved in the central regulation of reproduction and sexual behaviour. In our recent study in the chicken we could clearly demonstrate that AVT is expressed in the adult BnST in males only. Even after osmotic stimulation AVT gene expression in neurones of the parvocellular BnST in hens was not upregulated. In this paper we focus on the ontogeny of sexual differentiation in the chicken AVT system and on the role of gonadal hormones in organizing these differences during development and maintaining them during adulthood. Parvocellular AVT neurons become discernible in the BnST of males and females during the second half of prenatal development. These cells undergo a profound and irreversible sexual differentiation during ontogeny. Our recent behavioural studies suggest that the sex dimorphic vasotocinergic system could be implicated in processes of social recognition in various behavioural contexts. Department of Functional Genomics and Bioregulation, Institute of Animal Science and Animal Behaviour, Federal Agricultural Research Centre (FAL), Höltystr. 10, 31535 Neustadt, Germany S 11-3 CRH AND VASOPRESSIN RECEPTORS AND THE STRESS RESPONSE R. Landgraf Both CRH and vasopressin receptors are critically involved in behavioral regulation, including stress coping strategies. Using antisense targeting, virally-mediated transgene transfer and antagonist approaches, we specifically and transiently modulated neuropeptide-receptor in-

S 382 teractions in adult rats. As measured in a variety of behavioral tests, CRH1 receptors are primarily involved in the regulation of emotionality, whereas the CRH2 receptor subtype is likely to play a role in stress coping. Similar to CRH1 receptors, the vasopressin V1a receptor mediates anxiogenic responses and plays a major role in cognition. To study the involvement of CRH and vasopressin receptor subtypes in anxiety-related behavior, stress coping, and neuroendocrine phenomena associated with anxiety and stress in more detail, we used Wistar rats, bred for extremes in innate anxiety. While CRH1 receptors failed to differ significantly between high (HAB) and low (LAB) anxiety animals, CRH2 receptor binding was elevated in the paraventricular nucleus of HAB compared to LAB rats. Administration of the novel CRH1 receptor antagonist R121919 resulted in a similar inhibition of ACTH secretion in HABs and LABs, but in reduced anxiety in the former only. Interestingly, innate hyper-anxiety predominantly appears to be due to vasopressin over-expression. Accordingly, this neuropeptide is synthesized and released to a higher extent in the PVN of HAB relative to LAB animals. While vasopressin V1a receptor binding failed to differ between both lines, administration of a specific receptor antagonist reduced hyper-anxiety and passive stress coping as well as normalized anxiety-linked neuroendocrine aberrations in HAB animals. Max Planck Institute of Psychiatry, Kraepelinstr. 2, 80804 Munich, Germany S 11-4 GASES AS MODULATORS OF THE PITUITARY-ADRENAL AXIS A. Grossman1, A. Costa2, M. Forsling3, P. Navarra4 In recent years there has been increasing interest in the possibility that the gases nitric oxide (NO) and, even more recently, carbon monoxide (CO), can function as intercellular messengers in a completely novel manner, providing a medium of cellular communication quite distinct from the classical neuroregulators. These agents will diffuse according to their physicochemical characteristics, in evanescent ‘puffs’ with spherical limits, unconstrained by formal cell boundaries. Most work has concentrated on their roles in inflammation and control of the vascular tree, but there is now extensive data attesting to a complex involvement in the neuroendocrine axis, particularly the hypothalamopituitary-adrenal (HPA) axis. We have established that both NO and CO are powerful and specific inhibitors to the release of corticotrophin releasing hormone (CRH) and vasopressin from the rat hypothalamus in vitro, and in particular in attenuating stimulation of these peptides by cytokines. The enzymes for both gases are also present in the nuclei of origin of these releasing hormones. Further studies demonstrated that exposure of hypothalami to lipopolysaccharide in vitro led to an unexpected fall in both CRH and vasopressin. However, blockade of NO synthase attenuated the fall in CRH, while CO antagonism led to a highly significant stimulation of vasopressin release. Blockade of NO synthase in vivo leads to the enhancement of ACTH and corticosterone release in response to inflammatory stressors, while antagonism of CO synthesis leads to an exaggerated rise in endotoxin-stimulated vasopressin, but not of corticosterone. These results suggest that both NO and CO are important components of the HPA axis, and may act to counter-regulate stimulation of this axis in response to inflammatory stimuli. In the human, it seems likely that a failure of formation of CO in acute porphyria may be responsible for some of its clinical manifestation, such as paroxysmal hypertension and the syndrome of inappropriate antidiuretic hormone. 1

Dept.of Endocrinology, St. Bartholomew’s Hospital, London EC1A 7BE, UK, 2Institute of Neurology, IRCCS C. Mondino, Pavia, Italy, 3 Dept.of Physiology, UMDS, London, UK and 4Dept. of Pharmacology, Catholic University, Rome, Italy S 12-1 CONTRASTING ROLES OF CORTICAL AND SPINAL NEURONS IN PREPARATION AND EXECUTION OF VOLUNTARY HAND MOVEMENTS. E.E. Fetz, SI.Perlmutter, Y. Prut, S. Votaw The activity of cortical neurons during voluntary movements has been extensively documented, but little is known about the responses of spinal interneurons (INs) during normal behavior. To study the contribution of cervical INs to voluntary hand movement we

documented their activity and correlational linkages to muscles in monkeys performing ramp-and-hold flexion-extension torques about the wrist. Task-related spinal INs increased their activity during flexion or extension, typically with a tonic or phasic-tonic response pattern in their preferred direction. Surprisingly, most INs showed some activity during both flexion and extension, in unexpected contrast to the strictly unidirectional activity of corticomotoneuronal (CM) cells and motoneurons (but similar to the bidirectional activity of rubromotoneuronal cells). INs that produced post-spike effects in muscles typically affected a smaller set of muscles than CM cells. Thus, CM cells represent functional muscle synergies more distinctly than spinal INs. To document the role of spinal INs in preparation for voluntary movement we trained monkeys to perform instructed delay tasks. Two monkeys performed isometric wrist flexion or extension after a visually cued delay period. The third monkey was instructed by transient visual or proprioceptive cues to perform appropriate wrist movements against an elastic load. One third of the tested INs showed significant modulation of firing rate during the delay period (SDM) following the visual cue, relative to the pre-cue rate. The SDM responses often differed from the INs’ responses during the subsequent active torque period. Excluding the perturbation response, the delay period activity for many INs was similar in visual and perturbation trials, consistent with motor preparation independent of the cue modality. However, other INs exhibited different SDM in the visual trials and the perturbation trials, suggesting that motor preparation initiated by visual or proprioceptive cues may be mediated in part by different spinal pathways. These results suggest that segmental INs are involved, with cortex, in the earliest stages of movement preparation instructed by various cue modalities. To investigate the relation of spinal INs to a repertoire of different muscle synergies, activity of INs was recorded from a macaque performing a multidirectional wrist task. The monkey generated isometric torques in flexion/extension (FE), radial/ulnar deviation (RU) and pronation/ supination (PS), and executed a power grip (G) that co-contracted wrist flexor and extensor muscles. Of 27 INs showing task-modulated activity, 14 had preferred directions in the FE-RU space, with cosinetuned activation; 6 other INs were broadly tuned for FE-RU. The tuning of these INs tended to be broader than the tuning of forearm muscle EMG. These results suggest that INs share some properties of spatial tuning with cortical neurons. These observations suggest that under normal behavioral conditions many spinal INs have reponse properties similar to those previously documented for cortical neurons in behaving animals, although there are differences in representation. [Supported by NIH grants NS12542, NS09189 & RR00166, and APA PBR2-9502] Department of Physiology & Biophysics and Regional Primate Research Center, University of Washington, Seattle WA 98195-7290, USA

S 12-2 CORTICOSPINAL FUNCTION IN MAN J.B. Nielsen Transcranial magnetic stimulation (TMS) and imaging techniques such as f-MRI and PET have made an investigation of corticospinal function in human subjects possible. Important advances have thus been made recently in our understanding of the contribution of the corticospinal tract to various motor behaviours. One example is the control of rhythmic motor activity such as walking and bicycling. TMS applied over the leg area of the motor cortex easily activates leg muscles during treadmill walking (Capaday et al. 1999, Schubert et al. 1997). The evoked responses (MEPs) are greatly modulated during the gait cycle. Using H-reflex conditioning Petersen et al. (1998) provided evidence that part of this modulation may be caused by modulation in the excitability of corticospinal cells with monosynaptic projections to the motoneurones of the active muscles. In a recent study Petersen et al. (2001), furthermore, demonstrated that weak TMS, which was adjusted to only activate local cortical inhibitory neurones, produced a profound depression of the on-going EMG activity during treadmill walking. This suggests that the corticospinal tract contributes significantly to the basic locomotor activity during walking. This is likely also the case during bicycling. In a PET study Christensen et al. (2000) found a very pronounced increase in the cerebral blood flow during bicycling. When subtracting the blood flow during passive

S 383 pedalling significant peaks of increased blood flow was found bilaterally in the leg area of the motor cortex. Similar to other types of behaviour increased blood flow was also found in the supplementary motor cortex when subjects imagined the bicycle movement. These data highlight that the corticospinal tract and the motor cortex do not contribute only to the control of complex and highly skilled movements, but also to relatively automatic motor tasks such as walking and bicycling. Capaday C et al. (1999). J Neurophysiol 81, 129-139; Christensen, LOD et al. (2000) Exp Brain Res 135, 66-72; Petersen N et al. (1998). J Physiol 513, 599-610; Petersen N et al. (2001). J Physiol 537, 651656; Schubert M et al. (1997). ExpBrain Res 115, 234-246. Department of Medical Physiology, University of Copenhagen, Panum, Blegdamsvej 3, DK_2200 Copenhagen N. Denmark. E-mail: [email protected]

S 12-3 CORTICOSPINAL CONTROL OF UPPER LIMB FUNCTION IN DIFFERENT PRIMATES R.N. Lemon There is vigorous debate as to the significance of differences in the organisation of the corticospinal tract (CST) across species; these differences are important because they may provide useful insights into the role of the CST in the control of upper limb function. The CST terminates widely within the cervical enlargement, with projections to almost all laminae, emphasising its multi-functional role. There are particularly striking species differences in the density of direct, corticomotoneuronal (CM) projections. Terminations arising from the hand area of motor cortex are distributed along the full extent of the cervical cord, not just to the segments innervating the hand. Neurophysiological studies show that CM connections, which appear to be unique to primates, are rather weak in the New World squirrel monkey, stronger in the Old World macaque monkey and probably strongest in humans. CM EPSPs are found in most macaque upper limb motoneurones (MNs), being largest in hand muscle MNs. Inhibitory effects from the CST are also very common, with up to 65% of MNs showing a disynaptic IPSP following the early CM-EPSP (Maier et al., 1998). These IPSPs are probably mediated by segmental inhibitory interneurones. On the basis of anatomical evidence, oligosynaptic excitation of MNs from the CST should be rather common: this is indeed the case in the squirrel monkey, but not in the macaque (Nakajima et al., 2000). We have searched for evidence of a C3-C4 propriospinal (PN) system in primates, similar to that in the cat, in which this system represents a major pathway for transmission of cortical excitation to forelimb MNs. We looked for oligosynaptic excitation from the CST in intracellular recordings from MNs in anaesthetised macaques, and from single motor units in awake or lightly sedated macaques. In both preparations such excitation was rare (18% of MNs showed late EPSPs to repetitive stimulation of the CST). It may have been masked by segmental inhibition: however, oligosynaptic excitation was not common either in MNs with ‘pure’ CM EPSPs, or in motor units with no suppression of firing from the CST. A C5 lesion interrupting CST input to the lower cervical segments, reducing both CM EPSPs and disynaptic inhibition, did not unmask additional C3-C4 PN excitation. Alstermark et al. (1999) confirmed our results in the macaque but went on to show that after intravenous strychnine to abolish all glycinergic inhibition, C3C4 PN excitation was present, suggesting that C3-C4 PNs may be under stronger feedforward inhibition in macaque than in the cat. We have not so far found evidence for strong inhibition of this kind, but it could play a role in ‘gating’ transmission through the C3-C4 system in particular motor acts. All studies have emphasised species differences in the organisation of CST control of the upper limb. We have speculated that the increasing domination of the CM vs C3-C4 PN pathways may be related to the degree of dexterous use of the upper limb. (Supported by the Wellcome Trust and International Spinal Research Trust).References : Alstermark, B. Isa, T. Ohki, Y. Saito, Y. (1999) J. Neurophysiol.82:3580-3585. Maier, M A. Illert, M. Kirkwood, P A. Nielsen, J. Lemon, R N. (1998) J. Physiol. 511:191-212. Nakajima, K. Maier, M A. Kirkwood, P A. Lemon, R N. (2000). J. Neurophysiol. 84: 698-709 Sobell Department of Motor Neuroscience and Movement Disorders, Institute of Neurology, UCL, London WC1N 3BG UK

S 12-4 THE ROLE OF THE PYRAMIDAL TRACT AND CORTICAL AREAS IN GRASPING M. Wiesendanger Modern knowledge on the pyramidal tract (PT) functions is based to a considerable extent on the pioneering work of Charles G. Phillips. Together with his colleagues, he emphasized particularly the tight connections of the PT with the motor apparatus of the hand, which plays a crucial role in the dexterous manipulations of subhuman and human primates. However, he also realized, way back in the sixties, that the commands sent down to the spinal cord have been (pre-) organized outside the motor cortex. I will emphasize this point by illustrating the manifold of cortical and subcortical structures that are implicated in the organization of manipulatory skills. This will be illustrated by assessing hand functions, especially the regulation of grip force and load forces. In unstable conditions, subjects much depend on anticipatory force recruitment (motor set). In patients with various brain lesions, one can detect the loss of subtle anticipatory adjustments and impaired adaptive force scaling. Functional imaging in healthy human subjects shows well the distributed arrangement of involved structures during various manipulations, both during planning and execution. Bern, Switzerland S 12-5 POSTLESIONAL RECOVERY OF CORTICOSPINAL FUNCTION IN MAN H.J. Freund In cases with acute damage of motor cortex or its descending fibres the severity of the initial motor deficit is still the best predictor for outcome. Those patients who improve do so within the first days and weeks following the event. This emphasises a critical role of reafferent feedback for regaining function. The hypothesis that the absence of residual function over a critical time span prevents functional restitution was derived from experiments on deafferented monkeys who were not using the respective limb many weeks after injury. These observations laid the ground for the concept of „learned nonuse“ and the „constraint induced“ therapy. The significance of afferent input is further supported by observations that cases with additional damage of the thalamus have distinctly poorer prognoses. This indicates that not only damage of the output modules is a critical factor but also of certain nodal points in the sensory-motor network. The results provide further evidence to support the view that reafferent activity from the partly compromised limb is critically required for tuning the remaining network into function. Alterations in the perilesional tissue associated with major changes of neuronal activity in a surprisingly large fringe area have been identified as an another major determinant of functional restitution. Intracortical excitability is increased, inhibition decreased and spontaneous activity and stimulus response characteristics are distinctly altered. Audioradiography is consistent with decreased inhibition as it shows downregulation of GABA and upregulation of glutamate receptor densities. Although these electrophysiological, audioradiographic and metabolic patterns are clearly defined the perilesional area cannot be visualised by imaging methods, including experimental 7 Tesla MRI scans performed on histologically defined pathological zones. These functional and structural aspects modify former mechanistic views envisaging the location and size of the destructed area as the determinant for the resulting functional impairment. Dep Neurology HHU Düsseldorf FRG S 13-1 NEUROMODULATION OF VOLTAGE-GATED CALCIUM CHANNELS W.A. Catterall Voltage-gated Ca2+ channels which couple membrane depolarization to calcium influx are composed of a large α1 subunit of approximately 200 to 250 kDa in association with a disulfide-linked complex of α2 and δ subunits of 143 kDa and 27 kDa, an intracellular β subunit of 55 kDa, and a transmembrane γ subunit of 30 kDa. The pore-forming α1 subunits are organized in four homologous domains (I through IV)

S 384 which each contains six transmembrane alpha helices (S1 through S6) and additional membrane-associated sequences. The S4 segments contain positively charged residues that serve as voltage sensors. The S5 and S6 segments and the short membrane-associated segments between them form the calcium-selective pore. We have studied the mechanisms of neuromodulation of two distinct types of calcium channels that have different modes of regulation — presynaptic P/Q-type calcium channels of the Cav2 family and L-type calcium channels of the Ca v1 family. P/Q-type (Ca v2.1) calcium channels containing α1 2.1 subunits are localized in high density at synapses and are responsible for calcium entry that initiates fast release of transmitters. These presynaptic calcium channels bind directly to the plasma membrane SNARE proteins syntaxin and SNAP-25 and the vesicle SNARE protein synaptotagmin through a synaptic protein interaction (synprint) site in the large intracellular loop connecting domains II and III of their α1 subunits, an interaction that is required for efficient synaptic transmission. These presynaptic Cav2.1 channels are subject to feedback regulation by SNARE proteins, G proteins, and calcium. Binding of SNAP-25 inhibits calcium channel activity while subsequent binding of the synaptic vesicle protein synaptotagmin relieves this inhibition and prepares the calcium channel for stimulation of exocytosis. G proteins inhibit calcium channel activity in response to activation of G protein-coupled receptors in nerve terminals by released neurotransmitters. Binding of the Gβγ subunit complex to a site including the intracellular loop between domains I and II inhibits calcium channel activation, and strong depolarization relieves this inhibition to yield facilitation. Calcium-dependent regulation involves binding of calcium, calmodulin, and calcium binding protein 1 to a site in the Cterminal domain. This binding interaction causes facilitation and enhanced inactivation. This calcium-dependent facilitation may be responsible for paired-pulse facilitation and post-tetanic facilitation of synaptic transmission. L-type calcium channels containing al subunits of the Cav1 family are responsible for initiation of contraction in muscle cells, secretion in endocrine cells, and gene transcription and other cellular processes in neurons. They are strongly up-regulated by activation of the cAMP signaling pathway through G protein coupled receptors such as the β-adrenergic receptor. Activation of cAMPdependent protein kinase (PKA) rapidly increases channel activity and is required for prepulse potentiation on the msec time scale, suggesting a close association of the kinase and the channel. We have found that effective regulation of L-type calcium channels by PKA requires targeting of the kinase by a novel A Kinase Anchoring Protein, AKAP-15. It binds PKA and targets it to the C-terminal domain calcium channel αl subunit by specific interaction via a modified leucine zipper motif. Peptides that disrupt the leucine zipper interaction prevent calcium channel regulation. These results show that even ion channel modulation by the diffusible second messenger cAMP is tightly organized in space and time by a novel kinase anchoring protein and leucine zipper motif. Evidently spatial organization of signal transduction and ion channel regulation is required for the normal physiological function of these channels. Department of Pharmacology, University of Washington, Seattle, WA 98195-7280 S 13-2 THE MOLECULAR REGULATION OF Kir3.x CHANNELS J.L. Leaney, A. Benians, L.V. Dekker, A. Tinker The Kir3.x family of G protein-gated inwardly rectifying potassium (K+) channels is a member of the superfamily of inwardly rectifying K+ channels. They are predominantly expressed in atrial myocytes, neurones and endocrine cells where they play a role in slowing the heart rate, mediating postsynaptic inhibition and in modulating hormone release, respectively. The channels are activated directly by Gβγ dimers in a membrane-delimited fashion. In native tissue this is in response to stimulation of Gi/o-coupled receptors. In addition to direct Gβγ-mediated channel activation, Kir3.x channels are also thought to be regulated by a number of other signalling molecules, including ATP, Na+ and phosphatidylinositol 4,5-bisphosphate. We have made a number of HEK cell lines stably expressing the cloned counterpart of the neuronal channel, Kir3.1+3.2A either alone or with a G protein-coupled receptor. We then transiently transfected signalling components into these cell lines, along with pEGFP-N1 for visual identification of transfected cells. Membrane currents were studied using the whole-cell patch clamp technique (~140 mM symmetrical K +, ATP- and GTPsupplemented internal solution). Using this heterologous expression

system we have investigated how Kir3.1+3.2A channels are regulated by a number of families of G protein-coupled receptors. We have used Gi/o-, Gs- and Gq/11-coupled receptors and have shown that, similarly to native tissue, Gi/o- but not Gs-coupled receptors, are able to activate the channels. Stimulation of Gq/11-coupled receptor leads to inhibition of channel activity through a mechanism involving a Ca2+-independent novel isoform of protein kinase C. Supported by the Royal Society, the Wellcome Trust and the British Heart Foundation. J.L.L. is a Royal Society Dorothy Hodgkin Fellow. Centre for Clinical Pharmacology, UCL, The Rayne Institute, London WC1E 6JJ, United Kingdom S 13-3 BIOGENESIS AND ASSEMBLY OF VOLTAGE-GATED Na + CHANNELS T. Zimmer, C. Biskup, S. Dugarmaa, C. Bollensdorff, K. Benndorf Voltage-gated Na+ channels are responsible for the upstroke of the action potential in electrically excitable cells. They are heteromeric protein complexes of the plasma membrane consisting of a poreforming α subunit and accessory β subunits. To gain further insight into the subcellular assembly of Na+ channel subunits, we fused spectrally distinct variants of the green fluorescent protein (GFP) to the α, β1, and β2 subunit of the human heart sodium channel (hHl) and expressed these constructs in a mammalian cell line. We observed a strict colocalization of α and β1 subunits within intracellular membrane structures belonging to the endoplasmic reticulum (ER). Fluorescence resonance energy transfer (FRET) measurements indicated a narrow distance between the attached fluorophores suggesting that α/β1 subunit assembly occurs already within the ER. In contrast, α/ β2 subunit complexes were not found in intracellular membrane compartments. To identify the structural elements in the β1 subunit that are responsible for the association with the α subunit within the ER, we tested a series of β 1/β 2 subunit chimeras and deletion variants, and found that the extracellular domain of the β1 subunit is sufficient to mediate a respective colocalization. Our data suggest that α and β1 subunits assemble within the ER by the interaction of their extracellular domains whereas the β2 subunit does not contain a corresponding protein motif interacting with the α subunit. Institute of Physiology II, Friedrich Schiller University Jena, Teichgraben 8, 07740 Jena, Germany

S 13-4 ROLE OF KCNE2 IN THE FUNCTION OF VOLTAGE-GATED K CHANNELS IN THE HEART: IMPLICATIONS FOR ARRHYTHMOGENIC MECHANISM IN LQT6 G.-N. Tseng The minK-related peptide 1 (MiRPl) was cloned in 1999. It is encoded by the KCNE2 gene on chromosome 21. Inherited mutations in KCNE2 gene have been identified in patients with long-QT syndrome (LQT6). Although these patients display only marginal QTC prolongation, they are prone to acquired LQT syndrome. These finding point to the importance of KCNE2 in maintaining the stability of cardiac electrical activity, and in modulating the sensitivity to agents that suppress K channel function in the heart. However, the precise role of KCNE2 in the function of cardiac ion channels, and the mechanism(s) by which KCNE2 mutations increase the risk of arrhythmias, are not clear. The original proposal that KCNE2 is an auxiliary subunit of the rapid delayed rectifier (IKT) channel in the heart has been seriously challenged by recent observations. We will present data to show: (1) KCNE2 can associate with Kv4.2 and Kv4.3, the major pore-forming (α) subunits of cardiac transient outward (Ito) channel, and exert profound effects on their gating kinetics. Importantly, KCNE2 induces an ‘overshoot’ of Kv4.x (x=2 or 3) peak amplitude during recovery from inactivation, resembling the overshoot phenomenon described for I to in human epicardial myocytes. (2) KCNE2 can bind to KCNQl (α subunit of slow delayed rectifier, IKS, channel) simultaneously with KCNEl (obligatory auxiliary subunit of IKS), leading to a decrease in IKS current amplitude without appreciable effects on current kinetics. (3) KCNE2 mRNA can be readily detected in some, although not all, single cardiac myocytes. (4) LQT6-associated mutations in KCNE2 can have differential effects on Ito and IKS: I57T and V65M increase

S 385 one current but decrease the other, while Q9E and M54T decrease one but not the other. We propose that KCNE2 is an important modulatory subunit of Ito and IKS in the heart: it enhances Ito but decreases IKS in a rate-dependent fashion. Differential effects of LQT6-associated mutations on Ito and IKS, superimposed on the intrinsic heterogeneity of these channels in the heart, may contribute to an increase in dispersion of repolarization and predispose the heart to arrhythmias under certain conditions. Department of Physiology, Virginia Commonwealth University, PO Box 980551 Richmond, VA, 23298 USA S 13-5 TUNING NEURONAL ACTIVITY BY A-TYPE CHANNEL TRANSCRIPTION B. Liss, O. Franz, S. Sewing, J. Roeper The activity of dopaminergic (DA) substantia nigra (SN) neurons is essential for voluntary movement control. An intrinsic pacemaker in DA SN neurons generates their tonic spontaneous activity that triggers dopamine release. The molecular mechanisms that define the exact pacemaker frequency within the physiological spectrum of 0.5 to 5 Hz are unknown, however a role for A-type potassium channels has been proposed. We show here, by combining multiplex and real-time quantitative single-cell RT-PCR with patch-clamp techniques in midbrain slice that A-type potassium channels -made up by Kv4.3 and Kchip3.1have a key role in controlling pacemaker activity of mouse DA SN neurons. The number of active A-type potassium channels is not only tightly associated with the pacemaker frequency of individual DA SN neurons, but is also highly correlated with their number of Kv4.3L and Kchip3.1 (long splice variants) mRNA molecules. Consequently, the variation in Kv4α and Kv4β transcript numbers is sufficient to explain the full spectrum of spontaneous pacemaker frequencies in identified DA SN neurons. This linear coupling between Kv4α and Kv4β mRNA abundance, A-type channel density and pacemaker frequency suggests a surprisingly simple molecular mechanism how DA SN neurons tune their variable firing rates by transcriptional control of a ion channel genes.

determining the final amount of sodium reabsorbed. ENaC is localized in the apical membrane of tight epithelia lining e.g. the alveolar cells of the lung. ENaC is a heteromultimeric protein made up of three homologous subunits (alpha, beta and gamma) encoded by three different genes (Scnn1a, Scnn1b, and Scnn1c). Assembly and expression of functional active sodium channels in vitro and in vivo are strictly dependent on the alpha ENaC subunit, the beta and gamma being unable, by themselves, to induce an amiloride-sensitive sodium current. Inactivation of the mouse alpha ENaC (Scnn1a) gene locus by gene targeting in mouse embryonic stem cells revealed a crucial role of this channel in lung liquid clearance at birth (Hummler et al., Nat. Genet. 1996). Alpha ENaC null mutant neonates developed respiratory distress and died within 40 hours after birth. In those animals, amiloridesensitive electrogenic sodium transport was completely abolished. When we reintroduced a rat alpha ENaC transgene into the alpha ENaC knockout background, we recovered about 15% wild type ENaC activity that recued the perinatal lethality and the animals did not longer die due to failure in lung liquid clearance (Hummler et al, Proc. Natl. Acad. Sci. 1997). However, when we put adult transgenic mice under hypoxic conditions (FIO2 0.08), these mice suffer from a defective transepithelial sodium transport and are more predisposed to develop pulmonary edema. To study the mechanisms more in detail, we established primary mouse tracheal cell cultures and measured the amiloride-sensitive potential difference which was lowest in tracheal cells from these transgenic rescue animals (~40%) and intermediate in heterozygous mutant (+/-) mice (60% activity of wild type). We propose that the hypoxia-induced lower ENaC-mediated sodium transport in our transgenic rescue mice might be responsible for increased edema susceptibility. Edith Hummler, Institut de Pharmacologie & Toxicologie, Rue du Bugnon 27, CH-1005 Lausanne, Switzerland

S 14-2

S 14-1

THE MOLECULAR BASIS OF FLUID TRANSPORT IN THE DEVELOPING LUNG. RE Olver, SJ Ramminger, SC Land, SM Wilson. Recent work in our laboratory has focussed on the way in which betaadrenoceptor agonists, glucocorticoid and thyroid hormones interact with alveolar oxygen, which rises sharply at birth, to bring about the phenotypic switch from net secretion to net absorption in the distal lung. We have shown that, in mature rat fetal distal lung epithelial (FDLE) cells, a shift in Po2 over the physiological range of 23 mmHg (fetal) –100 mmHg (neonatal), mimicking the change in the lungs at birth, is associated with an increase in Isc but, whereas the increase in the redox-sensitive transcription factor, NF-kB, and sodium pump capacity are early events (initiated < 6 hr), activation of the alphaENaC promoter is not seen until after 24 hours, reaching a maximum (together with apical Na + conductance; G Na) at 48 hr. In order to determine how the effects of O 2 and hormones known to influence lung epithelial ion transport are integrated to control lung fluid absorption, we have undertaken experiments in mature rat FDLE using a minimal defined serum-free culture medium (MDSF). These studies established that addition of a combination of dexamethasone and T3 increases basal Isc and GNa irrespective of Po2. At fetal and neonatal Po 2 , isoprenaline stimulates G Cl irrespective of hormonal status but the addition of Dexamethasone and T3 is required to enable isoprenaline to upregulate GNa, an effect which is accompanied by a further increase in Glibenclamide-sensitive G Cl. Pump activity at a neonatal, but not fetal, Po2 is stimulated by Dexamethasone and T3 individually and in combination. Thus, it appears that the early increase in fluid absorptive capacity due to the rise in alveolar Po2 at birth is primarily the result of an increase in sodium pump capacity and that the increase in G Na is a secondary event. Both components of the response are enhanced by glucocorticoid and thyroid hormones which are also required for beta-adrenoceptor mediated control of G Na. Although the ENaC promoter contains a NF-kB binding site, and blocking NF-kB activation reduces the O 2-evoked rise in G Na , the discrepancy in the time course of its activation and the rise in ENaC expression indicates that the role of this transcription factor is more complex than at first thought.

ENaC AND LUNG LIQUID CLEARANCE E.Hummler The epithelial sodium channel (ENaC) plays an important role in

Lung membrane Transport Group, Tayside Institute of Child Health, University of Dundee, Ninewells Hospital and Medical School, DD1 9SY

MRC Anatomical Neuropharmacology Unit, Department of Pharmacology, Oxford University, Mansfield Road, OX1 3TH Oxford, UK

S 13-6 EFFECTS OF BUPIVACAINE ON Kv CHANNELS J. Nilsson1, M. Madeja2, P. Arhem1 The molecular mechanisms of anaesthetic action are largely unknown. Local anaesthetics, however, are known to act on voltage-gated channels by binding to the channel protein in a complex state-dependent way. A direct effect on the inactivation gate has been suggested to be essential. Recent studies, however, have challenged this view. In the present study we have investigated this issue as well as the nature of the local anaesthetic binding sites by analysing the effects of bupivacaine on non-inactivating (Kvl.l, 1.2, 1.5, 3.1, 3.2) and inactivating (Kv3.4) K channels. The study was made with the voltage clamp technique on cloned channels expressed in Xenopus oocytes. The results show that the channels are blocked by open-state dependent binding to sites in the internal vestibule, K d values being approximately equal for the non-inactivating channels (110 and 240 µM). By using the estimated affinity values for the non-inactivating channels, we show with numerical simulations that the action on inactivating channels can be explained without any direct effect on the inactivation gate. 1

Nobel Institute for Neurophysiology and Department of Neuroscience, Karolinska Institutet, S-177 77 Stockholm, Sweden; 2Institut fur Physiologie, D-48149 Münster, Germany

S 386 S 14-3

S 14-5

OXYGEN-EVOKED CHANGES IN TRANSCRIPTIONAL ACTIVITY OF THE αENaC PROMOTOR D. L. Baines, M. Janes The lung epithelium is exposed to changes in PO2 at the time of birth, in disease states (such as Chronic Obstructive Pulmonary Disease) and at high altitude. In fetal distal lung epithelial (FDLE) cells, an increase in P O 2 from 23mmHg (fetal) to 100mmHg (post natal alveolar) or 142 mmHg (incubator air) has been shown to increase amiloride-sensitive sodium transport and raise cellular protein and mRNA levels of the α-subunit of the amiloride-sensitive sodium channel ENaC. PO 2 also evokes an increase in transcriptional activity of a 5' flanking region of the αENaC gene and the activity of the redox-sensitive transcription factor nuclear factor-κB (NF-κB) (Baines et al. 2001). Using αENaC-luciferase reporter constructs transiently transfected into the human adult alveolar cell line (A549), we have investigated the relationship between P O 2, NF-kB activation and αENaC transcriptional activity. Transiently raising PO 2 from 23-142 mmHg for 24 hours evoked an increase in NF-κB DNA binding activity and transactivation of a NF-κB driven luciferase construct (pGLNF-κBpro) (P<0.05, n = 4) which was blocked by pre-treatment with 5mM sulphasalazine (SAS), an NF-κB inhibitor. Raising PO2 also increased transcriptional activity of αENaC reporter constructs containing the 3' transcriptional initiation site (TIS) and NF-κB consensus sequence (P<0.05, n = 6). This effect was not blocked by SAS suggesting that the PO2-evoked rise in promotor activity was not mediated by NF-κB in these cells. Conversely, the activity of a construct lacking the 3' TIS but retaining the 5' TIS and NF-kB consensus sequence was significantly suppressed by raising PO2 (P<0.001, n = 4). Pre-treatment with SAS reversed this effect suggesting that activation of NF-κB mediated a PO2-evoked suppression of transcription from the 5' TIS of αENaC. Baines, D. L., Ramminger, S. R., Collet, A., Haddad, J. J., Best, O. G., Land, S. C., Olver, R. E. and EWilson, S. W. (2001). Oxygen-evoked Na+ transport in rat fetal distal lung epithelial cells. J. Physiol., 532.1, 105-113.

REGULATION OF TRANSPORT PATHWAYS BY PURINERGIC STIMULATION IN NATURAL AIRWAY EPITHELIUM O. Frederiksen Extracellular nucleotides have been shown to activate apical Ca 2+dependent Cl- channels (CaCC) and Cl- secretion in amiloride-treated primary cultures of normal and cystic fibrosis human airway epithelia. It has been suggested that nucleotides may also regulate Na+ absorption. In this study we investigated the effects of extracellular nucleotides on paracellular and cellular transport pathways in isolated natural airway epithelium from rabbit nasal septum. Short circuit current (ISC), epithelial conductance (G), and tracer fluxes of Na +, Cl -, and mannitol were measured under short circuit conditions in Ussing chambers, and cell volumes of calcein-loaded epithelia were determined by confocal microscopy. Mucosal (but not serosal) ATP and UTP (up to 200 µM) initially increased and subsequently decreased ISC to a sustained inhibited level. Ion fluxes demonstrated that these changes reflected a small increase in Cl- secretion and a large decrease in amiloride-sensitive Na+ absorption. Calculations based on changes in ISC and G by mucosal amiloride demonstrate that paracellular G contributes by ∼80% and apical Na+ conductance by only ∼3% to total G. Since ATP and UTP not only inhibit apical Na + channels but also stimulate CaCC the inhibitions of epithelial G by ATP (16%) or UTP (31%) have to involve a decrease in paracellular G. This was supported by flux measurements demonstrating a decrease in passive (paracellular) Clflux without changes in mannitol and passive Na+ fluxes. Ionomycin acted in the same way as ATP and UTP on ISC and G and pretreatment with ionomycin largely prevented the nucleotide effects. Nucleotides caused a decrease in cell volume. The results suggest that nucleotides released to the airway surface liquid exert an autocrine regulation of epithelial NaCl absorption via P2Y receptors and an increase in [Ca2+]i, not only by inhibiting amiloride-sensitive ENaC channels but also by a down regulation of paracellular G and anion selectivity. The paracellular effect may be caused by a decrease in epithelial cell volume elicited by e.g. extracellular hypertonicity, inhibition of ENaC, or increased levels of [Ca2+]i.

St. George’s Hospital and Medical School, Tooting, London SW17 0RE. UK

Department of Medical Physiology, The Panum Institute, Blegdamsvej 3, DK-2200 Copenhagen, Denmark. S 15-1

S 14-4 THE AMPHIBIAN LUNG PROVIDES A PHYSIOLOGICALLY INTACT MODEL FOR INVESTIGATION OF ALVEOLAR ION TRANSPORT REGULATION M. Schnizler, A. Berk, M. Fronius, W.G. Clauss Electrophysiological experiments with native alveolar epithelia from higher vertebrates are hampered by the ramified morphology of their lungs. Therefore, lung cell cultures were established and meanwhile electro-physiologically well characterized. Of course, it cannot be concluded from data obtained in such cell lines without any restriction to physiological conditions in the intact tissue. Alveolar epithelia are covered by a mucus with defined electrolyte composition. An assemblage of different cell types use this mucus for para- and/or autocrine mechanisms which regulate transepithelial ion conductances. The baggy morphology of amphibian lungs allows an easy preparation of their alveolar epithelium for electrophysiological experiments in Ussingtype chambers. We used lungs from Xenopus laevis which were pretreated with adrenocorticotropic hormone (ACTH). ACTH achieved more consistency in electrophysiological parameters and higher reproducibility of experiments. We focused on control of electrogenic Na+ uptake and the amiloride-sensitive part of it and identified several substances with regulatory effects. For example, transepithelial Na + current INa was affected by β-adrenergic transduction mechanisms as well as by intercellular messengers as histamine or prostaglandins. Extracellular trinucleotides and adenosine turned out to be important in regulation of ion transport in the amphibian lung. The Xenopus respiratory epithelium proved to be a valuable model especially for G protein-coupled receptor mediated control of alveolar electrolyte transport. This project was supported by the DFG. Institut für Tierphysiologie, Justus-Liebig-Universität, Wartweg 95, 35392 Giessen

INTRODUCTION TO SPINAL INTERNEURON DEVELOPMENT J.C. Glover Interneurons provide the spinal cord with the capacity to integrate, coordinate, and transmit sensory information and motor action with great flexibility and plasticity. Yet despite the all-important role spinal interneurons play in linking the brain and body, their anatomy, physiology, and development have remained relatively poorly characterized. In recent years, progress towards a more complete understanding of spinal interneurons has accelerated in concert with the tremendous advances being made in neuronal imaging and molecular genetics technology. It is now possible to study the origins, differentiation, synaptic connections, and functional organization of spinal interneurons with increasing power and precision. In this symposium, ”Spinal interneurons: from genesis to function”, we provide an overview of recent advances with a primary focus on the spinal cord of higher vertebrates. Department of Physiology, University of Oslo, PB 1103 Blindern, 0317 Oslo, NORWAY

S 15-2 NEW INSIGHTS INTO THE ORGANIZATION OF THE MAMMALIAN LOCOMOTOR CPG O. Kiehn Local circuits in the spinal cord that generate rhythmic locomotor activity in vertebrates are referred to as central pattern generators or CPGs. The precise oganisation of the mammalian CPG is so far unknown. In this talk I will review recent findings that have identified inhibitory CPG neurons involved in left-right alternation during locomotion. I will also review findings showing that midline gate keeping

S 387 controlled by the EphA4 receptor tyrosine kinase and its cognate ligand ephrinB3 is necessary for correct coordination of the spinal locomotor CPG. These findings identify molecular cues controlling the organisation of the mammalian central pattern generator and reveal the first molecular marker for an excitatory component of the mammalian spinal CPG. Department of Neuroscience, Karolinska Institutet, Retzius vag 8, 171 77 Stockholm, Sweden S 15-3 INTERNEURONS AND THE DEVELOPMENT OF PATTERNED ACTIVITY IN THE SPINAL CORD. M.J. O’Donovan During development, spinal networks of the chick embryo are spontaneously active in recurring episodes of rhythmic bursting. Episodes last from a few seconds to a minute with interepisode intervals ranging from 2-20 minutes. During these episodes, hindlimb flexor and extensor motoneurons alternate their discharge in a manner resembling their mature motor function. We have investigated the role of spinal motoneurons and interneurons in the initiation and patterning of motoneuron discharge. At the onset of a spontaneous episode, calcium and voltage-sensitive dye imaging indicates that the earliest activity occurs in motoneurons. Electrical recordings from the muscle nerves and ventral roots have confirmed the imaging experiments, and revealed that some motoneurons begin to fire at about 1/3 of the inter-episode interval. The motoneuron discharge activates a population of interneurons (R-interneurons) that appear to be the avian homolog of the mammalian Renshaw cell. Spike triggered averaging has revealed that R-interneurons make monosynaptic – primarily depolarizing GABAergic - connections with motoneurons. Calcium imaging has shown that R-interneurons are among the first to be activated at the onset of spontaneous episodes suggesting that they may transmit the initial motoneuron discharge to the rest of the spinal network. The origin of the pre-episode motoneuron discharge is unknown but may derive, in part, from a GABA-mediated interepisode depolarization. In addition, just before an episode, motoneurons receive large depolarizing transients that are accompanied by a spatially diffuse optical signals from interneurons labelled with calcium dyes. The alternating pattern of flexor and extensor discharge arises because both sets of motoneurons are simultaneously depolarized during each cycle of activity. The flexor motoneurons fire on the rising and falling phases of the depolarization whereas the extensor motoneurons fire only during the peak depolarization. The pause in flexor firing is reduced by injection of neurotransmitter antagonists into the motor nucleus, including those to acetylcholine, glutamate and glycine and is abolished by GABAa antagonist injections. Further evidence indicates that the pause in flexor firing is mediated by a synaptic conductance shunt and that flexor motoneurons are more susceptible to shunting than extensor motoneurons. NINDS, NIH, Bethesda, MD 20892 USA

which appears to be derived from the glial cells. The endothelial cell lipid composition can be altered by incubation with DHA, and this can be shown to modulate the expression and activity of the P-glycoprotein, an important efflux transporter at the BBB. Studies of the mechanisms underlying pathological brain endothelial modulation include investigating the way in which the toxicant fluorocitrate modulates hexose transport by GLUT-1, the site and mode of action of betaamyloid (Aβ) on cell physiology, and the mechanisms by which circulating antibodies in Lupus patients may modulate and damage the BBB. These studies suggest some general principles underlying BBB modulation in health and disease.Supported by the Wellcome Trust, KCL BBB Consortium with Industry (Lilly, AstraZeneca, GlaxoSmithKline, Aventis, Mindset), and Lupus UK. Centre for Neuroscience Research, King’s College London, London SE1 1UL, UK S 16-2 STRUCTURE AND FUNCTION OF BLOOD-BRAIN BARRIER TIGHT JUNCTIONS 1 H. Wolburg, S. Liebner, A. Lippoldt, K. Wolburg-Buchholz, B. Engelhardt, 1G. Rascher-Eggstein Tight Junctions (TJs) are responsible for the restriction of the transendothelial permeability in the blood-brain barrier (BBB). In freeze-fracture replicas, they form strands of particles across the junctional membrane, which represent the most complex network in the body’s vasculature. The particles are associated with the protoplasmic (P-face) or external fracture face (E-face), whereas all endothelial cells (ECs) outside the CNS are predominantly connected by E-face-associated TJs. The BBB TJs are characterized by the highest P-face association among all EC TJs in the body. In culture, they lose their high P-face association and resemble TJs of peripheral blood vessel ECs (Wolburg et al. 1994, J Cell Sci 107: 1347). The molecular composition of TJs is dominated by occludin and the claudins (Tsukita et al., 2001 Nature Rev Mol Cell Biol 2: 285). In peripheral blood vessels, claudin-5 is the only claudin. In BBB ECs, claudin-3 and possibly claudin-1 are expressed together with claudin-5. Since claudin-1 and claudin-3 are associated with the P-face and claudin-5 with the Eface, we suggest that the ratio of E- and P-face-associations corresponds to that of the expression of claudin-1/3 and claudin-5. This ratio is regulated under physiological and dysregulated under pathological conditions such as cell culture (Liebner et al., 2000 EJCB 79: 707), glioma (Liebner et al., 2000 Acta Neuropathol 100: 323), and EAE (Hamm et al., submitted). On the other side, in the hypertensive SHRSP rat, the E-face-association of BBB TJs was enhanced without an alteration of junctional proteins as monitored by means of immunocytochemistry (Lippoldt et al., 2000 Brain Res 885: 251). In this case, we suggest the involvement of second messenger pathways regulating the linkage of TJs proteins with the cytoskeleton. Apparently, another essential role in TJ regulation plays the extracellular matrix (ECM). In leaky glioma vessels, occludin and/or claudin-1 is downregulated together with the heparan sulfate proteoglycan agrin whereas tenascin is upregulated suggesting that the ECM controls BBB permeability properties (Rascher et al., submitted).

S 16-1 1

MODULATION OF THE CEREBRAL ENDOTHELIUM: PHYSIOLOGY AND PATHOLOGY N.J. Abbott The cerebral endothelium forming the blood-brain barrier (BBB) regulates molecular traffic between blood and brain, and contributes to homeostasis of the brain microenvironment. It is characterised by extremely tight zonulae occludentes (tight junctions), and expresses a number of transport systems for specific uptake and efflux. However, the properties of the barrier are not fixed, but are subject to modulation, both in normal physiology and in pathological situations. Recent work from our laboratory has investigated the physiological modulation of endothelial intracellular calcium and junctional tightness by receptormediated processes. A number of receptors including bradykinin and nucleotide receptors are present on brain endothelial cells, and the receptor phenotype/profile appears to be influenced by the differentiation state of the cells. Parallel studies of intracellular calcium and transendothelial electrical resistance permit dissection of signal transduction mechanisms underlying tight junction modulation. When brain endothelial cells are co-cultured with astrocytic glial cells, they show increased levels of the fatty acid docosahexaenoic acid (DHA),

Inst. Pathol. Univ. Tübingen, Liebermeisterstr. 8, 72076 Tübingen, Germany S 16-3 MULTIDRUG EFFLUX PUMPS IN THE CEREBRAL ENDOTHELIUM M.A.Barrand.,G.C. Bennett, H.C.Cooray, R.A.Felix and O.M. Nwaozuzu Of the different cell layers separating peripheral and central compartments, the brain capillary endothelium constitutes the largest surface area for interchange of material between brain and blood. Various transporters assist in this interchange eg organic anion transporters (OATs), organic cation transporters (OCTs), but multidrug transporter, P-glycoprotein (Pgp) and multidrug resistance associated proteins (MRPs) are probably the main efflux transporters. The importance of Pgp at the blood-brain barrier (BBB) is now well established. Studies on knock-out mice have shown that lack of this transporter has a profound effect on the pharmacokinetics of many drugs with radical changes in drug absorption, elimination and particular distribution to the brain. The precise location of Pgp is still a matter of debate with

S 388 some suggesting it is located on glia and on astrocyte foot processes, others detecting it primarily on endothelial cells lining brain microvessels. Pgp may not only expel unwanted xenobiotics but also play a role in movement of endogenous substances with intriguing suggestions that it may, by so doing, discourage cells from undergoing apoptosis. Pgp expression can be profoundly affected by damaging species eg NO, TNF-α and H2O2 that are produced in various pathological situations. The pathways by which Pgp is upregulated during oxidant stress are currently being explored. Less certain is the presence and roles of MRPs in BBB function. MRPl is often expressed in cultured cells so its detection in vitro in brain endothelial cells cannot be taken as proof of its existence in vivo in brain microvessels. Immunological staining can provide some evidence but relies heavily on availability of suitable antibodies. We are currently employing laser capture microdissection to lift microvessels directly from sections of frozen brain, extract RNA and thereafter analyse expression of specific mRNA species via RT-PCR or microarrays. This combination of techniques can be used to establish whether any or all of these transporters are actually present in vivo but in addition allows analysis of changes in transporter expression in microvessels in different brain regions eg in vasculature supplying brain tumours and following pathological events eg ischemia/reperfusion. Pgp and MRPs are members of the ATP Binding Cassette (ABC) family. There is now evidence from mapping of the human genome for the existence of as many as 51 different ABC proteins. Whether other proteins in this family are present at the BBB and act there as efflux transporters is still not established. Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD UK S 16-4 GLUCOSE TRANSPORT ACROSS THE BLOOD-BRAIN BARRIER IN HEALTH AND DISEASE W. Kuschinsky, R. Duelli The existence of a blood-brain barrier which comprises tight junctions between the endothelial cells necessitates the transcellular transport of glucose from the blood to the brain through the luminal and the abluminal membranes of the brain endothelial cells. Among the 6 sodium independent types of glucose transporters which exist in the organism, 2 are relevant for the brain: GLUT1 which exists in both membranes of the endothelial cells and in the astrocytic endfeet and GLUT3 which covers the membranes of the neurons. GLUT1 is tightly associated with the blood-brain barrier: All brain structures which contain a blood-brain barrier are equipped with GLUT1 in their endothelial cells whereas GLUT1 cannot be found in any non barrier structure. These transporters GLUT1 and GLUT3 are heterogeneously distributed in the brain. Their densities correlate with the local glucose utilization and, in addition, with the capillary density. The density of GLUT1 can be moderately decreased by 3 weeks’ experimental hyperglycemia whereas a moderate reduction of GLUT3 is measured after one week of hypoglycemia. On a local level a moderate up- and downregulation of glucose transporters could be demonstrated: Local increases of glucose utilization by nicotine infusion for one week or by dehydration for 3 days resulted in parallel increases in the densities of GLUT1 and GLUT3 (between 19 and 39%); local downregulation of glucose transporters is minute (-1 to -7%) in structures of the visual system after one week of visual deprivation. Conclusions: The spatial heterogeneity of energy metabolism in the brain results in spatial heterogeneities of capillary density and glucose transporter density. The long term demand for glucose in each brain structure determines its glucose transporter density. Chronic changes in the local or global energy metabolism of the brain are followed by moderate changes in the glucose transporter densities. Department of Physiology and Pathophysiology, Im Neuenheimer Feld 326, D-69120 Heidelberg S 16-5 LEUCOCYTE-INDUCED BRAIN ENDOTHELIAL CELL SIGNALLING MECHANISMS CONTROL LEUCOCYTE MIGRATION THROUGH THE BLOOD-BRAIN BARRIER J. Greenwood, P. Adamson We have previously reported that ICAM-1 on brain endothelial cells (EC) not only serves as a leucocyte adhesion molecule, but upon

engagement results in EC intracellular signalling responses leading to endothelial facilitation of transendothelial lymphocyte migration. These EC signalling pathways include the tyrosine phosphorylation of the actin binding protein cortactin, focal adhesion kinase (FAK), paxillin and p130cas (Etienne et al, 1988). In addition, there is also an increase in the formation of actin stress-fibres, activation of c-jun kinase (JNK) (Etienne et al, 1998), PLCγ activation and Ca 2+ mobilisation (Etienne-Manneville et al., 2000). The efficient transduction of ICAM-1 mediated signalling responses in brain EC, and consequently transendothelial migration of T-lymphocytes, has also been shown to be critically dependent on the small GTP-binding protein Rho (Etienne et al., 1998; Adamson et al., 1999). To function correctly, Rho proteins require posttranslational prenylation and we have shown that pharmacological inactivation of brain EC Rho proteins, with inhibitors of protein prenyltransferases, effectively reduces lymphocyte migration in vitro and attenuates experimental autoimmune encephalomyelitis (EAE) in Biozzi ABH mice. Furthermore, we have now explored the effect of pharmacologically inhibiting the cholesterol pathway that generates the precursors of these isoprenoid groups. By inhibiting the enzyme 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase, the rate limiting step of cholesterol synthesis, with statins (a class of pharmacological agent commonly used to reduce cholesterol in patients) we have demonstrated both in vitro and in vivo that lymphocyte migration across the vascular barrier can be severely attenuated. It is envisaged that these studies may lead to new clinical therapeutic strategies for treating neuroinflammation. Etienne, S., Adamson, P., Greenwood, J., Strosberg, A.D., Cazaubon, S., Couraud, P-O. (1998). ICAM-1 signaling pathways associated with Rho activation in microvascular brain endothelial cells. J. Immunol. 161:5755-5761. Adamson, P., Etienne, S., Couraud, P.O., Calder, V. and Greenwood, J. (1999). Lymphocyte migration through brain endothelial cell monolayers involves signaling through endothelial ICAM-1 via a Rho-dependent pathway. J. Immunol. 162:2964-2973. Etienne-Manneville, S., Manneville, J.B., Adamson, P., Wilbourn, B., Greenwood, J. and Couraud, P-O. (2000). ICAM-1-coupled cytoskeletal rearrangements and transendothelial lymphocyte migration involve intracellular calcium signaling in brain endothelial cell lines. J. Immunol. 165:3375-3383. Endothelial and Epithelial Cell Biology Research Unit, Division of Cell Biology, Institute of Ophthalmology, UCL, London EC1V 9EL, UK S 17-1 SHORT TERM REGULATION OF SODIUM BALANCE: STUDIES IN CONSCIOUS DOG AND MAN P. Bie, NCF Sandgaard, MS Rasmussen Contributions of the individual components of the complex homeostatic response to salt loading can be identified (i) by clamping or elimination of one or more controllers, or (ii) by use of stimuli designed to allow separation by threshold and time of individual elements. The latter approach was used to investigate the composite reaction to intravenous salt (NaCl) loading. Conscious, trained, hydropenic dogs were studied at low-sodium diet (0.5 mmol/d/kg b.wt.) or high sodium diets. Student volunteers were investigated under similar conditions. NaCl was administered at infusion rates of 6 and 20 µmol/min/kg b.wt. either as isotonic or as hypertonic solutions during simultaneous measurements of haemodynamic, hormonal, and renal variables. Glomerular filtration rate (GFR) was quantified by clearance of exogenous creatinine (dog) or 51Cr-EDTA (man). Loads of 0.54 mmol/kg over 90 min (≈ 1 % of exchangeable sodium) augmented plasma vasopressin in dogs without other measurable effects. Hypertonic loads of 0.6 mmol/kg in 30 min in high-sodium dogs and isotonic loads of 1.2 mmol/kg in 60 min in euhydrated, low-sodium man decreased plasma angiotensin II and initiated marked natriuretic responses (increasing to 5-15 fold elevations) without changes in arterial blood pressure or GFR. Where measured, changes in plasma renin activity and plasma aldosterone concentrations paralleled changes in plasma angiotensin II. In dogs, the response was blunted by low-sodium diet. Plasma concentrations of atrial natriuretic peptide remained constant or showed minimal increase. In man, urinary excretion rates of cGMP and nitrite/nitrate (commercial kits) did not change. The data show that (i) elevations of <1 % in exchangeable sodium activates the vasopressin system, while increases in 1-3 % are required to deactivate the renin system probably through neuro-humoral modulation of renin secretion, (ii) the latter occurs at constant arterial pressure and GFR,

S 389 and (iii) the deactivation mechanisms seem sensitised by high-sodium diet. The primary response to modest, acute sodium loading seems to be neurohumoral deactivation of the renin system. Dept. Physiology and Pharmacology, University of Southern Denmark, 21 Winslowparken, DK-5000, Odense, Denmark. E-mail: [email protected]

evidence for angiotensin II being involved in setting the level at which this reflex operates. Together, these studies indicate that elevation of dietary sodium intake over the growing phase enhances the influence of angiotensin II in modulating the reflex neural regulation of the kidney and its function. Department of Physiology, University College Cork, Cork, Ireland

S 17-2

S 17-4

RENAL GENE EXPRESSION AND FLUID HOMEOSTASIS : STUDIES IN CONSCIOUS ANIMALS B. Nafz*, P.B. Persson Several decades ago Guyton and his co-workers established the concept of “body fluid pressure control”. According to this model lasting increases or decreases in systemic blood pressure (BP) establish a hallmark of long- term changes in fluid homeostasis. In the recent years, there have been a vast number of studies focusing on the interdependence between fluid homeostasis and blood pressure control in order to determine which intrarenal mechanisms interrelate long- term BP with renal fluid excretion and total body water (TBW). Nonetheless, the exact pathways which can explain the aforementioned phenomenon remain unclear. One reason for this can be seen in the fact that changes in TBW and systemic BP inevitably modulate many systems of cardiovascular control. Remarkably, it has been recognised that BP elevations coincide with changes in the dynamic properties of BP. This may be explained by the fact that several systems which are known to play an important role in the regulation of mean BP are also involved in buffering fast BP fluctuations. In spite of the widespread investigations on long- term mean values of BP and fluid excretion, astonishingly little is known about the importance of these short- term BP waves for cellular mechanisms of BP regulation and fluid homeostasis. This overview summarises first efforts to unravel the influence of graded reductions in renal perfusion pressure (RPP) and of RPP oscillations on different components of the renal renin- angiotensin- system, local hemodynamics and on renal fluid handling in conscious animals. It seems that induced oscillations in RPP can play a major role for local hemodynamics, renal gene expression and renal fluid and electrolyte excretion.

RELATIONSHIP BETWEEN PLASMA VOLUME AND CARDIOVASCULAR HOMEOSTASIS IN MAN: EFFECT OF DIETARY SALT, EXERCISE AND SLEEPING POSTURE R. Hainsworth Standing motionless upright results in accumulation of blood in dependent veins and loss of plasma through dependent capillaries. Cardiac output is reduced but, initially at least, blood pressure is normally maintained. Some apparently healthy subjects, however, are unable to tolerate orthostatic stress and experience troublesome attacks of posturally-related syncope. We hypothesised that a subject’s orthostatic tolerance would be partly dependent on his (normalized) plasma volume and that procedures which would increase plasma volume would also increase orthostatic tolerance. We assessed orthostatic tolerance as the time to presyncope in a test combining head-up tilting and lower body suction, and plasma volume from the dilution of Evans blue dye. In an initial study of 49 patients and volunteers we showed that there was indeed a highly significant correlation between orthostatic tolerance and plasma volume. We also found that, in 178 patients with unexplained attacks of syncope, those consuming higher quantities of salt, as indicated by a daily sodium excretion of over 170 mmol, had significantly greater orthostatic tolerance than those consuming less salt. On this basis it seemed logical to adopt measures which would be expected to increase plasma volume. In a placebo-controlled trial, we found that salt loading (120mmol/day) usually increased plasma volume and, if it did, orthostatic tolerance also increased. Exercise training is another measure known to increase blood and plasma volumes. Patients who successfully trained increased both plasma volume and orthostatic tolerance, although results in normal fit subjects were more variable. Another measure which is sometimes used in the management of patients with posturally-related syncope is to arrange for them to sleep head-up (12o). We found that this too usually increased plasma volume and, it if did it, also increased orthostatic tolerance. This research has established a strong link between orthostatic tolerance and plasma volume and has indicated that measures that increase plasma volume also increase orthostatic tolerance and are likely to be of benefit to patients with posturally-related syncope.

Medizinische Fakultät der Humboldt Universität (Charité), Johannes Müller - Institut für Physiologie, Tucholskystraße 2, 10117 Berlin, email: [email protected] S 17-3 INFLUENCE OF DIETARY SODIUM INTAKE ON THE NEURAL CONTROL OF THE KIDNEY E. J. Johns A number of human trials have reported a weak relationship between the level of dietary sodium intake and the predisposition to hypertension. Consequently, the exact manner by which sodium intake can determine whether a chronic elevation of blood pressure occurs has not been resolved. A number of animal studies have shown that elevating dietary sodium intake, particularly during the growing and maturation phase of life, causes an increased blood pressure in the adult associated with a raised autonomic control. Moreover, administration of a low dose slow infusion of angiotensin II into the brain (icv) caused a sustained increase in blood pressure and a renal nerve-dependent antinatriuresis. Importantly, these effects were not evident in mature animals exposed to a similar high sodium diet challenge. The aim of our investigations was to examine the role of the brain renin-angiotensin system in mediating the autonomic control of the kidney when animals were subjected to a high sodium intake in the growing phase. Groups of rats were placed on either a low (0.03%Na), normal (0.25%Na) or high (3.1%Na) from 4 weeks of age for either 4 or 8 weeks. Four weeks of high sodium diet was found to increase blood pressure and to elevate baroreflex gain control of renal sympathetic nerve activity, but not heart rate, which was normalised by icv captopril and restored by coadministration of angtiotensin II icv indicating an important role for this peptide in the brain. Activation of the somato-sensory system, by giving capsaicin sc, in rats on a normal sodium intake resulted in a renal nerve -dependent antinatriuresis which could be prevented by icv losartan. In rats exposed to either 4 or 8 weeks of the high sodium diet, twice as much losartan had to be given suggesting a greater influence of angiotensin II in these states. These observations provide further

Institute for Cardiovascular Research, University of Leeds, Leeds, LS2 9JT, UK S 18-1 PARACRINE FACTORS REGULATING FETAL PLACENTAL BLOOD FLOW A.M. Carter, R.I. Jensen, O. Skott , B.L. Jensen The umbilical arteries branch across the fetal surface of the placenta to supply some 20-40 villous trees. The fetal capillaries are found within the terminal villi of these trees. Blood returns to the fetus through a single umbilical vein. Uniquely, all these vessels are devoid of innervation. Regulation of fetal placental blood flow is by endocrine or paracrine mechanisms. Angiotensin-II and atrial natriuretic peptide, respectively, are able to constrict or dilate fetal placental arteries. However, the normal concentrations of these hormones in fetal plasma are far below the effective concentrations required for half-maximal response in vitro. Paracrine control of fetal placental blood flow may require a balance between vasoconstrictor molecules, such as endothelin-1, and vasodilators like nitric oxide. There is evidence that nitric oxide synthase activity in the endothelium of placental vessels is responsive to alterations in shear stress. In addition, several peptides produced locally within the placenta dilate placental vessels at low concentration. These include corticotrophin releasing factor, adrenocortico-trophic hormone and urocortin. Recently, we examined the expression in sheep placenta of another vasodilator peptide, adrenomedullin (ADM). The relative abundance of ADM mRNA, determined by ribonuclease protection assay with an ovine cRNA probe, was many fold higher in placenta than in any fetal organ

S 390 examined. ADM protein expression in placenta, especially in fetal mesenchyme, was shown by immunohistochemistry, using a polyclonal rat anti-ADM antibody. ADM was present at high concentration in fetal plasma (457±20 pg/ml) and possibly placental in origin. This suggests a potential role for ADM in maintaining vessel patency in the ovine placental circulation. In human placenta, either ADM or the closely related molecule, calcitonin-gene related peptide, may play a similar role. Department of Physiology and Pharmacology, University of Southern Denmark, Winsloewparken 21, Third Floor, DK-5000 Odense, Denmark S 18-2 HCG DILATES RESISTANCE ARTERIES BY DIFFERENT NONENDOTHELIAL MECHANISMS M. Hermsteiner Human chorionic gonadotropin (hCG), one of the most important hormones in human pregnancy, which is secreted in large quantities by the syncytiotrophoblast, dilates uterine as well as systemic resistance arteries within one to three minutes in a dose-dependent manner (EC50 in the range of 50 mU/ml for both types of vessel). The complete mechanical removal of the vascular endothelium does not reduce the efficacy or potency of hCG. In addition, when the vessels are exposed for 30 to 60 minutes, hCG attenuates the vasoconstrictor response of systemic resistance arteries to angiotensin II, obviously by an endothelial mechanism involving the increased production of NO and prostacyclin. These two vascular effects of hCG implicate a major contribution of the hormone not only to the maintenance of a sufficient utero-placental perfusion throughout gestation, but also to the marked early post-conceptional decrease of the total peripheral resistance in the maternal circulation. Objectives: To examine whether the acute vasodilatory effect of hCG is mediated by the classical heptahelical Gs protein coupled membrane receptor for luteinizing hormone (LHR), and to investigate the intracellular response to hCG by a functional method. Material and methods: The endothelium of isolated second and third order mesenteric arteries of virgin female Sprague-Dawley rats was ”brushed off” by a horse hair fed through the lumen. The vessels were then cannulated in the arteriograph, and kept at a constant transmural pressure of 50 mm Hg. The effect of various modulator substances on vessel diameter was studied by means of a videoelectronic servo-control unit. Results: Vasodilation by hCG was turned into moderate, and hardly dose-dependent constriction by either applying NF449 (1.5x10-5 mol/l), a Gsα−selective G protein antagonist, or 2´5´-dideoxyadenosine (5x10 -6 mol/l), an inhibitor of the membrane-bound adenylyl cyclase (AC), or H-89 (1x10-7 mol/l), a selective blocker of proteinkinase A (PKA). A similar effect was seen in the presence of methylene blue (1x10 -5 mol/l), an inhibitor of soluble guanylate cyclase (GC), or L-NAME (1x10-4 mol/l), an inhibitor of NO-synthase (NOS). The vasoconstriction, observed under the various conditions, could always be prevented by U-73122 (2.5x10-6 mol/ l), a specific inhibitor of phospholipase C (PLC). Glibenclamide (1x106 mol/l), a selective blocker of KATP-channels, reduced the maximum dilatory response to hCG by 56% and shifted EC50 from 27±3.4 mU/ml to 104±13.5 mU/ml. The response to hCG could not further be reduced by addition of TEA (10 mmol/l), a non-selective K-channel blocker. Conclusion: HCG acutely dilates systemic resistance arteries by a LHR mediated mechanism. Stimulation of AC by Gsα leads to elevation of intracellular cAMP-levels, by which PKA is activated. PKA in turn opens KATP-channels causing membrane hyperpolarization. In parallel, hCG seems to induce the synthesis of NO within the vascular smooth muscle, thereby activating the cGMP-pathway. We assume extensive cross-talk between cAMP- and cGMP-pathways at different levels. Dept. of Obstetrics & Gynecology, Justus Liebig University, Klinikstr. 32, D-35392 Gießen, Germany S 18-3 PLASMA VOLUME AND NITRITE OUTPUT DURING PREGNANCY IN WILDTYPE MICE AND TRANSGENIC ENOS-/- MICE W. Moll, L. Zhang, E. Tartler, A. Gödecke, T. Spruss During pregnancy, plasma volume increases up to 50% in species so far studied (man, guinea pig, rat). The cause of plasma volume increase is not yet established; the effect of blockade of nitric oxide synthase on

plasma volume in the pregnant rat (Zhang and Kaufman 2000) suggests a role of one or several nitric oxide synthases in gestational plasma volume control. In order to see the plasma volume in an animal where maternal weight doubles during pregnancy and to learn whether endothelial nitric oxide synthase (eNOS) is obligatory for plasma volume expansion, we measured plasma volume and urinary nitrate output in pregnant wild type mice (C57 Bl/N6) and in a mouse strain derived from transgenic mice with defective eNOS gene (eNOS/-; Gödecke et al. 1998). Plasma volume was determined in anaethetized mice from the concentration of Evans blue injected into the retrobulbar plexus (50 µl, 2µg/µl). Urinary nitrite output was determined from the urinary nitrate concentration (assayed after reduction to nitrite according Sohn and Fialy (2000) with Griess reagent) and urine volume (derived from creatine output over creatinine urine concentration). We found that plasma volume increases 50% in pregnant mice, from (46 ±5) ml/kg in prepregnant mice to (68±5) ml/kg maternal weight in late pregnant mice (mean±SD). There was no statistically significant difference between wild type and transgenic mice. Nitrate output was around (1.6±1) µmol/d in wildtype and (0.8±0.5) µmol/d in transgenic mice. There was no change in pregnancy in either group. The low nitrate output shows decreased NO formation in the transgenic group even if the occurrence of eNOS positive individuals in the strain was not excluded. We conclude that gestational plasma volume increase in mice is similar to that in other species and that eNOS, if absent from the very early stage of life, is no obligatory factor of plasma volume increase in mice. Physiologisches Institut der Universität Regensburg, Universitätsstr. 31, D-93040 Regensburg S 18-4 EFFECTS OF ANTENATAL GLUCOCORTICOIDS ON CEREBRAL BLOOD FLOW (CBF) IN THE OVINE FETUS AT DIFFERENT GESTATIONAL AGES M. Löhle1, C. Wicher2, T. Müller2, H. Schubert2, M. Schwab1 Objective: We have shown betamethasone (βM) at the dose clinically used to enhance fetal lung maturation decreases CBF by 30-50% in fetal sheep at 128 dGA (0.83 of gestation) (JPhysiol 528: 619-632, 2000). The goal of the present study was to determine the effects of βM on fetal CBF before the increase of endogenous plasma cortisol levels during late gestation (Endocrinology 107: 155-159, 1980). Methods: Starting at 110 dGA (0.75 of gestation) βΜ (10 µg/h, n = 6) or vehicle (n = 6) was infused over 48 h directly to the external jugular vein of chronically instrumented ovine fetuses. βM concentration in the fetal plasma reached values similar to human umbilical vein concentration 24h after maternal βM treatment. Regional CBF was estimated in 10 brain regions using fluorescent microspheres before, 24 h, and 48 h after onset of βM or vehicle treatment. Results: Total CBF at 110 dGA was 53% of that measured at 128 dGA (92 vs. 174 ml l00g1 min-1, p < 0.001). Regional CBF at 110 dGA ranged from 67 ml l00g1 min-1 in the cerebral cortex to 170 ml l00g-1 min-1 in the brain stem. CBF decreased between approximately 40% (cerebral cortex) and 50% (brain stem) within 24 h after onset of βM infusion (p < 0.05) and remained decreased during βM infusion. This decrease of CBF during βM infusion was associated to an increase in vascular resistance (p < 0.05) and not significantly different to that found at 128 dGA. Conclusions: Effects of βM on CBF were not different at 110 and 128 dGA and, thus, independent of maturation of the hypothalamic-pituitary-adrenal axis. 1 Department of Neurology, 2Institute of Laboratory Animal Science Friedrich Schiller University, 07740 Jena, Germany

S 18-5 ALTERED BLOOD PERFUSION IN FETAL LIVER FOLLOWING OCCLUSION OR DILATATION OF THE DUCTUS VENOSUS CAN REGULATE CELL PROLIFERATION IN FETAL ORGANS M. Tchirikov1, S. Kertschanska2, H.J. Schröder1. Objective: To investigate whether fetal growth is regulated by liver blood perfusion. Methods: In 9 ewes, which had twin pregnancies at gestational ages of 119±2 days, a dilating stent (diameter 4 mm) was placed into the ductus venosus (DV) in one twin (DVstent group). In 17 near term sheep with twin (n=11) or singleton (n=6) pregnancies the

S 391 DV was blocked with an embolisation coil (DVcoil group) for about one week. Umbilical and DV blood flow rate were measured using ultrasound Doppler (Acuson Aspen). Cell proliferation rate (pKi-67) was determined in liver, heart, skeletal muscle, kidneys and placenta (labeling with Mib-1). In 5 animals of DV coil group IGF-I and IGF-II mRNA content in the liver were estimated by in-situ hybridization (courtesy of V. Han). Results: Dilatation or occlusion of the DV did not change placental perfusion on the 1st day or later after surgery. The liver blood supply was decreased in DVstent group from 499±371 to 278±219 ml min-1 (n=4), and increased twofold in DVcoil group. (p<0.05). Relative liver weight (% body weight) was decreased from 3.9±0.6% (control twin) to 3.0-±0.2% (n=3) in DVstent group. Occlusion of the DV led to the increase of relative liver weight from 3.4±0.8% to 4.3±0.8% (n=11, p<0.05). The increased liver blood perfusion following occlusion of the DV was associated with increased cell proliferation in the liver (sixfold, n=8, p<0.005) and in heart muscle, skeletal muscle and the kidneys (twofold, p<0.05), but unaltered proliferation in the placenta. Reduced liver blood supply in DVstent was associated with reduction of cell proliferation in the liver (from 12.4±2.3 to 6.5±0.6), in heart (from 1.1±0.03 to 0.9±0.02) and skeletal muscle (from 0.82±0.05 to 0.5±0.01 (n=3, p<0-05, numbers of Mib-1 positive nuclei per µm2 10-4). Increased liver blood supply in DVcoil group increased mRNA expression for IGF-I and IGF-II in the liver, however, the concentration of IGF-I and IGF-II in fetal plasma were not different between fetuses with occluded ductus venosus and control twin fetuses [70±56 and 70±58 (ng ml-1)for IGF-I, and 588±310 and 733±344(ng ml-1)for IGF-II, respectively]. Conclusion: Liver blood perfusion regulates cell proliferation in fetal sheep.

nial pressure (ICP) in response to severe brain trauma is higher in the immature than in the mature brain. In the current study we asked whether this may be attributed to developmental changes in the intracranial pressure volume relation. In order to answer this question we measured the cerebral compliance (CC) and calculated the pressure volume index (PVI: volume causing a tenfold increase in instantaneous ICP) in three groups (n=6 per group) of newborn (1-2 days), 10d (8-12 days old), and 7w (6-8 weeks old) anesthetized piglets, using the modified bolus injection method. In addition, the effect of hypercapnia induced changes in cerebrovascular tone on CC and PVI was studied in 10d and 7w piglets. We found an age dependent increase in the PVI with 0.34±0.09 ml, 0.71±0.08 ml and 1.11±0.09 ml in 1d, 10d and 7w piglets, respectively (p<0.05). However, when PVI was related to brain weight, only 1d piglets showed a significantly lower PVI (p<0.05) compared to 10d and 7w piglets. Hypercapnic induced cerebral vasodilatation led to an increased PVI when ICP was kept in the normal range. Our data indicate that the total volume buffering capacity of the craniospinal compartment in the developing piglet - as in humans – is a function of age, making this animal model suitable for studying effects of cerebral insults on the intracranial pressure volume relation. However, equivalent increases in brain volume (e.g. due to brain edema) may be more injuring in the newborn brain compared to other age groups. Furthermore, CC depends also on cerebrovascular tone, and may be increased even in the presence of an increased cerebral blood volume. Supported by BMBF B378-01010 Institut für Pathophysiologie, Klinikum der Friedrich-Schiller-Universität, D-07740 Jena, Germany ([email protected])

1 Obstet./Gynecol, Universitätsklinkum Hamburg-Eppendorf, 2Obstet./ Gynecol, Universitätsklinkum Hamburg-Eppendorf Hamburg, Abtl. Exper. Med., 20246, Germany

S 18-6 EFFECT OF INTRAUTERINE GROWTH RESTRICTION ON CEREBROVASCULAR REGULATION IN NEWBORNS R. Bauer, B. Walter, U. Zwiener Brain weight development is only less altered in consequence of asymmetrical intrauterine growth restriction (IUGR) compared to other organs and tissues (brain sparing effect). Therefore, an increased fraction of the cardiac output is required for sufficient brain blood supply. In addition, IUGR is associated with an increased incidence of perinatal asphyxia, which results in cardiocirculatory redistribution. However, until now it is unknown whether an altered cerebrovascular regulation in IUGR newborns exists. Therefore, we studied (i) the effects of gradual reduction of cerebral perfusion pressure (CPP), (ii) the effects of normocapnic hypoxia, and (iii) the effects of hypercapnic hypoxia on regional cerebral blood flow (CBF) and cerebral oxidative metabolism (CMRO2) in 1-day old anesthetized normal (n=46) and IUGR (n=47) newborn piglets. We found a reduced CBF and CMRO2 in normal piglets even during moderate CPP (~58% of control) (P < 0.05) whereas IUGR showed an impaired brain oxidative supply only on pronounced CPP reduction (~45% of control) (P < 0.05). In contrast, a difference in brain oxidative supply did not occur between normal and IUGR newborn piglets in consequence of moderate and severe systemic hypoxia. However, there was marked difference in the effectiveness of brain oxygen delivery during hypercapnic hypoxia: Whereas normal piglets required a pronounced CBF increase (due to reduced O2 extraction rate) (P < 0.05), IUGR piglets showed a similar CBF increase, as shown during normocapnic hypoxia. Thus, newborn IUGR piglets exhibited an improved cerebrovascular/ metabolic adaptation overcoming periods with reduced CPP and an increase compensatory capacity to survive asphyxiated events. Institut für Pathophysiologie, Klinikum der Friedrich-Schiller-Universität, D-07740 Jena, Germany ([email protected]) S 18-7 THE EFFECT OF NEURAL DEVELOPMENT AND CEREBRAL VASCULAR TONE ON THE INTRACRANIAL PRESSURE VOLUME RELATION IN THE DEVELOPING PIGLET A. Nurlubayeva, B. Walter*, R. Bauer, Ch. Kranich, U. Zwiener The incidence of diffuse brain swelling and secondary rise in intracra-

S 18-8 THE WILMS’ TUMOR TRANSCRIPTION FACTOR WT1 IS EXPRESSED IN THE CORONARY VASCULATURE AFTER MYOCARDIAL INFARCTION K.D. Wagner1, N. Wagner1, A. Bondke1, B. Nafz1, B. Flemming1, H. Theres2, H. Scholz1 In addition to its role as a suppressor of malignant tumor growth, the Wilms’ tumor transcription factor Wt1 is critical for the development of different organs. Among other malformations, mouse embryos with targeted inactivation of the Wt1 gene exhibit a failure of normal cardiac development. Thinning of the muscular walls of the Wt1-/hearts is thought to result from disrupted formation of the epicardium, which normally expresses Wt1 in wild-type animals. Since Wt1 is required for normal cardiac myocyte growth during development, we reasoned that it may have a role also in cardiac growth in the adult organism. In order to test this hypothesis, we studied the expression of Wt1 in different models of cardiac hypertrophy in rats. No significant differences in Wt1 expression were detected by RNase protection assay between the hearts of normal rats and the hypertrophied left ventricles of spontaneously hypertensive rats (SHR) and rats harboring angiotensinogen and renin as transgenes (TGR), respectively. However, Wt1 transcripts were increased more than 3-fold in the left ventricular tissue between 1 day and 9 weeks after experimental myocardial infarction. Most interestingly, Wt1 expression was detected by in situ mRNA hybridization and immunohistochemistry in the walls of coronary vessels in the border zone of the infarcts. No Wt1 (mRNA and protein) was present in the coronary vasculature of the noninfarcted right ventricles. Double-immunofluorescent labeling of tissue sections from rat hearts with myocardial infarction revealed that Wt1 was co-localized with the endothelial cell proliferation marker PECAM-1 and proliferating cell nuclear antigen (PCNA) in coronary vessels, which normally do not express Wt1. Up-regulation of Wt1 in the vasculature of ischemic hearts was mimicked by exposure of rats to normobaric hypoxia (8% O2) and 0.1% carbon monoxide, respectively. Within the hypoxic hearts, Wt1 was co-expressed in coronary vessels with the hypoxia inducible factor HIF-1α, which is a master regulator of hypoxia-dependent gene expression. These findings demonstrate for the first time that Wt1 can be activated in adult hearts by hypoxic exposure. The expression of Wt1 in proliferating coronary endothelial cells after myocardial infarction suggests a role for the Wilms’ tumor transcription factor in coronary vessel formation in the (post-)ischemic myocardium.

S 392 1 Johannes-Müller-Institut für Physiologie and 2Klinik für Innere Medizin I, Humboldt-Universität, Charité, Tucholskystrasse 2, 10117 Berlin, Germany

opposing effects, both of which are secondary to the defect in proximal tubular endocytosis, probably determines whether there will be hypercalciuria and kidney stones.

S 19-1

Zentrum für Molekulare Neurobiologie, Martinistrasse 85, D-20246 Hamburg. * Present address: Klinik für Neurochirurgie, Universität Lübeck, Ratzeburger Allee 160, 23538 Lübeck

DEFECTS IN GLUCOSE HOMEOSTASIS IN MICE WITH GENETIC INACTIVATION OF THE FACILITATED DIFFUSION GLUCOSE TRANSPORTER GLUT2 B. Thorens, R. Burcelin The glucose transporter GLUT2 is present in intestine, liver, kidney and pancreatic ß cells. The consequences of GLUT2 gene inactivation in the mouse was studied on the mechanism of glucose sensing linking changes in blood glucose concentrations to insulin secretion by pancreatic ß cells. We showed that in the absence of GLUT2, glucose signalling to insulin secretion, gene transcription and mRNA translation were suppressed. This was associated with early diabetes and premature death of the animals. Reexpression by transgenesis of GLUT1 or GLUT2 selectively in the ß cells completely normalized glucose signaling and allowed mouse survival and breeding. We thus used these rescued mice (RIPG1xG2-/-) to study GLUT2 role in extrapancreatic glucose sensors. We showed that activation of the hepatoportal sensor by a portal-peripheral glucose gradient, which increases glucose utilization by peripheral tissues, was also dependent on GLUT2 expression. More recently, we demonstrated that glucagon secretion during hypo(2.5 mM) or hyperglycemic (10, 20 mM), hyperinsulinemic clamps was no longer increased or suppressed, respectively, in the RIPG1xG2/mice. In addition, these mice were hyperglucagonemic in the fed state due to an increased tone of the autonomic nervous system (ANS) to alpha cells. From these data, we conclude that: i) in the absence of GLUT2 there is an impaired control of glucagon secretion by low or high glucose, ii) this impaired secretory activity is not due to absence of GLUT2 from alpha since these cells do not normally express this transporter, iii) this dysregulation results from inactivation of GLUT2dependent glucose sensors located outside of the endocrine pancreas, iv) the absence of GLUT2 leads to an increased activity of the ANS to the alpha cells. Together, these data demonstrate the role of GLUT2 in multiple glucose sensors controlling a variety of physiological function.

S 19-3 GABA A -RECEPTOR SUBTYPES: DISSECTING THEIR PHARMACOLOGICAL FUNCTIONS U. Rudolph, F. Crestani, K. Löw, R. Keist, H. Möhler The enhancement of GABA-mediated synaptic transmission underlies the pharmacotherapy of various neurological and psychiatric disorders. GABA A receptors are pluripotent drug targets that display an extraordinary structural heterogeneity. In order to differentiate defined GABAA receptor subtypes on the basis of function, we introduced histidine to arginine point mutations which render the respective receptor subtypes insensitive to diazepam into the mouse α1, α2 and α3 subunit genes, respectively, by homologous recombination in embryonic stem cells. These mutations abolish diazepam binding. Behavioural studies revealed that while the sedative action of diazepam (measured as decrease of horizontal motor activity) is mediated by GABAA receptors containing the α1 subunit, its anxiolytic-like activity (as determined in the light/dark choice test and the elevated plus maze test) is mediated by GABAA receptors containing the α2 subunit but not the α3 subunit. Thus, benzodiazpine-induced behavioural responses are mediated by specific GABAA receptor subtypes in distinct neuronal circuits, which is of interest for drug design. In addition, the high density of the α2 subunit on somata and axon initial segments of principal cells in cerebral cortex, hippocampus and amygdala may outline the neuronal circuit which mediates anxiolytic activity. References : Rudolph, U. et al. (1999) Benzodiazepine actions mediated by specific γ-aminobutyric acidA receptor subtypes. Nature 401, 796800. Löw, K. et al. (2000) Molecular and neuronal substrate for the selective attenuation of anxiety. Science 290, 131-134. Rudolph, U. et al. (2001) GABAA receptor subtypes: dissecting their phamacological functions. Trends Pharmacol. Sci. 22, 188-194.

Institute of Pharmacology and Toxicology, University of Lausanne, 27, rue du Bugnon,1005 Lausanne, Switzerland

Institute of Pharmacology and Toxicology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich

S 19-2

S 19-4

THE CLC-5 CHLORIDE CHANNEL AND RENAL ENDOCYTOSIS – LESSONS FROM A MOUSE MODEL FOR DENT´S DISEASE W. Günther*, N. Piwon, M. Bösl, M. Schwake, T. J. Jentsch Dent’s disease is an X-chromosome linked disorder characterized by urinary loss of low-molecular-weight proteins, phosphate and calcium, resulting in nephrocalcinosis and nephrolithiasis. The disorder is due to mutations in ClC-5, a renal chloride channel that is expressed in endocytotic vesicles of proximal tubular cells and in intercalated cells of the distal nephron. We generated a mouse model of this disease by inactivating the clcn5 gene. KO mice showed proteinuria due to strongly reducing apical proximal tubular endocytosis. This included both receptor-mediated and fluid-phase endocytosis. The amount of megalin, an apical recycling receptor of proximal tubular cells that binds many luminal proteins and targets them to the degradative pathway, was reduced in a cell-autonomous manner. In addition, PTH-stimulated internalisation of the apical ion transporters NaPi-2 and NHE3 was slowed, but not abolished. When KO or WT mice were kept under low phosphate diet, the NaPi-2 transporter was found in the brush border membrane in both cases. Under standard phosphate diet, however, this localization was only observed in the first segment of the KO proximal tubule, whereas it was predominantly apically in all PT segments of WT kidneys. This can be explained by an increased stimulation of luminal parathyroid hormone receptors in the KO due to an increased luminal PTH concentration that results from decreased tubular reabsorption of PTH. Indeed, urinary PTH concentration was increased in the KO. The increase in luminal PTH should also increase the activity of the enzyme that hydroxylates 25(OH) vitamin D3 to the active hormone. This was indeed observed. However, this was counteracted by a urinary loss of the precursor 25(OH) vitamin D3 bound to its low molecular weight carrier protein DBP. The balance between these

EPITHELIAL SODIUM CHANNEL: NOVEL INSIGHTS FROM GENE TARGETING EXPERIMENTS E. Hummler The highly amiloride-sensitive epithelial sodium channel (ENaC) is a membrane constituent of many salt-reabsorbing epithelia that facilitates Na+ movement across the tight epithelia that line the distal renal tubule, the distal colon, the ducts of salivary and sweat glands, and the lung. ENaC is a heterodimer composed of three homologous subunits (α, β, and γ). ENaC expression has also been found in epithelia like the skin where the channel seems implicated in differentiation processes. Gene targeting is defined as the introduction of site-specific modifications into the genome. It allows the in vivo analysis of diverse aspects of gene function. The classical gene-targeting approach leads to inactivation of a gene in all tissues of the body, from the onset of developments through the entire life. Complete gene activation of all three Scnn1a genes (encoding for αENaC, βENaC, and γENaC) leads to early postnatal death. For example, the constitutive inactivation of the α subunit of ENaC revealed an important role of the channel in lung liquid clearance around birth (Hummler et al. Nat. Genet. 1996), and measurements of the amiloridesensitive tranepithelial potential differences in αENaC knockout mice demonstrated that ENaC activity was completely abolished. Because consequences of ENaC deficiency in adult tissues as kidney, skin, or lung of adult mice cannot be addressed, we decided to generate a conditional allele at the gene locus encoding the Scnn1a (αENaC subunit) gene using the Cre-loxP-mediated recombination system (Hummler et al., Genesis 2002). Gene activation can be restricted to a particular cell type in vivo by crossing a mouse strain harboring the floxed allele with a transgenic strain expressing Cre recombinase under the control of kidney-, lung-, colon- or skin-specific promoters. This strategy is therefore useful to dissect the role of ENaC in vivo.

S 393 Institut de Pharmacologie & Toxicologie, Rue du Bugnon 27, CH1005 Lausanne, Switzerland

Aventis Pharma, DG Cardiovascular, Industriepark Höchst, Bld. H821, D-65926 Frankfurt/Main

S 19-5 MOUSE MODELS FOR HUMAN DISEASE: VALUABLE TOOLS FOR INTEGRATIVE PHYSIOLOGY AND MOLECULAR BIOLOGY M. Bleich The advent of transgenic animals has opened a completely new field of physiological and pharmacological research. Knock-in/out of specific genes by homologous recombination for the first time provided the chance to investigate the function of a gene without the use of pharmacological inhibitors. Even more, the search for new drugs which starts at the validation of the functional relevance of a target gene, often originates in the pathophysiological analysis of genetically engineered animals. Others and we have collected a body of experience in the investigation of knockout mice and learnt several lessons. In this EKRA workshop we will cover the discussion of our recent work on aldosterone receptor knock-out mice as well as external work on mice with targeted cardiac genes. After a brief introduction into the physiological background of the RAAS and mouse cardiac electrophysiology we will evaluate the outcome and relevance of the present studies. Then we will go into details, pitfalls and technical limitations: Careful breeding and genotyping is essential to prevent the risk of inhomogeneous genetic background and misinterpretation. In lethal phenotypes rescue with treatments according to the functional defects can be feasible (The use of inducible knockouts is an elegant alternative). Surgery in mice requires refined microscopic methods, anaesthesia and tight control of body temperature. Simple methods for in vivo transepithelial measurements, renal clearance techniques, and blood pressure analysis are available. Telemetric measurements are possible but still hampered by the size of transducers in small animals. ECG in conscious mice is challenging since filters have to be adapted to the high heart rate on a noisy background signal. Taken together, analysis of knockout mice will provide pathophysiological understanding as well as pharmacological targets for the treatment of human disease. Some phenotypes might be hidden behind compensatory mechanisms and require experimental challenges for their discovery.

S 19-6 TUBULOGLOMERULAR FEEDBACK IN THE KIDNEY: INSIGHTS FROM TRANSGENIC MICE V. Vallon As described by Golgi about a century ago, in every nephron the thick ascending limb of the loop of Henle returns with invariable constancy to its capsule of origin, i.e., it passes through the angle formed by the afferent and efferent arterioles and contacts the vascular pole of the glomerulus. The juxtaglomerular apparatus (JGA) comprises all structures that are present within this specific tubuloglomerular contact area and represents the anatomical basis for the tubuloglomerular feedback (TGF). The TGF refers to a series of events whereby changes in the concentration of Na+, Cl- and K+ in the tubular fluid are sensed by the macula densa which then causes reciprocal changes in nephron filtration rate. The net effect of this negative feedback mechanism is a relative low and constant delivery of fluid and electrolytes to the distal nephron, which in these nephron segments allows fine adjustment of reabsorption and excretion according to hormonal stimulation reflecting body needs. Although micropuncture experiments in rats and studies in the isolated perfused JGA of the rabbit established significant insights on TGF, many aspects including the signal transduction and modulation in the JGA as well as the role of TGF in physiology and pathophysiology are still incompletely understood. This is in part due to unavailability of selective drugs to inhibit or activate a given molecule of interest as well as limited accessibility of the target. Adaptation of the respective single nephron methods to mice represents a technical challenge. When successfully applied to mice which, e.g., lack a given gene of interest, however, it provides a promising approach to gain further insights on TGF. Department of Pharmacology, University of Tübingen, D-72074 Tübingen