Histochem Cell Biol (1997) 108:167–178

© Springer-Verlag 1997

O R I G I N A L PA P E R

&roles:Jarin Hongpaisan · Godfried M. Roomans

Use of low temperature and high K+ incubation media for in vitro tissue preparation for X-ray microanalysis

&misc:Accepted: 24 April 1997

&p.1:Abstract Incubation of tissue slices in physiological buffers gives rise to significant changes in the intracellular ion concentrations, which may disturb subsequent Xray microanalysis. In the present study it was attempted to design incubation conditions that retain the in vivo conditions better. The following variables were investigated: (1) exchange of Na+ in the incubation medium for K+, and exchange of Cl– for the less permeable gluconate anion; (2) incubation at 4°C rather than at 37°C; and (3) addition of dextran to the incubation medium. Brief exposure (a few seconds) of liver slices to a buffer causes changes in the intracellular Na, Cl and K concentrations, depending on the ionic composition of the buffer. Incubation in a normal physiological (high NaCl) buffer at 37°C results in a further increase of Na and Cl and a further decrease in K in liver cells. The changes reach a maximum at 30 min and the concentrations then remain stable throughout a 2-h incubation. Incubation in sodium gluconate medium or addition of dextran to the physiological buffer somewhat reduces the changes in the intracellular ion composition (compared to the standard physiological incubation medium). Incubation in potassium gluconate medium results in a decrease in cellular Na and an increase in K. Quantitative morphological studies show that tissue oedema is observed to the same extent in hepatocytes incubated in sodium gluconate, potassium gluconate and physiological buffer containing 10% dextran. However, these buffers cause significantly less cell oedema than the physiological (high NaCl) buffer. Incubation of liver, cerebral cortex or submandibular gland slices in physiological (high NaCl) solutions at 4°C for 4 h caused a more extensive increase in Na+ and decrease in K+ than incubation at 37°C for 2 h. This suggests inhibition of the Na+, K+-ATPase under these conditions. As compared to incubation at 37°C for 2 h, tissues incubated in potassium gluconate buffer at 4°C for 4 h have a J. Hongpaisan · G.M. Roomans (✉) Medical Ultrastructure Research Unit, Department of Human Anatomy, University of Uppsala, Box 571, S-75123 Uppsala, Sweden Tel. +46 18 174114; Fax +46 18 551120&/fn-block:

cellular K concentration closer to the in situ value. Cholinergic stimulation of tissue slices from cerebral cortex and submandibular gland at room temperature for 1 min shows the best physiological response in tissue slices preincubated at 4°C for 4 h in high KCl, potassium gluconate and high NaCl, in this order. The response can, however, only be seen, when cholinergic stimulation is carried out in a standard physiological buffer with a high NaCl concentration. It is concluded that in vitro storage of tissue for X-ray microanalysis is best carried out at 4°C in a solution with a high K+ concentration.&bdy:

Introduction Most applications of X-ray microanalysis in pathology have dealt with the analysis of firmly bound elements, predominantly particulates (Shelburne et al. 1989), whereas a minority has dealt with diffusible elements. However, the number of investigations on pathological changes in the distribution of diffusible elements has increased progressively. Studies involving X-ray microanalysis of diffusible elements in human or experimental pathology have been reviewed by LeFurgey et al. (1988) and by Roomans and von Euler (1996). Use of human biopsy material for X-ray microanalysis is not without complications. These are particularly important when analysis of diffusible substances is to be carried out (Roomans 1991). For a reliable X-ray microanalytical study on the localisation of diffusible substances, cells and tissues in a defined functional state should be immobilised by cryofixation as rapidly as possible; in situ freezing is to be preferred to the use of dissected tissue (von Zglinicki et al. 1986; Spencer and Roomans 1989; Tvedt et al. 1989; Hongpaisan et al. 1994; Hongpaisan and Roomans 1995). In situ freezing is, however, difficult to carry out when human tissue is involved because of the damage to the surrounding tissue. It has therefore been suggested that it would be advantageous to use an in vitro system, where the biopsy

168

procedure and the freezing procedure would be uncoupled both in time and in place as this would allow the use of the most sophisticated freezing techniques in the laboratory (Roomans 1991). In vitro systems used for X-ray microanalysis include dissected tissue (Hongpaisan and Roomans 1995), isolated cells in suspension (Warley 1986), cell cultures (von Euler et al. 1993; Borgmann et al. 1994; Warley 1994; Warley et al. 1994) and isolated glandular epithelia (see, e.g. Hongpaisan et al. 1996). Dissection itself causes a relatively minor increase in Na and Cl, and a decrease in the K concentration. However, even a brief exposure of the dissected tissue to a physiological buffer results in marked changes in intracellular ionic composition (Harvey and McIlwain 1969; Hongpaisan et al. 1994; Hongpaisan and Roomans 1995). Incubation for up to 2 h in a physiological buffer results in a further increase in Na and Cl and a further decrease in K in brain and liver tissue, but in recovery of the in vivo K/Na ratio in the pancreas and submandibular gland (Hongpaisan et al. 1994; Hongpaisan and Roomans 1995). Also, isolated thymocytes in suspension (Warley 1986), rat liver cells and papillary collecting ducts (Zierold and Schäfer 1988) and cardiac myocytes (Ward and Warley 1990) may recover during appropriate incubation in a physiological medium. Isolated submandibular gland acini, maintained in a physiological buffer at 37°C, show recovery of the intracellular elemental content, whereas this recovery does not occur in isolated human sweat glands (Hongpaisan et al. 1996). Evidently, the response to dissection, isolation and incubation is very tissue specific. Several strategies can be conceived to improve the maintenance of the physiological ion concentrations in tissue incubated in vitro. One possibility is to carry out the incubation at a lower temperature to minimise ion fluxes. It would also be possible to incubate tissue in a medium with a composition close to that of intracellular fluid, i.e. with a high K+/Na+ ratio. This method has been used in some biochemical studies (Elliott 1969) and is, for example, used in the storage of tissue for transplantation (Wahlberg et al. 1986). Similarly, Sjöstrand (1992) has suggested that the so-called infarct-reflux damage to heart tissue can be partially prevented by perfusion with K+-rich instead of Na+-rich fluids. Finally, the presence Table 1 Standard Krebs-Ringer buffer and substitutions for the liver&/tbl.c:&

Substances

of high molecular weight compounds may be beneficial. The colloid dextran has been shown to prevent the development of interstitial oedema in the liver (Ar’Rajab et al. 1991) and, hence, reduce cell swelling. In a previous study we have demonstrated that incubation of tissue slices at 7°C (instead of 37°C) or exchanging NaCl in the physiological medium for sodium gluconate has only relatively minor effects on the intracellular K+/Na+ ratio (Hongpaisan and Roomans 1995). Exchanging NaCl for potassium gluconate results in a high intracellular K/Na ratio throughout the incubation. However, it is questionable whether tissue incubated in a medium with a high K+/Na+ ratio can be used for physiological and pharmacological studies. In the present paper, we investigated the following experimental variables: incubation at low temperature (4°C); addition of dextran to the incubation medium; and incubation in a medium with a high K+/Na+ ratio followed by stimulation in a physiological medium with a high Na+/K+ ratio.

Materials and methods Tissue For experiments on liver and submandibular gland, Sprague-Dawley rats (both male and female, 8–9 weeks old, 190–220 g) were used. The animals were anaesthetised with pentobarbital (45 mg/kg body weight). In some animals, the liver and submandibular gland were frozen in situ with liquid nitrogen-cooled brass clamps. In other animals, the liver or submandibular gland was dissected into 1 to 1.5mm-thick slices and incubated in the experimental solutions. For the experiments on brain tissue, NMRI (Naval Medical Research Institute) mice (11–13 weeks old, 30–35 g) were used. The animals were anaesthetised with pentobarbital (85 mg/kg body weight). Part of the calvarium was removed, taking care not to damage the underlying tissue. In some animals, the cerebral cortex was frozen in situ by immersing the head of the animal directly into liquid propane cooled by liquid nitrogen. In other animals, the brain was dissected and 1 to 1.5-mm-thick slices were incubated in the experimental solutions. The dissection of the cerebral cortex took about 5–7 min (from opening the skull to the start of the incubation). Incubation media The composition of the solutions used in the present study is summarised in Tables 1–3. All solutions were adjusted to pH 7.4. Liver was incubated in a physiological Krebs-Ringer buffer (KRB)

Solution Sodium

NaCl (mM) Sodium gluconate (mM) KCl (mM) Potassium gluconate (mM) HEPES (mM) MgCl2 (mM) CaCl2 (mM) D-Glucose (mM) Dextran (g) H2O (ml) &/tbl.:

Potassium

NaCl

Gluconate

KCl

Gluconate

140 – 5 – 5 1 1.5 5 − 1000

– 140 5 – 5 1 1.5 5 − 1000

5 – 140 – 5 1 1.5 5 − 1000

5 – – 140 5 1 1.5 5 − 1000

10% dextran 140 – 5 – 5 1 1.5 5 100 1000

169 Table 2 Krebs-Ringer bicarbonate buffer for the submandibular gland&/tbl.c:& Substances

NaCl (mM) KCl (mM) Potassium-gluconate (mM) KH2PO4 (mM) NaHCO3 (mM) MgSO4 (mM) CaCl2 (mM) D-Glucose (mM) β-Hydroxybutyrate (mM) Nicotinamide (mM) Inosine (mM) Adenine (mM) H2O (ml)

Solution NaCl

KCl

Potassium gluconate

118.5 4.7 – 1.2 24.9 1.2 2.7 2.8 5 10 10 0.5 1000

4.7 118.5 – 1.2 24.9 1.2 2.7 2.8 5 10 10 0.5 1000

4.7 – 118.5 1.2 24.9 1.2 2.7 2.8 5 10 10 0.5 1000

&/tbl.:

Fig. 1 Scanning electron micrograph of a freeze-dried 16-µmthick cryosection of the submandibular gland used for X-ray microanalysis. Bar 100 µm&ig.c:/f

Table 3 Artificial cerebro-spinal fluid for the cerebrum&/tbl.c:& Substances

NaCl (mM) KCl (mM) Potassium gluconate (mM) NaH2PO4 (mM) NaHCO3 (mM) MgCl4 (mM) CaCl2 (mM) D-Glucose (mM) H2O (ml)

Solution NaCl

KCl

Potassium Gluconate

120 3.1 – 1.3 26 2 2 10 1000

3.1 120 – 1.3 26 2 2 10 1000

3.1 – 120 1.3 26 2 2 10 1000

&/tbl.: with high NaCl, or in a high K+ buffer where NaCl was exchanged for potassium gluconate, or in modified KRB where NaCl was exchanged for sodium gluconate, or in KRB (high NaCl) supplemented with 10% dextran (Table 1). The tissue slices were incubated for 2 h at 37°C and oxygenated with 95% O2 and 5% CO2; samples were taken at half-hour intervals and frozen in liquid propane cooled with liquid nitrogen. The 0-min sample was obtained by incubating the tissue for a few seconds in the experimental solutions. The incubation solution for submandibular gland slices was KrebsRingers bicarbonate buffer containing high NaCl (Table 2). The brain slices were incubated in artificial cerebro-spinal fluid (ACSF) (Ballyk and Goh, 1992) containing high NaCl or in modified ACSF where NaCl had been exchanged for potassium gluconate (Table 3). The tissue slices were incubated for 2 h at 37°C and oxygenated with 95% O2 and 5% CO2. Samples were taken after a 2-h incubation and frozen as described above. Incubation at 4°C was carried out by cooling the experimental solutions (oxygenated with 95% O2/5% CO2) with ice on a shaking board for 4 h. For the experiments with cholinergic stimulation, the tissue slices (submandibular gland or cerebral cortex) were first incubated for 4 h in one of the experimental solutions described above or in modified KRB where the NaCl had been exchanged for KCl, and then for 1 min in either physiological buffer (high NaCl), modified physiological buffer where the NaCl had been exchanged for KCl or in modified physiological buffer where the NaCl had been exchanged for potassium gluconate. Cholinergic stimulation was carried out by exposing the tissue slices to 20 µM carbachol at room temperature for 1 min. In control experiments, the carbachol was omitted. The specimens were frozen as described above.

Fig. 2 In situ frozen rat liver tissue compared with liver slices incubated briefly in physiological buffer (high NaCl), sodium gluconate, potassium gluconate and dextran-containing physiological buffer. Data (in mmol/kg dry weight) are given as mean and standard error (bars) for four to five animals in each group (eight measurements per animal)&ig.c:/f X-ray microanalysis For X-ray microanalysis, 16-µm-thick cryosections were cut on a conventional cryostat at –30°C, mounted on a carbon planchet, freeze-dried inside the cryostat and coated with a thin carbon layer to prevent charging in the electron microscope (Wróblewski et al. 1978 as modified by McMillan and Roomans 1990). Sections

170

Morphological studies Morphological studies were carried out on aldehyde-fixed liver tissue. As the equivalent of in situ frozen tissue, tissue fixed by perfusion was used. The anaesthetised animals were perfused with fixative, containing 2% glutaraldehyde and 1% (para)formaldehyde in 0.15 M phosphate buffer (300 mosmol) at pH 7.4, through a cannula placed in the left ventricle. Incubated liver slices were fixed by immersion in the same fixative. Tissue was postfixed in osmium tetroxide (2% in phosphate buffer), dehydrated in a graded ethanol series and embedded in Agar 100 epoxy resin (Agar Scientific, Stansted, UK). Ultrathin sections were stained with uranyl acetate and lead citrate and viewed at 75 kV in a Hitachi 7100 electron microscope. For the study of effects of the incubation meFig. 3 Effect of incubation on the elemental content of liver cells incubated at 37°C for 2 h in a standard (high NaCl) physiological buffer, as well as in buffers containing sodium gluconate, potassium gluconate or high NaCl+10% dextran. Concentrations are given in mmol/kg dry weight, incubation time in minutes. Data are given as mean and standard error (bars) for four to six animals in each group (eight measurements per animal)&ig.c:/f

Fig. 4A–F Transmission electron micrographs of liver tissue. A After perfusion fixation. B After dissection and brief exposure to a physiological (high NaCl) solution (immersion fixation). C After 2 h incubation in NaCl solution and immersion fixation. D After 2 h incubation in a sodium gluconate solution and immersion fixation. E After 2 h incubation in a potassium gluconate solution and immersion fixation. F After 2 h incubation in a NaCl solution supplemented with 10% dextran and immersion fixation. Bars 2 µm&ig.c:/f

dia on cell volume, 2-µm-thick sections of embedded tissue were cut and counterstained with toluidine blue. Tissue sections were analysed under a light microscope at ×40 magnification. Areas of hepatocytes were measured directly on line by an image analyser (VIDS V; Synoptics, Cambridge, UK) equipped with a CCD camera (FCD-12; Ikegami, Japan). Only hepatocytes showing clear nuclei and clear borders of the cells were measured (five random cells from each light microscopic image; 20–30 random images from each experimental animal).

Results Figure 1 shows a scanning electron microscope image of a 16-µm-thick freeze-dried cryosection used for X-ray microanalysis in this study.

▲

were cut a few cell layers away from the dissected edge of the tissue block. The cryosections were analysed in a Philips 525 scanning electron microscope (Philips Electron Optics, Eindhoven, The Netherlands) in the secondary electron mode at 20 kV with a LINK AN10000 energy-dispersive spectrometer system (Oxford Instruments ISIS, Oxford, UK). All analyses were carried out with a stationary beam (probe size 100 nm). However, due to the spreading of the electron beam in the specimen, the resolution of analysis is at the cellular level (Wróblewski et al. 1978). Quantitative analysis was carried out based on the ratio of characteristic counts to background intensity in the same energy region (P/B ratio) (Roomans 1988). P/B ratios obtained on the samples were compared with those obtained on standards consisting of a gelatine/glycerol matrix containing mineral salts in known concentrations (Roomans 1988). Statistical significance between groups was determined by analysis of variance (ANOVA) or Student’s t-test.

171

172 Table 4 Changes of heaptocyte cell volume during excision and incubation. Data given as percentage of control, mean±standard deviation of mean; number of animals was three; 100 or 150 cells measured per animal. Controls are cells treated with perfusion fixation&/tbl.c:&

Perfusion

Dissection and brief NaCl

NaCl 2h

Sodium gluconate 2h

Potassium gluconate 2h

NaCl and 10% dextran 2 h

Cell volume 100.0±0.3 (% of control)

110.8±0.3 ***

133.8±0.4 ***

119.1±0.4 *** +++ NS

120.7±0.3 *** +++ NS

119.1±0.3 *** +++ NS

Statistical analysis was made by two-tailed paired t-test: compared with perfusion, *** P<0.001; and compared with NaCl, +++ P<0.001, and ANOVA compared within incubation in sodium gluconate, potassium gluconate and NaCl supplemented with 10% dextran for 2 h; NS non-significant &/tbl.:

Effects of the extracellular medium on tissue slices Briefly (a few seconds) dipping dissected liver tissue into a high NaCl solution resulted in a dramatic increase in Na and Cl, and a marked decrease in K concentration, as compared to the in situ level (Fig. 2). Brief exposure of the tissue to a sodium gluconate buffer resulted in a less pronounced increase in Na+ and a less striking decrease in K+ concentration (Fig. 2). Immersion of the excised tissue into a potassium gluconate solution caused a minor decrease in Na and a marked increase in K content (Fig. 2). The intracellular Cl concentration was moderately decreased after dissection and brief incubation in the sodium gluconate and potassium gluconate solutions (Fig. 2). Brief exposure of the tissue to a high NaCl buffer supplemented with 10% dextran resulted in a less pronounced increase in Na and a less striking decrease in K concentration (Fig. 2). Effects of NaCl, sodium gluconate, potassium gluconate and dextran Incubation of the liver slices in high NaCl solution (at 37°C) resulted in a further increase of intracellular Na and Cl, and a decrease of the K concentration (Fig. 3). Incubation in a solution with sodium gluconate resulted in a slightly, but significantly, lower increase in cellular Na+ concentration and, similarly, a somewhat less pronounced decrease in K+ concentration (Fig. 3). The intracellular Cl concentration decreased during the incubation in the sodium gluconate medium (Fig. 3). After incubation in the potassium gluconate medium, intracellular Na and Cl levels decreased and the K level increased significantly. Addition of 10% dextran reduced the increase in Na and Cl concentrations caused by incubation in the physiological (high NaCl) buffer (Fig. 3). A slight decrease of S could be noted in all incubations; this effect was most pronounced for the sodium gluconate, potassium gluconate and dextran-containing high NaCl solutions (Fig. 3).

incubated briefly as well as tissue incubated for 2 h in incubation solutions was compared to tissue fixed by perfusion. Dissection and brief exposure to the high NaCl KRB solution resulted in a minor, but significant, increase in the cell volume, as compared to tissue fixed by perfusion (Table 4). When the tissue was incubated in the high NaCl solution for 2 h, hepatocytes showed significant further swelling (Table 4). The increase in cell volume was reduced in tissue incubated in the sodium gluconate, potassium gluconate or dextran-containing high NaCl solutions, but cell volume was still significantly larger than in tissue fixed by perfusion (Table 4). However, according to ANOVA, the cell volume of cells incubated in sodium gluconate, potassium gluconate or dextran-containing high NaCl solutions was not significantly different (Table 4). Effects of temperature during incubation in high NaCl medium Incubation of liver and brain slices in high NaCl buffers at 37°C for 2 h resulted in a dramatic increase in Na and Cl, a marked decrease in K, and a minor decrease in S compared to the in situ content (Fig. 5). Incubation of submandibular gland slices in a high NaCl solution at 37°C for 2 h resulted in a minor increase in Na and Cl, and no significant change in K and S concentrations, compared to the in situ content (Fig. 5). In all tissue samples, incubation in high NaCl solutions at 4°C for 4 h caused a greater increase in Na+ and a greater decrease in K concentration than in tissues incubated at 37°C for 2 h (Fig. 5). Only in liver and brain tissue was the S concentration slightly decreased after incubation at 4°C for 4 h (Fig. 5). Effects of temperature during incubation in potassium gluconate buffer

Incubation of liver and brain tissue in potassium gluconate solution at 37°C for 2 h resulted in a decrease in Na, Cl and S and an increase in K concentration (Fig. 6). Incubation at 4°C for 4 h resulted in a K+ content approximately similar to the in situ level (Fig. 6). Incubation of Water absorption submandibular gland slices in a potassium gluconate medium at 4°C for 4 h resulted in a decrease in Na and Cl Tissue oedema in the liver slices was confirmed quantita- and a minor increase in K concentration, compared to the tively by morphological studies (Fig. 4). Dissected tissue in situ content, while the S content was unaffected (Fig. 6).

173

Fig. 5 In situ frozen tissue compared with tissue slices incubated in physiological (high NaCl) buffers at 37°C for 2 h and at 4°C for 4 h. Liver, cerebral cortex (cortex) and submandibular gland (subman) were used for the experiments. Data (in mmol/kg dry weight) are given as mean and standard error (bars) for three to six animals in each group (8–15 measurements per animal)&ig.c:/f

Fig. 6 In situ frozen tissue compared with tissue slices incubated in potassium gluconate buffer at 37°C for 2 h and at 4°C for 4 h. Liver, cerebral cortex (cortex) and submandibular gland (subman) were used for the experiments. Data (in mmol/kg dry weight) are given as mean and standard error (bars) for three to six animals in each group (8–15 measurements per animal)&ig.c:/f

Effects of ionic replacements in the incubation medium

high K+ solution (4 h) and the medium was then changed to either a high K+ or a high Na+ solution (incubation for 1 min), this resulted in a dramatically lower Na+ and higher K+ concentration compared to cells incubated in a high NaCl solution only (4 h+1 min) (Fig. 7). The Cl concentration in the liver slices incubated in either ice-

These experiments were carried out to investigate whether incubation in high K+ medium would protect the cells against the effects of subsequent incubation in high NaCl medium. When liver slices were incubated in an ice-cold

174 Fig. 7 Effects of incubation in high NaCl, high KCl and potassium gluconate media at 4°C for 4 h followed by incubation in the same medium or in a high NaCl medium at 37°C for 1 min. Liver, cerebral cortex (cortex) and submandibular gland (subman) were used for the experiments. Data (in mmol/kg dry weight) are given as mean and standard error (bars) for three to four animals in each group (15 measurements per animal)&ig.c:/f

cold high NaCl or high KCl medium was markedly higher compared to slices incubated in a high potassium gluconate buffer (Fig. 7). Cells incubated first in high KCl solution and then in high NaCl solution showed an increase in Na, a decrease in K, and no significant changes in Cl–, compared to cells incubated in high KCl solution only (4 h+1 min) (Fig. 7). Incubation in potassium gluconate medium and then in high NaCl solution caused a marked increase in intracellular Na and Cl, and also a marked decrease in K concentration, as compared to incubation in potassium gluconate solution only (4 h+1 min) (Fig. 7). Experiments with cerebral cortex slices and submandibular gland slices showed results similar to those obtained for liver slices (Fig. 7).

Fig. 8 Comparison of the effects of cholinergic stimulation on cerebral cortex slices. The tissue was preincubated in high NaCl, high KCl or potassium gluconate solutions at 4°C for 4 h and subsequently stimulated cholinergically at room temperature for 1 min, either in the same solutions as used for the preincubation or in standard physiological buffer (high NaCl). Concentrations of Na+, Cl–, K+ and Ca2+ were determined. Data (% of control) are given as mean and standard error (bars) for three animals in each group (15 measurements per animal). The significance of the difference between treated tissue and control is indicated by asterisks (*P<0.05; **P<0.01)&ig.c:/f

175

that had been stimulated with carbachol in a high NaCl medium (Fig. 9).

Discussion

Fig. 9 Comparison of the effects of cholinergic stimulation on submandibular gland slices. The tissue was preincubated in high NaCl, high KCl or potassium gluconate solutions at 4°C for 4 h and subsequently stimulated cholinergically at room temperature for 1 min, either in the same solutions as used for the preincubation or in standard physiological buffer (high NaCl). Concentrations of Na+, Cl–, and K+ were determined. Data (% of control) are given as mean and standard error (bars) for three animals in each group (15 measurements per animal). The significance of the difference between treated tissue and control is indicated by asterisks (*P<0.05; **P<0.01)&ig.c:/f

Physiological response to stimuli after incubation Cholinergic stimulation of cerebral cortex in vitro after incubation in high NaCl solution resulted only in a significant increase in Ca concentration (Fig. 8). Increase in Ca concentration, sometimes significant, sometimes not, was also seen under other experimental conditions, except in cells incubated in a potassium gluconate medium followed by cholinergic stimulation in high NaCl solution (Fig. 8). A significant increase in Na+ was observed in cells incubated in high KCl solution followed by cholinergic stimulation in high NaCl solution (Fig. 8). Stimulation of submandibular gland tissue with the cholinergic agonist carbachol resulted in a significant decrease in the K concentration under all conditions (Fig. 9). A significant decrease in Cl concentration was seen only in cells stored in high KCl solution (4 h) and stimulated with carbachol in a high NaCl environment (Fig. 9). An insignificant decrease in Cl content was observed in cells incubated in potassium gluconate medium

Although bulk specimens (16-µm-thick sections) were used for X-ray microanalysis in this study, the risk of contribution by the extracellular compartment during analysis is low, since the K/Na ratio found in the thick sections is close to that found in thinner sections of the same tissue, as previously discussed by Hongpaisan and Roomans (1995). The changes in Na, Cl and K observed therefore reflect changes in the intracellular concentrations of these ions. It should be stressed that in the present study, carried out on freeze-dried sections, ion concentrations are given on a dry weight basis and that changes in water content were not considered. However, the quantitative morphological data showed that the changes in intracellular ion content are indeed accompanied by cell swelling. The swelling is likely to reduce the density of the tissue, which would slightly increase the penetration of the electron beam in the sample. This effect is probably responsible for the small decrease in S concentrations observed in some of the experiments. Although in Fig. 1 ice crystal formation can be observed, the size of the ice crystals is small compared to the analysed volume which has a diameter of about 10 µm (Wróblewski et al. 1978). In principle, only in situ freezing can give values for elemental concentrations close to the living state. For biopsy or in vitro specimens, dissection itself causes relatively minor increases in Na and Cl concentrations and a decrease in the K concentration. The effects of rapid dissection on the elemental content of liver and pancreas are relatively small, but in excitable tissue such as brain tissue an increase in Na and Cl and a decrease in K could be noted as a consequence of dissection (Hongpaisan and Roomans 1995). The changes in diffusible elements in in vitro tissue preparations can depend on other experimental conditions. Zierold et al. (1994) were able to prepare isolated tubules of trout kidney with a high K/Na ratio (14.3) by cutting cryosections from intact cells far away from the tubule end damaged by excision. Takemura et al. (1991) prepared collagenase-dissociated coils from rhesus monkey palmar skin with a K/Na ratio of approximately 3. For isolated brain, more damage was evident histologically in chopped tissue than in sliced tissue (even when the resulting tissue pieces had similar dimensions) and chopped tissue showed less response to physiological stimulation; the degree of damage depended on the sharpness of the blade (McIlwain 1985). Another factor that can affect the changes in the concentrations of diffusible elements in in vitro tissue preparation is the composition of the incubation medium. Exposure of dissected tissue to incubation medium with a composition corresponding to that of the extracellular fluid results in a rapid and dramatic increase of intracellular Na and Cl and a decrease of intracellular K (Hong-

176

paisan et al. 1994; Hongpaisan and Roomans 1995). In the present study, these results were confirmed and, in addition, it was shown quantitatively that, after dissection and brief incubation in high NaCl solution, the cell volume was significantly increased in parallel with an increase in intracellular Na and Cl. As also shown in Fig. 2, changes in the concentrations of diffusible elements after brief exposure of dissected liver tissue depended on the ionic composition of the incubation solution. This means that the cell membrane effectively has lost its control over the ionic composition of the cytoplasm, possibly due to hypoxia during tissue dissection. These changes in the intracellular ion concentrations are largely determined by diffusion and the volume of the extracellular fluid. The Na+ and Cl– ions entering the cells come predominantly from the incubation medium, although some may come from the intercellular spaces in the tissue itself. The mechanisms by which these ion fluxes take place were not investigated. It could be speculated that, apart from mechanisms under metabolic control, more unspecific mechanisms such as passive diffusion could also play a role. It would be of interest to improve the preservation of tissue slices with respect to the ionic content of the cells. A theoretical possibility would be to store the tissue in a solution with a K/Na ratio closer to the intracellular ratio (instead of in a solution resembling the extracellular environment). The results of the present study on liver slices again confirmed our previous study on brain tissue, submandibular gland and pancreas (Hongpaisan and Roomans 1995) in that exchange of Cl– for the less permeable gluconate ions inhibits the influx of Na+ from the extracellular compartment. Although the intracellular Na+ concentration after 2 h incubation in sodium gluconate medium is significantly lower than after incubation in the regular high NaCl medium, the improvement is not sufficient for practical purposes. Incubation in potassium gluconate solution results in a high intracellular K+/Na+ ratio. Cl– slowly diffuses out of the cells into the gluconate medium. Quantitative morphological studies showed that despite the high intracellular K/Na ratio, significant oedema occurs even in the potassium gluconate medium and it has to be concluded that the cell membrane, under the conditions of the experiment, is not in full control of the ion and water fluxes. It is likely that K+ diffuses into the cell from the extracellular medium (see Fig. 3; compare ionic concentrations during incubation with their in situ counterparts), possibly following the electrochemical gradient of K+. The influx of K+ ions and H2O molecules (which may enter together into cells by forming hydrogen bonds around K+, i.e. as hydration water) results in an increase in the cell volume, i.e. cell oedema. In summary, the high intracellular K+/Na+ ratio in cells incubated in a high K+ medium does not prove that the cells are in optimal physiological condition. Swelling of brain tissue during incubation is known to occur and the swelling may be quantitatively different for different parts of the brain (Harvey and McIlwain 1969). Our results show that the cell volume of hepa-

tocytes increased by about 34% after incubation in high NaCl solution, i.e. a physiological solution. The cell volume increased approximately 20% after incubation in sodium gluconate or potassium gluconate medium. These results are in agreement with data on brain slices which, during a 1-h incubation, absorb up to 30–40% of their weight of water from ordinary saline media in the presence of O2 (reviewed by Elliott 1969). On the other hand, in the presence of high K+, brain slices absorbed more than 40% water (Elliott 1969). The only explanation for this discrepancy is that the less permeable anion gluconate used in our study can reduce the increase of water absorption compared to Cl–. Also, as shown in this study, addition of dextran to the high NaCl solution can reduce the influx of Na+ and Cl–, and efflux of K+, parallel to a reduction in the increase of the liver cell volume. This agrees with data that dextran can prevent the development of interstitial oedema in the liver (Ar’Rajab et al. 1991). Nevertheless, incubation with dextran results in significant changes in intracellular ion content. In general, hypothermia is supposed to reduce tissue oedema by slowing down the cellular metabolism (Elliott 1969; Wahlberg et al. 1986). The results of our experiments with incubation at a temperature of 7°C for 2 h were not very encouraging with respect to the use of incubated tissue for studies of the nervous system (Hongpaisan and Roomans 1995). In the present study, tissues were, however, incubated in a colder (4°C) solution for 4 h. This procedure may be suitable in practice for human tissue that needs to be stored after surgery until the start of the experiment. Also, this procedure imitates the conservation of organs for transplantation (Wahlberg et al. 1989). In the present study, incubation of tissue at 4°C for 4 h in the high NaCl solution caused a more pronounced increase in Na, and a more pronounced decrease in K concentration, as compared to incubation at 37°C for 2 h (Fig. 5), suggesting inhibition of the Na+, K-ATPase at 4°C. However, hypothermic conditions gave more promising results when the incubation was carried out in potassium gluconate solution, because the K concentration in cells incubated at 4°C for 4 h was closer to the in situ K level than in cells incubated in potassium gluconate solution at 37°C, as shown by the data on liver and cerebral cortex slices (Fig. 6). After incubation of tissue slices at 4°C for 4 h followed by a brief incubation at room temperature (this experiment served as a control for the effects of cholinergic stimulation), the changes in the concentration of diffusible elements were less than those observed in freshly dissected tissue after brief exposure to the incubation medium (compare Fig. 2 and Fig. 7). This suggests that, even after prolonged incubated in a cold solution, cells in vitro can maintain their membrane permeability characteristics better than freshly excised cells. Possibly, after incubation in cold high K+ solution, cells could regain their metabolic activity. Basically, cells in vivo maintain low Na+ and Cl– and high K+ concentrations in cytoplasm and nucleus. This environment is of crucial importance for, for example,

177

cellular osmotic control, electrolyte balance and currents, stability of polyelectrolytes, DNA, membrane structure and uptake of organic metabolites (Fraústo da Silva and Williams 1991). In later experiments, we therefore further investigated the effects of high KCl and potassium gluconate solutions (at 4°C) on the functional response of the incubated tissues to physiological and pharmacological substances. Cholinergic agonists were used at a concentration calculated to give maximal stimulation. After preincubation of brain tissue at either room temperature or at 37.5°C for 30–40 min in a glucose-bicarbonate medium, stimulation with excitatory amino acids results in an increase in Na+ and either an unchanged or decreased K+ concentration after 1 min stimulation (Bradford and McIlwain 1966; Harvey and McIlwain 1968). In the present study, cholinergic stimulation of brain slices for 1 min, after preincubation in high NaCl medium at 4°C for 4 h, resulted only in a significant increase in Ca2+ concentration, suggesting the influx of Ca2+ ions, possibly via Ca2+ channels. An increase of Ca2+ influx in neurons is very important for the secretion of neurotransmitters (Rubin 1970). It is unclear why the concentrations of Na+ and K+ were not affected in this experiment. However, after preincubation in high KCl buffer, a significant increase in Na+ parallel to a small increase in Ca2+ content was seen when cholinergic stimulation was carried out in a high NaCl environment (Fig. 8). This may suggest competition between Na+ and Ca2+ influx (Rubin 1970). The observed changes in ion content in submandibular gland acinar cells after cholinergic stimulation are a consequence of the activation of several ion transport processes: secretion of Cl– by apical chloride channels, secretion of K+ by basolateral potassium channels, and uptake of Na+, K+ and Cl– ions, presumably by a basolateral Na+-K+-Cl–-cotransporter (Petersen 1980; Poulsen and Kristensen 1982; Novak and Young 1986; Pirani et al. 1987). Under these conditions, there is a net efflux of K+ through the basolateral potassium channels, resulting in hyperpolarisation in the cell membrane. Cl– efflux takes place through the apical Cl– channels down its electrochemical gradient due to hyperpolarisation. By Xray microanalysis it has previously been shown that cholinergic stimulation results in a decrease in intracellular Cl– and K+, while the Na+ concentration is increased, decreased or unaffected in the submandibular gland in vivo (Hongpaisan and Roomans 1995; Mörk et al. 1996) and in isolated submandibular gland acini (Hongpaisan et al. 1996). After incubation of submandibular gland slices in high NaCl buffer at 37°C for 2 h, stimulation with a cholinergic agonist results in a decrease in cellular Cl and K concentrations (Hongpaisan and Roomans 1995). In the present study, cholinergic stimulation of submandibular gland acinar cells stored in the ice-cold high NaCl solution resulted only in a decrease in K concentration. Cl– secretion was seen in cells preincubated in ice-cold high K+ solution and stimulated with carbachol in a high

NaCl solution. This suggests that cells suffer more damage when preserved in high Na+ media than in high K+ media (see changes of Na+, Cl– and K+ in Fig. 5). As reviewed by Rubin (1970), when Cl– is replaced by nitrate or sulphate the ability of the salivary gland to respond to pharmacological stimulation is depressed, and it has therefore been postulated that electrogenic transport of Cl– into the cells is the primary step in the secretion of saliva. Cells preincubated with KCl showed more Cl– secretion than cells preincubated with potassium gluconate. This indicates that accumulation of intracellular Cl– is also important to the activity of Cl– secretion (compare Cl– concentrations in cells incubated in the two solutions in Fig. 7). A disadvantage of the in vitro systems seems to be the inherently larger spread in the data, presumably due to the fact that the changes provoked by dissection and incubation may vary for different cells in the tissue. However, our findings clearly demonstrate that, in the presence of NaCl, the response to cholinergic stimulation of the dissected tissue from cerebral and submandibular gland can be seen. Cholinergic stimulation at room temperature for 1 min shows the best physiological response in tissue previously incubated at 4°C for 4 h in high KCl, potassium gluconate, and high NaCl, media. Under conditions of excess K+, the respiratory rate in brain tissue can be increased (Elliott 1969). Liver slices incubated in the presence of a high K+ medium showed a slow rate of glycogenolysis compared to similar slices incubated in a high Na+ medium (Cahill et al. 1957). Figure 7 shows that cells after incubation in high KCl for 4 h, followed by exposure to high NaCl, maintain their intracellular K/Na ratio reasonably well. This is in constrast to freshly dissected tissue exposed for a short time to high NaCl (Hongpaisan and Roomans 1995); there a marked decrease of the intracellular K/Na ratio occurs. Also, cells kept in cold KCl and then transferred to a physiological buffer react normally to cholinergic stimulation. Hence, our results suggest that ice-cold high K+ solutions are useful to maintain the in vitro tissue specimens for X-ray microanalysis. However, physiological stimulation should be done in high NaCl solutions. &p.2:Acknowledgements The technical assistance of Mrs. Marianne Ljungquist, Mr. Anders Ahlander and Mr. Leif Ljung is gratefully acknowledged. This study was supported by a grant from the Swedish Medical Research Council (project 07125).

References Ar’Rajab A, Ahren B, Sundberg R, Bengmark S (1991) The function of a colloid in liver cold-storage preservation. Transplantation 52:34–38 Ballyk BA, Goh JW (1992) Elevation of extracellular potassium facilitates the induction of hippocampal long-term potentiation. J Neurosci Res 33:598–604 Borgmann S, Granitzer M, Crabbe J, Beck FX, Nagel W, Dorge A (1994) Electron microprobe analysis of electrolytes in whole cultured epithelial cells. Scanning Microsc Suppl 8:139–147

178 Bradford HF, McIlwain H (1966) Ionic basis for the depolarization of cerebral tissues by excitatory acidic amino acids. J Neurochem 13:1163–1177 Cahill GF Jr, Ashmore J, Zottu S, Hastings AB (1957) Studies on carbohydrate metabolism in rat liver slices. J Biol Chem 224:237–250 Elliott KAC (1969) The use of brain slices. In: Lajtha A (ed) Handbook of neurochemistry. Plenum Press, New York, pp 103–114 Euler A von, Pålsgård E, Vult von Steyern C, Roomans GM (1993) X-ray microanalysis of epithelial and secretory cells in culture. Scanning Microsc 7:191–201 Fraústo da Silva JJR, Williams RJP (1991) The biological chemistry of the elements: the inorganic chemistry of life. Clarendon Press, Oxford Harvey JA, McIlwain H (1968) Excitatory acidic amino acids and the cation content and sodium ion flux of isolated tissues from the brain. Biochem J 108:269–274 Harvey JA, McIlwain H (1969) Electrical phenomena and isolated tissues from the brain. In: Lajtha A (ed) Handbook of neurochemistry. Plenum Press, New York, pp 115–136 Hongpaisan J, Roomans GM (1995) Use of post mortem and in vitro tissue specimens for X-ray microanalysis. J Microsc 180:93–105 Hongpaisan J, Mörk AC, Roomans GM (1994) Use of in vitro systems for X-ray microanalysis. Scanning Microsc Suppl 8:109–116 Hongpaisan J, Zhang A-L, Mörk A-C, Roomans GM (1996) Use of primary cell cultures and intact isolated glandular epithelia for X-ray microanalysis. J Microsc 184:22–34 LeFurgey A, Bond M, Ingram P (1988) Frontiers in electron probe microanalysis: application to cell physiology. Ultramicroscopy 24:185–219 McIlwain H (1985) Cerebral subsystems and isolated tissues. In: Lajtha A (ed) Handbook of neurochemistry. Plenum Press, New York, pp 325–341 McMillan EB, Roomans GM (1990) Techniques for X-ray microanalysis of intestinal epithelium using bulk specimens. Biomed Res (India) 1:1–10 Mörk A-C, Zhang A-L, Martinez JR, Roomans GM (1996) Chloride secretion in the submandibular gland of adult and early postnatal rats studied by X-ray microanalysis. Histochem Cell Biol 105:297–303 Novak I, Young JA (1986) Two independent anion transport systems in rabbit mandibular salivary glands. Pflügers Arch 407:649–656 Petersen OH (1980) The electrophysiology of gland cells. Monographs of the Physiological Society, no. 36. Academic Press, London Pirani D, Evans LA, Cook DI, Young JA (1987) Intracellular pH in the rat mandibular salivary gland: the role of Na-H and Cl-HCO3 antiports in secretion. Pflügers Arch 408:178–184 Poulsen JH, Kristensen L (1982) Is stimulation-induced uptake of sodium in rat parotid acinar cells mediated by a sodium/chloride cotransport system? In: Case RM, Garner A, Thurnberg LA (eds) Electrolytes and water transport across gastrointestinal epithelia. Raven Press, New York, pp 199–208

Roomans GM (1988) Quantitative X-ray microanalysis of biological specimens. J Electron Microsc Tech 9:19–43 Roomans GM (1991) Cryopreparation of tissue for clinical applications of X-ray microanalysis. Scanning Microsc Suppl 5:S95–106 Roomans GM, Euler A von (1996) X-ray microanalysis in cell biology and cell pathology. Cell Biol Int 20:103–109 Rubin RP (1970) The role of calcium in the release of neurotransmitter substances and hormones. Pharmacol Rev 22:389–428 Shelburne JD, Tucker JA, Roggli VL, Ingram P (1989) Overview of applications in medicine. In: Ingram P, Shelburne JD, Roggli VL (eds) Microprobe analysis in medicine. Hemisphere, Washington, pp 55–78 Sjöstrand FS (1992) Information and misinformation regarding ischemia of heart muscle tissue. The cause of cell death during blood reperfusion and reactivation of heart muscle tissue after prolonged ischemia. Scanning Microsc 6:1041–1058 Spencer AJ, Roomans GM (1989) X-ray microanalysis of hamster tracheal epithelium. Scanning Microsc 3:505–509 Takemura T, Sato F, Saga K, Suzuki Y, Sato K (1991) Intracellular ion concentrations and cell volume during cholinergic stimulation of eccrine secretory coil cells. J Membr Biol 119:211–219 Tvedt KE, Kopstad G, Halgunset J, Haugen OA (1989) Rapid freezing of small biopsies and standard for cryosectioning and X-ray microanalysis. Am J Clin Pathol 92:51–56 Wahlberg J, Southard JH, Belzer FO (1986) Development of a cold storage solution for pancreas preservation. Cryobiol 23:477–482 Wahlberg J, Jacobsson J, Tufveson G (1989) Relevance of additive components of University of Wisconsin cold-storage solution. An experimental study in the rat. Transplantation 48:400–403 Ward JP, Warley A (1990) Subcellular elemental composition in isolated cardiac myocytes from rabbit following chronic potassium depletion and acute repletion, studied by X-ray microanalysis. Exp Physiol 75:855–858 Warley A (1986) Elemental concentrations in isolated rat thymocytes prepared for cryofixation in the presence of different media. J Microsc 144:183–191 Warley A (1994) The preparation of cultured cells for X-ray microanalysis. Scanning Microsc Suppl 8:129–137 Warley A, Cracknell KP, Cammish HB, Twort CH, Ward JP, Hirst SJ (1994) Preparation of cultured airway smooth muscle for study of intracellular element concentrations by X-ray microanalysis: comparison of whole cells with cryosections. J Microsc 175:143–153 Wróblewski R, Roomans GM, Jansson E, Edström L (1978) Electron probe X-ray microanalysis of human muscle biopsies. Histochemistry 55:281–292 Zglinicki T von, Bimmler M, Purz HJ (1986) Fast cryofixation technique for X-ray microanalysis. J Microsc 141:79–90 Zierold K, Hentschel H, Wehner F, Wessing A (1994) Electron probe X-ray microanalysis of epithelial cells: aspects of cryofixation. Scanning Microsc Suppl 8:117–126 Zierold K, Schäfer D (1988) Preparation of cultured and isolated cells for X-ray microanalysis. Scanning Microsc 2:1775–1790