Introduction: a definition

Signal Transduction and Protein Kinases: The Long Way from the Plasma Membrane into the Nucleus Ferdinand Hucho, Klaus Buchner Freie Universita¨t Berlin, Institut fu¨r Biochemie, AG Neurochemie, Thielallee 63, D-14195 Berlin, Germany

All living cells must be able to receive information from the extracellular space and to react to it by processing and converting it into intracellular effects. If the properties of cells are to change in the long term, some signals must reach the nucleus in order to bring about changes in gene transcription. Three of the pathways, beginning with an extracellular signal and ending with the nucleus, serve to illustrate some principles of signal transduction such as signal conversion, signal cascade, cross-talk, and on/ off switch. One element common to most of the pathways is the activation of protein kinases. One example of these kinases, the protein kinase C, is discussed as a vehicle of signal transport toward the nucleus and as a means of cross-talk between different signaling pathways.

Correspondence to: F. Hucho

Naturwissenschaften 84, 281–290 (1997)

© Springer-Verlag 1997

Signal transduction is as fundamental a feature of life as metabolism or self replication. All living cells receive information from the extracellular space; they react to it by processing and converting it into intracellular effects. The mechanism of this processing and conversion is the subject of much current biochemical and cell biological research. “Extracellular” for the unicellular organism is its environment, which supplies nutrition and energy. For the multicellular organism this environment is supplemented (and complicated) by neighboring cells, by cells of the same organism far away, which send messages important for concerted growth and action, and by the matrix which holds them together and guides their growth. For all these organisms signal transduction can be defined as the complete pathway of extracellular physical or chemical/molecular signals into the cytoplasm and/or into the nucleus, including its conversion into an effect. By this definition it is difficult to find an important cellular event in which no signal transduction is involved: a comprehensive review (omitting prokaryotes and unicellular eukarykotes) therefore would cover everything form the fertilization of the egg, through the cell cycle and development to the growth and death of individual cells and of tissues. It would have to include topics such as proliferation, differentiation, metabolic regulation, mechanical work, adaptation, “plasticity” and information storage in the nervous system, and the immune response. It would have to deal with cancer, apoptosis, and a great variety of diseases. It would also have to address specialities not only with respect to the plant and animal realms but also to the variability of phenomena in different cells, tissues, and organisms. No such comprehensive treatise in intended here. 281

We wish to summarize rather a few basic principles and molecules with a certain bias toward the nucleus, to protein kinases as universal vehicles of signal transport and for neuronal cells which are the signal processing units per se. Among the protein kinases we place special emphasis on the protein kinase C (PKC) because of its molecular and compartmental variability.

Two principle pathways of signal transduction Basically two functional principles are observed: The extracellular signal itself penetrates the plasma membrane and finds its way to the nucleus. This is the case with steroid hormones or with the neutrophil signal peptide lactoferrin [1, 2]. We call this the “short way.” Alternatively, the signal remains outside the cell and is converted at the plasma membrane into intracellular signals. Biochemistry textbooks describe this pathway as the “first messenger/second messenger concept.” In the context of this article we refer to it as the “long way.”

Sites of signal transduction In signal transduction by the long way the extracellular signal is the first messenger, which cannot permeate the plasma membrane, but rather is converted into a different intracellular signal called the second messenger. The site of this conversion is the plasma membrane. Further sites are in the cytoplasm, at organelle membranes such as the nuclear envelope, and inside the nucleus.

Molecules participating in signal transduction The signal or first messenger (leaving aside physical signals such as photons, mechanical tension, and pressure) may be a hormone, odorant, antigen, neurotransmitter, or the surface of another cell. Accordingly, the next type of molecule involved are the receptors, and the effector system, for example, the second messenger producing ion channels (for the second messenger Ca2+) and enzymes (adenylyl and guanylyl cyclases for the second messengers cAMP and cGMP, phospho282

Fig. 1. The signal transduction triad. The components of the triad may be separate proteins or domains of a single molecule [3]. R, receptor, or signal-receiving moiety; E, effector or effector moiety (e.g., the ion channel in the case of a ligand-gated ion channel); T, transducer, which in the case of a ligand-gated ion channel is an integral feature of the protein complex [4]. In the case of the heptahelical receptors T is one of the G proteins

lipases for inositol-1,4,5-trisphosphate (IP3), diacylglycerol (DAG), and arachidonic acid). Receptors and effectors are often functionally coupled by G proteins (also called “transducers”), coupling them to the activated receptors. The triad receptor/transducer/effector depicted as RTE in Fig. 1 is the machinery which transports the extracellular signal through the plasma membrane into the cell, thereby converting it from a first to a second messenger. Proceeding to the cytoplasm, the “message” carried by the second messenger must be executed. This task is performed in most cases by protein kinases (in collaboration with protein phosphatases), and by targets of the signal which may be enzymes of metabolism, intracellular receptors, special transport “vehicles” (see below) and finally (when the genome is the target) transcription factors.

Concepts of signal transduction Signal transduction operates by several principles which are common to the otherwise diverse pathways. We referred above to the concept of signal conversion. Another concept is the cascade: Normally, signal transduction (especially via the “long way”) consists of many steps which allow for signal amplification, specific targetting, regulation, and “cross-talk,” the mutual interference of various signaling pathways. Finally, signals must be limited in time and space. A basic principle in every signal transduction phenomenon is the occurrence of at least one on/off switch.

Fig. 2. Signal transduction pathways to the nucleus. As examples, the cAMP cascade (A ), the MAP kinase cascade, (B ), and the JAK-STAT pathway are depicted (see text for further description)

Examples Figure 2 shows three typical signal transduction cascades which illustrate the principles mentioned above. Starting with the fundamental discoveries by Sutherland, the cAMP cascade (Fig. 2 A) is the most thoroughly investigated one (and the one studded with the most Nobel prizes). It is made up of several steps which by their catalytic nature amplify the signal, for example, epinephrine, by at least four orders of magnitude. Signal conversion takes place at the plasma membrane of, for example, liver and muscle cells. The triad receptor/transducer/effector is represented by the b-adrenergic receptor/Gs protein/adenylyl cyclase complex. Its specific target are enzymes of glycogen metabolism, which is linked to muscle contraction and to transcription control at the genome. A pivotal role is played by the cAMP-activated protein kinase A (PKA). This enzyme’s two catalytic subunits are inhibited by two regulatory subunits which dissociate after cAMP binding, thereby rendering the catalytic subunits active. The catalytic subunits have multiple functions. They activate glycogen breakdown. At the same time they inactivate glycogen synthesis, and they translocate into the nucleus where they phosphorylate a protein called cAMP response element binding protein (CREB). This in turn, when activated by phosphorylation, acts as a transcription factor which controls transcription of “immediate early genes” via a regulatory sequence of the genome (the cAMP response element, CRE). They code for transcription factors regulating the ex-

pression of “delayed response genes” which are important for the phenotype of the respective cell. On-switches of this cascade are G proteins which are active when complexed with GTP. Receptors act as GDP/GTP exchange factors. Off-switches can be seen in the GTPase activity of the G protein, in the phosphodiesterase hydrolyzing the second messenger cAMP, and in the protein phosphatases which reverse the activation of the respective enzymes and factors by protein kinases. Regulation and “cross-talk” become evident when we discuss the other cascades. One point should be stressed: the complexity of modulatory effects allows for an impressive fine tuning. Various mechanisms, such as allosteric control, covalent modification, substrate availability, expression, and proteolytic removal of the proteins involved, prevent all-or-nothing effects. Rather, they cause a continuum, ranging from total inactivation to total activity. An example of this is the glycogen synthase, which is inactivated by phosphorylation while glycogen phosphorylase is activated. This enzyme (i.e., the glycogen synthase) has at least nine phosphorylation sites which are the subtrates of a whole variety of protein kinases, ranging from PKA, casein kinase II, cG kinase, to PKC [5, 6]. This allows for ample modulation of signals and cross-talk. Speaking in computer terminology: Signal transduction is not digital but rather analog. Our second example is the MAP kinase cascade, also called the Ras-Raf cascade (Fig. 2 B), which processes and transduces signals from growth hormones and neurotrophins [7]. The four presently known 283

neurotropins nerve growth factor (NGF), brain-derived neutrotrophic factor, and neurotropins 3 and 4/ 5 act through the family of receptors called tyrosine receptor kinase (Trk) A, B, C, (and a low-affinity receptor p 75, see below) [8, 9]. Upon ligand binding at the extracellular signal-receiving domain, these receptors dimerize and activate a protein tyrosine kinase located on their intracellular domain. Among others, this kinase can phosphorylate itself at specific tyrosine residues (it actually phosphorylates the neighboring receptor within the dimer). The phosphorylated tyrosine residues become the docking sites for proteins with src homology (SH) 2 domains such as the Grb 2/SOS complex. SOS is a classical “nucleotide-exchange protein” which activates the small G protein Ras by replacing its bound GDP by GTP. Ras acts as one of the on/off switches for this cascade. In its activated (GTP bound) state it activates the serine/threonine protein kinase Raf, which in turn activates by phosphorylation the serine and tyrosine protein kinase mitogen-activated protein kinase kinase (MAPKK, also known as MEK). Active (phosphorylated) MAPKK phosphorylates the mitogen-activated kinases, (MAPK, also called ERK-1 and ERK-2 for extracellular signal regulated kinase) at threonine and tyrosine residues. The MAPK then translocates into the nucleus where it phosphorylates transcription factors such as cFos, cJun, Elk-1 and CREB (obviously one site for cross-talk with the cAMP cascade), regulating the expression of immediate early genes [9]. A bifurcation leading to CREB activation has been recently described [10]; MAPK may also activate CREB kinase, a serine kinase previously known as pp 90RSK. CREB kinase has also been found to shuttle between the cytoplasm and the nucleus. In addition to protein phosphorylation, noncovalent protein-protein interactions seem to be involved in signal transport. In addition to receptor demerization and adapter protein docking, dimerization of Raf, mediated by the 14-3-3 proteins, is observed. Raf is known to be activated by Ras (which helps to translocate it to the plasma membrane) and by phosphorylation. However, especially in the absence of the latter, 14-3-3 may secure Raf activation through dimerization (or oligomerization) at the membrane [11]. As noted above, Ras, the small G protein, acts an an on/off switch of this signal transduction pathway. SOS plays the role of a GDP/GTP exchange factor (GEF), helping its “on-function.” The off-switch, as with the heterotrimeric G proteins represented by its GTPase activity, is also regulated by protein-protein interaction: GTPase-activating proteins (GAPs) stimulate the enzymatic hydrolysis of GTP and thereby the termination of the signal. The interplay between 284

GEF and GAP secures the fine-tuning of the cascade. The same cascade is used by growth hormones such as the epidermal growth factor (EGF) and the plateled-derived growth factor (PDGF). However, while the neurotrophins are involved primarily in differentiation and survival of their target cells, EGF and PDGF are mitogens, promoting proliferation. It is not yet fully understood how these different and specific effects are brought about via the same cascade. One idea is that the respective signal kinetics are different. A short and transient signal may trigger the cascade to a different degree than a long-lasting one does. NGF is a “target-derived survival factor.” To this end it must be involved in retrograde signaling. Retrograde signaling involves binding of NGF to Trk A at the nerve ending, internalization, and retrograde axonal transport to the neuronal soma [12]. Little is known about this process, but another kinase, phosphatidylinositol 3 kinase (PI 3-kinase), seems to be involved. This lipid kinase is activated by TrkA [13] – whether directly of indirectly is unknown – and activated PI 3-kinase plays a role in vesicle trafficking [14]. The low-affinity neurotrophin receptor p 75 seems to be associated with all Trks. p 75 itself does not contain a tyrosine kinase activity, but it enhances the affinity of TrkA for its ligand NGF and stimulates its autophosphorylation and activation. Interestingly, in addition to these functions which enforce signaling through Trks, p 75 has a signaling cascade of its own: In the absence of TrkA p 75 has been shown to trigger apoptosis in neurons (e.g., Barde and coworkers found this in an early developmental stage of retina neurons). It shares this property with proteins of the tumor necrosis factor receptor (TNF-R) family. As the other members of this familily p 75 contains a “death domain” thought to be essential in the apoptosis cascade. Proteins associated with p 75’s death domain have not yet been identified, and the mechanism leading to cell death are largely obscure. However, the dual function of the “survival factor” NGF in neuronal survival and death is highly intriguing and will surely soon be explained. A third signaling cascade which we would like to briefly present is the Jak-STAT pathway [15–17] (Fig. 2 C). Jak is a family of protein tyrosine kinases, while STAT represents a group of transcription factors which are substrates of the Jaks. In neurons and glia cells this pathway is activated by the ciliary neurotrophic factor (CNTF) [18]. Despite its name CNTF is not structurally related to the neurotrophins, and its mechanism of action is completely different. It binds at the plasma membrane to a trimeric recep-

tor composed of a-, b- (LIF-b), and gp 130 subunits, none of which contains a tyrosine kinase activity. CNTF shares the latter two components (LIF-receptor b and gp 130) with the hematopoietic interleukin (IL)6 receptor. Following CNTF binding to its receptor the Jak–tyrosine kinase is activated, which phosphorylates a cytosolic STAT [9]. STAT contains a SH 2 domain which enables docking to a second tyrosine phosphorylated STAT, resulting in homo- or heterodimeric STAT which translocates to the nucleus and acts as a transcription factor for immediate early genes (a different set than the one activated by the neurotrophins). Again, we see the same principles: signal conversion at the plasma membrane, amplification via a catalytic cascade, and the on/off switch represented by phosphorylation/dephosphorylation. An unresolved problem with the Jak-STAT pathway is specificity: with only six (or seven) known STATs and many dozens of peptide ligands ranging from CNTF through several ILs to leptin activating one or several of them, it is by no means clear how specific patterns of gene activation leading to a variety of characteristic phenotypes are brought about.

Cross-talk Signal transduction pathways do not exist in isolatation from each other; rather, there is ample cross-talk between them. Again, instead of a comprehensive treatise of these interactions, only a few examples are given to illustrate the stages at which cross-talk can take place. Recently an example of cross-talk between PKA and PKC pathways, which in general takes place at many levels (for review see [19]), was described by Lohse and coworkers. They found that PKC activates b-receptor kinase (bARK) by enhancing its translocation to the plasma membrane [20]. At the plasma membrane bARK and other members of its family, by phosphorylating receptors, cause uncoupling between receptors and G proteins, leading to cessation of signal transduction (desensitization; Fig. 3). PKC can also act in an inhibitory way on EGF receptor signaling (Fig. 3). Phosphorylation by PKC leads to inhibition of the EGF receptor’s tyrosine kinase activity [22] and thus induces desensitization of this pathway [23]. Cross-talk takes place at the other end of signal pathways as well. Transcription factors are in many cases the subject of regulation by several protein kinases belonging to different signal pathways. CREB, for example, as mentioned above, is a substrate for PKA, but it can also be phosphorylated by other ki-

Fig. 3. PKC in the cross-talk between signal transduction pathways. PKC can be activated via both G protein coupled receptors and tyrosine kinase receptors. Activated PKC can act on other pathways, for example, the cAMP cascade, by activating a b-receptor kinase, thus finally leading to receptor desensitization. Influence on tyrosine kinase receptor pathways may be inhibitory by phosphorylation of receptors, or stimulatory by activating Raf. Another possibility for stimulatory action is the (direct or indirect) activation of the tyrosine kinase PYK 2, which in turn activates the MAP kinase signaling pathway [21] (see text for further details)

nases such as CaMK II and IV, and at least in vitro by PKC. A second example is Fos. This transcription factor is phosphorylated by a number of kinase such as p 34cdc 2, CK II, PKA, and PKC [24]; however, the effects of the individual phosphorylation events and their interplay are not yet understood [25]. Protein kinases can also inhibit or activate signal pathways at intermediate levels between the plasma membrane and the nucleus. Activation of PKC, for example, can lead to phosphorylation and thereby activation of Raf, thus triggering the MAPK cascade (Fig. 3) [26, 27]. On the other hand, phosphorylation of Raf by PKA in fibroblasts leads to an inhibition of the MAPK pathway [28, 29]. However, this is not the case in all cell systems: in pheochromocytoma cells of neuroendocrine origin (PC 12 cells) cAMP has no inhibitory effect on growth factor stimulation of MAPK activity [30, 31]. Thus PKC seems to mediate the interplay between signal transducing cascades on all levels (Fig. 3). It connects various pathways and is a key element of the complex intracellular network of functional and regulatory cross-talk. Because of this central importance we discuss PKC and its role in signal transduction in more detail.

285

Protein kinase C: a vehicle of signal transport Molecular properties of PKC isozymes PKC is indeed one of the most important protein kinases involved in the intracellular transduction of signals. Since it can be activated by both G protein coupled receptors and tyrosine kinase receptors (Fig. 3), many signals can result in activation of PKC. This is accomplished by the activation of different isoforms of PI-specific phospholipase C, resulting in the formation of the physiological PKC activator DAG and IP3 [32]. IP3 in turn leads to the release of Ca2+ from intracellular stores. PKC was initially characterized as a Ca2+- and phospholipid-activated protein kinase [33] and is at its highest concentration in the nervous system. There it fulfills important functions in the regulation of transmitter release, function of receptors and ion channels, and longer lasting events such as synaptic plasticity and differentiation of neurons [34–36]. Presently at least 11 different PKC isozymes are known in mammals [37]. Some of these no longer meet the original definition as Ca2+- and phospholipid-dependent kinases. On the basis of their structural differences and the resulting enzymatic properties the PKC isozymes are divided into several groups. The conventional or “classical” isozymes comprise the isozymes a, bI/II (bI and bII are alternative splice variants), and c. The other nonconventional isozymes can be further divided into the novel PKCs [PKC d, e, g (L) and h)] and the atypical PKCs [PKC f, i (k)]. For a detailed discussion of structural features of the various PKC isoforms, the reader is referred to a number of excellent reviews (e.g. [38–40]).

Activation of PKC Activation of the conventional PKCs at membranes takes place in two steps: first Ca2+ induces a transloction to the (plasma) membrane; the PKC which is membrane-associated via phosphatidylserine is then activated by interaction with diacylglycerol. This membrane association is reversible: after removal of Ca2+ PKC falls off the membrane again [41]. In addition to this reversible association with membranes, PKC can also be present in a so-called “membraneinserted” form [42]. This form is found after longlasting stimulation of stimulation with phorbol ester. The inserted PKC cannot be removed from the mem286

brane by removal of Ca2+ but only by applying detergents. Another possibility for PKC activation is proteolytic cleavage into the regulatory and catalytic domains. As noted above, the catalytic domain (PKM) is active permanently, i.e., it is independent of Ca2+ and phospholipid. The physiological meaning of this proteolytic cleavage is not entirely clear, as cleavage into the two domains is followed by complete proteolytical degradation. Treatment of cells with phorbol ester for a prolonged time thus makes the PKC completely vanish [43]. However, PKM is detected in neurophils [44], and recent results suggest that proteolytic activation of PKC may be critically involved in neurite outgrowth in neuroblastoma cells [45]. The model for activation of PKC, established for the conventional isozymes, is surely a strong simplification, as PKC, for example, binds not only to the plasma membrane but also to intracellular membranes. Furthermore, PKC can also associate with elements of the cytoskeleton and after activation can phosphorylate a number of cytoskeleton-associated proteins such as MAP-2, plectin, and annexin [46].

PKC in signal transduction toward the nucleus A large part of the interest which PKC attracted early, was in the identification of this kinase as the receptor for the strong tumor promotor phorbol ester [47]. Direct activation of PKC by phorbol esters illustrates that PKC is critically involved in the regulation of proliferation and differentiation of cells (for reviews see [35, 48, 49]). Since these processes depend on the control of events in the cell nucleus, it is obvious that to enalbe the PKC growth-regulatory actions some signal must reach the nucleus after activation of PKC. Several possibilities are conceivable as to how PKC, once activated, can bring about the transduction of signals toward and within the cell nucleus. For some of these possibilities PKC remains in the cytoplasm and causes the nuclear effects indirectly. As mentioned above, PKC can trigger the nucleus-directed MAPK cascade by activating Raf. Another example of signal transfer to the nucleus induced by cytoplasmic PKC is provided by the transcripiton factor nuclear factor (NF)jB. Activation of PKC leads to phosphorylation of the inhibitor of NFjB (IjB), the cytoplasmic anchoring protein of NFjB. This phosphorylation is the signal for rapid proteolytic degradation of IjB. NFjB is released and translocates into

the nucleus. Phosphorylation of IjB in vivo is probably not achieved by PKC directly but by a yet unknown kinase [50]. This kinase could, for example, be activated by reactive oxygen intermediates, which can be generated after PKC activation [50]. Release from a cytoplasmic anchoring protein followed by nuclear translocation can also be observed for the catalytic subunits of PKA, which via the regulatory subunits are tethered to the specific binding protein AKAP [51]; this appears to be a widespread phenomenon [52]. Other signal pathways imply a direct role for PKC within the nucleus. To achieve this the following possibilities exist (Fig. 4): (a) PKC itself translocates to the cell nucleus. This may happen after activation in the cytoplasm, but it is also conceivable that PKC is translocated in an inactive state followed by an activation in the nucleus. (b) Nuclear PKC is activated by a messenger from the cytoplasm reaching the cell nucleus. (c) Nuclear PKC is activated by an activator generated in the nucleus in the course of another signal cascade. These possibilities implicate, at least temporarily, the presence and activity of PKC inside the cell nucleus. Such a direct nuclear function of PKC was originally controversial, but substantial evidence for it has accumulated over recent years [53].

Localization of PKC isozymes in the nucleus As stated above, PKC was initially discovered in brain tissue, where it is present with high activity and is involved in a variety of neural functions (for review see [36]). Together with the extensive neuroanatomical work on the localization within different brain areas, reports on the subcellular distribution showed that various PKC isozymes can be present in the cell nucleus [54–59]. The other line of evidence for nuclear association of PKC was obtained from subcellular fractionation studies, revealing the nuclear location of several PKC isozymes (for review see [53]). Reports on isolated nuclei from tissues suggest that PKC is present constitutively at and in the cell nucleus and therefore must be activated or inactivated there. In cultured cells it is possible to compare the subcellular distribution of PKC isoforms within resting and stimulated cells. Such studies have repeatedly reported that PKC is present at the cell nucleus only after stimulation, implicating a translocation of PKC toward the cell nucleus (e.g., [60]). A variety of signals have been shown to trigger the translocation of PKC toward the cell nucleus. These include artificial activators such as phorbol esters

Fig. 4. PKC in nuclear signaling. PKC may bring about the transduction of signals toward and within the cell nucleus either by acting in the cytoplasm (see text) or by acting within the cell nucleus. PKC itself may translocate into the nucleus, or nuclear PKC is activated by a messenger from the cytoplasm reaching the nucleus, or by a messenger generated inside the nucleus. Asterisks, activated molecules; A1, A2, activators of PKC; B, factor that can lead to the formation of a PKC activator

and physiological messengers such as insulin-like growth factor (IGF) 1, PDGF, a-thrombin, angiotensin II, and vitamin D3 [53]. The physiological stimuli inducing nuclear translocation of PKC comprise signals binding to phosphotyrosine kinase receptors (IGF-1, PDGF) and to G protein coupled receptors (a-thrombin, angiotensin II). All of these, by activating different isoforms of phospholipase C or phospholipase D, can lead to formation of DAG, the most important endogenous activator of PKC. In addition to those on nuclear PKC in stimulated cells, there are also reports on several PKC isoforms in nuclei of resting cells in culture (for review see [53]). Two detailed studies have reported complex patterns of PKC isoform distribution [61, 62]. In cardiac myocytes PKC a, b II, and f were cytosolic in nonstimulated cells and translocated to the nuclear envelope after stimulation with phorbol myristate acetate or norepinephrine [62]. In contrast, PKC d and e were found intranuclear in resting cells, and after stimulation translocated to the nuclear envelope and to myofibrils, respectively [62]. In our study on the subcellular distribution of PKC isoforms in neuroblastoma× glioma hybrid NG 108-15 cells [61], we showed that in undifferentiated cells PKC a is located in the cytoplasm and in the nucleus. PKC d was found in the cytoplasm and was especially enriched in the nucleoli, in differentiated cells particularly in the neurites. PKC e is colocalized with the nuclear pore complexes at the nuclear envelope. In differentiated cells a partial translocation ot the plas287

ma membrane and to the nuclear envelope was observed after stimulation with phorbol ester [61]. These two reports show that PKC isoforms can be distributed differentially within subnuclear compartments. They also illustrate that it is not possible to establish a general rule for subcellular distribution of PKC isoforms. Rather, each cell type seems to express a defined set of PKC isozymes which are specifically located within the cell [63]. This also includes the fact that in different cell types different PKC isoforms can be found at the nucleus and in subnuclear compartments. In many instances a stimulus-dependent relocation to various subcellular compartments takes place. This observation supports the role of PKC as a shuttle for signals within the cell.

Regulation of nuclear PKC As noted above, the fact that PKC can be present at the nucleus raises the question as to how the enzyme is activated there. An activator must either reach the nucleus or be formed there. The main physiological activator of PKC, DAG, can be produced at or in the nucleus, either by the action of PI-specific phospholipase C associated with the nuclear matrix [64] or by the action of a phosphatidyl choline hydrolyzing activity not yet defined at the nuclear envelope [65, 66]. Recently a nuclear phospholipase D activity which generates DAG has been shown [67]. In all cases the signal transduction pathway from the plasma membrane to the nuclear phospholipid-cleaving enzyme remains unclear.

Nuclear calcium One candidate for a messenger which can pass easily and rapidly throughout the cell, and which may activate enzymes such as PKC and the phospholipases mentioned above is Ca2+. Both the source and the regulation of nuclear Ca2+ are currently topics of intensive debate (e.g., [68–69]). Irrespective of whether nuclei are able to regulate their Ca2+ content independently from the cytosol in some way, it is important to emphasize that Ca2+ concentrations can be high enough for activation of Ca2+-dependent processes in the nucleus [70]. Activation of PKC and of the enzymes mentioned above is certainly only one aspect of the diverse actions of nuclear Ca2+, which include processes mediated by calmodulin [71, 72], Ca/CaM kinase [73], and other Ca2+-binding proteins [74, 75]. 288

Other activators Recently a novel PKC activator, selective for PKCbII, was demonstrated to be present at the nuclear envelope of human promyelocytic (HL 60) leukemia cells [76]. However, this nonproteinaceous activator, believed to be a nuclear membrane lipid or lipid metabolite, still awaits identification.

Function of nuclear PKC From the nuclear PKC substrates known so far one can deduce possible functions of nuclear PKC in the regulation of lamina structure, regulation of nuclear transport, and modulation of the properties of DNA regulatory proteins. The most prominent nuclear PKC substrate is lamin. Phosphorylation by PKC in vitro leads to an enhanced solubility of these intermediate filament proteins [77] and may therefore be important for the breakdown of the nuclear envelope at the beginning of mitosis [77, 78]. Another function of the lamin phosphorylation is conceivable in the context of a possible reorganization of the chromatin network by intranuclear lamin [79, 80]. The colocalization of a PKC isozyme with the nuclear pore complexes points to a role of PKC in regulation of nuclear transport. This is supported by the observation that activation of PKC by phorbol ester leads to an inhibition of mRNA transport through the nuclear envelope [81]. However, phosphorylation of nuclear pore complex proteins by PKC has not been shown. Another group of nuclear PKC substrates are DNA regulatory proteins such as DNA topoisomerases and DNA polymerases, and transcription factors such as CREB and Fos (for review see [53]). Phosphorylation of these proteins modulates their DNA binding or regulatory properties and may represent a connection between activation of PKC and regulation of transcription. However, in many cases in vivo phosphorylation by PKC inside the nucleus has not been shown unequivocally. In muscle, nuclear PKC appears to exert an important role in the regulation of muscle-specific genes. In denervated chicken muscle electrical stimulation leads to an increase of PKC in the nucleus and to inactivation of genes for the subunits of the nicotinic acetylcholine receptor [82]. The muscle-specific transcription factor myogenin is regulated in its DNA binding properties by phosphorylation by PKC [83]. In contrast to this, the DNA binding properties of the transcription factor Fos, which is phosphorylated by PKC in vitro at least, are not changed by PKC phos-

phorylation [24]. Fos is phosphorylated by several Ser/Thr kinases [24], and it is rather difficult to analyze which kinase phosphorylates Fos under physiological conditions. Additionally, phosphorylation of Fos appears to depend on the phosphorylation status, whether it is present as a dimer, and whether it is bound to DNA [25]. This example illustrates that PKC is only one of several components which bring about a complex network of phosphorylation and dephosphorylation reactions in the nucleus [84]. It seems likely that not all components of this network have yet been identified. Especially little is known about the counterparts of the kinases, the phosphatases in the nucleus. Some of the kinases, and very probably some of the phosphatases, are located in the nucleus only temporarily, depending on the cell cycle or the stimulation of the cell by signals at the plasma membrane. Therefore, in addition to the analysis of nuclear substrates, the analysis of mechanisms and regulation of the subcellular distribution of kinases and phosphatases are of great importance. In the case of PKC recent observations indicate that the transport of PKC to the nuclear pore complexes may differ from the transport of nuclear localization signal-bearing proteins [85]. Since PKC does not have a canonical nuclear localization signal, transport into the nucleus may occur via PKC-binding proteins which can function as carriers. Several PKC binding proteins have been characterized so far [86–88], but none of them is a good candidate for the postulated nuclear carrier. Inside the nucleus we recently identified PKC binding proteins (Rosenberger et al., manuscript in preparation), which, in addition to their other functions, may also play a role in keeping the PKC inside the nucleus once it is there. Acknowledgements. Work from the authors’ laboratory was supported by the Deutsche Forschungsgemeinschaft and the Fonds der chemischen Industrie.

1. He, J., Furmanski, P.: Nature 373, 721 (1995) 2. Baeuerle, P.A.: ibid. 373, 661 (1995) 3. Hucho, F., in: Neurotransmitter Receptors, New Comprehensive Biochemistry, p. 3. (F. Hucho, ed.). Amsterdam 1992 4. Hucho, F., Tsetlin, V.I., Machold, J.: Eur. J. Biochem. 239, 539 (1996) 5. Cohen, P.; in: The Enzymes, Vol 17, Control by Phosphorylation, p. 462 (P.D. Boyer, E.G. Krebs, eds.). New York 1986 6. Roach, P.J., in: The Enzymes, Vol. 17; Control by Phosphorylation, p. 500 (P.D. Boyer, E.G. Krebs, eds.). New York 1986 7. Seger, R., Krebs, E.G.: FASEB J. 9, 726 (1995) 8. Barbacid, M.: Curr. Opin. Cell Biol. 7, 148 (1995) 9. Segal, R.A., Greenberg, M.E.: Annu. Rev. Neurosci. 19, 463 (1996) 10. Xing, J., Ginty, D.D., Greenberg, M.E.: Science 273, 959 (1996)

11. Farrar, M.A. Alberola-Ila, J., Perlmutter, R.M.: Nature 383, 178 (1996) 12. Curtis, R., DiStefano, P.S.: Trends Cell Biol. 4, 383 (1994) 13. Ohmichi, M., Decker, S.; Saltiel, A.: Neuron 9, 769 (1992) 14. De Camilli, P., Emr, S.D., McPherson, P.S., Novick, P.: Science 271, 1533 (1996) 15. Ihle, J.N.: Nature 377, 591 (1995) 16. Ihle, J.N.: Cell 84, 331 (1996) 17. Schindler, C., Darnell Jr., J. E.: Annu. Rev. Biochem 64, 621 (1995) 18. Ip, N.Y., Yancopoulos, G.D.: Annu. Rev. Neurosci 19, 491 (1996) 19. Housley, M.D.: Eur. J. Biochem. 195, 9 (1991) 20. Winstel, R., Freund, S., Krasel, C., Hoppe, E., Lohse, M.J.: Proc. Natl. Acad. Sci. USA 93, 2105 (1996) 21. Lev, S., Moreno, H., Martinez, R., Canoll, P., Peles, E., Musacchio, J. M., Plowman, G. D., Rudy, B., Schlessinger, J.: Nature 376, 737 (1995) 22. Davis, R.: J. Biol. Chem. 263, 9462 (1988) 23. Livneh, E., Dull, T., Prywes, R., Ullrich, A., Schlessinger, J.: Mol. Cell. Biol. 8, 2302 (1988) 24. Abate, C., Marshak, D.R., Curran, T.: Oncogene 6, 2179 (1991) 25. Abate, C., Baker, S.J., Lees-Miller, S.P., Anderson, C.W., Marshak, D.R., Curran, T.: Proc. Natl. Acad. Sci. USA 90, 6766 (1993) 26. Kolch, W., Heidecker, G., Kochs, G., Hummer, R., Vahidi, H., Mischak, H., Finkenzeller, G., Marme, D., Rapp, U.R.: Nature 364, 249 (1993) 27. Sozeri, O., Vollmer, K., Liyanage, M., Frith, D., Kour, G., Mark, G.E., Stabel, S.: Oncogene 7, 2259 (1992) 28. Cook, S.J., McCormick, F.: Science 262, 1069 (1993) 29. Wu, J., Dent, P., Jelinek, T., Wolfman, A., Weber, M.J., Sturgill, T.W.: Science 262, 1065 (1993) 30. Frodin, M.; Peraldi, P., Van-Obberghen, E.: J. Biol. Chem. 269, 6207 (1994) 31. Vaillancourt, R.R., Gardner, A.M., Johnson, G.L.: Mol. Cell Biol. 14, 6522 (1994) 32. Nishizuka, Y.: FASEB J. 9, 484 (1995) 33. Takai, Y., Kishimoto, A., Iwasa, Y., Kawahara, Y., Mori, T., Nishizuka, Y.: J. Biol. Chem. 254, 3692 (1979) 34. Conn, P.J., Sweatt, J.D., in: Protein kinase C, p 199 (J.F. Kuo, ed.). New York 1994 35. Goodnight, J., Mischak, H., Mushinski, J.F.: Adv. Cancer Res. 64, 159 (1994) 36. Tanaka, C., Nishizuka, Y.: Annu. Rev. Neurosci. 17, 551 (1994) 37. Dekker, L.V., Parker, P.J.: Trends Biochem. Sci. 19, 73 (1994) 38. Stabel, S., Parker, P.J.: Pharmac. Ther. 51, 71 (1991) 39. Hug H, Sarre, F.: Biochem. J. 291, 329 (1993) 40. Newton A. C.: J. Biol. Chem. 270, 28495 (1995) 41. Quest, A. F. G., Bell, R. M., in: Protein Kinase C, p 64 (J. F. Kuo, ed.). New York 1994 42. Bazzi, M. D., Nelsestuen, G. L.: Biochemistry 27, 7589 (1988) 43. Rasmussen, H., Calle, R. A., Ganesan, S., Smallwood, J. I., Throckmorton, D. C., Zawalich, W.S., in: Protein kinase C. Current concepts and future perspectives, p 183 (D. S. Lester, R. M. Epand, eds.). New York 1992 44. Pontremoli, S., Michetti, M., Melloni, E., Sparatore, B., Salamino, F., Horecker, B. L.: Proc. Natl. Acad. Sci. USA 87, 3705 (1990) 45. Shea, T. B., Cressman, C. M., Spencer, M. J., Beerman, M. L., Nixon, R. A.: J. Neurochem. 65, 517 (1995) 46. Jaken, S., in: Protein kinase C. Current concepts and future perspectives, p 237 (D. S. Lester, R. M. Epand, eds.). New York 1992 47. Ashendel, C. L.: Biochim. Biophys. Acta 822, 219 (1985) 48. Clemens, M. J., Trayner, I., Menaya, J.: J. Cell Sci. 103, 881 (1992) 289

49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.

290

Nishizuka, Y.: Science 258, 607 (1992) Baeuerle, P. A., Henkel, T.: Annu. Rev. Immunol. 12, 141 (1994) Mochly-Rosen, D.: Science 268, 247 (1995) Hunt, T.: Cell 59, 949 (1989) Buchner, K.: Eur. J. Biochem. 228, 211 (1995) Wood, J. G., Girard, P. R., Mazzei, G. J., Kuo, J. F.: J. Neurosci. 6, 2571 (1986) Saito, N., Kikkawa, U., Nishizuka, Y., Tanaka, C.: ibid. 8, 369 (1988) Huang, F. L., Yoshida, Y., Nakabayashi, H., Young III, W. S., Huang, K.-P.: ibid. 8, 4734 (1988) Kose, A., Saito, N., Ito, H., Kikkawa, U., Nishizuka, Y., Tanaka, C.: ibid. 8, 4262 (1988) Tsujino, T., Kose, A., Saito, N., Tanaka, C.: Ibid. 10, 870 (1990) Merchenthaler, I., Liposits, Z., Reid, J. J., Wetsel, W. C.: J. Comp. Neurol. 336, 378 (1993) Leach, K. L., Powers, E. A., Ruff, V. A., Jaken, S., Kaufmann, S.: J. Cell Biol. 109, 685 (1989) Beckmann, R., Lindschau, C., Haller, H., Hucho, F., Buchner, K.: Eur. J. Biochem. 222, 335 (1994) Disatnik, M.-H., Buraggi, G., Mochly-Rosen, D.: Exp. Cell. Res. 210, 287 (1994) Asaoka, Y., Nakamura, S., Yoshida, K., Nishizuka, Y.: Trends Biochem. Sci. 17, 414 (1992) Divecha, N., Irvine, R. F.: Cell 80, 269 (1995) Leach, K. L., Ruff, V. A., Jarpe, M. B., Adams, L. D., Fabbro, D., Raben, D. M.: J. Biol. Chem. 267, 21816 (1992) Jarpe, M. B., Leach, K. L., Raben, D. M.: Biochemistry 33, 526 (1994) Balboa, M. A., Balsinde, J., Dennis, E. A., Insel, P. A.: J. Biol. Chem. 270, 11738 (1995) Al-Mohanna, F. A., Caddy, K. W. T., Bosolver, S. R.: Nature 367, 745 (1994) Albritton, N. L., Oancea, E., Kuhn, M. A., Meyer, T.: Proc. Natl. Acad. Sci. USA 91, 12458 (1994)

70. Beckmann, R., Lindschau, C., Haller, H., Buchner, K.: J. Neurosci. Res. 40, 571 (1994) 71. Bachs, O., Lanini, L.; Serratosa, J., Coll, M. J., Bastos, R., Aligue, R., Rius, E., Carafoli, E.: J. Biol. Chem. 265, 18595 (1990) 72. Corneliussen, B., Holm, M., Waltersson, Y., Onions, J., Hallberg, B., Thornell, A., Grundstro¨m, T.: Nature 368, 760 (1994) 73. Ohta, Y., Ohba, T., Miyamoto, E.: Proc. Natl. Acad. Sci. USA 87, 5341 (1990) 74. Gilchrist, J. S. C., Pierce, G. N.: J. Biol. Chem. 268, 4291 (1993) 75. Dedhar, S., Rennie, P. S., Shago, M., Hagesteijn, C.-Y. L., Yang, H., Filmus, J., Hawley, R. G., Bruchovsky, N., Cheng, H., Matusik, R. J., Gigue`re, V.: Nature 367, 480 (1994) 76. Murray, N. R., Burns, D. J., Fields, A. P.: J. Biol. Chem. 269, 21385 (1994) 77. Hocevar, B. A., Burns, D. J., Fields, A. P.: ibid. 268, 7545 (1993) 78. Goss, V. L., Hocevar, B. A., Thompson, L. J., Stratton, C. A., Burns, D. J., Fields, A. P.: ibid. 269, 19074 (1994) 79. Moir, R. D., Goldman, R. D.: Curr. Opin. Cell Biol. 5, 408 (1993) 80. Moir, R. D., Montag-Lowy, M., Goldman, R. D.: J. Cell Biol. 125, 1201 (1994) 81. Schro¨der, H. C., Rottmann, M., Wenger, R., Bachmann, M., Dorn, A., Mu¨ller, W. E. G.: Biochem. J. 252, 777 (1988) 82. Huang, C.-F., Tong, J., Schmidt, J.: Neuron 9, 671 (1992) 83. Li, L., Zhou, J., James, G., Heller-Harrison, R., Czech, M. P., Olson, E. N.: Cell 71, 1181 (1992) 84. Meek, D. W., Street, A. J.: Biochem. J. 287, 1 (1992) 85. Schmalz, D., Kalkbrenner, F., Hucho, F., Buchner, K.: J. Cell Sci. 109, 2401 (1996) 86. Hyatt, S. L., Liao, L., Chapline, C., Jaken, S.: Biochemistry 33, 1223 (1994) 87. Mochly-Rosen, D., Khaner, H., Lopez, J.: Proc. Natl. Acad. Sci. USA 88, 3997 (1991) 88. Staudinger, J., Zhou, J., Burgess, R., Elledge, S. J., Olson, E. N.: J. Cell Biol. 128, 263 (1995)