Differential Regulation of Proline-rich Tyrosine Kinase 2/Cell Adhesion Kinase β (PYK2/CAKβ) and pp125FAK by Glutamate and Depolarization in Rat Hippocampus

The mechanisms by which stimuli that raise cytosolic free Ca2+ concentrations in neurons can increase protein tyrosine phosphorylation are not known. Using rat hippocampal slices and cortical synaptosomes, we have examined the regulation of two highly related cytoplasmic tyrosine kinases, pp125 focal adhesion kinase (pp125FAK) and proline-rich tyrosine kinase 2/cell adhesion kinase β (PYK2/CAKβ). Membrane depolarization increased tyrosine phosphorylation of PYK2/CAKβ and pp125FAK. These effects were blocked by EGTA or by protein kinase C inhibitors (RO31-8220; GF109203X) and mimicked by ionomycin or phorbol 12-myristate 13-acetate, in the case of pp125FAK, or their combination in the case of PYK2/CAKβ. Glutamate and specific agonists of ionotropic (α-amino-3-hydroxy-5-methyl-4-isoxazole propionate and N-methyl-D-aspartate) or metabotropic (trans-1-aminocyclopentane-1,3,-dicarboxylate) glutamate receptors stimulated the phosphorylation of pp125FAK, but not of PYK2/CAKβ. Glutamate effects were prevented by GF109203X. Thus, in hippocampal slices, tyrosine phosphorylation of pp125FAK and PYK2/CAKβ are regulated differentially by pathways involving Ca2+ and protein kinase C. pp125FAK and PYK2/CAKβ may provide specific links between neuronal activity, increases in cytosolic Ca2+ and protein tyrosine phosphorylation, which may be important for neuronal survival, and synaptic plasticity.

Increases in cytosolic free Ca 2ϩ are critical for many neuronal functions including neurotransmitter synthesis and release, and neuronal survival, while excess in Ca 2ϩ can lead to neuronal death. The induction of two well-studied types of synaptic plasticity, long term potentiation (LTP) 1 and long term depression (LTD) requires also a marked increase in cytosolic Ca 2ϩ , which results usually from the stimulation of glutamate NMDA receptors (see Ref. 1 for a review). Many of the effects of Ca 2ϩ are mediated by the activation of serine/ threonine kinases, including classical PKCs (2) and Ca 2ϩ /calmodulin kinases (3), and of a serine/threonine phosphatase, calcineurin (4). In addition, cytosolic Ca 2ϩ is able to activate protein tyrosine phosphorylation and signaling cascades involving Ras and MAP kinase which result in alterations in gene transcription (see Ref. 5 for a review). Depolarization and neurotransmitter agonists increase tyrosine phosphorylation of several proteins in synaptosomes (6), hippocampal slices (7), neurons in culture (7), and cell lines with neuronal characteristics (8,9). This link between tyrosine phosphorylation and Ca 2ϩ in neurons opens exciting perspectives since tyrosine phosphorylation appears necessary for the establishment of LTP (10) and LTD (11) and plays an important role in the control of neuronal differentiation and survival (5). However, the molecular basis for the stimulation of tyrosine phosphorylation in response to Ca 2ϩ is not known. Evidences obtained in PC12 cells indicate that Src (8) and PYK2 (9) are involved in the Ca 2ϩ -induced activation of tyrosine phosphorylation and downstream signaling cascades. Yet, the mechanism by which Ca 2ϩ activates Src and PYK2 in PC12 cells is not known, and it is not clear which tyrosine kinase is activated first. In addition, the regulation of these, or related, kinases has not been demonstrated in normal neuronal tissue. Here, we have investigated, in rat hippocampus, the possible regulation by Ca 2ϩ of two structurally related cytoplasmic tyrosine kinases, pp125 FAK and PYK2/CAK␤, which could be associated with Src family kinases in neurons.
pp125 FAK is a 125-kDa cytosolic tyrosine kinase devoid of SH2 or SH3 domains, which is associated with focal adhesions (12,13). pp125 FAK is phosphorylated on tyrosine in response to integrin engagement and to stimulation of various G proteincoupled receptors (14). Autophosphorylated pp125 FAK binds to the SH2 domain of Src or Fyn (15). Phosphorylation of pp125 FAK by Src on multiple residues (16) allows the recruitment and activation of phosphatidylinositol-3-kinase (17,18) and the binding of Grb2, leading to the activation of MAP kinase cascade (19). pp125 FAK is highly expressed in nervous tissue during development, a period at which it is enriched in neuronal growth cones (20). In the brain of adult rats, pp125 FAK is expressed at higher levels than in most other tissues, especially in the hippocampus, and the cerebral cortex (20). However, pp125 FAK immunoreactivity is rather diffuse in adult neurons (20), and its precise localization and function are not known. Interestingly, pp125 FAK appears to be a major phosphoprotein altered in Fyn knock-out mice (21), which dis-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ Supported by Comité Evaluation Orientation de la Coopération Scientifique and European Commission Contract ERBCI1*CT940038.
Recently, a 110-kDa tyrosine kinase which shares a high degree of sequence homology with pp125 FAK has been cloned independently by several groups and named proline-rich tyrosine kinase 2 (PYK2) (9), cell adhesion kinase ␤ (CAK␤) (23), related adhesion focal tyrosine kinase (24), or pp125 FAK2 (25). In transfected COS-7 cells, PYK2/CAK␤ is not localized in focal adhesions, but in regions of cell-cell contacts (23). In PC12 cells, PYK2 is phosphorylated on tyrosine and activated in response to various stimuli which raise intracellular Ca 2ϩ and stimulate protein kinase C (9), although the precise mechanism of this activation is not known. In addition, PYK2/CAK␤ induces the phosphorylation on tyrosine of several signal-transducing proteins and ion channels and activates the MAP kinase pathway (9). Since PYK2/CAK␤ is highly expressed in adult brain (23,24), these observations make it an ideal candidate for coupling depolarization and/or activation of neurotransmitter receptors to tyrosine phosphorylation pathways in neurons.
To assess the potential role of PYK2/CAK␤ and pp125 FAK in the nervous system, we have investigated the ability of glutamate agonists and membrane depolarization to regulate tyrosine phosphorylation of these two kinases in rat hippocampal slices, one of the most widely used model for studying synaptic plasticity. Our results show that membrane depolarization stimulates dramatically the phosphorylation of PYK2/CAK␤, whereas agonists of three different types of glutamate receptors increase specifically phosphorylation of pp125 FAK . In addition, we show that phosphorylation of PYK2/CAK␤ and pp125 FAK requires an active PKC. These results demonstrate a high degree of selectivity in the activation of PYK2/CAK␤ and pp125 FAK in mature nervous systems where these kinases may be important to connect neuronal activity and tyrosine phosphorylation pathways.

EXPERIMENTAL PROCEDURES
Rat Hippocampal Slices-Rat hippocampal slices (300 m) were prepared from male Sprague-Dawley rats (100 -150 g) with a McIlewain tissue chopper and incubated (3 slices per tube) in 1 ml of artificial cerebrospinal fluid (ACSF), at 35°C for 50 min before pharmacological treatments, as described previously (7). To avoid indirect effects due to neuronal firing, experiments were all carried out in the presence of 1 M TTX, which was added at the beginning of slice incubation. Treatments with NMDA were carried out in Mg 2ϩ -free ACSF. Neither TTX, nor the absence of Mg 2ϩ , had any effect on protein tyrosine phosphorylation by themselves (data not shown). At the end of the experiment, ACSF was aspirated, and the slices were frozen and kept at Ϫ80°C. Slices were either used for immunoprecipitation (see below) or sonicated in 200 l of a hot solution of 1% (w/v) SDS and 1 mM sodium orthovanadate, in water, and boiled for 5 min.
Subcellular Fractionation-Subcellular fractionation was carried out as described (26). Synaptosomes were prepared from P2 over a Percoll density gradient (27) and incubated in ACSF. After treatment, they were collected by centrifugation and homogenized by sonication as described above for slices.
Antibodies-Antiphosphotyrosine mouse monoclonal antibody 4G10 was from UBI. Serum SL38 was prepared by immunizing a rabbit against the amino-terminal fragment of rat pp125 FAK (31) (residues 1-376) expressed in Escherichia coli as a hexahistidine fusion protein. Serum 539754 was obtained by immunizing a rabbit against a 17-amino acid peptide encompassing residues 2-18 of rat PYK2/CAK␤. Serum 670 -716 was obtained by immunizing a rabbit against a glutathione S-transferase-fusion protein including residues 670 -716 of PYK2/ CAK␤ and affinity-purified (23).

RESULTS AND DISCUSSION
Membrane depolarization results in Ca 2ϩ influx by opening voltage-sensitive Ca 2ϩ channels in neurons, while stimulation of specific glutamate receptors may lead to increases in cytosolic Ca 2ϩ by several mechanisms (see Ref. 32). NMDA receptors, in the absence of extracellular Mg 2ϩ , are highly permeant to Ca 2ϩ , and stimulation of AMPA receptors may result in Ca 2ϩ influx by opening neighboring voltage-sensitive Ca 2ϩ channels. In addition, AMPA receptors of the GluR2 subtype, which are highly expressed in hippocampus, are permeable to Ca 2ϩ (32). Finally, metabotropic glutamate receptors are coupled to phospholipase C and can release Ca 2ϩ from intracellular stores (33). Total slice lysates, boiled in 1% SDS, were immunoblotted with antiphosphotyrosine antibodies, followed by autoradiographic detection of chemiluminescence generated by horseradish-coupled secondary antibodies. Following stripping of these antibodies, the membrane was divided into halves which were immunoblotted either with anti-pp125 FAK monoclonal antibody (2A7, left panel), or with affinity-purified anti-PYK2/ CAK␤ polyclonal antibodies (serum 670-716, right panel). B, following homogenization, samples were successively immunoprecipitated with antibodies against pp125 FAK (SL48, left panel) or PYK2/CAK␤ (670-716, right panel). The immunoprecipitates were analyzed by antiphosphotyrosine immunoblotting. The lack of response of PYK2/CAK␤ to glutamate agonists was observed in all experiments, independently of the order of immunoprecipitation, whereas the effects of KCl on pp125 FAK phosphorylation were variable (see Fig. 2).
increased tyrosine phosphorylation of several proteins, including major components of 120 -130 kDa (Fig. 1A). Membrane depolarization induced by a 2-min increase in the extracellular concentration of K ϩ ions (40 mM KCl) resulted also in a marked stimulation of tyrosine phosphorylation of several proteins (Fig. 1A), confirming and extending our previous results (7). As already reported, the effects of KCl were more pronounced than those of glutamate agonists on a 110-kDa protein (Fig. 1A). Following stripping of the membrane from antiphosphotyrosine antibodies, immunoblotting with specific antibodies revealed that pp125 FAK comigrated with a major 125-kDa band phosphorylated in response to glutamate agonists, whereas PYK2/CAK␤ comigrated with a 110-kDa band phosphorylated in response to KCl-induced depolarization (Fig. 1A).
PYK2/CAK␤ and pp125 FAK were immunoprecipitated from slice homogenates with specific antibodies, and their state of phosphorylation was studied by antiphosphotyrosine immunoblotting. A marked increase in tyrosine phosphorylation of pp125 FAK , but not of PYK2/CAK␤, was observed in response to glutamate agonists (Fig. 1B). It should be noted that a 110-kDa protein phosphorylated in response to glutamate agonists was visible on immunoblots of total homogenates (Fig. 1A). Since it was not immunoprecipitated by two different antibodies against PYK2/CAK␤ ( Fig. 1B and data not shown), it is likely to be a different protein. However, we cannot rule out formally the existence of a form of PYK2/CAK␤ not recognized by our antibodies because of strong interaction with other proteins or of post-translational modifications. In contrast, KCl-induced depolarization increased dramatically tyrosine phosphorylation of PYK2/CAK␤ (Fig. 1B, lane 6). The effects of depolarization on pp125 FAK phosphorylation were less pronounced than on PYK2/CAK␤ (Fig. 2B) and failed to be observed in about onefourth of the slice preparations (compare Figs. 1B and 2). Neither glutamate agonists nor depolarization altered significantly the amounts of pp125 FAK or PYK2/CAK␤ recovered in the immune precipitates, as estimated by a phosphorylation assay (data not shown).
The marked difference observed between the phosphorylation of pp125 FAK and PYK2/CAK␤ in response to glutamate agonists and depolarization raised the possibility that they may have different locations. For instance, PYK2/CAK␤ could be selectively enriched at the presynaptic level and pp125 FAK at the postsynaptic level, where most glutamate receptors are located. However, fractionation experiments demonstrated that only low amounts of the two kinases were present in the crude synaptosomal fraction (P2), in cerebral cortex (Fig. 3A), and in hippocampus (data not shown). Moreover, when purified synaptosomes were incubated in vitro and depolarized by the application of KCl, tyrosine phoshorylation of pp125 FAK was increased (Fig. 3B), whereas no change in PYK2/CAK␤ phosphorylation was detected (data not shown). These results suggest that the selective sensitivity of PYK2/CAK␤ phosphorylation to KCl-induced depolarization is not due to its enrichment in nerve terminals.
The role of Ca 2ϩ in the effects of depolarization on pp125 FAK and PYK2/CAK␤ tyrosine phosphorylation were examined. External Ca 2ϩ was absolutely required for the tyrosine phosphorylation of PYK2/CAK␤ in response to depolarization, which was completely blocked in the presence of 3 mM EGTA applied 3 min before KCl (Fig. 2). The application of EGTA by itself enhanced tyrosine phosphorylation of pp125 FAK in hippocampal slices (Fig. 2) and in cortical synaptosomes (Fig. 3B), but prevented any further increase upon depolarization in both preparations. The basis for the effect of EGTA on pp125 FAK basal phosphorylation is not known and may be related, in part, to nonexocytotic release of endogenous neurotransmitters (34). To examine further the action of Ca 2ϩ , we studied the ability of a Ca 2ϩ ionophore and a phorbol ester, alone or in FIG. 3. Subcellular distribution of PYK2/CAK␤ and pp125 FAK and regulation of pp125 FAK phosphorylation in isolated nerve terminals. A, rat cerebral cortex was homogenized and fractionated by differential centrifugation, as described in Ref. 26. The presence of PYK2/CAK␤ and pp125 FAK was detected in each fraction by immunoblotting of equal amounts of protein with anti-pp125 FAK monoclonal antibody (2A7) or affinity-purified anti-PYK2/CAK␤ polyclonal antibodies (serum 670-716). The lanes correspond to: total homogenate; P1, crude nuclear fraction; P2, crude synaptosomal fraction; P3, crude microsomal fraction; S3, cytosol. B, synaptosomes were purified from P2 by centrifugation on a Percoll gradient, incubated for 45 min in ACSF at 35°C. They were exposed for 2 min to 40 mM KCl, in the absence or presence of 3 mM EGTA, added 3 min prior to KCl. Immunoprecipitation of pp125 FAK was carried out with serum SL38, and its tyrosine phosphorylation was assessed by antiphosphotyrosine immunoblotting .   FIG. 2. Effects of membrane depolarization on tyrosine phosphorylation of PYK2/CAK␤ and pp125 FAK . Hippocampal slices were incubated as described in the legend to Fig. 1 and exposed for 2 min to 40 mM KCl, in the absence or presence of 3 mM EGTA, added 3 min prior to KCl. A, PYK2/CAK␤ and pp125 FAK were immunoprecipitated with specific antibodies (539 754 and SL38, respectively), and their tyrosine phosphorylation was assessed by antiphosphotyrosine immunoblotting. B, quantification of phosphorylation was achieved with computer-assisted measurement of optical density of immunoreactive bands on autoradiograms. Data correspond to the mean Ϯ S.E. of 3-16 samples per condition (statistical analysis was done by analysis of variance followed by Fisher's least significant difference test; *, different from controls p Ͻ 0.02). combination, to reproduce the effects of depolarization in hippocampal slices. A 10-min application of 0.1 M PMA or a 2-min application of 1 M ionomycin was sufficient to increase the phosphorylation of pp125 FAK , but had no consistent effect on PYK2/CAK␤ (Fig. 4). However, the combination of PMA and ionomycin treatments resulted in a synergistic effect on PYK2/ CAK␤ phosphorylation (Fig. 4). These results suggest strongly that activation of PKC is sufficient to induce tyrosine phosphorylation of pp125 FAK . In contrast, PYK2/CAK␤ phosphorylation required the simultaneous addition of Ca 2ϩ ionophore and phorbol ester, an observation which emphasizes the difference in the regulation of these two tyrosine kinases.
To test the potential role of PKC or other Ca 2ϩ -activated enzymes in mediating the effects of depolarization on tyrosine phosphorylation of pp125 FAK and PYK2/CAK␤, various pharmacological inhibitors were used. Pretreatment of slices with antagonists of calmodulin (calmidazolium, trifluoroperazine) or inhibitors of Ca 2ϩ /calmodulin-dependent kinases (KN62) or of calcineurin (FK506, cyclosporin A, alone or in combination) did not prevent the stimulation of protein tyrosine phosphorylation in response to depolarization (data not shown). In contrast, two inhibitors of PKC, RO31-8220 (35) (Fig. 5) or GF109203X (36) (data not shown) inhibited markedly the effects of depolarization on tyrosine phosphorylation of pp125 FAK and PYK2/CAK␤. Moreover, GF109203X prevented also the effects of glutamate on pp125 FAK tyrosine phosphorylation (Fig. 6). These observations indicate that an active PKC is required for the stimulation of PYK2/CAK␤ tyrosine phosphorylation in response to depolarization and of pp125 FAK tyrosine phosphorylation in response to depolarization and glutamate.
Our results show that tyrosine phosphorylation of PYK2/ CAK␤ and pp125 FAK is increased in response to extracellular signals in rat hippocampal slices. However, we found a marked preferential response of PYK2/CAK␤ to depolarization. Although we cannot rule out that PYK2/CAK␤ is activated by neurotransmitters in some circumstances, the situation in hippocampal slices contrasts with that reported in PC12 cells, in which an increase in PYK2 phosphorylation occurred in response to various stimuli (9). Thus, our results suggest the existence of a more specific mechanism of activation of PYK2/ CAK␤ in mature nervous system than in cell lines. This specific activation cannot be attributed to an enrichment of PYK2/ CAK␤ in nerve terminals, but may be related to a preferential localization of PYK2/CAK␤ in the vicinity of voltage-activated Ca 2ϩ channels. In contrast, pp125 FAK phosphorylation was stimulated not only following depolarization, but also in response to stimulation of glutamate receptors. These results argue for the presence of pp125 FAK at the postsynaptic level, where these receptors are mostly found, in agreement with our previous immunocytochemical observations (20). Nevertheless, we found that pp125 FAK is also present in nerve terminals where it is phosphorylated in response to depolarization. It should be noted that, since phosphorylation of pp125 FAK is enhanced in brain in response to several neurotransmitters including glutamate (this study), acetylcholine, 2 and anandamide (37), it is possible that the increase in pp125 FAK phospho- FIG. 4. Effects of phorbol ester and Ca 2؉ ionophore on PYK2/ CAK␤ and pp125 FAK tyrosine phosphorylation. Hippocampal slices were incubated as described in the legend to Fig. 1 and exposed for 10 min to 0.1 M PMA, or 2 min to 1 M ionomycin, or the combination of the two treatments. A, PYK2/CAK␤ and pp125 FAK were immunoprecipitated with specific antibodies (539 754 and SL38, respectively), and their tyrosine phosphorylation was assessed by antiphosphotyrosine immunoblotting. B, quantification of phosphorylation and statistical analysis were done as indicated in the legend to FIG. 5. Blockade by a PKC inhibitor of the effects of membrane depolarization on PYK2/CAK␤ and pp125 FAK tyrosine phosphorylation. Hippocampal slices were incubated as described in the legend to Fig. 1 and exposed for 2 min to 40 mM KCl, in the absence or presence of 100 M RO31-8220 added 40 min prior to KCl. PYK2/CAK␤ and pp125 FAK were immunoprecipitated with specific antibodies (539 754 and SL38, respectively), and their tyrosine phosphorylation was assessed by antiphosphotyrosine immunoblotting. The data presented are representative of three independent experiments which gave similar results.
FIG. 6. Blockade by a PKC inhibitor of the effects of glutamate on pp125 FAK tyrosine phosphorylation. Hippocampal slices were incubated as described in the legend to Fig. 1 and exposed for 5 min to 100 M glutamate, in the absence or presence of 100 M GF109203X added 40 min prior to glutamate. pp125 FAK was immunoprecipitated with specific antiserum SL38, and its tyrosine phosphorylation was assessed by antiphosphotyrosine immunoblotting. The data presented are representative of three independent experiments which gave similar results. rylation in response to depolarization results in part from the release of endogenous neurotransmitters.
PYK2/CAK␤ and pp125 FAK tyrosine phosphorylation in response to depolarization required the presence of extracellular Ca 2ϩ and is likely to result from Ca 2ϩ influx. In addition, several observations support a critical role for PKC. First, PKC activation, by a phorbol ester alone in the case of pp125 FAK , or in combination with a Ca 2ϩ ionophore in the case of PYK2/ CAK␤, increased tyrosine phosphorylation of these proteins. Second, among all the various inhibitors of Ca 2ϩ -activated enzymes that we have tried, only PKC inhibitors were able to decrease significantly the phosphorylation of PYK2/CAK␤ and pp125 FAK resulting from depolarization and the phosphorylation of pp125 FAK induced by glutamate. However, we cannot exclude that PKC plays only a permissive role and that direct stimulation of another Ca 2ϩ -activated protein is responsible for the activation of pp125 FAK and/or PYK2/CAK␤ phosphorylation.
Phosphorylation of pp125 FAK and PYK2/CAK␤ provides a link between increases in Ca 2ϩ and tyrosine phosphorylation pathways, which may be involved in neuronal survival and synaptic plasticity (5). In light of the known importance of Ca 2ϩ -activated processes, PKC and tyrosine phosphorylation in LTP and LTD (10,11,38), it is tempting to speculate that pp125 FAK and/or PYK2/CAK␤ are involved in the signaling events leading to synaptic plasticity. However, only pp125 FAK was clearly sensitive to stimulation of NMDA receptors which plays a critical role in synaptic plasticity (1,38). On the other hand, pp125 FAK phosphorylation was also increased in response to stimulation of AMPA and metabotropic glutamate receptors, which appears unable to induce LTP or LTD in hippocampus (1,38), indicating that phosphorylation of pp125 FAK is not sufficient for generating these long term changes in synaptic efficacy. Thus, both pp125 FAK and PYK2/ CAK␤ appear to be involved in the Ca 2ϩ -induced activation of tyrosine phosphorylation in nervous tissue. However, in spite of the similarities between these two tyrosine kinases, their regulation and, probably, their role appear different.