Calcineurin is essential for depolarization-induced nuclear translocation and tyrosine phosphorylation of PYK2 in neurons Free

Proline-rich tyrosine kinase 2 (PYK2), also referred to as cell-adhesion kinase β (CAKβ) (Sasaki et al., 1995), RAFTK (Avraham et al., 1995) or CADTK (Yu et al., 1996), is a non-receptor tyrosine kinase, closely related to focal adhesion kinase (FAK). PYK2 is highly enriched in neurons and thought to play an important role in neuronal plasticity (Girault et al., 1999; Henley and Nishimune, 2001). PYK2 is phosphorylated on tyrosine after depolarization (Siciliano et al., 1996) and its activation appears to be necessary for the induction of LTP (Huang et al., 2001). In addition, PYK2 is activated in the hippocampus following brain ischemia and convulsions (Tian et al., 2000) suggesting a role in these pathologies. PYK2 is also involved in many other biological functions including activation of macrophages (Okigaki et al., 2003) and osteoclasts (Lakkakorpi et al., 2003), as well as in pathological conditions including cancer (Lipinski et al., 2003) and cardiac hypertrophy (Hirotani et al., 2004).

The most striking characteristic of PYK2 is its activation following increases in cytosolic free Ca²⁺ (Lev et al., 1995). The mechanism of this activation, however, is not understood. In many cell types activation of PYK2 appears to involve protein kinase C (PKC) (Lev et al., 1995; Siciliano et al., 1996) and, in some cases, a Ca²⁺-calmodulin-dependent protein kinase (Zwick et al., 1999). The phosphorylation reactions mediating directly or indirectly the activation of PYK2 remain to be identified. Increases in Ca²⁺ lead to PYK2 autophosphorylation on Tyr402, creating a Src-homology-2 (SH2) binding site that recruits Src family kinases, which phosphorylate other tyrosine residues of PYK2 and associated proteins (Dikic et al., 1996; Park et al., 2004). Interactions between the activated PYK2-Src module and proteins such as the Grb2-Sos complex, p130Cas, paxillin and Graf regulate multiple intracellular signaling pathways (reviewed by Avraham et al., 2000).

PYK2 possesses a C-terminal focal adhesion targeting (FAT) domain very similar to that of FAK (61% identity), although in most cells PYK2 is not enriched at focal contacts. Transfected PYK2 is generally expressed in the cytoplasm (Schaller and Sasaki, 1997; Zheng et al., 1998), whereas endogenous PYK2 can be observed in perinuclear punctuate structures (Sieg et al., 1998) or, for a fraction, at focal adhesions (Du et al., 2001). In neurons, PYK2 is mostly localized in perikarya and dendritic shafts (Corvol et al., 2005; Menegon et al., 1999). Although it is likely that PYK2 plays an important role in post-synaptic densities, because of its interaction with NMDA receptors and associated scaffold proteins (Bongiorno-Borbone et al., 2005; Cheung et al., 2000; Heidinger et al., 2002; Liu et al., 2001; Seabold et al., 2003), it does not appear to be enriched in spines. By contrast, deleted or mutated forms of PYK2 accumulate in the nucleus of transfected COS-7 cells (Aoto et al., 2002) and PYK2 is localized in the nucleus in chondrocytes and keratinocytes (Arcucci et al., 2006; Schindler et al., 2007). In rat brain, following ischemia or convulsions, PYK2 immunoreactivity appears to be in part nuclear (Tian et al., 2000). However, the significance of these observations is not known.

We show here that depolarization induces PYK2 translocation to the nucleus of neurons and PC12 cells. PYK2 redistribution to the nucleus is independent of its tyrosine phosphorylation and tyrosine kinase activity. By contrast, inhibition of the Ser/Thr phosphatase calcineurin blocked both depolarization-induced PYK2 tyrosine phosphorylation and nuclear translocation, providing evidence of a key role of calcineurin in the regulation of PYK2.

We and others previously reported that depolarization of hippocampal slices induces a robust tyrosine phosphorylation of PYK2 (Derkinderen et al., 1998; Huang et al., 2001; Siciliano et al., 1996). In this preparation, KCl-induced PYK2 phosphorylation peaked around 2 minutes and returned to a basal state in 20 minutes (Corvol et al., 2005). PYK2 phosphorylation at Tyr402 occurred primarily in cell bodies and dendritic shafts of hippocampal neurons (Corvol et al., 2005). In the course of these experiments, we noticed that depolarization induced an accumulation of PYK2 immunoreactivity in the nuclei of neurons and we undertook a systematic study of PYK2 localization in rat hippocampal slices by immunohistofluorescence.

In the absence of depolarization, there were few neurons with nuclear PYK2 immunoreactivity in the CA1 region of hippocampus (Fig. 1A). When the extracellular concentration of K⁺ ([K⁺]_o) was raised by 40 mM for 2 minutes a striking accumulation of PYK2 was observed in the nuclei of neurons in CA1. Counting PYK2-immunoreactive nuclei showed that approximately 40% of nuclei were labeled in depolarized slices (Fig. 1B). PYK2 nuclear translocation was rapid and transient since it was observed within 2 minutes after the onset of depolarization and disappeared after 10 minutes (Fig. 1C).

Depolarization of hippocampal slices increases PYK2 autophosphorylation on Tyr402 and its phosphorylation by Src-family kinases (Corvol et al., 2005; Siciliano et al., 1996). We investigated the role of Src recruitment in the depolarization-induced nuclear translocation of PYK2, using PP2, a Src-family inhibitor that inhibits tyrosine phosphorylation of PYK2 in depolarized hippocampal slices, without altering phosphorylation of Tyr402 (Corvol et al., 2005). Pretreatment of hippocampal slices with PP2 decreased dramatically depolarization-induced PYK2 tyrosine phosphorylation and Src autophosphorylation (Fig. 1D). However, this treatment did not alter high [K⁺]_o-induced PYK2 translocation to the nucleus (Fig. 1E,F), showing that Src-family kinases activation was not necessary for PYK2 redistribution.

We next examined whether synaptic activation of neurons in response to electrical stimulation of afferent fibers (Schaeffer collaterals) in the CA1 region of mouse hippocampus altered PYK2 subcellular localization. In unstimulated slices, PYK2 immunoreactivity examined by confocal microscopy was localized mainly in the cytoplasm and dendritic shafts (Fig. 2A upper panel). After high [K⁺]_o, PYK2 immunostaining in the nucleus was greater or equal to that in the cytoplasm in 75% of PYK2-immunoreactive cells (Fig. 2A upper panels, B). Electrical stimulation of afferent fibers of CA1 pyramidal cells (100 Hz for 1 second, four times at 10-second intervals) induced a potentiation of synaptic transmission (Fig. 2C) and triggered nuclear accumulation of PYK2 immunoreactivity (Fig. 2A,B). In the same experimental conditions, we studied phosphorylation of Tyr402, using phosphorylation state-specific antibodies (Fig. 2A lower panels). Virtually no immunofluorescence was observed in control slices, whereas a dramatic increase in pY402-PYK2 immunoreactivity was observed after high [K⁺]_o-induced depolarization or electric stimulation of Schaeffer collaterals. Immunofluorescence was observed in dendrites and perikarya, but was more intense in the nuclei. These results revealed that synaptic activation of hippocampal pyramidal neurons, as well as direct depolarization, induced a nuclear translocation of PYK2 and that its autophosphorylated form also accumulated in the nucleus.

To determine whether depolarization-induced PYK2 nuclear accumulation occurred in other experimental conditions we first used hippocampal neurons in primary culture. Analysis by fluorescence microscopy showed that high [K⁺]_o induced the nuclear accumulation of PYK2 immunoreactivity (Fig. 3A). PYK2 subcellular distribution was predominantly cytoplasmic in ∼70% of the neurons in primary cultures in basal conditions, whereas 3 minutes after depolarization, it was detected predominantly in the nucleus in ∼85% of the neurons (Fig. 3B). This high [K⁺]_o-induced nuclear accumulation of PYK2 was confirmed by confocal microscopy analysis (Fig. 3C). As the major consequence of depolarization is the opening of voltage-gated Ca²⁺ channels and the increase in intracellular Ca²⁺ levels, we examined if PYK2 nuclear translocation was Ca²⁺-dependent in hippocampal neurons in culture. Pretreatment with EGTA prevented the change in PYK2 localization (Fig. 3A,B), demonstrating that PYK2 nuclear translocation required Ca²⁺ entry.

We then used PC12 cells, a cell line with neuronal characteristics, to further investigate the mechanisms involved in the depolarization-induced nuclear translocation of PYK2 (Fig. 4A). Subcellular distribution of endogenous PYK2 was classified into three classes (n<c, n=c and n>c) depending on whether nuclear labeling intensity was less, equal or greater than cytoplasmic labeling, respectively. In this model, high [K⁺]_o depolarization triggered PYK2 translocation, as evidenced by its colocalization with DAPI (Fig. 4A,B). Depolarization also increased pY402-PYK2 immunoreactivity, which was detected in the cytoplasm and nuclei of PC12 cells (Fig. 4C,D). As in hippocampal slices, in depolarized PC12 cells PP2 caused a dramatic decrease in tyrosine phosphorylation of PYK2 (supplementary material Fig. S1A,B) but did not alter high [K⁺]_o-induced PYK2 translocation to the nucleus (supplementary material Fig. S1C,D).

Many cytoplasmic proteins undergo a constant cyto-nuclear shuttling, which is unmasked by treatment with leptomycin B, an antibiotic blocking CRM1 (also known as exportin 1)-mediated nuclear export (Nishi et al., 1994). When PC12 cells were treated with leptomycin B for 3 hours, an increase in PYK2 nuclear staining was observed (Fig. 4A,B). However, quantification showed that the effects of leptomycin B on the nuclear accumulation of PYK2 were much weaker than those of depolarization (Fig. 4B).

We then studied the intracellular localization of a green fluorescent protein-PYK2 fusion protein (GFP-PYK2) in transfected PC12 cells. GFP-PYK2 was mainly cytoplasmic in normal [K⁺]_o conditions, with no overlap with DAPI nuclear staining in most cells (Fig. 5A). By contrast, 3 minutes after high [K⁺]_o, GFP-PYK2 fluorescence was partly colocalized with DAPI (Fig. 5A) and clearly located in the nucleus when viewed by confocal microscopy (Fig. 5B). Depolarization-induced nuclear translocation of GFP-PYK2 in PC12 cells was prevented by EGTA pretreatment (Fig. 5A,C). These results showed that GFP-PYK2 behaved like endogenous PYK2 in PC12 cells in response to depolarization.

Our above results in hippocampal slices showed that PYK2 translocation was independent from Src-family kinases. To investigate the role of PYK2 autophosphorylation and kinase activity we used mutant GFP-PYK2, in which we replaced Tyr402 or Lys457, a critical ATP-binding residue, with Phe (Y402F) or Ala (K457A), respectively. Immunoblotting with anti-phospho-Tyr402-PYK2-specific antibody showed that high [K⁺]_o induced a strong phosphorylation of wild-type GFP-PYK2-Tyr402 (Fig. 6A), whereas this phosphorylation was abolished in GFP-PYK2-Y402F and GFP-PYK2-K457A (Fig. 6A). In the same samples, Tyr402 in endogenous PYK2 was weakly phosphorylated in response to depolarization in all cases (Fig. 6A). Although the total amounts of endogenous PYK2 and GFP-PYK2 were similar, Tyr402 phosphorylation of GFP-PYK2 appeared stronger. A possible explanation is that PYK2 concentration was higher in transfected cells, which were only 5-10% of the total. This high concentration is expected to favor autophosphorylation that occurs through an intermolecular mechanism (Park et al., 2004).

We then examined the subcellular localization of transfected wild-type or mutant GFP-PYK2 fusion proteins in response to depolarization (Fig. 6B). Wild type, Y402F and K457A GFP-PYK2 displayed a similar nuclear translocation after K⁺ depolarization (Fig. 6B,C), demonstrating that PYK2 autophosphorylation and kinase activity are not necessary for depolarization-induced PYK2 nuclear translocation.

In the course of these experiments, we noticed on immunoblots a slight downward shift of the endogenous PYK2 band in samples from K⁺-treated cells (Fig. 6A, see also Fig. 1D in slices). To further investigate this observation, small amounts of protein from control or K⁺-treated PC12 cells were loaded on a 7% acrylamide-bisacrylamide gel and migration time was increased. This resulted in a distinct shift of endogenous PYK2 in depolarized PC12 cells, towards a lower apparent molecular mass (Fig. 6D). Several mechanisms could account for this accelerated gel mobility, including dephosphorylation. As depolarization induces a marked increase in PYK2 tyrosine phosphorylation, such dephosphorylation would be expected to occur on Ser/Thr residues.

Depolarization-induced Ca²⁺ influx can activate protein phosphatase 2B, also known as calcineurin, a Ser/Thr phosphatase regulated by Ca²⁺ and calmodulin. To test whether calcineurin might be involved in the activation of PYK2, we used cyclosporin A, a potent calcineurin inhibitor (Liu et al., 1991). Pretreatment of PC12 cells with cyclosporin A (10 μM, 30 minutes) had no effect by itself (data not shown) but dramatically decreased depolarization-induced PYK2 tyrosine phosphorylation (Fig. 7A,B,). We then examined the effects of cyclosporin A pretreatment on PYK2 intracellular localization (Fig. 7C). Pretreatment of PC12 cells with cyclosporin A before depolarization prevented PYK2 nuclear translocation (Fig. 7C,D).

In order to rule out a non-calcineurin-mediated effect of cyclosporin A, we used FK506 an immunosuppressant structurally unrelated to cyclosporin A, which inhibits calcineurin by interacting with a different immunophilin (Rusnak and Mertz, 2000). Pretreatment of PC12 cells with FK506 (1 μM, 30 minutes before high K⁺) prevented the depolarization-induced increase in endogenous PYK2 tyrosine phosphorylation (Fig. 7E,F) and specifically the phosphorylation of Tyr402 (Fig. 7E,G). Depolarization-induced translocation of GFP-PYK2 was also completely prevented in FK506-pretreated PC12 cells (Fig. 7H,I). Similar results were obtained with hippocampal slices pretreated with FK506 (supplementary material Fig. S2). Since we have shown above that nuclear translocation of PYK2 did not require its tyrosine phosphorylation, these results strongly indicate that Ca²⁺-induced activation of calcineurin is necessary for two independent events, the autophosphorylation of PYK2 and subsequent recruitment of Src-family kinases, and the nuclear translocation of PYK2.

Calcineurin is a heterodimer of a catalytic (CnA) subunit and a regulatory (CnB) subunit (Klee et al., 1998). CnA encompasses several domains: a catalytic domain, a CnB-binding domain, a calmodulin-binding domain and an autoinhibitory domain. To test the role of calcineurin in PYK2 tyrosine phosphorylation we used a Flag-tagged CnA construct (1-397) deleted from its autoinhibitory domain and bearing a point mutation in its catalytic site (H151Q, phosphatase-dead, PD-CnA) which behaves as a dominant-negative inhibitor of calcineurin (Kahl and Means, 2004). We co-transfected GFP-PYK2 and PD-CnA or vector, and cells were treated with a control solution or a high [K⁺]_o depolarizing solution (Fig. 8). Depolarization-induced tyrosine phosphorylation of GFP-PYK2 was dramatically decreased in PD-CnA co-transfected cells (Fig. 8A,B). Since cotransfection of PD-CnA slightly decreased the expression of GFP-PYK2, we normalized the amount of tyrosine-phosphorylated GFP-PYK2 to the amount of total GFP-PYK2 protein: depolarization-induced increase in tyrosine phosphorylation was 80±18% in mock-cotransfected cells and 24±3% in PD-CnA-cotransfected cells. Thus, these results strongly supported the implication of endogenous calcineurin in this response.

We then investigated the role of calcineurin in PYK2 nuclear translocation by examining the effects of constitutively active CnA (1-397, Flag-CA-CnA) or phosphatase-dead CnA (Flag-PD-CnA) on endogenous localization of PYK2 in PC12 cells (Fig. 9A). These two plasmids gave similar levels of Flag-protein expression (Fig. 9B). PC12 cells were transfected with Flag-PD-CnA, Flag-CA-CnA or, as control, an unrelated protein subcloned into the same expression vector (Flag-protein), and treated or not with high [K⁺]_o. Endogenous PYK2 localization was examined by immunofluorescence in the Flag-positive cells. In PC12 cells transfected with Flag-PD-CnA the effects of K⁺ on PYK2 nuclear translocation were markedly decreased as compared with Flag-protein-transfected control cells, showing that inhibition of endogenous calcineurin altered PYK2 nuclear translocation (Fig. 9A,C). By contrast, in cells transfected with CA-CnA, we observed an increase in the number of cells with nuclear PYK2, in the absence of depolarization (Fig. 9A,C). High [K⁺]_o did not further increase that proportion. These results showed that blockade of calcineurin with a dominant-negative mutant prevented depolarization-induced PYK2 nuclear accumulation, whereas a constitutively active mutant of calcineurin was sufficient to trigger that translocation.

Although the non-receptor tyrosine kinase PYK2 was initially described as a cytoplasmic protein, activated by increases in cytosolic free Ca²⁺ (Avraham et al., 1995; Sasaki et al., 1995; Yu et al., 1996), in some cells, PYK2 immmunoreactivity is nuclear (Arcucci et al., 2006; Schindler et al., 2007; Tian et al., 2000). The significance of these observations was not clear and our results provide the first evidence for a physiological translocation of PYK2 from the cytoplasm to the nucleus since we show that depolarization or electrical stimulation induced a nuclear accumulation of PYK2 in hippocampal slices. We observed similar effects in neurons and PC12 cells in culture. Several proteins that appear to be located in the cytoplasm undergo a constant cyto-nuclear shuttling, revealed by treatment with leptomycin B, an antibiotic blocking CRM1/exportin1-mediated nuclear export (Nishi et al., 1994). The effects of leptomycin B on PYK2 nuclear accumulation were weak, indicating that PYK2 undergoes a limited basal CRM1-dependent nucleo-cytoplasmic shuttling in PC12 cells. In COS-7 cells, a more pronounced nuclear accumulation of transfected PYK2 after treatment with leptomycin B has been reported (Aoto et al., 2002), suggesting that high levels of expression of PYK2 may facilitate its nuclear accumulation, perhaps by saturating cytoplasmic anchors for this protein. In PC12 cells, depolarization-induced nuclear accumulation of PYK2 was rapid (2-3 minutes), transient, and more intense than following leptomycin B. Thus the effects of depolarization on PYK2 subcellular localization are unlikely to result only from an interference with CRM1-mediated nuclear export. Although we cannot exclude that PYK2 also undergoes a CRM1-independent nuclear export, these results suggest that depolarization acts by releasing PYK2 from a cytoplasmic anchor and/or by promoting its nuclear import.

Increases in cytosolic free Ca²⁺ trigger PYK2 autophosphorylation and the recruitment of Src family kinases, which in turn phosphorylate multiple tyrosines of PYK2 (Lev et al., 1995). Since nuclear translocation of PYK2 was also Ca²⁺-dependent and PYK2 activation and nuclear translocation occurred concomitantly, we tested whether nuclear translocation involved PYK2 tyrosine phosphorylation. Clearly, PYK2 nuclear translocation did not require its tyrosine phosphorylation. By contrast, nuclear translocation of PYK2 was prevented by pretreatment of PC12 cells with two calcineurin inhibitors, cyclosporin A and FK506, which act through binding to different immunophilins. Calcineurin inhibition also blocked depolarization-induced PYK2 tyrosine autophosphorylation and subsequently tyrosine phosphorylation by Src family kinases. The role of calcineurin was confirmed by transfection of dominant-negative or constitutively active forms of this enzyme in PC12 cells. These results strongly indicate that Ca²⁺-activated calcineurin simultaneously triggers two independent events: autophosphorylation of PYK2 on Tyr402 and nuclear translocation of PYK2. As tyrosine phosphorylation of PYK2 is not necessary for its nuclear translocation, it is likely that calcineurin acts upstream from these two events.

Cyto-nuclear shuttling of many proteins is regulated by Ser/Thr phosphorylation events (Poon and Jans, 2005). A well characterized example is the translocation of the nuclear factor of activated T cells (NFAT) triggered by calcineurin-mediated dephosphorylation of Ser/Thr residues located near a nuclear localization sequence (Crabtree and Olson, 2002). Our results reveal that calcineurin also plays a critical role in the regulation of PYK2 nuclear redistribution. Calcineurin may dephosphorylate Ser/Thr residues in PYK2 or in associated proteins. PYK2, as the related FAK, contains many potential serine and threonine phosphorylation sites, some of which are known to be phosphorylated (Grigera et al., 2005; Wissing et al., 2006). Regulation of PYK2 by serine/threonine phosphorylation is likely to be complex since several protein kinases including protein kinase C and Ca²⁺-calmodulin-dependent kinases appear to be involved (Lev et al., 1995; Siciliano et al., 1996; Zwick et al., 1999). The mechanism of action of dephosphorylation can only be speculative at this time. In transfected COS-7 cells nuclear accumulation of PYK2 was promoted by mutation of Pro859 to Ala (Aoto et al., 2002), whereas in chondrocytes a N-terminally truncated form of PYK2 was detected in the nucleus (Arcucci et al., 2006). Possible interpretations of these findings are that mutation or truncation releases PYK2 from a cytoplasmic anchor and/or modifies the exposition of nuclear localization sequences. Our results suggest that such processes could be triggered by Ser/Thr dephosphorylation, in response to a physiological stimulus raising intracellular Ca²⁺. PYK2 does not have a canonical nuclear localization sequence, indicating that its nuclear import involves a different targeting motif or association with other proteins. Studies are in progress to clarify these issues.

Altogether, our results provide novel information regarding PYK2 regulation and function. Although this tyrosine kinase was initially described as activating cytoplasmic signaling pathways, our results show that it may have an important function in the nucleus. Interestingly, in keratinocytes PYK2 is constitutively nuclear and appears to be involved in Jun-D and Fra1 expression (Schindler et al., 2007). It has also been reported that PYK2 and calcineurin are involved in the control of serum response element by muscarinic receptors (Lin et al., 2002). Thus, it is tempting to speculate that, in neurons, PYK2 regulates nuclear functions important for long-lasting plasticity, in response to Ca²⁺ entry. Interestingly, the related kinase FAK can also undergo a nuclear translocation in some specific conditions or cell types (Yi et al., 2006), and can be sumoylated (Kadare et al., 2003), a post-translational modification which usually takes place in the nucleus. These observations suggest that PYK2 and possibly FAK, under specific circumstances, play a role in the control of nuclear functions such as RNA transcription or processing.

Another important implication of the present study is that inhibition of PYK2 must be considered among the relevant targets of calcineurin inhibitors. These drugs are powerful and useful immunosuppressants, known to act by preventing NFAT activation in lymphocytes. As PYK2 is involved in maturation of specific B cell populations (Guinamard et al., 2000) and in macrophage activation (Okigaki et al., 2003), prevention of its activation through inhibition of calcineurin may contribute to the immunosuppressive effects of FK506 and cyclosporin A. Interestingly calcineurin inhibitors have also been reported to exert protective effects in cardiac hypertrophy, attributed to NFAT activation (Heineke and Molkentin, 2006). As PYK2 is also a critical player in some forms of cardiac hypertrophy (Hirotani et al., 2004), our results raise the intriguing possibility that some of the cardioprotective effects of calcineurin inhibitors are mediated by PYK2 inhibition. Finally, it should be pointed out that FK506 or cyclosporin A may possibly be a useful means of indirectly inhibiting PYK2 activation in pathological conditions, including osteoclastic bone resorption and in some cancers.

4-Amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (PP2), leptomycin B, and cyclosporin A were from Calbiochem. Anti-phosphotyrosine (4G10) mouse monoclonal antibody (blot: 1:4000) was from Upstate Biotechnology. Affinity purified rabbit antibodies anti-PYK2 residues 2–18 (blot: 1:200) were described previously (Siciliano et al., 1996). Affinity-purified rabbit anti-phospho-Tyr402-PYK2 antibodies (blot: 1:1000) and affinity-purified rabbit anti-phospho-Tyr418-Src antibodies (blot: 1:2500) were from Biosource. Anti-flag M2 mouse monoclonal antibody was from Sigma (blot: 1:1000, immunofluorescence: 1:250). Alexa Fluor 488- or Cy3-coupled secondary antibodies (1:400) were from Molecular Probes.

Hippocampal slices were prepared from desflurane-anesthetized B6CBA mice (older than 60 days). Transverse slices (400 μm) were cut from the middle portion of each hippocampus with a vibroslicer in artificial cerebrospinal fluid (ACSF, 4°C, bubbled with 95% O₂ and 5% CO₂, pH 7.4), placed in a humidified interface chamber at 30±1°C and perfused with ACSF containing 2 mM CaCl₂. Orthodromic stimuli (50 μseconds, <300 μA, 0.1 Hz) were delivered alternately through two tungsten electrodes, in the stratum radiatum of the CA1 region. Extracellular synaptic responses were monitored by two ACSF-filled glass electrodes placed in the corresponding synaptic layers. After obtaining stable synaptic responses in both pathways (0.1 Hz stimulation) for at least 10-15 minutes, the slices were exposed to one of the following three procedures: (1) in the control group synaptic responses were monitored during low frequency stimulation (0.1 Hz) to ascertain the viability of the slices and fixed; (2) in the high [K⁺]_out group, synaptic responses were monitored before and during exposure to 40 mM K⁺; (3) in the tetanization group synaptic responses were monitored before and 4-6 minutes following a tetanization procedure, which consisted of 1-second 100 Hz stimulation given four times (10-second interval) alternatively to each of the radiatum pathways. The stimulation strength used for tetanization was well above threshold for generation of a population spike in response to a single test shock. Synaptic efficacy was assessed by measuring the slope of the synaptic field excitatory post-synaptic potentials (fEPSPs) in the middle third of their rising phase, and normalized to stable control recordings. At the end of the recordings, slices were fixed as described below.

Rat hippocampal slices (300 μm) were prepared from young male Sprague-Dawley rats (100–150 g) with a McIlwain tissue chopper and incubated as previously described (Corvol et al., 2005; Siciliano et al., 1996). Slices from rat or mouse were placed in paraformaldehyde (4% weight/vol.) in PBS at 4°C overnight, and then stored at 4°C in PBS. Sections (30-μm) were cut with a microtome (Leica) from agarose-embedded slices and kept at –20°C in a solution containing 30% ethylene glycol, 30% glycerol, and 0.1 M phosphate buffer. Immunolabeling procedures were as described previously (Valjent et al., 2000) using Alexa Fluor 488- or Cy3-coupled secondary antibodies. Sections were mounted in Vectashield with 4α, 6-diamidino-2-phenylindole (DAPI) counterstain (Vector Laboratories) and studied using laser scanning confocal microscopy (SP2, Leica).

Low-density primary cultures of hippocampal neurons were prepared from rat embryos (E18). Hippocampal cells were dissociated and plated at a density of 20,000/cm² on polylysine-coated glass coverslips in Neurobasal-B27 (Neurobasal, Invitrogen). PC12 cells were grown on type I collagen-coated dishes (BD Biosciences) in RPMI medium (GIBCO) containing 10% horse serum, 5% fetal calf serum. Transfections were done with Lipofectamine 2000 (Invitrogen). For fluorescence analysis, PC12 cells were grown in RPMI on type I collagen-coated glass coverslips after incubation with poly-L-lysine (Sigma). For 40 mM K⁺-induced depolarization of cells in culture, half of the cell culture medium was removed and replaced for 3 minutes by a solution containing 1 mM MgCl₂, 2 mM CaCl₂, 25 mM Hepes and either 135 mM NaCl, (control solution), or 55 mM NaCl, 80 mM KCl (high [K⁺]_o). In both cases, the osmolarity of the solution remained 300 mOsm/l.

Cells were fixed for 15 minutes in a solution containing 4% (weight/vol) paraformaldehyde, permeabilized on ice with methanol-acetone (vol./vol.) for 12 minutes. Cells were washed with PBS, blocked and incubated for 2 hours with primary antibodies [anti-PYK2 antibody (1/100), anti-phospho-Tyr402 (1/200) and mouse anti-flag M2 monoclonal antibody (1/250)]. After washes, cells were incubated with Alexa Fluor 488- or Cy3-coupled secondary antibodies (1/400) for 45 minutes, washed and mounted in Vectashield with DAPI. PC12 cells transfected with GFP-PYK2 constructs were fixed, washed and mounted in Vectashield with DAPI. Images were acquired with a digital camera CCD Micromax (Roper Scientific) or with a confocal microscope Leica SP2.

PC12 cells were lysed on ice, in RIPA buffer [1% Triton X-100, 0.5% (weight/vol.) deoxycholate, 0.1% (weight/vol.) SDS, 50 mM Tris (pH 7.4), 150 mM NaCl, 1 mM sodium orthovanadate, 50 mM NaF, 10 mM sodium pyrophosphate and protease inhibitors (Complete Boehringer 1/20)]. Immunoprecipitations and immunoblot were carried out as described previously (Derkinderen et al., 2001). Quantifications were carried out by scanning the autoradiograms and measuring relative optical density with Scion Image software, or by direct measurement using the Odyssey imaging system (Li-Cor Bioscences, Lincoln, NE). Data were normalized to the mean value of untreated controls in the same autoradiograms.

A plasmid encoding the N terminus of rat PYK2 (1-364) was fused to green fluorescent protein by ligating the BamHI-SalI fragment of PYK2 into the BglII-SalI sites of pEGFP-C1 (Clontech). Full-length PYK2 was obtained by ligating the ScaI-BamHI fragment of PYK2 into the ScaI-BamHI sites of this construct. The Y402F and K457A mutants of PYK2 were prepared by site-directed mutagenesis (QuikChange, Stratagene). The HindIII fragment of PYK2 was subcloned into the HindIII site of pBlueScript KS (Fermentas). Oligonucleotides to change Y402 to phenylalanine and K457 to alanine were GCATAGAGTCAGACATCTTTGCAGAGATTCCTGATGAGACCC and GGGAAAAAATTAATGTGGCCGTCGCGACCTGTAAGAAAGATTGTACCC, respectively. Mutations were verified by DNA sequencing. Plasmids encoding constitutively active calcineurin A (CnA) (residues 1-397, CA-CnA), and mutated CnA (H151Q) were a gift from A. Means and were subcloned into the BglII-SmaI sites of FLAG-CMV2 expression vector (Sigma). The H151Q CnA mutant was truncated and the phosphatase-dead form used in the study was PD-CnA 1-397 H151Q.

Cells were classified in three categories by observers blind to the treatment: cytoplasmic fluorescence more intense than nuclear fluorescence (c>n), cytoplasmic equal to nuclear fluorescence (c=n) or cytoplasmic inferior to nuclear fluorescence (c<n). The boundaries of the nuclei were determined by DAPI staining. The percentage of cells in each category was determined for each coverslip (approximately 50-100 cells per coverslip in 10-20 fields). Data are from at least three independent experiments each in duplicate. Statistical analysis was done using GraphPad Prism 3.02.

This work was supported in part by Inserm, UPMC, and grants ACI-BCMS from the Ministry of Research, Appel non thématisé from Agence Nationale de la Recherche (ANR), and Association pour la Recherche contre le Cancer (ARC). The authors thank A. Means and M. Sanchez Matamales for providing constructs.