In blood platelets, stimulation of G protein-coupled receptors (GPCRs) by thrombin triggers the activation of Src family kinases (SFKs), resulting in the tyrosine-phosphorylation of multiple substrates, but the mechanism underlying this process is still poorly understood. In the present study, we show that the time-dependent protein-tyrosine phosphorylation triggered by thrombin in human or murine platelets was totally suppressed only upon concomitant chelation of intracellular Ca2+ and inhibition of SFKs. Thrombin-induced activation of SFKs was regulated by intracellular Ca2+ and accordingly the Ca2+ ionophore A23187 was sufficient to stimulate SFKs. A23187 also triggered the phosphorylation and activation of the Ca2+-dependent focal adhesion kinase Pyk2 and Pyk2 activation by thrombin was Ca2+-dependent. Stimulation of SFKs by thrombin or A23187 was strongly reduced in platelets from Pyk2 knockout (KO) mice, as was the overall pattern of protein-tyrosine phosphorylation. By immunoprecipitation experiments, we demonstrate that Lyn and Fyn, but not Src, were activated by Pyk2. Inhibition of SFKs by PP2 also reduced the phosphorylation of Pyk2 in thrombin or A23187-stimulated platelets. Analysis of KO mice demonstrated that Fyn, but not Lyn, was required for complete Pyk2 phosphorylation by thrombin. Finally, PP2 reduced aggregation of murine platelets to a level comparable to that of Pyk2-deficient platelets, but did not have further effects in the absence of Pyk2. These results indicate that in thrombin-stimulated platelets, stimulation of Pyk2 by intracellular Ca2+ initiates SFK activation, establishing a positive loop that reinforces the Pyk2/SFK axis and allows the subsequent massive tyrosine phosphorylation of multiple substrates required for platelet aggregation.
The essential role of blood platelets in haemostasis and thrombosis relies on their ability to efficiently respond to extracellular signals upon vascular injury and to organize a complex network of signal transduction pathways leading to secretion, aggregation and thrombus formation . Early in the nineties, a number of observations documented the occurrence of a rapid and robust phosphorylation of several proteins on tyrosine residues upon stimulation of platelets and megakaryocytes [2–4]. Accordingly, not only megakaryocytes, but also the circulating anucleated platelets have been found to express a high number of different receptor and cytosolic tyrosine kinases, including Syk, Btk, FAK, Pyk2 and at least six members of the Src family kinases (SFKs) Src, Fyn, Lyn, Frg, Lck and Yes [5–7]. Among these, Src, Lyn and Fyn have been recognized to play a crucial role in many signal transduction pathways associated with platelet activation . Fyn and Lyn bind to the cytosolic domain of GPVI and upon collagen stimulation are responsible for FcR γ-chain phosphorylation and initiation of a protein-tyrosine phosphorylation cascade leading to phospholipase Cγ2 (PLCγ2) and phosphatidylinositol 3-kinase β activation . Src, and also Fyn, are found to be physically associated with integrin β3 cytosolic tail and are involved in the initiation of the outside-in signalling during platelet adhesion and aggregation [10,11]. A role for Lyn in GPIb-IX-V signalling upon binding of von Willbrand factor has also been documented .
In addition to these well-characterized models, SFKs are also activated by stimulation of G protein-coupled receptors (GPCRs) and protein-tyrosine phosphorylation in platelets was actually first documented upon stimulation with thrombin [2,3]. Although an early report suggested that SFKs are not required for thrombin-induced platelet activation , subsequent studies have indicated that SFKs may differently modulate platelet responses, depending on the mechanism of activation . For instance, Lyn, activated downstream of Gq stimulation, supports platelet activation by thrombin [15–17], whereas Fyn, activated downstream of G12/13 exerts an inhibitory effect . SFKs are also activated downstream of Gi and participate in granule secretion and TxA2 production .
Although models to understand SFKs role in integrins or GPVI signalling have been proposed, the mechanism for their activation upon GPCRs stimulation is largely unclear. Gαi has been shown to directly interact with SFKs and such an association has also been documented in epinephrine-stimulated platelets, which however causes only a weak protein-tyrosine phosphorylation [19,20]. The robust SFKs activation and protein-tyrosine phosphorylation produced by thrombin is mediated by Gq stimulation, which activates PLCβ and leads to an increase in intracellular Ca2+ concentration and protein kinase C activation . Previous reports have demonstrated that SFK activity evaluated as phosphorylation of a critical tyrosine residue in the catalytic domain, is promoted by Ca2+ signals in thrombin-stimulated platelets [17,21]. Nevertheless, SFKs are not reported to sense and directly bind Ca2+ and thus the mechanism by which SFKs are regulated by Ca2+ is still unknown.
In addition to SFKs, platelets possess a number of different protein kinases, including the focal adhesion kinase Pyk2, whose expression is restricted to neurons, epithelial and haematopoietic cells [22,23]. Pyk2 was originally named as Ca2+-dependent tyrosine kinase, to highlight its unique property to be activated by intracellular Ca2+ . In platelets, Pyk2 is activated by thrombin and von Willebrand factor [24–26] and recent findings have demonstrated that Pyk2 plays a major role in integrin outside-in signalling, and is required for thrombin-induced aggregation and thrombus formation [27–30].
In the present study, we hypothesized that Pyk2 could represent a Ca2+ sensor in thrombin-stimulated platelets, capable of initiating the conversion of Ca2+ signals into a tyrosine kinase-based pathway. The results indicate that Pyk2 actually mediates Ca2+-promoted activation of SFKs to support protein-tyrosine phosphorylation in thrombin-stimulated platelets.
Thrombin, apyrase grade I and grade VII, prostaglandin E1, acetylsalicylic acid, BAPTA-AM were from Sigma. PP2 was from Alexis (Vinci Biochem). Anti-Src (L4A1) monoclonal antibodies were from Cell Signaling Technology (Euroclone). The rabbit polyclonal antibodies against Pyk2 (N-19), Lyn (44), Fyn (FYN3) were from Santa Cruz Biotechnology (Tebu-Bio). Appropriate peroxidase-conjugated anti-IgG antibodies were from Bio-Rad. Mouse trueBlot®, peroxidase conjugated anti-IgG antibodies were from Rockland (Tebu-Bio). The enhanced chemiluminescence (ECL) substrate, Fura-2-AM and anti-P-Src (Tyr416; clone 9A6) were from Millipore. AG17 was from Calbiochem (Merck Millipore). Pyk2 knockout (KO) mice generation was described elsewhere . All the procedures involving the use of C57BL/6 wild-type (WT) and Pyk2 KO mice were approved by the Ethics Committee of the University of Pavia. Fyn and Lyn KO mice were purchased by Jackson Laboratory and used as previously described .
Preparation of washed platelets
Human platelets were obtained from healthy volunteers essentially as previously described with minor modifications . Briefly, whole blood in citric acid/citrate/dextrose (ACD) (152 mM sodium citrate, 130 mM citric acid, 112 mM glucose) was centrifuged at 120 g for 10 min at room temperature. Apyrase grade I (0.2 units/ml), prostaglandin E1 (1 μM) and acetylsalicylic acid (1 mM) were then added to the platelet-rich plasma (PRP). Platelets were recovered by centrifugation at 720 g for 15 min, washed with 10 ml of PIPES buffer (20 mM PIPES, 136 mM NaCl, pH 6.5) and finally gently resuspended in HEPES buffer (10 mM HEPES, 137 mM NaCl, 2.9 mM KCl, 12 mM NaHCO3, pH 7.4).
For mouse platelet preparation, blood was withdrawn from the abdominal vena cava of anaesthetized animals in syringes containing ACD–3.8% sodium citrate (2:1) as anti-coagulant. Anti-coagulated blood was diluted with HEPES buffer up to 2 ml and centrifuged for 7 min at 180 g to obtain PRP. PRP was then transferred to new tubes and the remaining red blood cells were diluted with HEPES buffer to a final volume of 2 ml and centrifuged again at 180 g for 7 min. The upper phase was added to the previously collected PRP and 0.02 units/ml apyrase grade I and 1 μM PGE1 and acetylsalicylic acid (1 mM) were added before centrifugation at 550 g for 10 min. The supernatant platelet poor plasma was removed and the platelet pellet was resuspended in 1 ml of PIPES buffer, centrifuged at 550 g for 10 min and finally resuspended in 500 μl of HEPES buffer.
In the reported experiments, stimulation of aspirin-treated human washed platelets was performed with 0.05–0.1 units/ml thrombin or with 0.5 μM A23187 for increasing times in the presence of 0.1 units/ml apyrase grade VII and samples were mixed, but not stirred, to avoid aggregation and subsequent integrin αIIbβ3 outside-in signalling.
Samples of resting and stimulated platelets (0.3 ml at 8×108 platelets/ml) were lysed with an equal volume of ice-cold immunoprecipitation buffer 2× (20 mM Tris/HCl, pH 7.2, 300 mM NaCl, 2 mM EGTA, 20 μg/ml leupeptin, 20 μg/ml aprotinin, 2 mM PMSF, 4 mM Na3VO4, 2 mM NaF, 2% Triton X-100, 0.2% SDS, 2% sodium deoxycholate). Insoluble material was removed by centrifugation and lysates were pre-cleared for 1 h at 4°C with 70 μl of protein G-Sepharose (50 mg/ml stock solution). The cleared supernatants were incubated with 3 μg of anti-Fyn, anti-Src or anti-Lyn antibodies for 2 h at 4°C and the immunocomplexes were precipitated by the addition of 70 μl of protein G-sepharose for 45 min. After brief centrifugation, immunoprecipitates were washed three times with 1 ml of immunoprecipitation buffer 1×, resuspended in 25 μl of SDS-sample buffer and heated at 95°C for 5 min.
Platelet samples (0.1 ml, 5×108 platelets/ml) were incubated at 37°C and typically stimulated with 0.1 units/ml thrombin or treated with an equivalent volume of HEPES buffer for increasing times. Reactions were stopped by the addition of 0.05 ml of SDS-sample buffer 3× (37.5 mM Tris, 288 mM glycine, pH 8.3, 6% SDS, 1.5% DTT, 30% glycerol, 0.03% Bromophenol Blue). Samples were heated at 95°C for 5 min and proteins from aliquots of identical volume were separated by SDS/PAGE, transferred to PVDF membranes and probed by immunoblotting with the antibodies indicated in the figure legends at a 1:500 dilution, as previously described . Images of reactive bands were acquired using a Chemidoc XRS apparatus (Bio-Rad) and quantification of band intensity was performed using the QuantityOne software. All the experiments were repeated at least three times.
Measurement of cytosolic Ca2+ concentration
Intracellular Ca2+ concentration was measured in Fura-2-AM loaded platelets essentially as previously described . Briefly, whole blood in ACD was centrifuged at 120 g for 10 min at room temperature and platelets in PRP were incubated with 3 μM Fura-2-AM for 40 min at 37°C in the dark. Apyrase and PGE1 were then added to PRP. Platelets were recovered by centrifugation at 720 g for 15 min, washed with 10 ml of PIPES buffer and finally gently resuspended in HEPES buffer+0.5% BSA. For these experiments, the final platelet concentration was adjusted to 2×108 cells/ml, stimulation was performed at 37°C on 0.4 ml of samples under gentle stirring in a Perkin Elmer Life Sciences LS3 spectrofluorometer in the presence of 1 mM CaCl2. All determinations were repeated at least four times with platelets from different donors.
Washed platelets from WT and Pyk2 KO mice (0.25 ml, 3×108 platelets/ml) were stimulated under constant stirring with 0.05 units/ml thrombin in a Chronolog Aggregometer (Mascia Brunelli). Platelet aggregation was monitored continuously over 5 min.
Data are expressed throughout as mean ± S.E.M. The results were analysed by t test (for comparisons of two groups) or one-way ANOVA with Bonferroni post-test (for multiple comparisons).
Role of Ca2+ and SFKs in thrombin-induced protein-tyrosine phosphorylation
In order to analyse the mechanism underlying protein-tyrosine phosphorylation, aspirin-treated human washed platelets were stimulated with 0.1 units/ml thrombin for increasing times in the presence of apyrase and the overall pattern of protein-tyrosine phosphorylation in whole cell lysates was evaluated by immunoblotting with anti-phosphotyrosine antibody. Figure 1(A) shows that thrombin induced a time-dependent phosphorylation of several substrates, mostly migrating as strong reactive bands of ~120, 100, 75 and 60 kDa. Intracellular Ca2+ chelation by BAPTA-AM dramatically reduced thrombin-induced protein-tyrosine phosphorylation and only a residual phosphorylation of the bands of 120 and 75 kDa was detectable after 5 min of stimulation. Control experiments on Fura-2-loaded platelets confirmed that the treatment with BAPTA-AM was efficient in completely preventing the increase in intracellular Ca2+ stimulated by 0.1 units/ml thrombin (Supplementary Figure S1). Inhibition of SFKs by PP2 also caused a strong inhibition of thrombin-induced protein-tyrosine phosphorylation, which however was totally suppressed only upon concomitant incubation with BAPTA-AM and PP2. These results suggest a co-operation between intracellular Ca2+ and SFKs in promoting protein-tyrosine phosphorylation by thrombin.
Thrombin-induced protein-tyrosine phosphorylation is regulated by Ca2+ and SFKs
A similar contribution of intracellular Ca2+ and SFKs was also documented in murine platelets. Figure 1(B) shows that the pattern of tyrosine phosphorylated proteins in thrombin-stimulated murine platelets was less pronounced than in human platelets, revealing a possible different sensitivity to the low dose of agonist used. Nevertheless, as for human, also in murine platelets protein tyrosine phosphorylation was similarly prevented by incubation with either BAPTA-AM or PP2.
Intracellular Ca2+ regulates SFKs
We next investigated whether SFKs and intracellular Ca2+ represent distinct mechanisms supporting protein-tyrosine phosphorylation or elements of the same pathway. We first analysed the role of SFKs in thrombin-induced intracellular Ca2+ increase in FURA-2-loaded platelets. Figure 2(A) shows that pre-incubation of platelets with PP2 did not significantly alter either the kinetics or the peak of intracellular Ca2+ concentration in thrombin-stimulated platelets, indicating that SFKs activity is not required for Ca2+ release.
Regulation of SFKs activation by intracellular Ca2+
SFKs activation was then evaluated by measuring the extent of autophosphorylation on a conserved tyrosine residue in the catalytic domain (Tyr416 in Src). The phospho-specific antibody used was generated against the phosphorylated Src (P-SrcTyr416), but it cross-reacts with the phosphorylated forms of all the SFK members expressed in platelets. Figure 2(B) shows that activation and autophosphorylation of SFKs was completely prevented by PP2, as expected, but was also inhibited by approximately 50% in the presence of BAPTA-AM, indicating that intracellular Ca2+ contributes to SFKs activity. According to this observation, we found that direct elevation of intracellular Ca2+ concentration by the ionophore A23187 was sufficient to promote SFKs autophosphorylation (result not shown; Figure 3D).
Pyk2 mediates Ca2±-dependent activation of SFKs
The focal adhesion kinase Pyk2 links intracellular Ca2+ to SFKs activation
Our results indicate that intracellular Ca2+ is a crucial regulator of SFKs activation and protein-tyrosine phosphorylation in thrombin-stimulated platelets. We hypothesized that the focal adhesion kinase Pyk2, which is known to be regulated by both Ca2+ and SFKs , could mediate Ca2+-induced activation of tyrosine kinases in stimulated platelets. Activation of Pyk2 was evaluated by measuring the autophosphorylation of Tyr402 with a specific anti-phospho-Pyk2 antibody. Figure 3(A) shows that the ionophore A23187 caused the time-dependent phosphorylation of Pyk2 in platelets, indicating that Pyk2 activation is actually regulated by intracellular Ca2+. Moreover, thrombin-induced Pyk2 activation was strongly reduced, albeit not abolished, upon intracellular Ca2+ chelation with BAPTA-AM (Figure 3B).
We next analysed the contribution of Pyk2 in Ca2+-induced protein-tyrosine phosphorylation using platelets from Pyk2 KO mice. Figure 3(Ci) shows that A23187 induced the phosphorylation of multiple substrates in murine platelets and that this process was dramatically impaired in platelets lacking Pyk2, indicating that Pyk2 is a major mediator of Ca2+-induced protein tyrosine phosphorylation. Similarly, also thrombin induced protein-tyrosine phosphorylation was reduced in the absence of Pyk2 (Figure 3Ci). To verify whether the role of Pyk2 in regulating Ca2+-promoted protein-tyrosine phosphorylation could be extended also to human platelets, we performed experiments with the Pyk2 inhibitor AG17. Figure 3(Cii) shows that in human platelets Pyk2 inhibition by AG17 reduced both A23187 and thrombin-stimulated tyrosine phosphorylation of some substrates. To investigate whether Pyk2 directly links intracellular Ca2+ elevation to SFKs activity, A23187-induced SFKs phosphorylation was compared in control and Pyk2-deficient platelets. Figure 3(D) shows that elevation of intracellular Ca2+ almost completely failed to cause SFKs activation in the absence of Pyk2. Equal loading control was performed with anti-p60Src and confirmed that comparable amounts of this member of the SFKs is expressed in WT and Pyk2-deficient platelets. Similarly, activation of murine platelets with thrombin induced a time-dependent phosphorylation of SFKs that was almost completely suppressed in the absence of Pyk2 (Figure 3E). These results indicate that Ca2+-mediated activation of SFKs requires the focal adhesion kinase Pyk2. To further strengthen our observations, we analysed SFKs activation and protein-tyrosine phosphorylation in GPVI-stimulated platelets, when SFKs activation is expected to represent an early event that occurs independently and upstream of Ca2+ mobilization . According to this, we found that the phosphorylation of SFKs, detected with the anti-pSrc(Y416) antibody, as well as the overall pattern of protein-tyrosine phosphorylation induced by 100 ng/ml convulxin was comparable in platelets from WT and Pyk2 KO mice (Supplementary Figure S2). These results indicate that the ability of Pyk2 to link Ca2+ signals to SFKs may be restricted to the stimulation of GPCRs.
Pyk2 regulates Fyn and Lyn, but not Src, in thrombin-stimulated platelets
We next tried to identify the specific Src kinase that is regulated by Pyk2 in thrombin-stimulated platelets focusing on those members of the family that are known to be highly expressed in platelets: Src, Fyn and Lyn. Since the pY416 Src antibody does not discriminate among these SFK members, immunoblotting was performed upon immunoprecipitation of the single proteins with specific antibodies. In order to limit the possible interference in the immunoblotting analysis caused by the immunoglobulin heavy chain, TrueBlot® was exploited for these experiments, instead of the routinely used secondary antibodies. Figure 4(A) shows that thrombin-induced phosphorylation of Src was comparable in both WT and Pyk2-deficient platelets. This observation is in line with results reported in Figure 3(D) showing that the SFK member whose phosphorylation is triggered by Ca2+ in WT but not in Pyk2-deficient platelets migrated with a molecular mass slightly different than that of p60Src. By contrast, the level of phosphorylation of Fyn (Figure 4B) and Lyn (Figure 4C) was strongly reduced in platelets lacking Pyk2. We conclude that Fyn and Lyn, but not Src are regulated by Pyk2 in thrombin-stimulated platelets.
Fyn and Lyn, but not Src are regulated by Pyk2
Regulation of Pyk2 by Fyn
Figure 3 documents the Ca2+-dependent activation of Pyk2 in thrombin-stimulated platelets, but also shows that a residual phosphorylation of Pyk2 persists even upon intracellular Ca2+ chelation by BAPTA-AM. Since, in T-cells, Pyk2 may also be regulated by SFK-mediated phosphorylation [23,35], we investigated the role of SFKs in thrombin-induced Pyk2 activation. Figure 5(A) shows that, like BAPTA-AM, inhibition of SFKs by PP2 strongly reduced Pyk2 phosphorylation, indicating a role for SFKs in Pyk2 activation. Interestingly, simultaneous inhibition of SFKs and intracellular Ca2+ chelation completely abolished Pyk2 phosphorylation (Figure 5A). To shed light on the identity of the SFKs responsible for Pyk2 activation, we compared thrombin-induced Pyk2 phosphorylation in control platelets and platelets from Fyn and Lyn KO mice. Figures 5(B) shows that phosphorylation of Pyk2 on Tyr402 was strongly reduced in the absence of Fyn, but occurred normally in Lyn-deficient platelets (Figure 5C), indicating that Fyn, but not Lyn, is involved in Pyk2 phosphorylation.
Fyn regulates Pyk2 activation
Pyk2 and SFKs are both required for thrombin-induced aggregation
We have previously shown the reduced aggregation of Pyk2-deficient platelets in response to a low dose of thrombin . Since we have now demonstrated the existence of a bi-directional link between Pyk2 and SFKs, we analysed the effect of PP2 on platelet aggregation. Figure 6 confirms that platelet aggregation in response to thrombin was strongly inhibited in the absence of Pyk2. Inhibition of SFKs by PP2 reduced the aggregation of control murine platelets to a level comparable to that observed in platelets from Pyk2 KO mice. Importantly, PP2 had no further inhibitory effect on aggregation of Pyk2-deficient platelets, indicating that the Pyk2-SFKs pathway is essential to elicit full platelet response.
Analysis of thrombin-induced platelet aggregation
In the present study, we have documented that the focal adhesion kinase Pyk2 integrates Ca2+ and SFK-mediated signals to promote protein-tyrosine phosphorylation in thrombin-stimulated platelets. These results shed light into the mechanism regulating SFKs activation downstream of GPCRs in platelets.
Previous studies have reported that thrombin-induced SFKs activation requires the stimulation of Gq and is mediated by the subsequent intracellular Ca2+ increase [17,21]. In our study, we have expanded these notions, by analysing the impact of intracellular Ca2+ and SFKs activity on the overall process of protein tyrosine phosphorylation. We found that, in human as well as in murine platelets, both Ca2+ and SFKs play a role in this process. In fact, simultaneous inhibition of SFKs and intracellular Ca2+ chelation completely abolished protein-tyrosine phosphorylation induced by low doses of thrombin. Interestingly, selected inhibition of Ca2+ signalling or SFKs causes very similar, albeit not identical, alterations of the protein-tyrosine phosphorylation pattern in thrombin-stimulated platelets, indicating that Ca2+ and SFKs organize a common signalling pathway, although they also operate through distinct and not redundant mechanisms. In the attempt to hierarchically organize these regulators of protein-tyrosine phosphorylation, we used a SFKs inhibitor and Ca2+ ionophore to demonstrate that, as suggested by previous reports [17,21], intracellular Ca2+ is upstream of SFKs activation, but that SFKs activity is not required for intracellular Ca2+ increase. We have not been able to detect any significant effect of SFKs inhibition of either early or delayed phases of thrombin-induced intracellular Ca2+ elevation, as suggested in previous studies . Although differences in the dose of the agonist and in the SFKs inhibitor used may explain this discrepancy, our results are in line with previous observation that PP2 does not alter the intracellular Ca2+ increase triggered by selective stimulation of either one of the thrombin receptors PAR1 and PAR4 . Nevertheless, it is remarkable that a residual protein-tyrosine phosphorylation persists upon either intracellular Ca2+ chelation or SFKs inhibition. This indicates that, in addition to participating in a common pathway, Ca2+ can trigger some phosphorylation events without the need for SFKs and vice versa. In addition, there might be some basal activation of SFKs that is not inhibited upon Ca2+ chelation.
In this work, we have addressed the currently unclear mechanism responsible for SFKs activation by intracellular Ca2+ and we have documented the unique and peculiar role of the focal adhesion kinase Pyk2 in this process. A key feature of Pyk2 that distinguishes this member of the focal adhesion kinase family from any other kinases expressed in platelets, including the highly related FAK, is its ability to be directly activated by intracellular Ca2+ [23,37]. Pyk2 is highly expressed in platelets and recent studies on KO mice have identified its important role in integrin outside-in signalling [27,28]. Pyk2 is also activated in thrombin-stimulated platelets [24,26] and Pyk2-deficient platelets show an impaired response to thrombin . In particular, we have previously shown that the absence of Pyk2 hampers platelet activation in response to low, but not high doses of thrombin . Since the present study was aimed to further characterize the molecular mechanisms for Pyk2 regulation of platelet response to thrombin, all the experiments were performed with a low dose of the agonist. In the present study, using a low dose of thrombin, we have recognized a novel role for Pyk2 as essential mediator of Ca2+-induced SFKs activation and protein-tyrosine phosphorylation stimulated by thrombin. We have documented that Pyk2 phosphorylation in platelets is directly triggered by Ca2+ ionophore and that thrombin-induced Pyk2 phosphorylation is prevented by intracellular Ca2+ chelation. In addition, we have shown that the activation of SFKs, which is promoted by the Ca2+ ionophore and occurs in a Ca2+-dependent mechanism in thrombin-stimulated platelets, is almost completely abolished in platelets from Pyk2 KO mice. To our knowledge this is the first direct evidence that Pyk2 can regulate SFKs activity. Moreover, we have found that Ca2+-induced protein tyrosine phosphorylation is almost completely suppressed in murine platelets lacking Pyk2. This observation supports the relevance of SFKs activation by Pyk2 for the Ca2+-driven protein-tyrosine phosphorylation in platelets. A reduction in thrombin-induced protein-tyrosine phosphorylation was also observed in human platelets treated with the Pyk2 inhibitor AG17, indicating that our model can be confidently translated from mice to humans. However, the effect of the Pyk2 inhibitor was clearly less pronounced than that observed upon genetic ablation of Pyk2. This difference may be related to the major responsiveness to thrombin of human platelets, also documented by Figure 1(A), but may also be indicative of a possible catalytically-independent contribution of Pyk2 to the regulation of protein tyrosine phosphorylation. The possible role of Pyk2 as a scaffold protein in platelets, however deserves further specific investigation. Although our results support the notion that intracellular Ca2+ is a major regulator of Pyk2 downstream of GPCRs, other mechanisms may be involved as well. We have recently demonstrated, for instance that Pyk2 phosphorylation by GPCRs requires G12/13 or, alternatively, integrin αIIbβ3 signalling . Therefore it is possible that, under physiological stimulation of membrane receptor, Ca2+ acts in combination with additional signalling pathways to stimulate Pyk2. This possibility certainly requires further investigation.
By means of immunoprecipitation experiments on Pyk2-deficient platelets, we have been able to identify Fyn and Lyn, but not Src, as the SFK members regulated by Pyk2. These results are consistent with the notion that Lyn plays a role in thrombin-induced platelet activation, whereas Src is mainly involved in integrin signalling [10,14–17]. In this context, it is interesting to note that a role for Lyn in TxA2 production by γ-thrombin has been documented and we have recently demonstrated that thrombin-induced TxA2 production is significantly inhibited in platelets from Pyk2 KO mice, thus causing a reduced aggregation . Accordingly, we have shown in the present study that inhibition of SFKs can reduce thrombin-induced aggregation of control murine platelets to a level comparable to that observed in platelets from Pyk2 KO mice, but has no further inhibitory effect in the absence of Pyk2. These results are consistent with our previous findings showing that the absence of Pyk2 does not affect thrombin-induced aggregation of aspirin-treated platelets . We concluded that the Ca2+-Pyk2-SFKs axis plays an important role in platelet response to low dose of thrombin.
SFKs are considered to be maintained in an inactive state by an intramolecular interaction between their SH2 domain and a phosphorylated tyrosine residue in the C-terminal tail and disruption of this interaction by competitive binding partners leads to kinase activation. The tyrosine residue 402 in the linker region of Pyk2, between the N-terminal FERM domain and the central catalytic domain, represents an autophosphorylation site, and, when phosphorylated, provides a binding site for the SH2 domain of SFKs [23,38]. According to a previously proposed model, Ca2+ and calmodulin can bind to the N-terminal FERM domain of Pyk2, causing Pyk2 dimerization and autophosphorylation on Tyr402 . Our results suggest that such a process in thrombin-stimulated platelets generates a binding site for the SH2 domains of SFKs Lyn and Fyn leading to their activation. Although a physical interaction between Pyk2 and members of the SFKs has been documented in nucleated cells, as well as in platelets [38,39], we have been unable, under our experimental conditions, to detect a specific co-immunoprecipitation between these kinases, indicating that the interaction may be transient or may not be preserved during sample handling.
Although we have, in the present study, shown that Pyk2 links Ca2+ signals to SFKs activation, previous reports have documented that SFKs can regulate Pyk2 activity in T-cells . We ourselves have recently reported that Pyk2 activation in integrin αIIβ3 outside-in signalling requires Src activity . In addition to tyrosine residues 578/579 in the catalytic domain, Tyr402 itself may also be a direct substrate for SFKs-directed phosphorylation . Therefore, a complex interplay among Ca2+, Pyk2 and SFKs exist in different cell types. In the present work, we a have also documented a novel positive feedback among Pyk2 and SFKs in platelets. In fact, we have shown that Ca2+ ionophore- as well as thrombin-induced Pyk2 phosphorylation on Tyr402 can be prevented by platelet treatment with the SFKs inhibitor PP2. Using platelets from KO mice, we clearly demonstrated that Fyn, but not Lyn, mediates Pyk2 phosphorylation upon stimulation of the thrombin receptor PAR4. These results are in agreement with previous findings that have identified Fyn as the SFKs responsible for Pyk2 activation in T-cells . The dual regulation of Pyk2 by either Ca2+ or SFKs has generally been found to operate in distinct cellular contexts. For instance, in TCR signalling Pyk2 activation is mainly mediated by SFKs activity , whereas stimulation of GPCRs, such as bradykinin receptor in PC12 cells, activates Pyk2 through increased Ca2+ concentration . Differently from these previous reports, our results demonstrated that in platelets both Ca2+ and Fyn concomitantly participate to achieve the full Pyk2 phosphorylation. The physiological relevance of Pyk2 feedback regulation by Fyn may be questioned, as no clear evidence for a major contribution of Fyn in platelet activation by thrombin has been reported. Notably, however, Quek et al.  documented a reduction in the late, but not early, phase of platelet aggregation induced by 0.1 unit/ml thrombin in Fyn KO mice, which suggests a role of Fyn in the reinforcement of platelet response consistent with our findings.
In conclusion, we propose a model (Figure 7) in which thrombin-induced intracellular Ca2+ increase initiates Pyk2 activation causing autophosphorylation on Tyr402. This generates a docking site for SFKs and leads to the activation of Lyn and Fyn. Activated Fyn further reinforces Pyk2 phosphorylation thus establishing a positive loop that leads to massive activation of Pyk2 and in turn, SFKs, an event that propagates tyrosine-phosphorylation of multiple substrates and participates in the development of platelet response to thrombin. These results shed new light on the mechanism regulating SFKs in platelets and identifies Pyk2 as a crucial regulatory element at the cross-roads between Ca2+- and tyrosine kinases-mediated signalling pathways.
A model for Ca2± induced SFKs activation
Ilaria Canobbio conceived and performed the experiments and analysed data. Lina Cipolla, Gianni Guidetti, Daria Manganaro and Caterina Visconte performed experiments and analysed data. Soochong Kim, Satya Kunapuli contributed vital reagents, performed experiments and edited the manuscript. Marco Falasca and Mitsuhiko Okigaki contributed vital new reagents and edited the manuscript. Mauro Torti designed research, analysed data and wrote the manuscript.
This work was supported by the Cariplo Foundation, Italy [grant number 2011-0436 (to M.T.)]; and the National Institutes of Health [grant numbers HL93231 and HL118593 (to S.P.K.)].