Mauro Torti, Department of Biochemistry, University of Pavia, via Bassi 21, 27100 Pavia, Italy. Tel.: +39 0382 987238; fax: +39 0382 987240. E-mail: firstname.lastname@example.org
Summary. Background: Platelet adhesion promoted by integrin α2β1 induces integrin αIIbβ3 activation through the phospholipase C (PLC)-dependent stimulation of the small GTPase Rap1b. Objective: To analyze the mechanism of PLC activation downstream of α2β1 that is required for regulation of Rap1b and αIIbβ3. Methods: Human and murine platelets were allowed to adhere to immobilized type I monomeric collagen through α2β1. Tyrosine phosphorylation of PLCγ2, PLC activation, accumulation of GTP-bound Rap1b and fibrinogen binding were measured and compared. Results: Integrin α2β1 recruitment induced an evident PLC activation that was concomitant with robust tyrosine phosphorylation of PLCγ2, and was suppressed in platelets from PLCγ2-knockout mice. Moreover, PLCγ2−/− platelets were unable to accumulate active Rap1b and to activate αIIbβ3 upon adhesion through α2β1. Inhibition of Src kinases completely prevented tyrosine phosphorylation of PLCγ2 in adherent platelets, but did not affect its activation, and both Rap1b and αIIbβ3 stimulation occurred normally. Importantly, αIIbβ3-induced phosphorylation and activation of PLCγ2, as well as accumulation of active Rap1b, were totally suppressed by Src inhibition. Integrin α2β1 recruitment triggered the Src kinase-independent activation of the small GTPase Rac1, and activation of Rac1 was not required for PLCγ2 phosphorylation. However, when phosphorylation of PLCγ2 was blocked by the Src kinase inhibitor PP2, prevention of Rac1 activation significantly reduced PLCγ2 activation, GTP-Rap1b accumulation, and αIIbβ3 stimulation. Conclusions: Src kinases and the Rac GTPases mediate independent pathways for PLCγ2 activation downstream of α2β1.
Integrins αIIbβ3 and α2β1 play a major role in platelet adhesion, activation, and aggregation . Inactive αIIbβ3 on resting platelets is conformationally converted into an active receptor able to bind soluble fibrinogen and to trigger platelet aggregation upon stimulation with many agonists. Inside-out activation of αIIbβ3 integrates many intracellular signals, and is critically regulated by the cytoskeletal protein talin and the small GTPase Rap1b [2,3]. A tyrosine kinase-based outside-in signaling pathway is subsequently initiated upon αIIbβ3 binding to fibrinogen, to consolidate platelet–platelet or platelet–matrix interaction [1,4].
Together with the glycoprotein (GP)VI–FcR γ-chain complex, α2β1 acts as a collagen receptor . Platelet adhesion through α2β1 initiates an outside-in signaling pathway characterized by the activation of tyrosine kinases and small GTPases, including Rap1b and Rac1 [6–9]. Agonist-induced inside-out activation of α2β1 has also been documented, but the mechanism regulating this process is still poorly understood [10,11].
There is evidence indicating complex and bidirectional cross-talk between αIIbβ3 and α2β1. For instance, agonist-induced α2β1 activation is regulated by αIIbβ3 outside-in signaling . Moreover, we have recently demonstrated that α2β1 promotes the activation of αIIbβ3 and induces fibrinogen binding to adherent platelets through a pathway that requires the active form of Rap1b generated downstream of phospholipase C (PLC) . Therefore, stimulation of PLC is certainly an early crucial event in the cross-talk between α2β1 and αIIbβ3.
Human platelets mainly express β2, β3 and γ2 isoforms of PLC . PLCβ2 and PLCβ3 are activated by Gq-coupled receptors for soluble agonists such as thrombin, ADP, and thromboxane A2 (TxA2) whereas PLCγ2 is typically activated downstream of the ITAM-bearing receptors GPVI–FcR γ-chain and FcγRIIA, and the GPIb–IX–V complex [14–16]. It is generally accepted that PLCγ2 activation occurs through tyrosine phosphorylation, and it has been found that members of the Src family of tyrosine kinases initiate this process downstream of ITAM-bearing receptors. PLCγ2 is also tyrosine phosphorylated in integrin outside-in signaling, and, for instance, it is essential for platelet spreading on collagen and fibrinogen [6–8,15,17].
In this study, we have investigated the role of tyrosine phosphorylation and activation of PLCγ2 in the signal transduction pathway for α2β1-dependent activation of αIIbβ3. We have found that PLCγ2, but not Src kinase activity, is essential for integrin cross-talk in collagen-adherent platelets, and we found that α2β1-promoted activation of PLCγ2 can occur independently of its tyrosine phosphorylation, through a novel pathway involving the small GTPase Rac1.
Materials and methods
Monomeric type I collagen was provided by M. E. Tira (University of Pavia, Italy). GFOGER peptide and collagen related peptide (CRP) were provided by R. Farndale (University of Cambridge, UK). PLCγ2-knockout mice were kindly provided by J. Ihle (St Jude Children’s Research Hospital, Memphis, TN, USA) through S. P. Watson (University of Birmingham, Birmingham, UK). NSC23766 was from Calbiochem (VWR International, Milan, Italy). The polyclonal antibodies against Rap1 and PLCγ2 were from Santa Cruz Biotechnology (Tebu-Bio, Magenta, Italy). The polyclonal antibody against phospho(Ser)-protein kinase C (PKC) substrates was from Cell Signaling Technology (Celbio, Pero, Italy). Anti-pleckstrin antibody was from Abcam (Cambridge, UK). Phosphotyrosine antibody 4G10, and anti-Rac1 antibody (23A8) were from Upstate Biotechnology (Lake Placid, NY, USA). All other reagents were obtained as previously indicated [6,8,18].
Preparation of human and murine platelets
Human and murine platelets were prepared as previously described [8,18], and finally resuspended in HEPES buffer (10 mm HEPES, 137 mm NaCl, 2.9 mm KCl, 12 mm NaHCO3, pH 7.4). Treatment of platelets with inhibitors was as follows: 1 mm aspirin or 30 μm BAPTA-AM for 30 min, 10 μm PP2 for 15 min, 5 μm Ro31-8220 or 100 μm NSC23766 for 5 min, and 2 U mL−1 apyrase for 1 min. A corresponding volume of vehicle (buffer or dimethylsulfoxide) was added to control samples. Stimulation of platelets in suspension was performed on 0.1-mL samples containing 3 × 108 cells mL−1 by addition of 50 ng mL−1 convulxin or by FcγRIIA clustering with 2 μg mL−1 monoclonal antibody IV.3 for 2 min followed by 30 μg mL−1 goat anti-mouse F(ab′)2. Stimulation was stopped after 1 min with 0.1 mL of ice-cold RIPA buffer (50 mm Tris–HCl, pH 7.4, 200 mm NaCl, 2.5 mm MgCl2, 1% Nonidet P-40, 10% glycerol, 1 mm phenylmethanesulfonyl fluoride, 1 μm leupeptin, 0.1 μm aprotinin, 0.1 mm Na3VO4).
Polystyrene dishes (60 mm) were coated overnight at room temperature with 50 μg mL−1 type I collagen diluted in 0.1 m acetic acid, 10 μg mL−1 GFOGER, or 100 μg mL−1 fibrinogen diluted in phosphate-buffered saline (PBS). Dishes were washed three times with 5 mL of PBS, blocked with 2 mL of 5% bovine serum albumin in PBS for 2 h at room temperature, and then washed three times with PBS. Human or murine platelets (0.5 mL, 2 × 108 platelets mL−1) were added to collagen-coated dishes in the presence of 2 mm MgCl2 and 1 mm RGDS or to fibrinogen-coated dishes in the presence of 2 mm CaCl2 for 30 min at room temperature. Non-adherent cells were removed, and dishes were washed three times with 5 mL of PBS. Adherent platelets were recovered by lysis with 1 mL of ice-cold RIPA buffer, and lysates were centrifuged at 18 000 × g for 10 min. Aliquots of the cleared supernatants containing the same amount of proteins were used for Rap1b or Rac activation assays, immunoprecipitation, or immunoblotting analysis. Quantification of murine platelet adhesion and spreading was performed upon staining with tetramethyl rhodamine isothiocyanate-conjugated phalloidin using imageJ software, as previously described .
Rap1b and Rac1 activation assay
Analysis of Rap1b and Rac activation in adherent cells was performed with the glutathione S-transferase (GST)-tagged Rap-binding domain of RalGDS or GST-tagged PAK to selectively precipitate active Rap1b or Rac1, respectively, from platelet lysates, according to a previously described procedure . Immunoblotting with anti-Rap1 or anti-Rac1 antibodies subsequently allowed the identification of the active GTPases in the precipitates, as well as the evaluation of the total amount of the GTPases in cell lysates.
Measurement of fibrinogen binding to αIIbβ3 in platelets adherent through α2β1 was performed using biotin-labeled fibrinogen as previously described .
Analysis of pleckstrin phosphorylation, immunoprecipitation, and immunoblotting
Aliquots of platelet lysates containing the same amount of proteins (typically 20 μg) were separated by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 5–15% polyacrylamide gel, and proteins were transferred to nitrocellulose. The extent of pleckstrin phosphorylation was analyzed by immunoblotting with anti-phospho(Ser)-PKC substrate and anti-pleckstrin antibodies, as previously described . PLCγ2 was immunoprecipitated from platelet lysates using 2 μg of anti-PLCγ2 antibody, as previously described , separated by SDS-PAGE on 7.5% acrylamide gels, transferred to nitrocellulose, and probed with anti-phosphotyrosine antibody. Upon stripping, the same blots were reprobed with anti-PLCγ2 antibody. All of these experiments were repeated at least three times, and comparable results were obtained.
We have previously described a signaling pathway activated downstream of α2β1 and leading to αIIbβ3 activation through the PLC-dependent stimulation of the small GTPase Rap1b . We now hypothesize that Src-mediated phosphorylation and activation of PLCγ2 could represent an early step in this signaling pathway for integrin cross-talk. Therefore, we first analyzed the effect of the Src kinase inhibitor PP2 on α2β1-induced activation of Rap1b. Figure 1A shows that accumulation of GTP-bound Rap1b in platelets adherent to monomeric collagen through α2β1 occurred normally when Src kinases were inhibited. To exclude any compensative contribution of PLCβ isoforms possibly activated by released TxA2 and ADP, aspirin-treated platelets were plated on monomeric collagen in the presence of the ADP scavenger apyrase. Figure 1A shows that inhibition of TxA2 production and neutralization of secreted ADP did not alter Rap1b activation and did not unmask any effect of PP2.
These intracellular signaling events in platelets adherent to monomeric collagen were triggered by α2β1, and are unlikely to involve either GPVI or αIIbβ3. As previously reported , we confirmed that FcR γ-chain was not phosphorylated in our experimental context (data not shown). Moreover, we supported our observations by analyzing platelets adherent to GFOGER peptide, a very well-characterized specific ligand for α2β1 [7,19] We found that the anti-α2β1 antibody 6F1 blocked platelet adhesion to monomeric collagen and GFOFER to similar extents (75% ± 7% and 81% ± 20%, respectively), but did not significantly alter adhesion to the GPVI ligand CRP (7% ± 17%), indicating that, as for GFOGER, α2β1 is the sole receptor for monomeric collagen. Moreover, adhesion to GFOGER activated the same pathways elicited by monomeric collagen. In fact, Fig. 1 shows that, upon platelet adhesion to GFOGER, Rap1b activation occurred, and was not prevented by Src kinase inhibition. Finally, any contribution of αIIbβ3 to the measured signaling events was excluded by analyzing Rap1b activation upon platelet adhesion to collagen in the presence of the blocking RGDS peptide (Fig. 1A). Moreover, RGDS did not affect platelet spreading on monomeric collagen, which was quantified as 98% ± 13% vs. control (n = 3), by measuring the average area of spread cells.
Src kinase activation was also dispensable for α2β1-dependent activation of αIIbβ3. Indeed, PP2 did not affect fibrinogen binding to platelets treated with aspirin and apyrase that were adherent to either monomeric collagen or GFOGER (Fig. 1B). As expected , however, fibrinogen binding was dramatically reduced, in both cases, by the neutralization of PLC effects through the simultaneous inhibition of PKC by Ro31-8220 and chelation of intracellular Ca2+ by BAPTA-AM.
To verify the efficacy of PP2 in PLCγ2 tyrosine phosphorylation, PLCγ2 was immunoprecipitated from collagen-adherent or GFOGER-adherent platelets and analyzed by immunoblotting with anti-phosphotyrosine antibodies. Figure 1C shows that tyrosine phosphorylation of PLCγ2 was unaffected by aspirin and apyrase, or by blockade of αIIbβ3 with RGDS peptide, but was suppressed by PP2. As these results document normal Rap1b and αIIbβ3 stimulation in the absence of PLCγ2 phosphorylation, we directly measured the phosphorylation of pleckstrin, the main PKC substrate, as an index of PLC activation. Surprisingly, we found that upon platelet adhesion to collagen, both in the absence and presence of RGDS, and upon platelet adhesion to GFOGER, pleckstrin phosphorylation was unaffected by inhibition of Src kinases with PP2, independently of aspirin and apyrase (Fig. 1D). These findings indicate that a PLC isoform is actually stimulated downstream of α2β1 and supports Rap1b and αIIbβ3 activation even under conditions in which inhibition of Src kinases prevents tyrosine phosphorylation of PLCγ2.
These results, however, question the real involvement of PLCγ2 in the signaling pathway linking α2β1 to Rap1b and αIIbβ3. In order to definitively clarify this point, we analyzed platelets from PLCγ2-knockout mice. PLCγ2−/− platelets adhered normally to collagen-coated coverslips, but showed impaired spreading (Fig. 2A,B). We measured PLC activation in adherent platelets by evaluating the phosphorylation of pleckstrin. Whereas in wild-type platelets robust pleckstrin phosphorylation occurred, adhesion of PLCγ2−/− platelets to monomeric collagen caused only minimal phosphorylation of pleckstrin (Fig. 2C), indicating that PLCγ2 is certainly the main PLC isoform activated downstream of α2β1. In agreement with this conclusion, we found that the lack of PLCγ2 almost completely prevented adhesion-dependent activation of Rap1b (Fig. 2D), and reduced by about 80% fibrinogen binding to αIIbβ3 (Fig. 2E).
To verify whether these findings reflect a peculiarity of α2β1-mediated outside-in signaling, we analyzed PLCγ2 tyrosine phosphorylation and activation in platelets stimulated with the GPVI ligand convulxin or by clustering of FcγRIIA. Figure 3 shows that under both conditions, PP2 totally prevented phosphorylation of PLCγ2, and also pleckstrin phosphorylation, as expected. In parallel, activation of Rap1b was also totally inhibited. PLCγ2 has also been implicated in αIIbβ3 outside-in signaling [15,17]. We found that in platelets adherent to immobilized fibrinogen, PP2 caused the simultaneous inhibition of PLCγ2 tyrosine phosphorylation, PLC activation, and Rap1b stimulation (Fig. 4). Altogether, these results indicate that only upon platelet adhesion to α2β1, tyrosine phosphorylation is dispensable for activation of PLCγ2 and for eliciting downstream events.
A recent study showed that PLCγ2 can be activated, in vitro and in a cell-free system, by the GTP-bound forms of the small GTPases Rac1, Rac2 and, to a lesser extent, Rac3, and that this process occurs independently of PLCγ2 tyrosine phosphorylation . However, whether or not this pathway has any physiologic significance was not clarified. Therefore, we investigated the possible involvement of Rac GTPases in the α2β1-mediated activation of PLCγ2. Rac inhibition was accomplished by using a cell-permeable specific inhibitor, NSC23766, which blocks Rac-GEF (GDP/GTP exchange factor) activity and inhibits both Rac1 and Rac2, thus preventing several Rac-dependent effects . We used an antibody against Rac1, the predominant Rac isoform expressed in platelets , to verify that α2β1-mediated adhesion actually induced Rac activation, and that this process was efficiently suppressed by NSC23766 (Fig. 5A). Interestingly, we also found that Rac activation was not affected by PP2 (Fig. 5A), and that NSC23766 did not alter α2β1-promoted phosphorylation of PLCγ2 (Fig. 5B). We next investigated the effect of Rac1 inhibition on phosphorylation of pleckstrin, Rap1b activation, and fibrinogen binding to αIIbβ3. Figure 5C,D shows that NSC23766 alone was not able to block either pleckstrin phosphorylation or Rap1b stimulation. However, when Src kinases were inhibited by PP2, NSC23766 caused evident reductions in both pleckstrin phosphorylation and accumulation of Rap1b-GTP. The results were quantified, and the effect of NSC23766 in combination with PP2 was found to be statistically significant (Table 1). Similarly, the simultaneous inhibition of Src kinases and Rac GTPases caused a significant reduction in α2β1-induced binding of fibrinogen to αIIbβ3 (Table 1).
Table 1. Quantitative analysis of PP2 and NSC23766 effects on phospholipase C activation, GTP-Rap1b accumulation and fibrinogen binding in platelets adherent through α2β1
Quantification of pleckstrin phosphorylation and Rap1b activation was performed by densitometric scanning of immunoblots such as those reported in Fig. 5. Band intensities were measured and analyzed with imageJ and graphpad prism4 software. Results from five different experiments were normalized, taking the intensity of the bands in platelets stimulated in the absence of inhibitors as 100%, and data are reported as mean ± standard deviation (SD). Specific binding of biotinylated fibrinogen to adherent platelets was performed as described under Materials and methods. Results are reported as mean ± SD of three different experiments. Statistical analysis was performed by the anova test, and the results, when significant, are indicated in parentheses.
98.0 ± 18.8
86.4 ± 10.1
86.6 ± 15
85.4 ± 30.4
73.5 ± 21.7
88.6 ± 8.9
NSC23766 + PP2
41.5 ± 6.7 (P < 0.05)
37.0 ± 14.1 (P < 0.01)
62.4 ± 6.7 (P < 0.001)
We have previously shown that platelet adhesion to collagen through α2β1 activates the small GTPase Rap1b through a pathway that depends on PLC activity . Both PLC and Rap1b activation are then required for the subsequent activation of αIIbβ3 .
By analyzing platelets from knockout mice, we have demonstrated in this work that the signaling pathway for Rap1b and αIIbβ3 activation initiated by α2β1 strictly requires the expression and stimulation of PLCγ2. Interestingly, minimal residual PLC activity, detected as phosphorylation of the main PKC substrate pleckstrin, as well as residual Rap1b and αIIbβ3 activation were observed in PLCγ2−/− mice. As identical results were also obtained with aspirin-treated platelets in the presence of apyrase (data not shown), it is unlikely that this effect is due to activation of a PLCβ isoform by secreted TxA2 or ADP. Rather, it may reflect a minimal contribution of PLCγ1, which is expressed in platelets at much lower levels than PLCγ2, but has also been implicated in the residual response of PLCγ2−/− platelets to GPVI stimulation . PLCγ2 has been previously documented to be required for platelet spreading upon adhesion to α2β1 . We have also observed reduced spreading of PLCγ2−/− platelets on monomeric collagen. In addition, our results extend the importance of this PLC isoform in outside-in signaling through α2β1 regulating Rap1b activation and αIIbβ3 stimulation.
PLCγ2 is implicated in many signaling pathways for platelet activation, including not only α2β1, but also αIIbβ3, GPVI–FcR γ-chain, FcγRIIA, and GPIb–IX–V [6–8,15–17]. Under all these circumstances, PLCγ2 is considered to be activated as a consequence of phosphorylation of specific tyrosine residues. Although different kinases, including Syk and Btk, have been proposed to be responsible for direct PLCγ2 phosphorylation, this process is typically controlled by members of the Src family of kinases, which act at earlier steps in the signal transduction pathway leading to PLCγ2 activation. We made here the unexpected observation that although Src kinase inhibition caused the complete suppression of PLCγ2 tyrosine phosphorylation induced by platelet adhesion through α2β1, activation of PLCγ2 was not affected. Moreover, in agreement with the persisting PLC activity, signaling events downstream of PLC, such as stimulation of Rap1b and αIIbβ3, occurred normally upon inhibition of Src kinases. A previous study found that α2β1-mediated platelet spreading, which largely depends on PLCγ2 activity, was significantly affected by inhibition of Src kinases . However, PLCγ2 activity was not directly measured in this study, and thus it remains possible that the effect of the Src kinase inhibitor on platelet spreading may involve additional targets.
We also found that the phosphorylation-independent activation of PLCγ2 was a peculiar feature of α2β1 outside-in signaling. In fact, in agreement with a very well-documented notion, we found that ITAM-mediated phosphorylation of PLCγ2 was found to be essential for its activation. Moreover, even upon platelet adhesion to fibrinogen, which recruits αIIbβ3, inhibition of PLCγ2 tyrosine phosphorylation was paralleled by the suppression of PLC activity as well as by the inhibition of Rap1b activation. Although we cannot yet provide a mechanistic explanation for the observed differences between α2β1 and αIIbβ3, our results support the notion that different integrins can adopt either common or peculiar signaling strategies.
With regard to the peculiar mechanism for stimulation of non-phosphorylated PLCγ2 downstream of α2β1, we provide here evidence that the small GTPase Rac may play a crucial role. The ability of the active GTP-bound forms of Rac1 and Rac2 to activate PLCγ2 has been documented both in vitro and in a cell-free system . However, the specific cellular conditions in which this possible regulation may take place and have physiologic implications were not determined. Platelets express both Rac1 and Rac2 proteins, and we have confirmed that Rac proteins are activated upon adhesion through α2β1. However, in contrast to a previous study , we found here that α2β1-mediated activation of Rac, which was completely inhibited by blocking the Rac-GEFs with the NSC23766 compound, was not altered by inhibition of Src kinases with PP2. We do not have a clear explanation for this discrepancy, which, however, may be related to the different inhibitor and doses employed. Importantly, we document here that inhibition of Rac had no effects on PLCγ2 activation in adherent platelets, as long as PLCγ2 tyrosine phosphorylation occurred. However, under conditions of prevention of PLCγ2 phosphorylation, the antagonist of Rac caused a significant reduction in PLCγ2 activation, and impaired some PLC-driven responses, such as Rap1b and αIIbβ3 stimulation. These results thus identify a specific physiologic context in which the activation of PLCγ2 by active Rac may be of significance. It is interesting to note, however, that the contribution of Rac to PLCγ2 activation appears to be negligible as long as tyrosine phosphorylation of PLCγ2 takes place normally. Therefore, we conclude that Src kinases and Rac mediate alternative pathways for PLCγ2 activation in α2β1-mediated outside-in signaling. Our results clearly document residual PLC activation, Rap1 stimulation, and fibrinogen binding upon concomitant inhibition of Src kinases and Rac GTPases that is significantly greater than that observed in PLCγ2−/− platelets. This indicates the existence of additional mechanisms regulating PLCγ2 downstream of α2β1. We have actually obtained preliminary evidence that phosphoinositide 3-kinase may contribute to α2β1-promoted PLCγ2 activation (data not shown), but this aspect is currently under investigation.
While this article was being written, Pleines et al. reported impaired PLCγ2 activation in Rac1−/− platelets upon stimulation of GPVI and CLEC-2, despite normal PLCγ2 tyrosine phosphorylation . This study concluded that, in mice, ITAM-dependent stimulation of PLCγ2 required both tyrosine phosphorylation and Rac1, but recognized the importance of these results being confirmed in human platelets . In this context, our results document a possible significant difference in humans, by identifying a physiologic context in which the two events are not concomitantly required, but represent alternative pathways. It should be noted that, as far as platelet stimulation downstream of ITAM-bearing receptors is concerned, our results are in line with those previously reported by Pleines et al., in that we have found that upon stimulation with convulxin or by clustering of FcγRIIA, PP2 inhibited not only tyrosine phosphorylation of PLCγ2, but also Rac activation (data not shown). Therefore, our results are consistent with the model of both Src kinases and Rac GTPases being necessary for PLCγ2 activation downstream of GPVI or FcγRIIA, but show a novel aspect of the mechanism of outside-in signaling through α2β1.
We thank J. Ihle and S. P. Watson for PLCγ2-knockout mice. This work was supported by grants from the Ministero dell’Istruzione, Università e Ricerca Scientifica (MIUR, PRIN 2006), and from the Consorzio Interuniversitario Biotecnologie (CIB). G. F. Guidetti is the recipient of a fellowship from Regione Lombardia (Progetto Ingenio).
Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.