Summary. The dynamics of the actin cytoskeleton, largely controlled by the Rho family of small GTPases (Rho, Rac and Cdc42), is critical for the regulation of platelet responses such as shape change, adhesion, spreading and aggregation. Here, we investigated the role of adenosine diphosphate (ADP), a major co-activator of platelets, on the activation of Rac. ADP rapidly activated Rac in a dose-dependent manner and independently of GPIIb/IIIa and phosphoinositide 3-kinase. ADP alone, used as a primary agonist, activated Rac and its effector PAK via its P2Y1 receptor, through a Gq-dependent pathway and independently of P2Y12. The P2Y12 receptor appeared unable to activate the GTPase per se as also observed for the adenosine triphosphate receptor P2X1. Conversely, secreted ADP strongly potentiated Rac activation induced by FcγRIIa clustering or TRAP via its P2Y12 receptor, the target of antithrombotic thienopyridines. Stimulation of the α2A-adrenergic receptor/Gz pathway by epinephrine was able to replace the P2Y12/Gi-mediated pathway to amplify Rac activation by FcγRIIa or by the thrombin receptor PAR-1. This co-activation appeared necessary to reach a full stimulation of Rac as well as PAK activation and actin polymerization and was blocked by a G-protein βγ subunits scavenger peptide.
Although adenosine diphosphate (ADP) is considered as a weak aggregating agent per se, it plays a central role in hemostasis by acting as a key cofactor of platelet activation, in vitro and in vivo. ADP is highly concentrated in dense granules and is released upon platelet stimulation to reinforce the level of activation to reach complete responses of these cells [1,2]. ADP by itself can induce platelet activation through stimulation of two metabotropic P2 receptors, the Gq-coupled P2Y1 receptor and the Gi-coupled P2Y12 receptor [1–4]. The analysis of the clinical profiles of patients with impaired response to ADP, the use of selective P2Y1 or P2Y12 antagonists  and the development of P2Y1 and P2Y12 knockout mice [5–7] allow a better understanding of the respective roles of these purinergic receptors in platelet function. It is now well established that the P2Y1 receptor is involved in shape change and in the initiation of ADP-induced platelet aggregation through calcium mobilization [2,3,5,8,9]. The P2Y12 receptor is responsible for ADP-induced adenylyl cyclase inhibition and is required, in cooperation with the P2Y1 receptor, for a full platelet aggregation in response to ADP alone [1,2,10]. An important aspect in thrombosis and hemostasis is the critical role of ADP, through its P2Y12 receptor, as a co-factor of platelet activation by other physiologic agonists [1,7,10]. This receptor is the target of the antithrombotic thienopyridine drugs [6,8,11]. Several groups have highlighted the major role of secreted ADP as a co-activator of PAR-1 thrombin receptor activating peptide (TRAP) [12,13] or FcγRIIa-mediated platelet stimulation in vitro [14,15]. In the case of TRAP stimulation, ADP stabilizes platelet aggregates through a mechanism involving phosphoinositide 3-kinase (PI 3-kinase) and its product phosphatidylinositol 3,4-bisphosphate (PtdIns(3,4)P2) . The implication of the P2Y12 receptor is crucial in platelet activation by FcγRIIa cross-linking as aggregation is totally abolished when the P2Y12 receptor activation is inhibited . In this case, a concomitant signaling pathway from P2Y12/Gi and FcγRIIa/tyrosine kinases is required for PLCγ2 activation through a mechanism involving the synthesis of the second messenger phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P3) and lipid rafts [15–17].
In addition to the P2Y1 and P2Y12 receptors, platelets also express the ionotropic P2X1 receptor that appears now as a receptor for ATP rather than ADP . Upon stimulation, P2X1 mediates a rapid calcium influx and leads to platelet shape change. Recent works suggest that the P2X1 receptor contributes to collagen-induced platelet activation via a PKC/Erk2 pathway  and that it contributes to the thrombosis of small arteries . Although ADP is known to mediate morphological changes in platelets through actin cytoskeleton reorganization [21,22], little is known about the molecular mechanisms initiated by this agonist to regulate the actin cytoskeleton dynamics. Members of the Rho family GTPase, including Rho, Rac and Cdc42, are highly expressed in platelets and are thought to play a major role in the control of this process [23–25]. Rac is rapidly activated after stimulation of the thromboxane A2 receptor  or the thrombin receptor PAR-1  and has been suggested to play a central role in thrombin receptor-induced actin assembly . The phosphatidylinositol 4-phosphate 5-kinase Iα (PI(4)P 5-kinase Iα) and the p21-activated kinase (PAK) are two effectors of Rac identified in platelets [27–29]. The product of PI(4)P 5-kinase Iα, phosphatidylinositol 4,5-bisphosphate PI(4,5)P2, is known to play an important role in the control of actin cytoskeleton dynamics through interactions with several actin-regulatory proteins . Hartwig et al.  previously demonstrated the role of this phosphoinositide in the regulation of actin assembly downstream of Rac in human platelets. The various isoforms of the serine/threonine kinase PAK have an N-terminal auto-inhibitory domain interacting with the C-terminal catalytic domain of the protein. The binding of Rac-GTP or Cdc42-GTP to PAK disrupts this interaction resulting in transphosphorylation and stimulation of PAK kinase activity. Activated PAK is able to positively regulate the serine/threonine-kinase LIM which in turn phosphorylates cofilin and inhibits its ability to depolymerize actin [31,32]. PAK can also participate in the downregulation of the myosin-light-chain kinase thus contributing to a decrease in actin stress fibers . The role of PAK-kinases is still poorly characterized in platelets but it has been shown that this kinase is involved in lamellae spreading during platelet adhesion [27,28].
The aim of the present work was to investigate Rac activation downstream of the various ADP receptors either in conditions where ADP is used alone, as a primary agonist, or when ADP is secreted and acts as a co-activator of other agonists. We show that when ADP acts as a primary agonist, the rapid activation of Rac in human and mouse platelets is GPIIb/IIIa and PI 3-kinase-independent, occurs through the P2Y1/Gq pathway and does not require the P2Y12 receptor. Conversely, when ADP is secreted and acts as a co-activator, the P2Y12/Gi pathway plays a central role in the potentiation of Rac activation in human platelets stimulated by FcγRIIa clustering or by TRAP. This P2Y12/Gi-dependent pathway is involved in PAK activation and actin polymerization and can be replaced by the α2A-adrenergic receptor/Gz pathway. In response to FcγRIIa clustering, this pathway is inhibited by heterotrimeric G-protein βγ subunits scavenger.
Materials and methods
The ATP analog N6-(2methyl-thioethyl)-2-(3,3,3-trifluoropropylthio)-β,γ-dichloromethylene ATP (AR-C69931MX) was a generous gift from Astra (Charnwood, UK). The SR 121566 compound was obtained from Dr P. Savi (Sanofi-Aventis, Toulouse, France), the GPIIb/IIIa RAM2 blocking antibody was a generous gift from Dr F. Lanza and Y27632 was from Calbiochem (VWR International, Pessac, France). The anti-Rac monoclonal antibody was purchased from Upstate Biotechnology (Lake Placid, NY, USA), the anti-phospho-PAK polyclonal antibody was from Cell Signaling-Ozyme (Saint Quentin Yvelines, France), the anti-Gβ monoclonal antibody was from BD Biosciences Pharmingen (San Diego, CA, USA). TRAP (SFLLRNP) was from BACHEM (Budendorf, Switzerland) and serotonin from Alexis Biochemicals (Coger, Paris, France). The monoclonal IV.3 antibody to FcγRIIa was from Medarex (Annandale, NJ, USA) and the specific F(ab′)2 fragments were from Jackson Immunoresearch Laboratories (West Grove, PA, USA). Human fibrinogen was from Kabi (Stockholm, Sweden). The N20K peptide (NRSYVISSFTELKAYDLLSK)  either biotinylated (Biot-N20K) or not was synthesized by NeoSystem (Strasbourg, France) and the control peptide (CNAGSVEQTPKKPGLRRRQT) was obtained from Santa Cruz (Santa Cruz, CA, USA). The other compounds, including A2P5P and MRS-2179, were from Sigma-Aldrich (Saint Quentin-Fallavier, France) unless specified.
Mutant mice deficient in the P2Y1 receptor were produced as described by Leon et al.  and both wild type and knockout mice were of pure C57BL/6 genetic background (9 back-cross). Gq-deficient mice were produced as described  and normal and Gαq knockout mice were of 129/Sv×C57BL/6 genetic background. Adult wild type or P2X1 receptor-deficient mice were produced as described by Mulryan et al. . These mice were of pure C57BL/6 genetic background.
Platelet preparation and stimulation
Washed mouse platelets were prepared from blood (6 volumes) drawn from the abdominal aorta of anesthetized mice into a plastic syringe containing acid citrate dextrose anticoagulant (1 volume) as previously described . Pooled blood was centrifuged at 175 g at 37 °C to obtain platelet-rich plasma which was then centrifuged at 1570 g at 37 °C. The platelet pellet was washed twice in a buffer (137 mm NaCl, 2 mm KCl, 12 mm NaHCO3, 0.3 mm NaH2PO4, 1 mm MgCl2, 2 mm CaCl2, 5.5 mm glucose, 5 mm HEPES, pH 7.3) containing 2 mm CaCl2 and 0.35% bovine serum albumin (BSA) and finally resuspended at a density of 5 × 108 platelets mL−1 in the same buffer in the presence of 0.02 U mL−1 of the ADP scavenger apyrase (adenosine 5′-triphosphate diphosphohydrolase, EC 220.127.116.11), a concentration sufficient to prevent desensitization of platelet ADP receptors during storage. Platelets were kept at 37 °C throughout all experiments. Concerning platelets from P2X1-deficient and WT control mice, a high concentration of apyrase (0.9 U mL−1) was used at all steps of platelet handling to avoid P2X1 receptor desensitization as previously described .
Human blood was collected from a forearm vein (6 blood volumes × 1 volume of acid/citrate/dextrose anticoagulant), and twice-washed platelet suspensions were prepared as described previously . Briefly, platelets were washed in a buffer (pH 6.5) containing 140 mm NaCl, 5 mm KCl, 5 mm KH2PO4, 1 mm MgSO4, 10 mm HEPES, 5 mm glucose, and 0.35% BSA (w/v). The same buffer containing 1 mm CaCl2 was added to the final suspension (5 × 108 platelets mL−1), and pH was adjusted to 7.4.
Aggregation was measured at 37 °C by a turbidimetric method in a dual-channel Payton aggregometer (Payton Associates, Scarborough, ON, Canada). A 450-μL aliquot of platelet suspension was stirred at 1100 rev./min and activated by addition of different agonists, with or without antagonists, in the presence of human fibrinogen (0.25 mg mL−1) when ADP was used as a primary agonist, in a final volume of 500 μL. For stimulation via FcγRIIa, the monoclonal antibody IV.3 was first added followed 1 min later by the F(ab′)2 fragments as described [15,16].
Platelet-loading procedure with fura-2/AM and intracellular calcium measurements were performed as previously described .
Determination of activated cellular Rac
Platelet suspension (500 μL at 5 × 108 platelets mL−1) was stimulated as indicated and the amount of activated cellular Rac was determined by precipitation with a fusion protein consisting of GST and the Rac-binding domain from human PAK1 (amino acids 67–150) as described . Clarified cell lysates (40 μL) were used to control total Rac. The beads were washed as described . Rac-GTP and total Rac were analyzed on a 12% SDS-PAGE, and blotted with the relevant antibody, peroxidase-conjugated secondary antibody, and enhanced chemiluminescence system.
Platelets were prepared without albumin to prevent immunodetection difficulties as albumin and PAK have about the same molecular weight. Resting and activated platelets (500 μL at 5 × 108 platelets mL−1) were immediately lyzed with Laemmli sample buffer. A fraction of the platelet lysate (equivalent of 1.8 × 107 platelets) was loaded on a 10% SDS-PAGE and blotted with a specific antibody against phospho-PAK, reflecting the activation of this kinase .
Cytoskeleton isolation and quantification of F-actin content
Platelet suspension (500 μL at 2 × 109 platelets mL−1) was stimulated as indicated. Cytoskeleton was extracted from activated or resting platelets by adding 1 volume of ice-cold twice-concentrated cytoskeleton buffer as previously described . Cytoskeletal proteins (corresponding to 2 × 108 platelets) were separated by SDS-PAGE (7.5%) and F-actin content was estimated by Coomassie blue staining and densitometric analysis.
Studies with permeabilized platelets
Human platelets (5 × 109 cells mL−1) were resuspended in a buffer containing 132 mm NaCl, 2.8 mm KCl, 0.86 mm MgCl2, 8.9 mm NaHCO3, 2 mm HEPES, 5.6 mm glucose, 12.2 mm Na3 citrate, and 10 mm Tris, pH 7.1, as described . One hundred microliters of this platelet suspension was mixed with 400 μL of buffer containing 120 mm KCl, 4 mm MgCl2, 25 mm NaCl, 1 mm NaH2PO4, 1 mm EGTA, 0.269 mm CaCl2 and 15 mm HEPES, pH 7.1. After 30 min at 37 °C, the N20K or the control peptides were added at the indicated concentration and 10 μg mL−1 saponin was immediately added to permeabilize platelets. After 2 min, FcγRIIa was cross-linked and Rac activation was determined 2 min later as described above.
To assess the binding of N20K peptide to the βγ subunits, platelets were lysed in an ice-cold buffer containing 20 mm Tris, pH 7.4, 150 mm NaCl, 1 mm EDTA, 1% NP40, 10 μg mL−1 each of leupeptin and aprotinin, 1 mm orthovanadate and 1 mm PMSF and biotinylated-N20K peptide (50 or 75 μm, as indicated) was added. Ten minutes later, streptavidin–agarose beads were added and after 45 min of incubation at 4 °C, the samples were centrifuged and the beads were washed three times with a buffer containing 20 mm Tris, pH 7.4, 150 mm NaCl, 1 mm EDTA, 0.1% NP40, 1 μg mL−1 each of leupeptin and aprotinin, 0.1 mm orthovanadate and 0.1 mm PMSF. The presence of the G-protein β subunit bound to the streptavidin–agarose beads was analyzed by Western blotting with a specific antibody.
ADP-induced Rac activation in human and mouse platelets
ADP-induced Rac activation was investigated using the GST-PAK1 binding assay to selectively precipitate the active, GTP-bound form, of this GTPase, as previously described [26,38]. Figure 1A shows a dose-dependent activation of Rac by ADP both in human and mouse platelets. As indicated by a time course experiment (Fig. 1B), Rac was very rapidly activated in response to ADP. Densitometric analysis of immunoblots indicated that, under our experimental conditions, the amount of Rac charged in human platelets stimulated with ADP (10 μm, 30 s) was 17% ± 5% relative to the total charging obtained with GTPγS (100 μm).
The relative contribution of the different platelet purinergic ADP receptors to Rac activation was first investigated using selective pharmacologic antagonists. MRS-2179 (Fig. 1B) and A2P5P (data not shown), two P2Y1-selective antagonists, were used to block ADP signaling through its Gq-coupled receptor. These compounds caused a dramatic decrease in Rac activation in human and mouse platelets. Conversely, the selective P2Y12 antagonist, AR-C66931MX, had no significant effect on ADP-induced Rac activation (Fig. 1B), although it nearly abolished platelet aggregation (data not shown). Accordingly, the kinase PAK, a downstream effector of Rac proposed to play an important role in actin cytoskeleton reorganization during platelet stimulation [27,28], was activated by ADP via the P2Y1 receptor (Fig. 1C). The activation of PAK was assessed using an anti-phospho-PAK antibody recognizing the autophosphorylation sites of PAK which are thought to maintain the kinase under an active form . These results strongly suggest that the P2Y1 receptor is essential for ADP-induced Rac and PAK activation, whereas the P2Y12 receptor is not required.
Figure 2A shows that the GPIIb/IIIa integrin was not involved in ADP-induced Rac activation, as the SR121566 (3.5 μm) compound, known to antagonize the binding of fibrinogen to GPIIb/IIIa , abolished ADP (10 μm)-induced human platelet aggregation (data not shown) without affecting Rac activation. Furthermore, ADP was still able to stimulate Rac in platelets from patients with type I Glanzmann's thrombasthenia (devoid of GPIIb/IIIa) which are unable to aggregate (data not shown). In agreement, GPIIb/IIIa blocking antibody RAM2 completely inhibited ADP (10 μm)-induced mouse platelet aggregation (data not shown) but did not affect Rac activation (Fig. 2A, right panel). As PI 3-kinase can function upstream of Rac in several models, we checked the effect of its inhibition by wortmannin (50 nM) (Fig. 2B) or LY294002 (25 μm) (data not shown) under conditions where the synthesis of its lipid products was blocked by 95%. PI 3-kinase inhibition did not affect the activation of Rac by ADP (10 μm). Overall these results indicate that Rac activation occurs as an early response upon ADP stimulation, independently of GPIIb/IIIa and PI 3-kinase.
ADP-induced Rac activation requires the P2Y1/Gαq pathway
To further evaluate the specific role of the P2Y1 receptor in Rac activation, we used P2Y1-deficient transgenic mouse platelets. As previously reported , platelet shape change and aggregation in response to ADP (10 μm) are completely abolished in P2Y1 knockout mice (Fig. 3A). ADP no longer activated Rac in P2Y1 knockout mouse platelets either at short time of stimulation (Fig. 3A) or at longer times (2 or 5 min, data not shown). As the P2Y1 receptor is known to couple to Gq and to stimulate phospholipase C and calcium mobilization [2,3], we used Gαq knockout mouse platelets to confirm this result. ADP failed to activate Rac in Gαq knockout platelets either after 10 s (Fig. 3A) or after 5 min (data not shown) of stimulation. These results support the data obtained with pharmacologic inhibitors (Fig. 1) and indicate that the P2Y1 receptor and its downstream effector Gαq are essential for ADP-induced Rac activation in platelets. They also demonstrate that P2Y12 is not able to stimulate the GTPase per se. To confirm that a receptor which couples exclusively to Gq is able to activate Rac, we stimulated human platelets with serotonin. This weak platelet agonist induced a modest but significant activation of Rac (Fig. 3B).
P2X1 receptor triggering fails to stimulate Rac by itself
The P2X1 purinergic receptor present on platelets preferentially binds ATP. As ADP preparations are often contaminated by about 10% of ATP, we checked whether this receptor would be involved in Rac activation upon ADP stimulation. Using high concentration of apyrase (0.9 U mL−1) to prepare washed mouse platelets, the P2X1 receptor was kept functional, as shown by αβMeATP (100 μm)-mediated Ca2+ influx in wild-type platelets (Fig. 4A). Under these conditions, ADP induced a similar activation of Rac in wild type and in P2X1 knockout mouse platelets, and this activation was abolished by the P2Y1 antagonist MRS-2179 (100 μm) (Fig. 4B). These results indicate that the P2X1 receptor is not required for ADP-induced Rac activation in mouse platelets in vitro under our experimental conditions.
To investigate whether the P2X1 receptor could activate the GTPase by itself, washed human platelets were prepared with apyrase (0.9 U mL−1) and stimulated with αβMeATP (100 μm). This selective and stable P2X1 receptor agonist induced a rapid Ca2+ influx (data not shown) but we did not detect Rac activation in the course of this Ca2+ influx (Fig. 4C). Thus, such a Ca2+ rise is not sufficient to activate the GTPase per se.
ADP potentiates TRAP- and FcγRIIa-induced Rac activation via the P2Y12 receptor
It is now clearly established that secreted ADP plays a key role as a co-activator of platelet activation by TRAP or FcγRIIa cross-linking [12–15]. Therefore, we investigated the impact of secreted ADP on Rac activation initiated by FcγRIIa cross-linking or by TRAP. Figure 5A shows that Rac activation induced by FcγRIIa clustering was strongly decreased in the presence of the P2Y12 receptor antagonist, AR-C69931MX. Conversely, when the P2Y1 receptor was blocked with the MRS-2179 compound (Fig. 5A) or with the A2P5P antagonist (data not shown), Rac activation was not significantly affected. Similar results were obtained with TRAP as a primary agonist (Fig. 5A, right panel). These results strongly suggest that ADP can potentiate TRAP or FcγRIIa-induced Rac activation through a P2Y12-dependent pathway.
α2A-adrenergic receptor triggering overcomes AR-C69931MX-mediated Rac, PAK and actin polymerization inhibition in platelets stimulated by FcγRIIa clustering or by TRAP
Figure 5B shows that the activation of Rac upon FcγRIIa cross-linking was accompanied by the autophosphorylation of its downstream effector PAK as well as actin assembly. These events were strongly reduced by the P2Y12 antagonist AR-C69931MX. In platelets, epinephrine is known to bind to the α2A-adrenergic receptor and to activate the Gz pathway leading to inhibition of adenylyl cyclase activity. Interestingly, epinephrine (10 μm) was able to overcome the inhibitory effects of AR-C69931MX (Fig. 5B) on FcγRIIa-induced Rac and PAK activation and actin assembly whereas it was not able to activate Rac by itself (data not shown). These results suggest that a Gi/Gz-dependent pathway is necessary for a full activation of Rac, PAK autophosphorylation and actin assembly in response to platelet stimulation by FcγRIIa clustering. Similar results were obtained upon TRAP stimulation (Fig. 5B, right panel), although independently of ADP, TRAP induced a fair amount of Rac activation and actin assembly. The implication of PI 3-kinase in Rac activation by the combined action of P2Y12 and FcγRIIa was also tested. Inhibition of PI 3-kinase by wortmannin (50 nm) or LY294002 (25 μm) strongly inhibited the activation of Rac (78% of inhibition), indicating a role of this lipid kinase in this process (data not shown). The effect of PI 3-kinase inhibition on TRAP-induced Rac activation was minor (data not shown).
A role of G-protein βγ subunits in FcγRIIa-mediated Rac activation
To further investigate the molecular mechanisms supporting the synergy between the Gi-coupled P2Y12 receptor and FcγRIIa in Rac activation, we investigated the potential role of G-protein βγ subunits. For this purpose, we used a peptide from phospholipase C β2 that interacts with G-protein βγ subunits with high affinity and can be used as a βγ subunit sequestering molecule as described previously . Indeed, Fig. 6A shows that this peptide interacted with human platelet G-protein β subunits. Interestingly, addition of this peptide to permeabilized platelets led to a dose-dependent inhibition of FcγRIIa-mediated Rac activation whereas an unrelated peptide was without effect (Fig. 6B). These results strongly suggest a role of G protein βγ subunits in Rac activation induced by the combined signals of FcγRIIa and P2Y12/Gi in human platelets.
ADP is a key physiological co-activator of platelet functions in vitro and in vivo and can strongly influence the morphological changes of these cells during their activation process [21,22]. In this study, we show that ADP alone rapidly induces Rac activation independently of GPIIb/IIIa integrin engagement and PI 3-kinase activity. In agreement, we previously found that pharmacologic inhibitors of fibrinogen binding to GPIIb/IIIa did not affect the rapid Rac activation induced by other physiologic platelet agonists such as thrombin or the TXA2 mimetic U46619 [26,38]. These results do not exclude a regulation of Rac by GPIIb/IIIa during the late steps of platelet activation or upon adhesion and spreading as previously suggested [27,41,42]. However, they clearly indicate that heterotrimeric G-protein-coupled receptors can activate Rac independently of integrin signaling. This was confirmed by the fact that triggering the serotonin receptor, exclusively coupled to Gq, activates Rac without inducing platelet aggregation. Using both pharmacologic- selective antagonists and P2Y1-deficient mouse platelets, we show that the P2Y1 receptor and Gαq are essential for the activation of Rac and of its downstream effector in response to ADP alone. Thus the P2Y1/Gq pathway is major for the activation of GTPase by ADP used as a primary agonist. Triggering of TXA2 receptor, coupled to Gq and G12/13 in platelets, also leads to Rac activation through a Gq-dependent pathway . This is consistent with the fact that calcium mobilization downstream of Gq and PLC activation is essential for Rac activation by various agonists in human and mouse platelets [26,38]. In other models, various Gq/G11 protein-coupled receptors have been shown to activate Rac, including bradykinin and M1 muscarinic acetylcholine receptors . The specific guanine nucleotide exchange factor (GEF) of Rac involved downstream of Gq and cytosolic Ca2+ in platelets remains unknown. The Rac GEF Tiam1 can be activated by ionophore-mediated Ca2+ entry in fibroblast cells potentially via a mechanism involving calcium/calmodulin kinase . However, the implication of Tiam in platelet activation has not been described and GEFs for small GTPases remain poorly characterized in general in this model. A recent study suggests that two GEFs of Rac, Vav1 and Vav2, are not necessary for the early activation of Rac in platelets, although Vav1 may act at later stages of platelet aggregation . Moreover, Vav1 is not phosphorylated upon ADP stimulation, suggesting that this GEF is not involved in ADP-induced Rac activation . Nevertheless, the third member of the Vav family, Vav3, has not yet been investigated in platelets and may compensate for the lack of Vav1 and Vav2 in mice.
Our results also show that αβMeATP, a stable analog of ATP inducing a transient rise in cytoplasmic Ca2+ via the P2X1 receptor, is unable to stimulate Rac under our experimental conditions. This observation indicates that a relatively weak and transient cytosolic Ca2+ rise is not sufficient per se to activate Rac.
Platelets contain members of the Gq, Gi, Gs and G12/13 families of G proteins and combined signal of members of these different families (i.e. Gi and Gq or G12/13 and Gi) appears to be sufficient to cause platelet aggregation [46–48]. Here, we show that the rapid ADP-induced Rac activation requires a P2Y1/Gq pathway and that the impact of P2Y12/Gi in this process is undetectable. However, a role of P2Y12 in potentiating the P2Y1 in ADP-induced platelet morphology changes has been reported . The interplay between the two ADP receptors may be important to modulate mechanisms cooperating with Rac to control the actin cytoskeleton dynamics.
When TRAP or FcγRIIa clustering agents are used as primary agonists, secreted ADP plays a major role in Rac activation via its P2Y12 receptor, which is not able to activate this GTPase per se as shown by the use of selective antagonists and P2Y1 knockout platelets. Combined signaling of P2Y12/Gi and PAR-1/Gq or FcγRIIa/tyrosine kinases is then necessary to activate Rac. This P2Y12-dependent effect is also observed on PAK activation and actin polymerization, two events known to involve Rac activation. Interestingly, the P2Y12 receptor, coupled to Gi, could be replaced by the α2A-adrenergic receptor coupled to Gz. Indeed, epinephrine was able to overcome the inhibitory effect of P2Y12 blockade. Stimulation of α2A-adrenergic receptor alone is insufficient to cause platelet secretion and aggregation in vitro  and does not induce Rac activation per se but epinephrine potentiates these events via Gz upon stimulation by other agonists. It is noteworthy that when Gi or Gz potentiate the activation of Rac in response to TRAP or FcγRIIa cross-linking, the extent of Rac activation is more important (three- to fourfold, data not shown) than that observed in response to ADP alone where the potentiation of P2Y1 by P2Y12 is undetectable. This suggests that a signaling mechanism induced by FcγRIIa and PAR-1, but not by P2Y1, is required to obtain the co-stimulatory effect of P2Y12 on Rac. The activation of Rac by the TXA2 receptor is also largely decreased by ADP scavengers . In contrast to P2Y1, this receptor, like PAR-1, is known to couple to Gq and G12/13. This could mean that activation of Rho via G12/13-proteins decreases the activation of Rac, an effect that could be overcome by Gi-mediated signaling. To further evaluate this hypothesis, we inhibited the major effector of Rho, Rho-kinase with the Y27632 compound, and observed a slight but reproducible increase in Rac activation (28% ± 6%, n = 3) induced by TRAP. The weak residual activation of Rac observed in the presence of the P2Y12 antagonist was also increased in the same range (16% ± 5%, n = 3) by Rho-kinase inhibition. These observations suggest that the Rho pathway partly decreases the level of Rac activation but the P2Y12 receptor is likely not capable to block this negative cross-regulation. Alternatively, P2Y1 could act as a negative regulator of P2Y12 as recently suggested by Hardy et al. . The complex interplay between the two ADP receptors requires further investigation to better understand this point.
Both αi and αz are capable of inhibiting adenylyl cyclase leading to a decrease in cyclic AMP concentration, which is a permissive event required for full platelet activation [47,51]. The βγ subunits of these G-proteins have been proposed to play a role in cell signaling, although their exact contribution in platelet activation is not clear yet. Our results show that the G-protein βγ subunits play a role in Rac activation induced by FcγRIIa clustering. Interestingly, P-Rex1, a Rac-specific GEF, has been shown to be directly and synergistically activated by βγ subunits and the PI 3-kinase product PtdIns(3,4,5)P3 . P-Rex1 is mainly expressed in hematopoietic cells, particularly in neutrophils, and in brain. As both βγ subunits and PI 3-kinase are implicated in Rac activation in response to FcγRIIa clustering, one can suggest that under these conditions a GEF such as P-Rex1 could integrate these signaling mechanisms to control Rac-GTP level. However, in the case of FcγRIIa triggering, PtdIns(3,4,5)P3 plays an important role in PLCγ2 activation and Ca2+ mobilization [15,16]. As Ca2+ is an important determinant of Rac activation in this model , PI 3-kinase is likely an indirect upstream regulator of this GTPase, through the control of PLCγ2 activation. Moreover, FcγRIIa-mediated PI 3-kinase activation, and in turn Ca2+ mobilization, are extremely dependent on ADP and P2Y12 receptors [15,16]. Therefore, blockade of this pathway by P2Y12 antagonists may explain the inhibition of Rac activation by FcγRIIa. Another Rac-GEF widely expressed is Ras-GRF2, which is active on Ras and Rac and is regulated by Ca2+ and G-protein βγ subunits . Although, so far, its expression has not been shown in platelets, such a GEF could also be implicated. Clearly, the identification of the specific Rac-GEF involved in the different conditions of platelet activation appears necessary as they may be potential targets for future drugs able to shrewdly modulate platelet responses.
In conclusion, our data demonstrate that the platelet P2 receptors play differential roles in regulating Rac. On the one hand, the P2Y1 receptor and its downstream effector Gq are responsible for the activation of the GTPase by ADP alone, independently of P2Y12, GPIIb/IIIa, and PI 3-kinase. On the other hand, the Gi-coupled P2Y12 receptor plays an essential role as co-activator of Rac, PAK and actin assembly when platelets are stimulated via other agonists such as TRAP or FcγRIIa clustering agents. Moreover, using a βγ subunit scavenger, we provide the first evidence for a role of these G-protein subunits in the activation of Rac via FcγRIIa.
Thanks are due to G.M. Bokoch for kindly providing reagents, C. Racaud-Sultan, H. Tronchère and F. Gaits-Icovoni for stimulating discussions. C.S. was supported by a grant from the Ministère de la Recherche et de la Technologie. This work was partly supported by grants from Association pour la Recherche Contre le Cancer (ARECA-Toulouse) and from CNRS/INSERM/Ministère de la Recherche joint programme (ACI) ≪Molécules et cibles thérapeutiques≫.