Phosphatidylinositol 3,4,5-Trisphosphate-dependent Stimulation of Phospholipase C-γ2 Is an Early Key Event in FcγRIIA-mediated Activation of Human Platelets*

Platelets express a single class of Fcγ receptor (FcγRIIA), which is involved in heparin-associated thrombocytopenia and possibly in inflammation. FcγRIIA cross-linking induces platelet secretion and aggregation, together with a number of cellular events such as tyrosine phosphorylation, activation of phospholipase C-γ2 (PLC-γ2), and calcium signaling. Here, we show that in response to FcγRIIA cross-linking, phosphatidylinositol (3,4,5)-trisphosphate (PtdIns(3,4,5)P3) is rapidly produced, whereas phosphatidylinositol (3,4)-bisphosphate accumulates more slowly, demonstrating a marked activation of phosphoinositide 3-kinase (PI 3-kinase). Inhibition of PI 3-kinase by wortmannin or LY294002 abolished platelet secretion and aggregation, as well as phospholipase C (PLC) activation, indicating a role of this lipid kinase in the early phase of platelet activation. Inhibition of PLCγ2 was not related to its tyrosine phosphorylation state, since wortmannin actually suppressed its dephosphorylation, which requires platelet aggregation and integrin αIIb/β3 engagement. In contrast, the stable association of PLCγ2 to the membrane/cytoskeleton interface observed at early stage of platelet activation was fully abolished upon inhibition of PI 3-kinase. In addition, PLCγ2 was able to preferentially interact in vitro with PtdIns(3,4,5)P3. Finally, exogenous PtdIns(3,4,5)P3 restored PLC activation in permeabilized platelets treated with wortmannin. We propose that PI 3-kinase and its product PtdIns(3,4,5)P3 play a key role in the activation and adequate location of PLCγ2 induced by FcγRIIA cross-linking.

Platelets express a single class of Fc␥ receptor (Fc␥RIIA), which is involved in heparin-associated thrombocytopenia and possibly in inflammation. Fc␥RIIA cross-linking induces platelet secretion and aggregation, together with a number of cellular events such as tyrosine phosphorylation, activation of phospholipase C-␥2 (PLC-␥2), and calcium signaling. Here, we show that in response to Fc␥RIIA cross-linking, phosphatidylinositol (3,4,5)-trisphosphate (PtdIns(3,4,5)P 3 ) is rapidly produced, whereas phosphatidylinositol (3,4)bisphosphate accumulates more slowly, demonstrating a marked activation of phosphoinositide 3-kinase (PI 3-kinase). Inhibition of PI 3-kinase by wortmannin or LY294002 abolished platelet secretion and aggregation, as well as phospholipase C (PLC) activation, indicating a role of this lipid kinase in the early phase of platelet activation. Inhibition of PLC␥2 was not related to its tyrosine phosphorylation state, since wortmannin actually suppressed its dephosphorylation, which requires platelet aggregation and integrin ␣ IIb /␤ 3 engagement. In contrast, the stable association of PLC␥2 to the membrane/cytoskeleton interface observed at early stage of platelet activation was fully abolished upon inhibition of PI 3-kinase. In addition, PLC␥2 was able to preferentially interact in vitro with PtdIns(3,4,5)P 3 . Finally, exogenous PtdIns(3,4,5)P 3 restored PLC activation in permeabilized platelets treated with wortmannin. We propose that PI 3-kinase and its product PtdIns(3,4,5)P 3 play a key role in the activation and adequate location of PLC␥2 induced by Fc␥RIIA cross-linking.
In addition to specific interactions involving their Fab domains, immunoglobulins G can interact with various membrane receptors by their Fc region. These so-called Fc receptors are coded by a heterogeneous family of at least eight genes forming a complex locus on chromosome 1. Based on both functional and structural criteria, these proteins can be differentiated into three groups (I-III) displaying various expression patterns in cells of the immune system (1,2). Platelets possess a single class of Fc␥ receptors, Fc␥RIIA. 1 As reviewed by Anderson et al. (3), clustering of Fc␥RIIA induces shape change, secretion, and aggregation, which are typical platelet responses contributing to their hemostatic function. In addition, there is a close link between secretion and aggregation, since released ADP was shown recently to be very critical for platelet aggregation evoked by Fc␥RIIA cross-linking (4). The precise role of platelet Fc␥RIIA is still obscure, although it explains how these cells are activated by specific antibodies directed against various membrane antigens, by immune complexes or by aggregated IgG (3). On a pathophysiological point of view, platelet Fc␥RIIA might involve an hemostatic response at inflammatory sites displaying IgG deposits, and there is increasing evidence for a direct involvement of Fc␥RIIA in heparin-associated thrombocytopenia occurring in patients under heparin therapy (3).
Fc␥RIIA is a 40-kDa polypeptide bearing two IgG-like domains in its extracellular region, a single transmembrane segment, and an ITAM, also known as Reth motif, in the cytoplasmic tail (5). ITAM sequences contain two variably spaced tyrosine residues; they are present in Fc␥ receptors, Fc⑀R, various subunits of both T cell and B cell antigen receptors, and the ␥-chain of Fc receptor in platelets (6); and they support the paradigm of immune cell activation (7)(8)(9)(10)(11)(12). Indeed, clustering of membrane receptors present in cells of the immune system promotes the phosphorylation, presumably by Src kinases, of the two tyrosine residues of ITAM. This results in the specific anchoring to ITAM sequences of tyrosine kinases bearing two SH2 domains, i.e. either Syk or ZAP-70. These two tyrosine kinases then induce a complex set of signaling events including calcium mobilization as well as activation of PI 3-kinase and mitogen-activated protein kinases, the latter ones via an upstream cascade involving Grb2-Sos and the small GTPase Ras (7)(8)(9)(10)(11)(12).
Phosphoinositide metabolism plays a key role during the stimulation of immune cell receptors and involves two types of enzymes, PLC and PI 3-kinase. Among various isoforms identified so far, PLC-␥2 is abundant in hematopoietic cells, it is activated downstream of tyrosine kinases, and it promotes both calcium mobilization and activation of protein kinase C via the two second messengers produced upon hydrolysis of PtdIns(4,5)P 2 (13,14). On the other hand, the heterodimeric IA class PI 3-kinase is also regulated by tyrosine kinases and promotes the accumulation of D3-phosphoinositides, which are considered as potential second messengers (15-17). Two recent studies have shown specific interactions between PLC-␥1 and PtdIns(3,4,5)P 3 , but depending on the authors this might involve either the N-terminal PH domain or the SH2 domains of the protein (18,19). However, no direct link between PLC-␥2, which is inactive toward the products of PI 3-kinase, and PI 3-kinase itself has been demonstrated so far in cells activated via Fc␥RIIA cross-linking.
Besides the fact that Fc␥RIIA might play a key role in the hemostatic and inflammatory function of platelets, these cells are interesting to consider in so far they contain a single class of Fc␥ receptor (Fc␥RIIA), in contrast to neutrophils or monocytes, for instance (1,3). Clustering of platelet Fc␥RIIA promotes phosphorylation of the two tyrosine residues of the ITAM motif, which is followed by the classical set of signaling events occurring under similar conditions, i.e. various tyrosine phosphorylations and calcium mobilization, through activation of PLC␥2 (3, 20 -22). Moreover, Chacko et al. (23) provided evidence that PI 3-kinase associates transiently with Fc␥RIIA upon platelet receptor clustering, probably via Syk. The present study was thus undertaken in order to determine possible changes of phosphoinositide metabolism occurring upon clustering of platelet Fc␥RIIA. In addition, taking advantage of the use of two specific and unrelated inhibitors (wortmannin and LY294002), we have focused our interest on signaling events occurring downstream of PI 3-kinase. Our present data unravel a causal relationship between PI 3-kinase and PLC-␥2, which might also function in the signaling cascade evoked by other membrane receptors involved in the immune response.
Preparation and Activation of Platelets-Platelets were isolated from concentrates obtained from the local blood bank (Etablissement de Transfusion Sanguine, Toulouse, France) essentially as described previously (24). They were washed in washing-buffer (pH 6.5) containing 140 mM NaCl, 5 mM KCl, 5 mM KH 2 PO 4 , 1 mM MgSO 4 , 10 mM Hepes, 5 mM glucose, 0.35% (w/v) bovine serum albumin. The same buffer plus 1 mM CaCl 2 was added to the final suspension, and pH was adjusted to 7.4. In experiments dealing with inositol lipid analysis, platelets were labeled with 0.5 mCi/ml [ 32 P]orthophosphate during 60 min in a phosphate-free washing buffer (pH 6.5) at 37°C. 32 P-Labeled platelets were then washed once in the same buffer and finally suspended at a final concentration of 1 ϫ 10 9 cells/ml (pH 7.4). Cross-linking of Fc␥RIIA was achieved by preincubation of platelets for 1 min with the monoclonal antibody IV.3 (2 g/ml) followed by addition of anti-mouse IgG F(abЈ) 2 (30 g/ml) at 37°C under gentle shaking as described previously (25).
Lipid Extraction and Analysis-Reactions were stopped by addition of chloroform/methanol (1/1, v/v) containing 0.4 N HCl, and lipids were immediately extracted following the modified procedure of Bligh and Dyer (26,27).
Measurement of Aggregation and 5-Hydroxytryptamine Secretion-Aggregation was monitored using a Chrono-log dual channel aggregometer with stirring at 900 rev/min at 37°C (5 ϫ 10 8 platelets/ml). 5-Hydroxytryptamine secretion was determined as described previously (29). Briefly, platelets labeled with 5-hydroxy[ 14 C]tryptamine were preincubated or not with increasing concentrations of wortmannin or LY294002 for 15 min and stimulated by Fc␥RIIA cross-linking during 3 min in presence of 5 M imipramine. Incubations were stopped by addition of 3% formaldehyde and 0.1 M EDTA, cooling on ice, and centrifugation. The radioactivity of 5-hydroxy[ 14 C]tryptamine released from platelet dense granules was determined by liquid-scintillation counting.
Gel Electrophoresis and Immunoblotting-Proteins were resuspended in electrophoresis sample buffer containing 100 mM Tris-HCl, pH 6.8, 15% (v/v) glycerol, 25 mM dithiothreitol, and 3% SDS, boiled for 5 min, separated on 7.5% SDS-polyacrylamide gel electrophoresis, and transferred onto a nitrocellulose membrane (Gelman Sciences). The nitrocellulose was blocked for 60 min at room temperature with 1% (w/v) milk powder, 1% (w/v) bovine serum albumin in a buffer containing 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.05% (v/v) Tween 20. Immunodetection was achieved using the relevant antibody, peroxidase-conjugated secondary antibody, and ECL system. The various bands were quantified by densitometric analysis measuring the pixel volume in each area (Gel Doc 1000, Bio-Rad).
Immunoprecipitation-Reactions were stopped by addition of twiceconcentrated ice-cold lysis buffer containing 80 mM Tris-HCl, pH 7.4, 200 mM NaCl, 200 mM NaF, 20 mM EDTA, 80 mM Na 4 P 2 O 7 , 4 mM Na 3 VO 4 , 2% (v/v) Nonidet P-40, and 10 g/ml each of aprotinin and leupeptin. After gentle shaking during 20 min at 4°C and centrifugation (12,000 ϫ g for 10 min at 4°C), the soluble fraction was collected and precleared for 30 min at 4°C with protein A-Sepharose CL4B. The precleared suspensions were then incubated overnight at 4°C with the adequate antibody, and immune complexes were then precipitated by addition of 10% (w/v) protein A-Sepharose CL4B for 1 h at 4°C and centrifugation (6,000 ϫ g for 5 min at 4°C). The immunoprecipitates were washed once in lysis buffer and twice in washing buffer containing 10 mM Tris-HCl, pH 7.4, 100 mM NaCl, 100 M Na 3 VO 4 , 1 g/ml each of aprotinin and leupeptin. Immunoprecipitated proteins were resolved by 7.5% SDS-polyacrylamide gel electrophoresis and analyzed by Western blotting.
Isolation of Cytoskeleton-Reaction was stopped and the cytoskeleton immediately extracted by adding one volume of ice-cold twice-concentrated cytoskeleton buffer containing 100 mM Tris-HCl, pH 7.4, 20 mM EGTA, 2 mM Na 3 VO 4 , 4 g/ml each of aprotinin and leupeptin, 2 mM phenylmethylsulfonyl fluoride, and 2% (v/v) Triton X-100 as described previously (30,31). After 10 min at 4°C, the cytoskeleton was pelleted by centrifugation (12,000 ϫ g for 10 min at 4°C), washed once in cytoskeleton buffer with 0.5% Triton X-100, and once with the same buffer without Triton X-100. Cytoskeleton was then immediately suspended in electrophoresis sample buffer (32).
Cytosol Depletion-After stimulation, platelets were centrifuged (3,000 ϫ g for 30 s) and suspended in 20 mM Pipes buffer (pH 6.8) containing 150 mM KCl, 2 mM EDTA, and 30 g/ml saponin. After 3 min at room temperature under shaking, supernatant and pellet fractions were separated by centrifugation (12,000 ϫ g for 40 s). The pellet was suspended in Laemmli's sample buffer, and PLC-␥72 was probed by Western blotting using a specific antibody as described above.

PI 3-Kinase and PLC Are Rapidly Activated upon Fc␥RIIA
Cross-linking-Cross-linking of platelet Fc␥RIIA has been shown to induce a transient association of PI 3-kinase to the ITAM sequences present in the cytoplasmic tail of the receptor; however, its consequences on a possible in vivo activation of the enzyme were not emphasized (23). Therefore, using an HPLC technique, we have first measured the time course of D3-phosphoinositide synthesis during Fc␥RIIA-mediated activation of 32 P-labeled platelets. We found that PtdIns(3,4,5)P 3 was rapidly produced, whereas PtdIns(3,4)P 2 accumulated with a slower time course (Fig. 1A).
In addition, we observed a rapid drop in the substrate of PLC, PtdIns(4,5)P 2 , followed by its resynthesis (Fig. 1B). This sharp decrease in PtdIns(4,5)P 2 was concomitant with the production of PtdOH (Fig. 1C), an event considered to reflect PLC activation in platelets. In these cells, the main part of diacylglycerol produced by PLC is converted into PtdOH by a diacylglycerol kinase, the contribution of phospholipase D being rel-atively minor (37,38). In agreement with this, the production of PtdOH evoked by Fc␥RIIA cross-linking was abolished by the PLC-specific inhibitor U73122 (data not shown). Finally, the radioactivity of phosphatidylinositol 4-phosphate (PtdIns(4)P) did not change significantly over the whole period of platelet stimulation by Fc␥RIIA cross-linking (Fig. 1D).
PI 3-Kinase Is Upstream of the Fc␥RIIA-dependent Activation of PLC-To determine whether PI 3-kinase was required for Fc␥RIIA-mediated physiological responses, wortmannin and LY294002, two unrelated PI 3-kinase inhibitors, were used in platelet secretion and aggregation assays. As reported previously (23), platelet aggregation induced by Fc␥RIIA crosslinking was fully inhibited by low doses of wortmannin, whereas, in similar conditions, platelet aggregation induced by TRAP became reversible and thrombin-induced aggregation was not significantly affected ( Fig. 2A). Similar data were obtained with LY294002, which was able to fully inhibit Fc␥RIIA-mediated aggregation (95 Ϯ 5% inhibition at 25 M). Interestingly, platelet secretion evoked by Fc␥RIIA cross-linking was also strongly inhibited by wortmannin or LY294002 in a dose-dependent manner (Fig. 2B). These results clearly indicated a critical role for PI 3-kinase in the early steps of Fc␥RIIAdependent platelet aggregation.
Under these conditions (10 nM wortmannin), the synthesis of both PtdIns(3,4,5)P 3 and PtdIns(3,4)P 2 was totally suppressed, while no change was observed for the other phosphoinositides (data not shown). However, interestingly, PtdOH production was almost abolished in platelets challenged with Fc␥RIIA cross-linking, whereas thrombin-induced accumulation of Pt-dOH remained insensitive to wortmannin (Fig. 3, A and B). Inhibition of PtdOH and D3-phosphoinositide production displayed very similar dose-response curves using the two inhibitors of PI 3-kinase, with IC 50 of 4 nM and 2 M for wortmannin and LY294002, respectively (Fig. 3C). These values are comparable to those determined for inhibition of serotonin secretion (6 nM and 5 M, see Fig. 2B). Finally, myrecithine, a natural flavonoid with inhibitory activity toward PI 3-kinase (39), also blocked PtdOH formation. 2 Specific inhibition of PI 3-kinase thus appeared to secondar-  ily block PLC activation evoked by Fc␥RIIA cross-linking, at a variance with the signaling pathway evoked by thrombin. As a main difference between thrombin and Fc␥RIIA cross-linking, the former activates PLC-␤2 and -␤3, which are regulated by heterotrimeric G-proteins and are present in significant amounts in platelets (40,41). In contrast, PLC-␥2 appears as an essential intermediate in the signaling pathway evoked downstream of Fc␥RIIA, which involves its tyrosine phosphorylation presumably by Syk (25,42). Indeed, inhibition of Syk by a specific inhibitor, piceatannol, was found to inhibit secretion, aggregation, and PtdOH production induced by Fc␥RIIA cross-linking, the same effects being reproduced with the specific PLC inhibitor U73122 (data not shown). This led us to explore the effect of wortmannin on the tyrosine phosphorylation of PLC-␥2.
PI 3-Kinase Regulates PLC-␥2 by a Mechanism Independent of Its Tyrosine Phosphorylation-We first investigated the whole pattern of phosphotyrosyl proteins in platelets stimulated by Fc␥RIIA cross-linking. As shown in Fig. 4A, this was not affected even at a high concentration of wortmannin (100 nM). In contrast to thrombin, Fc␥RIIA cross-linking promoted its own transient tyrosine phosphorylation, as reported by oth-  32 P-Labeled platelets were incubated or not with wortmannin or LY294002 (15 min, 37°C) and activated by Fc␥RIIA cross-linking or by thrombin (1 IU/ml) during 3 min. Lipids were immediately extracted, PtdOH was separated by TLC and visualized by autoradiography (A). In B and C, the radioactivity corresponding to PtdOH was quantified by PhosphoImager analysis. In B, data are means Ϯ standard errors (four experiments). The dose-response curves shown in C correspond to one representative determination. ers (43,44). Again, this was insensitive to wortmannin.
PLC-␥2 also displayed a transient tyrosine phosphorylation in response to Fc␥RIIA cross-linking. This reached a maximum at 1 min and decreased rapidly thereafter (Fig. 4, B, panel a,  and C). Surprisingly neither wortmannin (10 nM) nor LY294002 (10 M) altered the early phase of PLC-␥2 tyrosine phosphorylation, but they abolished the secondary dephosphorylation step (Fig. 4B, panels b and c, and C). A clearer interpretation of the latter data became possible when we observed the same effects with the tetrapeptide RGDS, which inhibits platelet aggregation by competing for fibrinogen binding to integrin ␣ IIb /␤ 3 or in the absence of shaking that also prevented aggregation (Fig. 5A). It thus appeared that dephosphorylation of PLC-␥2 is a relatively late signaling event occurring downstream of integrin ␣ IIb /␤ 3 engagement, which was shown previously to activate or to induce the subcellular relocation of at least two protein-tyrosine phosphatases (45)(46)(47)(48). It is thus logical that wortmannin and LY294002, which inhibit platelet aggregation, albeit at a step upstream of integrin ␣ IIb /␤ 3, also perturb dephosphorylation of PLC-␥2.
However, in strong contrast to wortmannin and LY294002, inhibition of aggregation by avoiding platelet shaking or upon addition of RGDS did not significantly affect Fc␥RIIA-mediated activation of PLC, as shown by the production of PtdOH (Fig.  5B). Altogether, these data indicated that tyrosine phosphorylation of PLC-␥2 is not sufficient to trigger its activation, suggesting that another crucial factor might be required. This prompted us to investigate in more detail the localization of PLC-␥2 in order to examine whether PI 3-kinase would play a role in its interactions with the membrane and (or) the cytoskeleton.
PI 3-Kinase Triggers the Stable Interaction of Cytosolic PLC-␥2 to the Platelet Membrane/Cytoskeleton-In a first series of experiments, the cytoskeleton from platelets stimulated by Fc␥RIIA cross-linking was isolated by precipitation in Triton X-100 according to established procedures (30,31). To focus on the initial signaling events occurring under these conditions and to avoid major cytoskeletal changes occurring downstream of ␣ IIb /␤ 3 engagement, aggregation was inhibited either by the lack of shaking or upon addition of RGDS. Immunoblotting of cytoskeletal pellet allowed us to observe the appearance of PLC-␥2 upon Fc␥RIIA cross-linking, thus reflecting the translocation of the enzyme from the cytosol to the cytoskeleton (Fig. 6A). However, this was abolished in the presence of wortmannin (Fig. 6A).
Very similar data were obtained using another experimental procedure. In this case, platelets were permeabilized with saponin under conditions allowing the release of over 92% of the cytoplasmic enzyme lactate dehydrogenase. It was then possible to quantify the amounts of PLC-␥2 remaining associated with the cells, presumably by interacting with the membrane and (or) the cytoskeleton. As shown in Fig. 6B, Fc␥RIIA crosslinking promoted platelet retention of PLC-␥2 only when PI 3-kinase was fully active. These various data suggested that PI 3-kinase products might be implicated in the behavior of PLC-␥2 in platelets stimulated by Fc␥RIIA cross-linking. The possibility that PLC-␥2 might directly interact with D3-phosphoinositides was thus examined.

)P 3 Preferentially Interacts with PLC-␥2 and Overcomes the Inhibition of Platelet PLC Subsequent to PI 3-Kinase Blockade-PLC-␥2
was immunoprecipitated from resting platelets or from the same cells stimulated for 1 min by Fc␥RIIA cross-linking. Upon in vitro incubation with a mixture of 32 P-labeled D3-phosphoinositides, the enzyme was found to specifically bind PtdIns(3,4,5)P 3 (Fig. 7A). Apparently, this interaction did not depend on the phosphorylation state of the protein, since identical data were obtained with PLC-␥2 iso- lated from resting or activated platelets. To see if this observation was of significant relevance, we then tested whether PtdIns(3,4,5)P 3 was able to relieve the inhibition of PLC-␥2 observed in wortmannin-treated platelets.
In a last series of experiments, 32 P-labeled platelets were permeabilized with saponin and challenged by Fc␥RIIA crosslinking. Although their responses were reduced compared with intact platelets, these conditions allowed us to detect a stimulation-dependent production of PtdOH. Apparently, permeabilization induced some spontaneous activation of PLC in resting platelets, resulting in a lower increase of PtdOH radioactivity upon Fc␥RIIA cross-linking (2.5-fold, as compared with about 10-fold in intact platelets). However, a low concentration of wortmannin (10 nM) again depressed PtdOH production (60% inhibition, Fig. 7B), whereas further addition of PtdIns(3,4,5)-P 3 (15 M) restored PLC activity to 80% of the value determined in control activated platelets (Fig. 7B). In contrast, PtdIns(3,4)P 2 was without effect. DISCUSSION Combined to other data previously reported in the literature, the results obtained in the present study allow us to describe a sequence of signaling events occurring during Fc␥RIIA-mediated platelet activation and illustrated in Fig. 8. Cross-linking of Fc␥RIIA leads immediately to tyrosine phosphorylation of its ITAM sequences, which allows recruitment and activation of the tyrosine kinase Syk (23,25). There is some evidence that Syk then interacts with PI 3-kinase (23), which is recruited to the membrane and activated, leading to the rapid accumulation of PtdIns(3,4,5)P 3 . Another downstream effector of Syk has been identified as PLC-␥2, which is converted into an active form upon tyrosine phosphorylation, although this might involve additional protein-tyrosine kinases such as Bruton's tyrosine kinase (Btk), as shown for the B cell receptor (49 -51). However, tyrosine phosphorylation is not sufficient to allow expression of PLC-␥2 activity, which requires its specific interaction with the membrane via PtdIns(3,4,5)P 3 generated by PI 3-kinase. Calcium mobilization and activation of protein kinase C, promoted by the two second messengers generated by PLC-␥2, will allow both platelet secretion and conformational change of integrin ␣ IIb /␤ 3 . This finally leads to platelet aggregation upon fibrinogen binding to its integrin receptor. In the same time, additional signals generated as a consequence of integrin ␣ IIb /␤ 3 engagement and of aggregation will activate (or simply modify subcellular localization) of a protein-tyrosine phosphatase, which dephosphorylates PLC-␥2. Additional negative feedback signals might also include dephosphorylation of PtdIns(3,4,5)P 3 into PtdIns(3,4)P 2 , for instance by a 5-phosphatase such as SH2 domain-containing inositol 5-phosphatase (SHIP) (52). All the data presented herein fit with these conclusions. 1) Inhibition of Syk by piceatannol or of PI 3-kinase by wortmannin or LY294002 both inhibit PLC activation as well FIG. 5. Inhibition of aggregation impairs tyrosine dephosphorylation without affecting PLC activity in platelets stimulated by Fc␥RIIA cross-linking. Platelets were incubated in the absence or in the presence of RGDS (200 g/ml) and stimulated by Fc␥RIIA crosslinking with or without shaking, which is necessary for aggregation to occur. A, PLC-␥2 was immunoprecipitated and checked by immunoblotting with the anti-phosphotyrosine 4G10 antibody as in Fig. 4. B, 32 P-labeled platelets were used and the radioactivity of PtdOH was determined as in Fig. 3. Data (-fold increase compared with nonstimulated platelets) are means of two independent experiments. as secretion and aggregation induced by Fc␥RIIA cross-linking; 2) accumulation of PtdIns(3,4,5)P 3 precedes the formation of PtdOH; 3) in the absence of PI 3-kinase inhibitors, phosphorylation of PLC-␥2 coincides with the time where the enzyme is fully active, as detected by the accumulation of PtdOH; 4) whatever the step involved (PI 3-kinase or integrin ␣ IIb /␤ 3 ), inhibition of aggregation abolishes the secondary dephosphorylation of PLC-␥2; 5) the inhibition of PI 3-kinase can be overcome by introduction, into permeabilized platelets, of PtdIns(3,4,5)P 3 which was found to bind specifically to PLC-␥2.
To ascertain our pharmacological evidence that PI 3-kinase plays a crucial role in the activation of PLC-␥2 evoked by Fc␥RIIA cross-linking, it is important to notice that wortmannin and LY294002, which act on PI 3-kinase by different mechanisms, were added at concentrations previously shown to be specific for PI 3-kinase and used to demonstrate a role of this enzyme in various cell responses including histamine secretion in RBL-2H3 cells or proliferation of aortic smooth muscle cells, for instance (53,54). In addition, an effect on PtdIns 4-kinase, which would have limited the amount of PtdIns(4,5)P 2 available to PLC-␥2, can be excluded for three reasons. This should have also inhibited PLC activity triggered by thrombin, which was not the case; we could not find any decrease in the labeling of both PtdIns(4)P and PtdIns(4,5)P 2 ; and there are two types of PtdIns 4-kinase, type I being insensitive to wortmannin and LY294002 (53,54), whereas type III requires wortmannin concentrations in the micromolar range for its blockade (55).
Two recent studies have shown specific interactions between PLC-␥1 and PtdIns(3,4,5)P 3 , but depending on the authors this might involve either the N-terminal PH domain or the SH2 domains of the protein (18,19). The present investigation was not aimed at identifying which region of PLC-␥2 participates in binding to PtdIns(3,4,5)P 3 but one may suppose that, actually, both domains might be involved. Whatever the precise mechanism would be, PtdIns(3,4,5)P 3 synthesis was found to be required for the association of PLC-␥2 with either Triton X-100insoluble cytoskeleton or the whole particulate fraction remaining after cell permeabilization. A critical role of cytoskeleton in the signaling events described in the present study is strongly suggested by the recent observation that the small GTPase Rho, which stabilizes focal adhesion plaques and promotes the generation of stress fibers (56), tightly regulates calcium signaling and phagocytosis evoked by macrophage Fc␥ receptors (57). It is difficult at this stage to conclude whether PLC-␥2 interacted with PtdIns(3,4,5)P 3 bound to cytoskeletal proteins, as repeatedly observed for this and other phosphoinositides (58), or whether some membrane domains remained associated with the platelet cytoskeleton, thus suggesting that platelets and incubated with a mixture of 32 P-labeled PI 3-kinase products, unlabeled phosphoinositides and phosphatidylserine as indicated under "Experimental Procedures." The phosphoinositides that remained associated with immunoprecipitated PLC-␥2 after several washing steps were extracted, separated by TLC, and detected by autoradiography. Shown is a representative result from three different experiments with very similar data. Lane 1, total 32 P-labeled phosphoinositides incubated with the different immune complex; lane 4, control with nonimmune complex. B, 32 P-labeled platelets were incubated with or without wortmannin (10 nM, 15 min, 37°C), then permeabilized with saponin (20 g/ml) and activated by Fc␥RIIA cross-linking in the absence or in the presence of diC16-PtdIns(3,4,5)P 3 or diC16-PtdIns(3,4)P 2 (15 M) as described under "Experimental Procedures." After 1 min of stimulation, lipids were extracted and the radioactivity of PtdOH was determined as in Fig. 3. Data (percentage of PtdOH increase, over control) are means Ϯ standard errors of three independent experiments. a, significant difference (p Ͻ 0.002) with platelets stimulated in the absence of wortmannin; b, significant difference (p Ͻ 0.041) with platelets stimulated in the presence of wortmannin (Student's t test).
FIG. 8. Postulated mechanism of PI 3-kinase-dependent activation of PLC-␥2 in platelets stimulated by Fc␥RIIA cross-linking. Upon clustering, Fc␥RIIA is tyrosine-phosphorylated and recruits Syk, allowing tyrosine phosphorylation of PLC-␥2. Presumably via Syk, Fc␥RIIA also recruits PI 3-kinase, whose subsequent activation promotes the rapid accumulation of PtdIns(3,4,5)P 3 . The latter phosphoinositide allows binding of PLC-␥2 to the membrane/cytoskeleton, enabling activation of the enzyme and generation of the second messengers inositol 1,4,5-trisphosphate (IP 3 ) and diacylglycerol (DAG). Calcium mobilization and activation of protein kinase C will promote platelet secretion as well as a conformational change of integrin ␣ IIb /␤ 3 , whose engagement results in aggregation. A negative feedback signal involves activation (or appropriate relocation) of a protein-tyrosine phosphatase (PTP), which dephosphorylates PLC-␥2. An additional negative feedback mechanism involving SHIP (52) is not indicated for the sake of clarity. Other targets of PtdIns(3,4,5)P 3 include the serine/threonine kinase Akt (62) or Bruton's tyrosine kinase (Btk), which is also a target of PtdIns(3,4,5)P 3 (63,64). This suggests that Btk could act together with Syk for the tyrosine phosphorylation of PLC␥2, as shown for B cell receptor (49 -51).
an appropriate localization of the enzyme might be at specific contact points between membrane and cytoskeleton. Identification of the fine structure of membrane/cytoskeleton microdomains targeting PLC to its substrate via PtdIns(3,4,5)P 3 will require additional investigations.
In studies mentioned above, interaction of PLC-␥1 with PtdIns(3,4,5)P 3 was used to explain how PI 3-kinase regulates the activity of PLC-␥1 in cells stimulated with platelet-derived growth factor, which involves a membrane receptor displaying intrinsic protein-tyrosine kinase activity (18,19). In the present work, we obtained the first evidence that PI 3-kinase is absolutely required for the activation of PLC-␥2 by Fc␥RIIA cross-linking, the tyrosine phosphorylation of PLC␥2 being necessary but not sufficient for its activation. It will thus be important to see whether this can be extended to different receptors acting by a similar mechanism in various immune cells, for instance other Fc␥R, Fc⑀R, T cell antigen receptors, and B cell antigen receptors. A recent report indicates that this is the case for the B cell surface glycoprotein CD19 induced PLC␥ isozyme activation (59). Another example giving indirect evidence of a similar mechanism, i.e. human neutrophils where calcium mobilization triggered by Fc␥R clustering also requires PI 3-kinase activity (60). However, upon antigen stimulation of RBL-2H3 mast cells, PI 3-kinase is required for activation, translocation and also tyrosine phosphorylation of PLC␥1 whereas it has no effect on PLC␥2 activation (61). As we have observed for PLC␥2 in platelets stimulated by Fc␥RIIA crosslinking, upon platelet-derived growth factor treatment of transfected COS-1 cells, PI 3-kinase does not regulate the activity of PLC␥1 at the level of tyrosine phosphorylation but rather allow a stable interaction of PLC␥1 to the membrane probably leading to efficient substrate hydrolysis (18). Bae et al. (19) have observed that in vitro PtdIns(3,4,5)P 3 is sufficient, at elevated concentration, to activate PLC␥1 and PLC␥2; they also noticed that production of PtdIns(3,4,5)P 3 within the cell does not always result in PLC␥ activation. Altogether, these results suggest that, according to the model and the agonist used, PLC␥ isozymes may be stimulated by different events including tyrosine phosphorylation, PtdIns(3,4,5)P 3 production and subtle compartmentalization mechanisms that seems to be critical for the modulation of their regulation.
In conclusion, we took advantage of the use of a simple cellular model, human platelets, which express a single class of Fc␥R, to elucidate a very crucial point of cell signaling involving tight regulation of PLC-␥2 by PtdIns(3,4,5)P 3 generated upon activation of PI 3-kinase. This mechanism is an obligatory step in Fc␥RIIA cross-linking-dependent platelet activation and might also function in the signaling cascade evoked by other membrane receptors containing ITAM.