Essential Role of Protein Kinase Cδ in Platelet Signaling, αIIbβ3 Activation, and Thromboxane A2 Release*

The protein kinase C (PKC) family is an essential signaling mediator in platelet activation and aggregation. However, the relative importance of the major platelet PKC isoforms and their downstream effectors in platelet signaling and function remain unclear. Using isolated human platelets, we report that PKCδ, but not PKCα or PKCβ, is required for collagen-induced phospholipase C-dependent signaling, activation of αIIbβ3, and platelet aggregation. Analysis of PKCδ phosphorylation and translocation to the membrane following activation by both collagen and thrombin indicates that it is positively regulated by αIIbβ3 outside-in signaling. Moreover, PKCδ triggers activation of the mitogen-activated protein kinase-kinase (MEK)/extracellular-signal regulated kinase (ERK) and the p38 MAPK signaling. This leads to the subsequent release of thromboxane A2, which is essential for collagen-induced but not thrombin-induced platelet activation and aggregation. This study adds new insight to the role of PKCs in platelet function, where PKCδ signaling, via the MEK/ERK and p38 MAPK pathways, is required for the secretion of thromboxane A2.

with collagen at low shear (7). This in turn induces platelet signaling and activation, which lead to aggregation through ␣ IIb ␤ 3 binding to fibrinogen (8).
Platelet signaling may be induced by the adhesion process itself or by several agonists leading to granule secretion and activation (9 -11). One of the most physiologic and potent agonists, thrombin, activates platelets through proteinase-activated receptors (12). Stimulation of proteinase-activated receptors 1 and 4 in human platelets induces inside-out signaling (13) through the activation of phospholipase C (PLC␤). Furthermore, collagen, the most abundant protein of the extracellular matrix, promotes the adhesion and activation of platelets through its binding to platelet integrin ␣ 2 ␤ 1 and GPVI (14,15). Binding of collagen to GPVI stimulates a non-receptor proteintyrosine kinase that phosphorylates and thereby activates PLC␥ (16 -18). Activated PLC hydrolyzes phosphatidylinositol 4,5bisphosphate, generating the second messengers diacylglycerol (DAG) and inositol 1,4,5-triphosphate. Inositol 1,4,5-triphosphate, in turn, mediates the release of Ca 2ϩ from intracellular stores, whereas DAG activates PKCs.
The PKC family comprises 12 related isoforms, classified into three structurally and functionally distinguished subgroups (19 -22). The conventional PKC isoforms (PKC␣, PKC␤I, PKC␤II, and PKC␥) are regulated by both DAG and Ca 2ϩ , the novel isoforms (PKC␦, PKC⑀, PKC, and PKC) are insensitive to Ca 2ϩ , and the atypical isoforms (PKC and PKC/) are Ca 2ϩ -and DAG-insensitive (23). Even though PKCs share a high degree of homology, it is believed that particular PKC isoforms have unique and specific functions within different cell types (24). In platelets, multiple PKC isoforms are expressed (25)(26)(27)(28), and it is presumed that each individual isoform plays one or more distinct roles. Indeed, several studies have underlined a role for PKCs in platelet function, such as aggregation and secretion of granular contents (29 -31). However, the role of individual PKC isoforms in platelet function remains to be elucidated. For instance, although it is well established that collagen-induced platelet activation and aggregation through GPVI is mediated by tyrosine kinase Syk and activation of PLC␥ 2 (32,33), there is limited information concerning the roles of specific PKC isoforms and their downstream effectors in this process.
This study was therefore undertaken to examine the role of PKCs, in particular PKC␦, in platelet activation and aggregation. Using thrombin and collagen, two physiological and potent platelet activators, we were able to show that PKC␦ is essential for collagen-induced but not thrombin-induced platelet aggregation and activation of integrin ␣ IIb ␤ 3 . This study also shows that PKC␦ is positively regulated by ␣ IIb ␤ 3 outside-in signaling and is involved in the release of TxA 2 through the MEK/ERK and p38 MAPK signaling pathways, which is required for collagen-induced but not thrombin-induced platelet activation and aggregation.
Preparation of Human Platelets-Venous blood was drawn from healthy volunteers, free from medication known to interfere with platelet function for at least 10 days before the experiment, in accordance with the guideline of the ethical committee of the Montreal Heart Institute. The platelets were prepared as previously described (35,36). Briefly, the platelets were obtained by series of differential centrifugations and adjusted to a final concentration of 250 ϫ 10 6 /ml using an automated cell counter in which purity exceeded 99%. The platelets were allowed to stand at 37°C for 30 min before further experiments.
Platelet Aggregation-Aggregation of washed platelets was monitored on a four-channel optical aggregometer (Chronolog Corp., Havertown, PA) (35,36). The platelets were preincubated with inhibitors for 5 min at 37°C. The samples were then FIGURE 1. Phospholipase C signaling is required for platelet activation and aggregation: role of PLC signaling in ␣ IIb ␤ 3 activation, P-selectin translocation, and platelet aggregation. Isolated human platelets (250 ϫ 10 6 /ml) were preincubated with the PLC inhibitor, U73122 (1-5 M), prior to stimulation by collagen (2 g/ml) or thrombin (0.1 unit/ml). ␣ IIb ␤ 3 activation and P-selectin translocation to the membrane (left and middle panels, respectively) were assessed by flow cytometry. Overlay plots (gray, base line; white, control; black, U73122 at 5 M) shown are presented as the number of events over the log of associated fluorescence and are representative of mean data presented in the lower panels (n ϭ 3; *, p Ͻ 0.05 versus base line). Collagen-and thrombin-induced platelet aggregation (right panel) was monitored in the presence or absence of increasing doses of U73122 (1-5 M) and traces shown are representative of mean data presented in the bottom panels (n ϭ 4; *, p Ͻ 0.05 versus base line). stimulated under continuous stirring (1000 rpm) at 37°C. Platelet aggregation was recorded and measured 5 min after the addition of the agonists.
SDS-PAGE and Immunoblotting-The platelets were stimulated for the appropriate time with stirring at 37°C in the presence or absence of inhibitors; the reaction was terminated by the addition of a 4 ϫ SDS-Laemmli sample buffer and then boiled for 10 min. The proteins were resolved by electrophoresis in 8 -10% SDS-polyacrylamide gels and subsequently transferred to nitrocellulose. The membranes were blocked with 5% bovine serum albumin at room temperature, incubated overnight at 4°C with the primary antibody, washed, and labeled with horseradish peroxidase-conjugated secondary antibody at room temperature. Following additional washing steps, proteins were detected by enhanced chemiluminescence (PerkinElmer Life Sciences).
Subcellular Fractionation of Platelets-The platelet reactions were terminated by the addition of an equi-volume of ice-cold Triton X-100 lysis buffer (2% Triton X-100, 10 mM EGTA, 2 mM phenylmethylsulfonyl fluoride, 20 g/ml leupeptin, 20 g/ml benzamidine, and 100 mM Tris-HCl, pH 7.4), and the samples were allowed to stand on ice for 30 min. The platelet lysates were subsequently centrifuged at 100,000 ϫ g for 2 h at 4°C to separate the soluble from the insoluble membrane fractions. The pellets (insoluble Triton X-100 membrane fractions) were solubilized in SDS sample buffer, boiled for 5 min, and both fractions were analyzed by SDS-PAGE immunoblotting as described above.
Flow Cytometry-␣ IIb ␤ 3 activation and P-selectin translocation were assessed by flow cytometry, as described previously (35,36), using monoclonal antibodies against P-selectin (clone AK4 phycoerythrin-conjugated; Santa Cruz) and the active form of ␣ IIb ␤ 3 (clone PAC-1 fluorescein isothiocyanate-conjugated; Becton Dickinson, Mississauga, Canada). The platelets (250 ϫ 10 6 /ml) were preincubated with inhibitors for 5 min at room temperature prior to cell stimulation. The samples were then fixed with 1% paraformaldehyde for 1 h at 4°C, washed, and labeled with saturating concentration of monoclonal antibody for 30 min. ␣ IIb ␤ 3 activation was determined by incubating PAC-1 antibody in platelet suspension prior to activation. All of the samples were analyzed on an Altra flow cytometer (Beckman Coulter, Mississauga, Canada). Nonspecific labeling was excluded using appropriately labeled isotype-matched control IgG. The platelets were identified and gated by their characteristic forward and side scatter properties, and 20,000 platelets were analyzed from each sample.
Confocal Immunofluorescence and Imaging-The platelets from the above experiments were fixed with 2% (v/v) paraformaldehyde in phosphate-buffered saline, washed twice, and allowed to immobilize on poly-L-lysine-coated coverslips over-night. Adhered platelets were subsequently blocked with 2% bovine serum albumin and incubated with wheat germ agglutinin with Alexa 488 in 1% bovine serum albumin for 1 h. The coverslips were then permeabilized with 0.1% Triton X-100 in 2% normal donkey serum, incubated with polyclonal anti-PKC␦ for 3 h, washed, labeled with anti-rabbit IgG-Alexa 555 secondary antibody for 1 h, and mounted on microscopic slides. A series of fluorescent confocal images (Z stacks) were acquired with a LSM 510 confocal microscope (Zeiss, Oberkochen, Germany). Anti-rabbit (donkey) Alexa 555 (Molecular Probes, Eugene, OR) and wheat germ agglutinin with Alexa 488 were visualized using a 543-nm helium-neon laser line and a 488-nm argon laser line, respectively. A 100ϫ/1.3 Plan-Neofluar objective (Zeiss) was used for magnification. Voxel size is 30 ϫ 30 ϫ 150 nm (X, Y, and Z). Z stacks were deconvolved with the Huygens Pro 2.6.5a software (Scientific Volume Imaging). Deconvolved Z stacks were saved in .tif image file format series and transferred to the LSM510 software.
Thromboxane A 2 Measurement-The platelets were prepared as described above, pretreated with antagonists or with vehicle solution, and stimulated under stirring conditions at 37°C. The reaction was terminated by the addition of an equivolume of ice-cold EDTA solution (10 mM) containing 10 M of indomethacin, and the samples were centrifuged at 3000 ϫ g for 5 min at 4°C. Thromboxane B 2 , the stable metabolite of TxA 2 , was measured using a correlate-EIA thromboxane B 2 enzyme immunoassay kit (Assay Designs, Inc, Ann Arbor, MI), according to the manufacturer's instructions.
Data Analysis-The experiments were performed at least three times on different donors. The results are presented as the means Ϯ S.E. Comparisons of multiple treatments in the same individuals were analyzed by repeated measures analysis of variance with Bonferroni t test for multiple comparisons, whereas treatment groups composed of different individuals were analyzed by one-way analysis of variance. The values with p Ͻ 0.05 were considered statistically significant.

Role of PLC in Platelet Activation and Aggregation-Collagen
activates platelets via a tyrosine kinase-dependent cascade that culminates in tyrosine phosphorylation of PLC␥. Activation of PLC␥ leads, in turn, to the generation of second messenger inositol 1,4,5-triphosphate and DAG, which mobilize intracellular Ca 2ϩ and activate certain PKC isoforms, respectively (14,15). Thus, it was important that we first confirm the relative role of PLC in collagen-and thrombin-induced platelet activation and aggregation. As depicted in Fig. 1, collagen-and thrombin-induced platelet activation (␣ IIb ␤ 3 activation and P-selectin translocation) and aggregation were inhibited, in a dose-dependent manner, by the PLC inhibitor U73122. The responses to collagen, but not to thrombin, were completely blocked by PP1 and Piceatannol, which selectively inhibit Src and Syk, respectively (data not shown). These results indicate that platelet activation and aggregation, in response to both collagen and thrombin, occur in a PLC-dependent manner.
Implication of PKCs in Platelet Activation and Aggregation-Because PLC is involved in platelet function and PKCs are downstream effectors, we sought to evaluate the specific role of individual PKC isoforms in platelet function. As illustrated in Fig. 2A, aggregation of platelets induced by collagen and thrombin was completely inhibited by the wide spectrum PKC inhibitor, chelerythrin. However, platelet aggregation was not affected by the PKC␣ and PKC␤ inhibitors HBDDE and Hispidin, respectively, or by Gö6976, an inhibitor of both the conventional PKC␣ and ␤ isoforms. In contrast, collagen-induced platelet aggregation was completely blocked by a selective PKC␦ inhibitor (Rottlerin), whereas aggregation of platelets by thrombin remained unaffected by Rottlerin. In addition, activation of ␣ IIb ␤ 3 in response to collagen was completely prevented by PKC␦ inhibition, whereas the response to thrombin was less affected (Fig. 2B). On the other hand, P-selectin translocation was slightly reduced by Rottlerin but only in response to thrombin. These results clearly indicate that activation of PKC␦ is essential for collagen-induced but not thrombin-induced platelet function.
Phosphorylation of PKC␦ and Regulation by ␣ IIb ␤ 3 -Having shown that PKC␦ is essential for collagen-induced but not thrombin-induced platelet activation and aggregation, we next evaluated its phosphorylation as an index of its activation. As expected, both collagenandthrombinwereableinatimedependent manner to phosphorylate PKC␦ (Fig. 3A). However, PKC␦ appears to be phosphorylated more rapidly by thrombin than by collagen, that is, 30 s after stimulation by thrombin as compared with 60 s by collagen. As shown in Fig.  3B, pretreatment of platelets with U73122 completely prevented the phosphorylation of PKC␦ by both collagen and thrombin, indicating that PLC is responsible for the activation of PKC␦. Surprisingly, Rottlerin, which selectively inhibits PKC␦, prevented its phosphorylation by collagen but not by thrombin.
Because the results obtained with the phosphorylation state of PKC␦ mainly correspond with those obtained with the activation and aggregation of platelets, we presumed that ␣ IIb ␤ 3 might regulate or be involved in the phosphorylation of PKC␦. To support this possibility, we showed that the blockade of ␣ IIb ␤ 3 using Reopro (a monoclonal antibody that specifically binds to ␣ IIb ␤ 3 ) or Aggrastat (a non-peptide antagonist of ␣ IIb ␤ 3 ) was able to completely abolish the phosphorylation of PKC␦ (Fig. 3B) induced by both collagen and thrombin. These results indicate that ␣ IIb ␤ 3 outside-in signaling may positively modulate the phosphorylation of PKC␦.
Translocation of PKC␦ to Cell Membrane and Regulation by ␣ IIb ␤ 3 -One particular and important aspect of PKC activation is the intracellular redistribution of the enzyme from the cytosol to the cell membrane. We studied this mechanism using confocal microscopy and subcellular fractionation of platelets into soluble and membrane fractions (Fig.  4). We found that in resting platelets, PKC␦ is distributed across the cytosol of the cell and redistributed to the membrane after activation by collagen and thrombin. We also found that in platelets pretreated with Rottlerin, U73122, or Reopro, PKC␦ was incapable of translocating to the platelet membrane upon activation by collagen. In contrast, thrombin-induced translocation of PKC␦ was not affected by Rottlerin but was inhibited by U73122 and Reopro (Fig. 4). Thus, the phosphorylation and translocation of PKC␦ to platelet membrane, both events being required for full enzymatic activation, require the engagement of ␣ IIb ␤ 3 .
Activation of the MEK/ERK and p38 MAPK Pathways by PKC␦-In platelets, both thrombin and collagen can activate MAPK signaling pathways, and it is presumed that the PKC family may be involved in this process (37)(38)(39). Thus, the MEK/ ERK and p38 MAPK pathways may be important downstream effectors of PKC␦ involved in platelet activation and aggregation. As illustrated in Fig. 5A, pretreatment with the MAPK kinase inhibitors, U0126 for MEK/ERK and SB202109 for p38, dose-dependently inhibited collagen-induced but not thrombin-induced platelet aggregation and activation of ␣ IIb ␤ 3 (Fig.  5B). The same results were obtained using two other MEK/ERK and p38 inhibitors, PD98059 and SB203580, respectively (data not shown). Furthermore, we assessed the phosphorylation of MEK1/2, ERK1/2, and p38 upon platelet activation. Fig. 6A shows that both agonists were able to time-dependently phosphorylate MEK1/2, ERK1/2, and p38. However, thrombin appears to phosphorylate these MAPKs faster than collagen, as observed with the phosphorylation of PKC␦. As expected, U0126 was able to prevent the phosphorylation of both MEK1/2 and ERK1/2, whereas SB202109 prevented the phos-phorylation of p38 (Fig. 6B). In addition, U73122 and Rottlerin inhibited the phosphorylation of these kinases, indicating that PLC and PKC␦ are involved in this signaling cascade. In contrast, Reopro slightly affected the phosphorylation state of these MAPKs. Thus, it appears that the MEK/ERK and p38 signaling pathways are important downstream effectors of PKC␦ in platelets.
Role of Secondary Mediators in PKC␦ Signaling-It is well established that collagen induces platelet aggregation through a pathway that is primarily mediated by the release of TxA 2 and ADP (40 -42) and that thrombin-induced platelet aggregation seems to be less dependent upon these mediators. It was therefore important that we determine the role of TxA 2 and ADP secretion in collagen-and thrombin-induced PKC␦ signaling. As expected, we were first able to confirm the importance of TxA 2 in collagen-induced but not thrombin-induced platelet aggregation (Fig. 7A) and activation of integrin ␣ IIb ␤ 3 (Fig. 7B), using the TxA 2 receptor antagonist SQ29548 and the COX inhibitor indomethacin. Moreover, the ADP scavenger apyrase also diminished collageninduced but not thrombin-induced platelet aggregation (Fig. 7A). Given that the secretion of TxA 2 appears to be essential for collageninduced platelet activation and aggregation, we next investigated the role of PKC␦ and the other signaling mediators involved in this process. As depicted in Fig. 7C, both collagenand thrombin-induced TxA 2 generation were completely blocked by PLC, PKC␦, MEK/ERK, and p38 inhibitors. However, blockade of the integrin ␣ IIb ␤ 3 receptor with Reopro showed little or no effect on the secretion of TxA 2 by platelets. Taken together, these data clearly indicate that PKC␦, in addition to its upstream and downstream effectors, PLC and the MEK/ERK/p38 pathways, respectively, are essential for collagen-and thrombin-mediated TxA 2 generation. However, the secretion of TxA 2 is necessary for the activation of ␣ IIb ␤ 3 and the aggregation of platelets only in response to collagen but not to thrombin.

DISCUSSION
In platelets, the PKC family is an important signaling mediator required for activation, secretion of granule contents, and aggregation. In the current study, we examined the role of individual PKC isoforms in platelet activation and aggregation induced by physiological agonists such as colla- gen and thrombin. The major findings are: 1) activation of PKC␦ is essential for collagen-induced but not thrombininduced platelet activation of ␣ IIb ␤ 3 and aggregation; 2) the activity of PKC␦, as measured by its phosphorylation and translocation to the platelet membrane upon activation by both collagen and thrombin, is positively regulated by ␣ IIb ␤ 3 outside-in signaling; and 3) PKC␦ triggers activation of the MEK/ERK and p38 MAPK signaling cascades and the subsequent release of TxA 2 , which are essential for collagen-induced but not thrombin-induced platelet activation and aggregation.
It is well established that collagen supports the adhesion and activation of platelets at sites of vascular injury via ␣ 2 ␤ 1 and GPVI (43,44). The binding of collagen to GPVI results in receptor clustering and phosphorylation of the tyrosine kinase Syk and the activation of several signaling pathways including PLC. Phospholipase C can then hydrolyze phosphatidylinositol 4,5-bisphosphate and generate DAG, which in turn, activates certain PKC isoforms. In platelets, many PLC isoforms have been identified; all of the evidence suggests that thrombin activates a G-protein-coupled PLC␤ pathway, whereas signaling through collagen is mediated by receptorbound tyrosine kinases and activation of PLC␥ (45,46). Indeed, PLC inhibition using U73122 completely prevented platelet aggregation, ␣ IIb ␤ 3 activation, and Pselectin translocation following platelet activation.
Having confirmed the involvement of PLC in collagen-and thrombin-induced platelet activation and aggregation, we next evaluated the role of individual PKC isoforms in platelet function. First, the implication of PKCs in general was assessed with Chelerytrin, a wide spectrum PKC inhibitor that dosedependently inhibited collagen-and thrombin-induced platelet aggregation, thus supporting a role for PKCs in this process. We were thereafter able to demonstrate that collagen-induced platelet aggregation was mainly mediated by PKC␦ and not by the conventional PKC␣ or PKC␤ isoforms, whereas thrombin-mediated platelet aggregation appeared to be dependent on multiple PKC isoforms. Indeed, inhibition of thrombin-induced platelet aggregation was only observed with the wide spectrum inhibitor Chelerytrin or when the conventional PKC isoforms (PKC␣ and PKC␤) and the novel PKC␦ isoform were blocked (data not shown). In contrast, it has been shown that PKC␦ can interact with the tyrosine kinase Fyn to down modulate platelet function (34). However, in that report, platelets were activated by   OCTOBER 6, 2006 • VOLUME 281 • NUMBER 40 JOURNAL OF BIOLOGICAL CHEMISTRY 30031 the snake venom alboaggregin-A, which is able to activate both GPIb-V-IX and GPVI on platelets. In our study, collagen, a physiological agonist, was employed to achieve platelet activation, and thus only signaling events mediated through GPVI and possibly ␣ 2 ␤ 1 were studied. It seems, therefore, that PKC␦ may play dichotomous roles in platelet aggregation, possibly involving differential regulatory mechanisms related to the signaling pathways triggered by various specific receptors.

PKC␦ in Platelet Function
Platelet aggregation is a process that is accomplished by the adhesive interaction of two active ␣ IIb ␤ 3 receptors on adjacent platelets, via a fibrinogen bridge. Given that PKC␦ is required to achieve this process, we next evaluated its involvement in the activation of integrin ␣ IIb ␤ 3 and in the translocation of P-selectin to the platelet surface upon activation. As expected, collagen-induced activation of ␣ IIb ␤ 3 was totally dependent on PKC␦, which explains the incapacity of platelets to aggregate in the presence of Rottlerin. On the other hand, PKC␦ inhibition had no effect on thrombin-induced platelet aggregation while slightly reducing ␣ IIb ␤ 3 activation. These observations suggest the presence of a ␣ IIb ␤ 3 receptor threshold level required to mediate platelet aggregation. P-selectin translocation to the platelet surface, a marker of ␣ granule secretion, can also be used as an index of platelet activation. The PKC family appears to be involved in this process through its capacity to phosphorylate a number of platelet proteins involved in the secretion of granule contents, such as platelet Sec 1 and syntaxin 4 (47)(48)(49). Here, we have demonstrated that PKC␦ does not appear to be required for collagen-induced P-selectin translocation and seems to have only a minor impact in this process in thrombin signaling. These results add new insights to the mechanisms of platelet P-selectin translocation; they also confirm our previous findings showing that the novel ⑀ and and atypical but not the conventional ␣ and ␤ and the novel ␦ PKCs are involved in this process (50).
Phosphorylation of PKC is an essential regulatory mechanism required for complete kinase activity. Hence, having showed the importance of PKC␦ in platelet function, it was important that we study the phosphorylation state of PKC␦ after activation of platelets by collagen and thrombin. Some investigators have reported PKC␦ to be phosphorylated upon stimulation by platelet activators such as thrombin, convulxin, and alboaggregin-A (25,34,51). Activation of PKC␦ by collagen, however, has yet to be demonstrated. As expected, both collagen and thrombin were able to time-dependently phosphorylate PKC␦, as soon as 30 s for thrombin and 60 s for collagen. These times happen to coincide with the onset of aggregation induced by both these activators. When evaluating the impact of PLC (U73122) and PKC␦ (Rottlerin) inhibition on the phosphorylation activity of PKC␦, as compared with the impact of these inhibitors on platelet aggregation, we found these two phenomena to correlate with one another. These observations lead us to believe that ␣ IIb ␤ 3 , the main platelet integrin responsible for the formation of a platelet aggregate, might be participating in the regulation of the phosphorylation activity of PKC␦. This hypothesis was further demonstrated with functional ␣ IIb ␤ 3 antagonists (Reopro and Aggrastat), which inhibit platelet aggregation, binding to fibrinogen and subsequent outside-in signaling. In the presence of Reopro and Aggrastat, we were unable to detect any given phosphorylation of PKC␦ after stimulation by either collagen or thrombin, thus supporting a role for ␣ IIb ␤ 3 outside-in signaling in the regulation of PKC␦. This regulatory mechanism most likely involves a positive feedback loop that maintains PKC␦ phosphorylated, as recently demonstrated for the platelet PKC isoform (52). Platelet function is normally accompanied by phosphorylation and dephosphorylation of many proteins. Phosphatase activities are important regulators of signal transduction in platelets. Indeed, a protein-tyrosine phosphatase inhibitor (vanadate) has been shown to stimulate protein tyrosine phosphorylation and to increase all platelet functions (53). Blockade of secondary ␣ IIb ␤ 3 outside-in signaling prevents PKC␦ phosphorylation, suggesting that such a signal is involved in the regulation of PKC␦, possibly via phosphatase inhibition. On the other hand, it has been reported that thrombin-induced activation of PKC␦ is achieved by a ␣ IIb ␤ 3 -independent pathway (51). However, the investigators in that study worked with very high thrombin concentrations (5 units/ml); it is likely that at these concentrations, PKC␦ dephosphorylates much more slowly, thereby allowing the detection of a phosphorylation even in the absence of ␣ IIb ␤ 3 outside-in signaling. In the present study, we presume that PKC␦ is activated independently of ␣ IIb ␤ 3 (the initial activation of PKC␦ precedes the activation of ␣ IIb ␤ 3 ) but that it may be regulated by the latter. ␣ IIb ␤ 3 outside-in signaling contributes to the adhesive strengthening of the platelet plug and allows irreversible aggregation, both mechanisms being essential for the formation of stable and adequate platelet aggregates (54). The importance of ␣ IIb ␤ 3 outside-in signaling in the regulation of PKC␦ was further pointed out when studying the translocation of the enzyme from the cytosol to the cell membrane upon activation (another important aspect of PKC activation). In resting platelets, PKC␦ appeared to be randomly distributed in the cytosol and to rapidly translocate to the cell surface upon activation by collagen and thrombin. In experiments in which ␣ IIb ␤ 3 outside-in signaling was blocked (Reopro), no translocation of PKC␦ to the cell membrane upon activation was observed. Thus, the regulation of PKC␦ by ␣ IIb ␤ 3 could be of physiological significance in platelet function.
Having established the role of PKC␦ in platelet activation and aggregation, we then evaluated the involvement of the MAPK signaling pathway in these platelet responses. It has been reported that physiological agonists such as collagen and thrombin can activate ERK1/2 and p38 in platelets; it also appears that the PKC family may be involved in this process (37)(38)(39). However, the exact role of the MAPK signaling cascade in platelet function remains unclear. Here, we have shown that collagen-induced but not thrombin-induced platelet aggregation and activation of ␣ IIb ␤ 3 are dependent on the MAPK signaling pathway, as demonstrated by the MAPK kinase inhibitors U0126 and SB202109. These results are in agreement with a previous study indicating that MEK1 is not involved in thrombin nor U46619 platelet response, although it plays an important role in collagen-induced platelet aggregation (55). In addition, we demonstrated that MEK1/2, ERK1/2, and p38 are activated by collagen and thrombin, and more importantly, established the requirement for PKC␦ and PLC activation in this process. Moreover, we found that inhibition of ␣ IIb ␤ 3 slightly affected the phosphorylation activity of MEK1/2, ERK1/2, and p38, which indicates that the regulation of PKC␦ phosphorylation exerted by ␣ IIb ␤ 3 outside-in signaling is not required for the activation and phosphorylation of PKC␦ substrates and downstream effectors, such as the MAPK signaling cascade. These results indicate, too, that the ␣ IIb ␤ 3 outside-in signaling cascade involved in the regulation of PKC␦ most likely involves a signaling pathway, downstream of PKC␦, other than that MEK/ERK and p38.
Secondary mediators, such as ADP and TxA 2 , have been reported to be involved in collagen-induced platelet aggregation; it was therefore important for us to establish the role of these mediators in PKC␦ signaling. As expected, we were first able to confirm the importance of ADP and most importantly, TxA 2 in collagen-induced but not in thrombin-induced platelet aggregation and activation of ␣ IIb ␤ 3 . Furthermore, we showed that both collagen-and thrombin-induced TxA 2 release by platelets strongly depends on PKC␦ signaling, in addition to its upstream (PLC) and downstream (MEK/ERK/p38) effectors. These data suggest that the PKC␦/MEK/ ERK/p38 signaling cascade is intimately involved in the generation and release of TxA 2 in platelets, which appears to be essential for collagen-induced but not thrombin-induced platelet function. Also, because blockage of ␣ IIb ␤ 3 showed little or no effect on the release of TxA 2 , we can again conclude that the ␣ IIb ␤ 3 outside-in signaling cascade involved in the regulation of PKC␦ does not lead to the activation of MEK/ERK/p38 and the subsequent generation of TxA 2 . In this connection, it has been recently shown that thrombin-mediated (via proteinase-activated receptor activation) TxA 2 release by platelets was dramatically blocked by PKC␦ inhibition (56). Here we confirm these observations and further elucidate the mechanisms by which this platelet phenomenon occurs, not only in thrombin-mediated but also in collagen-mediated signaling.
In conclusion, this study adds new insights to the role of PKC␦ in platelet signaling and function (Fig. 8). After the binding of platelet receptors to collagen, phosphorylation of the tyrosine kinases Syk and Src lead in part to the activation of PKC␦, through the action of PLC␥. PKC␦ then triggers activation of the MEK/ERK and p38 signaling pathways, which ultimately result in the generation and release of TxA 2 . Secretion of TxA 2 in platelets is well known to cause recruitment and activation of platelets at sites of vascular injury. In this study, we have shown that the release of TxA 2 is essential for collageninduced ␣ IIb ␤ 3 activation and platelet aggregation. In thrombin-mediated platelet signaling, PKC␦ is also required for the release of TxA 2 , and this platelet response occurs in a MEK/ ERK/p38-dependent manner. In contrast to collagen-induced platelet signaling, this release does not appear to be essential for platelet function. Finally, binding of fibrinogen to ␣ IIb ␤ 3 trig- FIGURE 8. Diagram summarizing the role of PKC␦ in platelet signaling. Collagen induces platelet signaling through it's binding to GPVI and ␣ 2 ␤ 1 , which represent the two main collagen receptors on the platelet surface. This leads, in turn, to sequential activation of the tyrosine kinases Src and Srk, which then phosphorylate PLC␥. The latter activates PKC␦ and subsequently the MEK/ERK and p38 MAPK signaling pathways, which are ultimately responsible for the synthesis and secretion of TxA 2 . Released TxA 2 can, in turn, bind to its receptor (TP) on the cell surface and activate the ␣ IIb ␤ 3 integrin through a pathway still to be elucidated. Finally, binding of fibrinogen to activated ␣ IIb ␤ 3 induces outside-in signaling that leads, in part, to a regulatory positive feedback loop, which maintains PKC␦ phosphorylated. In thrombin-induced platelet signaling, PKC␦ is also responsible for the synthesis and release of TxA 2 , primarily through activation of the MEK/ERK/p38 pathway. However, secretion of TxA 2 does not appear to be required for the activation of ␣ IIb ␤ 3 or the aggregation process. Activation of ␣ IIb ␤ 3 appears to be regulated by more than one PKC isoform ( Fig. 2A). In thrombin-stimulated platelets, outside-in signaling via fibrinogen binding to ␣ IIb ␤ 3 also contributes to maintain PKC␦ phosphorylated.
gers outside-in signaling, which leads, in part, to a regulatory positive feedback loop, which in turn maintains PKC␦ phosphorylated. However, this second wave of PKC␦ phosphorylation does not seem to involve the MEK/ERK/p38-dependent TxA 2 release.