Direct demonstration of involvement of protein kinase Calpha in the Ca2+-induced platelet aggregation.

Platelets play critical roles in hemostasis and thrombosis through their aggregation following activation of integrin alphaIIbbeta3. However, the molecular mechanism of the integrin activation inside platelets remains largely unknown. Pharmacological experiments have demonstrated that protein kinase C (PKC) plays an important role in platelet aggregation. Because PKC inhibitors can have multiple substrates and given that non-PKC-phorbol ester-binding signaling molecules have been demonstrated to play important roles, the precise involvement of PKC in cellular functions requires re-evaluation. Here, we have established an assay for analyzing the Ca2+-induced aggregation of permeabilized platelets. The aggregation of platelets was inhibited by the addition of the arginine-glycine-aspartate-serine peptide, an integrin-binding peptide inhibitor of alphaIIbbeta3, suggesting that the aggregation was mediated by the integrin. The aggregation was also dependent on exogenous ATP and platelet cytosol, indicating the existence of essential cytosolic factors required for the aggregation. To examine the role of PKC in the aggregation assay, we immunodepleted PKCalpha and beta from the cytosol. The PKC-depleted cytosol lost the aggregation-supporting activity, which was recovered by the addition of purified PKCalpha. Furthermore, the addition of purified PKCalpha in the absence of cytosol did not support the aggregation, whereas the cytosol containing less PKC supported it efficiently, suggesting that additional factors besides PKC would also be required. Thus, we directly demonstrated that PKCalpha is involved in the regulation of Ca2+-induced platelet aggregation.

Platelets play critical roles in hemostasis and thrombosis through their aggregation following activation of integrin ␣ IIb ␤ 3 . However, the molecular mechanism of the integrin activation inside platelets remains largely unknown. Pharmacological experiments have demonstrated that protein kinase C (PKC) plays an important role in platelet aggregation. Because PKC inhibitors can have multiple substrates and given that non-PKC-phorbol ester-binding signaling molecules have been demonstrated to play important roles, the precise involvement of PKC in cellular functions requires re-evaluation. Here, we have established an assay for analyzing the Ca 2؉ -induced aggregation of permeabilized platelets. The aggregation of platelets was inhibited by the addition of the arginine-glycine-aspartate-serine peptide, an integrin-binding peptide inhibitor of ␣ IIb ␤ 3 , suggesting that the aggregation was mediated by the integrin. The aggregation was also dependent on exogenous ATP and platelet cytosol, indicating the existence of essential cytosolic factors required for the aggregation. To examine the role of PKC in the aggregation assay, we immunodepleted PKC␣ and ␤ from the cytosol. The PKC-depleted cytosol lost the aggregation-supporting activity, which was recovered by the addition of purified PKC␣. Furthermore, the addition of purified PKC␣ in the absence of cytosol did not support the aggregation, whereas the cytosol containing less PKC supported it efficiently, suggesting that additional factors besides PKC would also be required. Thus, we directly demonstrated that PKC␣ is involved in the regulation of Ca 2؉ -induced platelet aggregation.
Platelets play critical roles in hemostasis and thrombosis through aggregation following activation of integrin ␣ IIb ␤ 3 (1)(2)(3). Although ␣ IIb ␤ 3 of platelets in the resting stage does not bind fibrinogen or von Willebrand factors (vWF), 1 the activation of platelets results in conformation changes, which allow ␣ IIb ␤ 3 to bind these ligands (1)(2)(3). When fibrinogen or vWF binds to ␣ IIb ␤ 3 , the ligand-occupied integrin signals downstream to stabilize platelet aggregation through reorganization of the actin cytoskeletal network and the release of bioactive substances stored in the granules (1)(2)(3). Thus, the process of platelet activation and aggregation consists of a series of orchestrated responses (1)(2)(3). However, the molecular mechanisms that underlie this process remain unclear because it is difficult to use molecular biology and biochemistry in platelets that do not synthesize new proteins. To overcome this difficulty, semi-intact assays using permeabilized platelets have been established for the study of granule secretion (4 -8). Although some semi-intact aggregation assays have now been developed, it has been difficult to demonstrate a cytosol dependence in these experiments (8).
Protein kinase C (PKC) family members are important signaling molecules regulating many cellular functions (9,10). Among the family members, conventional PKCs (cPKC), which include PKC␣, ␤I, ␤II, and ␥, have regulatory Ca 2ϩ -and phorbol ester-binding domains (9,10). The involvement of cPKCs in the cellular functions has been analyzed mainly pharmacologically using cell-permeable small compounds of inhibitors and stimulators such as phorbol esters. However, these experiments are somewhat indirect because 1) no inhibitors have absolute specificity and 2) important signaling molecules other than PKCs have also been demonstrated to contain phorbol ester-binding C1-domains, which were first identified in cPKC (9,10). For example, Ras-guanyl nucleotide-releasing protein (11,12) contains the C1 domain at its C terminus and acts as a stimulator for small GTPase Ras involved in the regulation of cell growth. Chimerin (13,14) also contains a C1 domain and acts as a GTPase-activating protein of small GTPase Rac involved in the regulation of cytoskeletal reorganization. Thus, the effects of phorbol esters could be through multiple pathways. Therefore, it is important to re-evaluate and demonstrate the involvement of PKCs in certain cellular functions in a direct fashion (15).
In platelets, PKC has been considered to play important roles in aggregation (16,17) and granule secretion (18). For the granule secretion, we have recently demonstrated the direct involvement of PKC (19). We have established a semi-intact secretion assay (19,20) where the secretion does not occur upon stimulation without adding exogenous cytosol, indicating the existence of cytosolic essential factors. We purified an essential factor for the secretion and identified it to be PKC␣ (19). On the other hand, it has been so far difficult to demonstrate the involvement of PKC in platelet aggregation without using small compounds of PKC inhibitors or phorbol esters. Here, we directly demonstrate the involvement of PKC␣ in the regulation of platelet aggregation by using a stable semi-intact aggregation assay with cytosol dependence.

EXPERIMENTAL PROCEDURES
Materials-Anti-PKC␣ and anti-pan-PKC␤ mouse monoclonal antibodies were purchased from Transduction Laboratories. This anti-PKC␣ antibody, which was described in the manufacturer's instruction to interact with PKC␤ but not with other PKCs, was used for immunodepletion experiments. Another anti-PKC␣ mouse monoclonal antibody purchased from Santa Cruz Biotechnology had no cross-reactivity with PKC␤ and was used for the immunoblotting. Anti-PKC␥, anti-PKC␦, and anti-PKC mouse monoclonal antibodies were from Transduction Laboratories, and anti-PKC␤I and anti-PKC␤II rabbit polyclonal antibodies were from Santa Cruz Biotechnology. A control mouse IgG used for the immunodepletion experiment was from Zymed Laboratories Inc. Horseradish peroxidase-labeled anti-mouse and anti-rabbit IgG polyclonal antibodies were from Amersham Biosciences, which were used as secondary antibodies for immunoblotting visualized by enhanced chemiluminescence method (Amersham Biosciences). Unless otherwise specified, all of the chemicals including peptides of arginineglycine-aspartate-serine (RGDS) and arginine-glycine-glutamate-serine (RGES) were purchased from Sigma with the exception of SLO, which was from Dr. Bhakdi (Mainz University, Mainz, Germany). Protein concentrations were determined by the Bradford method (Bio Rad) (21) or densitometric scanning of the Coomassie Blue-stained band in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (22) using bovine serum albumin as a standard.
The Aggregation Assay with Permeabilized Platelets-The aggregation assay was established by modification of our previous assay (23). Unless otherwise specified, the standard assay method was used as follows. Washed human platelets from healthy donors were prepared (24), resuspended in ice-cold Buffer A (50 mM Hepes/KOH, pH 7.2, 78 mM KCl, 4 mM MgCl 2 , 0.2 mM CaCl 2 , 2 mM EGTA, 1 mM dithiothreitol, and the calculated free [Ca 2ϩ ] was ϳ20 nM (25)) containing 4 mg/ml bovine serum albumin and 20 ng/ml prostaglandin E 1 and kept at 4°C. The platelets were incubated with 0.6 g/ml SLO at 4°C for 10 min and washed once to remove unbound SLO (19,20,26,27), The treated platelets were resuspended in ice-cold Buffer A containing 4 mg/ml bovine serum albumin at a density of 5 ϫ 10 8 /ml, quantified with a Coulter Counter, and incubated at 30°C for 5 min to make holes in their plasma membrane (19,20,26,27). The permeabilized platelets were kept on ice for 15-30 min with 2 mg of proteins/ml human platelet cytosol (or rat brain cytosol), an ATP-regenerating system (19,20,26,27), and tested substances. Because fibrinogen has been added in previously established aggregation assays using washed platelets at 0.4 mg/ml by Kinlough-Rathbone et al. (28) and 0.38 mg/ml by Santos et al. (29), we also added fibrinogen (Sigma) in our assay at the concentration of 0.4 mg/ml. The platelets then were incubated at 37°C for 3 min and stimulated by the addition of calcium chloride to make final [Ca 2ϩ ] at 200 M (25). The aggregation was measured by a light transmission aggregometer, MCM HEMA TRACER 212 (MC Medical). For the morphological analysis, permeabilized platelets were incubated with 200 M or 20 nM Ca 2ϩ for 20 min in the standard assay condition. The samples then were immediately examined with a phase-contrast microscopy.
Cytosol Preparation-For generation of platelet cytosol, platelets from healthy donors were washed and resuspended in ice-cold Buffer A containing a protease inhibitor mixture (Sigma). The platelets then were sonicated, and the supernatant after the low speed centrifugation at 1000 ϫ g for 10 min was further centrifuged at 100,000 ϫ g for 30 min. The final supernatant was dialyzed extensively with Buffer A and kept as platelet cytosol at Ϫ80°C until use. The cytosol of rat brain was prepared in a similar way except using a Potter-type blender instead of sonication. For generating the PKC-depleted cytosol, the human platelet cytosol (0.4 mg of proteins) was incubated with protein A-agarose beads (Roche Diagnostics) coated with 25 g of the anti-PKC␣ antibody (Transduction Laboratories) or the control mouse IgG for 6 h at 4°C. After the beads were removed by centrifugation, the supernatants were used as the PKC-depleted cytosol and the control cytosol, respectively. PKC␣ was purified from rat brain cytosol as described previously (19).

Establishment of an Assay Analyzing the Ca 2ϩ -induced Aggregation of Permeabilized Platelets-We have established an
aggregation assay with platelets permeabilized by SLO, which has been shown to form pores of ϳ30 nm in diameter in the membrane (27). In this method, we permeabilized only the plasma membrane judging from observations that the intracellular membrane structures of the platelets appeared to be intact morphologically (30) and that vWF stored in ␣-granules did not leak out (19,20). Because the condition used here with 0.6 g/ml SLO induced leakage of 80% cytosolic lactate dehydrogenase from platelets (19,20), it was presumed that most of ATP and cytosol were also lost by diffusion through the pores in the plasma membrane. Therefore, we exogenously added ATP and cytosol in the assay to reconstitute the aggregation. We also added 0.4 mg/ml fibrinogen in our assay, which would bridge the activated integrin ␣ IIb ␤ 3 on both sides of platelets to be aggregated (2,3). Because calcium ionophore has been demonstrated to induce platelet aggregation (31, 32), we used calcium ions at a calculated concentration at 200 M (25) as a stimulus.
We first examined morphologically whether aggregates of permeabilized platelets were indeed generated in the assay. After confirming the Ca 2ϩ -induced aggregation by the lighttransmission aggregometer (data not shown), the samples were subjected to observation with a phase-contrast microscopy. A typical set of photographs showed that many platelet aggregates were formed upon stimulation with 200 M Ca 2ϩ , whereas the permeabilized platelets incubated with 20 nM Ca 2ϩ remained unaggregated (Fig. 1A). The quantification of the platelets in the images revealed that unaggregated platelets were reduced upon Ca 2ϩ stimulation (Fig. 1B) and that the aggregates consisting of Ͼ10 platelets were drastically increased (Fig. 1C). Thus, the formation of the aggregates in the assay was confirmed morphologically.
We next examined the effects of Ca 2ϩ concentrations on the platelet aggregation. Although Ca 2ϩ at 20 and 200 nM did not induce the platelet aggregation, Ca 2ϩ at 2-200 M efficiently induced the aggregation (Fig. 2). Because it has been shown that [Ca 2ϩ ] in resting platelet cytosol is around 10 nM and that it increases to 1-10 M upon platelet activation (33), the Ca 2ϩ sensitivity of the aggregation in the assay using permeabilized platelets was similar to physiological conditions. Aggregation of Permeabilized Platelets Was Mediated by the Integrin-It has been well known that platelet aggregation is mediated by activated integrin ␣ IIb ␤ 3 (2, 3). We examined whether it was the case for the aggregation of permeabilized platelets. RGD is the integrin-binding sequence (34), which is present in both fibrinogen and vWF, and the aggregation of intact platelets has been shown to be inhibited by the addition of the RGD-containing peptide (35,36). As shown in Fig. 3, the RGDS peptide inhibited the Ca 2ϩ -induced aggregation of permeabilized platelets, whereas the control RGES peptide did not. Furthermore, when fibrinogen was omitted from the assay, the aggregation was also inhibited (data not shown). Taken together, the aggregation in the assay was mediated by the integrin.
Platelet Aggregation Was Cytosol-dependent-ATP and cytosol in the permeabilized platelets would be lost by diffusion through the pores. Without the addition of ATP, the permeabilized platelets did not aggregate upon the Ca 2ϩ stimulation (data not shown), indicating that ATP is essential for the aggregation. When the cytosol was not added exogenously into the assay, the permeabilized platelets did not aggregate upon the Ca 2ϩ stimulation in the condition where ATP and fibrino-gen were sufficiently supplemented (Fig. 4A). On the other hand, the aggregation was reconstituted by the addition of platelet cytosol in a concentration-dependent manner (Fig. 4A).
These results indicated the existence of cytosolic essential factor(s). We next tested rat brain cytosol for the reconstitution of the aggregation. The rat brain cytosol also supported the aggregation as efficiently as the human platelet cytosol (Fig. 4B).

FIG. 4. The Ca 2؉ -induced aggregation was cytosol-dependent.
A, the permeabilized platelets were first incubated for 15 min at 4°C with indicated concentrations of platelet cytosol, and the Ca 2ϩ -induced aggregation was analyzed as described under "Experimental Procedures." B, the permeabilized platelets were first incubated for 15 min at 4°C with rat brain cytosol or human platelet cytosol at 1.5 mg of proteins/ml, and the Ca 2ϩ -induced aggregation was analyzed as described under "Experimental Procedures." The data shown are the representative of four independent experiments with similar results.
These results indicated that cytosolic essential factor(s) were expressed ubiquitously.
Involvement of PKC in the Regulation of Platelet Aggregation-The cytosol dependence of the aggregation indicates that some cytosolic factors are required for the platelet aggregation. Although the identity of these factors is unknown, one important factor could be cPKC. As shown previously with intact platelets (37), GF109203X, an inhibitor of cPKCs, also affected the Ca 2ϩ -induced aggregation in our semi-intact assay in a concentration-dependent manner (Fig. 5).
To examine the involvement of PKC directly, we first prepared PKC-depleted platelet cytosol. As shown in Fig. 6A, PKC␣ was completely depleted from the platelet cytosol with the anti-PKC␣ antibody-coated beads while PKC␣ stayed in the cytosol after the same procedure with control IgG-coated beads. Among other cPKCs, PKC␤I and PKC␤II were detected in platelets, whereas PKC␥, a neuronal specific cPKC, was not (data not shown). By the immunodepletion, PKC␤ was also completely depleted because of the cross-reactivity of the antibody (Fig. 6A). When we used lower amounts of the anti-PKC␣ antibody for the immunodepletion, PKC␤ was completely depleted while PKC␣ still remained in the cytosol (data not shown), suggesting that platelet cytosol contained more PKC␣ than PKC␤. As expected, although PKC␦ and PKC, both of which are classified as novel PKCs, were detected in platelets, they were not affected by the immunodepletion procedure either with anti-PKC␣ antibody or control IgG (Fig. 6A).
The PKC-depleted cytosol lost the aggregation activity, whereas the Ca 2ϩ -induced platelet aggregation was efficiently reconstituted with the cytosol treated with the control IgG (Fig.  6C). When PKC␣ purified from rat brain (Fig. 6B) (19) was supplemented to the PKC-depleted cytosol, the aggregation activity was recovered (Fig. 6C), indicating that cPKC, possibly PKC␣, is an essential cytosolic factor for the platelet aggregation. We next examined whether PKC␣ is a sufficient cytosolic factor for the aggregation. As shown in Fig. 7, purified PKC␣ (50 nM) alone was not sufficient to support the Ca 2ϩ -induced platelet aggregation. On the other hand, platelet cytosol at 0.6 mg of proteins/ml, which contained 15 nM PKC␣ determined by Western blot with the anti-PKC␣ antibody using purified PKC␣ as a control (data not shown), efficiently induced platelet aggregation (Fig. 7). Thus, PKC␣ is not a sufficient cytosolic factor for platelet aggregation. Furthermore, the addition of purified PKC␣ to the low concentration of cytosol (0.6 mg of proteins/ml) strongly enhanced the platelet aggregation, suggesting that PKC␣ is a limiting factor for the Ca 2ϩ -induced platelet aggregation. Thus, PKC␣ is an essential but not sufficient cytosolic factor for the Ca 2ϩ -induced platelet aggregation. DISCUSSION Here we established an aggregation assay system using permeabilized platelets, and using this assay, we first directly demonstrated that PKC is an essential but not sufficient factor in the cytosol for platelet aggregation.
Because platelets lack the protein-producing activity, it is difficult to apply molecular biology for investigating the molecular mechanism of aggregation and granule secretion inside activated platelets. Therefore, the research in these fields has been performed mainly pharmacologically. To overcome this difficulty, much effort has been made to establish semi-intact assay systems using permeabilized platelets. In the research of platelet granule secretion, several semi-intact assays have been established (4 -8, 19, 20). However, for platelet aggregation, only a few semi-intact aggregation assays have been established (8), and as far as we know, no stable assays with cytosol dependence have been established. at 4°C for 6 h, and the supernatant after removing the beads was examined by immunoblotting using another PKC␣-specific antibody (Sigma) and anti-pan-PKC␤, anti-PKC␦, and anti-PKC antibodies as described under "Experimental Procedures." B, PKC␣ was purified from rat brain cytosol and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis stained by Coomassie Blue as described under "Experimental Procedures." C, the permeabilized platelets were first incubated for 30 min at 4°C with 2 mg of proteins/ml of the treated platelet cytosol and/or purified PKC␣ at 50 nM as indicated in the figure, and the Ca 2ϩ -induced aggregation was then analyzed as described under "Experimental Procedures." The data shown are the representative of four independent experiments with similar results.
We have previously established an aggregation assay system with SLO-permeabilized platelets and demonstrated that small GTPase Rho plays an important role in thrombin-induced aggregation (23). However, the assay did not demonstrate cytosol dependence since a low concentration of SLO (0.1 g/ml) was used for the permeabilization and the aggregation was induced without adding exogenous cytosol (23). Here, we have established another semi-intact aggregation assay by modifying the previous one (23). The aggregation of the permeabilized platelets in our in vitro assay appears physiological since the time course, the Ca 2ϩ sensitivity and the involvement of the integrin are similar to those of intact platelets. Since the cytosol was extensively depleted from the permeabilized platelets in our semi-intact aggregation assay using a higher concentration of SLO than that used previously (23), permeabilized platelets did not aggregate in response to calcium stimulation without adding exogenous cytosol. This cytosol dependence would widen the application of the assay to investigate the molecular mechanism of platelet aggregation.
The cytosol dependence also indicated the existence of essential cytosolic factor(s) for aggregation. A cytosolic protein PKC has been shown to play an important role in platelet aggregation by pharmacological experiments using cell-permeable small compounds of inhibitors and stimulators such as phorbol esters (16,17). However, the results obtained from such experiments appear somewhat indirect because the specificity of inhibitors is not absolutely strict and important signaling molecules containing the phorbol ester-binding C1 domain other than PKC have been recently identified such as Ras-guanyl nucleotide-releasing protein (11,12) and chimerin (13,14). Munc13-1 present in the presynapse also contains the C1 domain (38), and it has very recently been demonstrated that the effect of phorbol ester in the neurotransmitter release is through Munc13-1 (39). Thus, at the moment, it is ambiguous whether phorbol esters exert their functions through PKC or other non-PKC-signaling molecules. Therefore, re-evaluation and direct demonstration are required in various cell functions where PKCs have been suggested to be involved (15,43).
Using the semi-intact aggregation assay, we have directly demonstrated the involvement of PKC␣ in the Ca 2ϩ -induced platelet aggregation. First, an inhibitor of conventional PKC affected the aggregation. Second, immunodepletion of PKC␣ and PKC␤ from the cytosol abolished the Ca 2ϩ -induced aggregation. Third, the aggregation-supporting activity of PKC␣/␤depleted cytosol was rescued by supplementation of purified PKC␣. Supplementation of PKC␣ alone to the PKC␣/␤-depleted cytosol was enough to reconstitute the aggregation, suggesting that PKC␣ but not PKC␤ is the essential factor or otherwise that the activity of cPKC, namely PKC␣ or PKC␤, is essential. Because PKC␣ and PKC␤ show similar substrate specificity in vitro (40), we cannot exclude a possibility that added PKC␣ covered the lack of PKC␤ activity in the assay. Although cPKC is an essential factor for the aggregation, it is not a sufficient cytosolic factor since the addition of purified PKC␣ alone without exogenous cytosol did not support the aggregation. Furthermore, platelet cytosol containing less PKC supported the aggregation efficiently, and purified PKC␣ strongly enhanced the aggregation in the presence of low concentration of cytosol, suggesting that PKC␣ is a limiting factor in the cytosol for the aggregation and that additional factors besides PKC would also be required for the aggregation.
PKC␣ and PKC␤ are known to be activated by Ca 2ϩ , diacylglycerol, and phosphatidylserine (9, 10). Although we did not add these stimulators besides Ca 2ϩ , the purified PKC␣ added to the assay was indeed active since the purified PKC␣ phosphorylated a PKC substrate efficiently in the similar assay condition used here (19). We speculate that the components inside the platelets, possibly including phosphatidylserine, help support the activity of PKC␣ (19). Furthermore, because PKC increases the intracellular Ca 2ϩ concentration by modulating Ca 2ϩ channels in the plasma membrane in neurons (41,42), it remains unclear whether PKC acts upstream and/or downstream of increased Ca 2ϩ . Because we used Ca 2ϩ as a stimulus, we could safely say that PKC␣ plays an important role at least at the downstream of increased Ca 2ϩ . Further investigation is required for elucidation of how PKC␣ activates the integrin ␣ IIb ␤ 3 and induces platelet aggregation. The assay established here will be a powerful tool for future experiments aimed at elucidating these mechanisms. FIG. 7. Purified PKC␣ alone without addition of cytosol did not support the Ca 2؉ -induced aggregation. Permeabilized platelets were first incubated for 15 min at 4°C with indicated concentrations of platelet cytosol and/or purified PKC␣, and the Ca 2ϩ -induced aggregation was analyzed as described under "Experimental Procedures." It is noted that 1 mg of proteins/ml platelet cytosol contains 25 nM PKC. The data shown are the representative of four independent experiments with similar results.