Identification of Protein Kinase Cα as an Essential, but Not Sufficient, Cytosolic Factor for Ca2+-induced α- and Dense-core Granule Secretion in Platelets*

Upon activation, platelets release many active substances. Here, we have analyzed the mechanism governing Ca2+-induced secretion of von Willebrand factor stored in α-granules and 5-hydroxytryptamine in dense-core granules in permeabilized human platelets. Both secretions were dependent on ATP and cytosol. An essential factor for both granule secretions was purified from rat brain cytosol and identified to be protein kinase Cα (PKCα) by partial amino acid sequencing. Purified PKCα efficiently stimulated both secretions in the presence of cytosol, whereas PKCα alone did not support the secretion of either type of granules, suggesting that PKCα is not a sufficient factor. Finally, in human platelet cytosol fractionated by a gel filtration column, the stimulatory activity for dense-core granule secretion paralleled with the concentration of PKC, suggesting that PKC could also be such a stimulatory factor in platelet cytosol. Thus, we identified PKCα as an essential, but not sufficient, cytosolic factor for the Ca2+-induced secretions of both α- and dense-core granules in platelets.

droxytriptamine (5-HT) 1 and ␣-granules contain protein factors such as von Willebrand factors (vWF). Despite the biological significance of the secretions in platelet function (7), the molecular mechanisms governing the secretions remain unclear. N-Ethylmaleimide-sensitive factor and ␣-SNAP, which are required for priming SNARE proteins in other types of cells (8), have also been shown to play important roles in granule secretions in platelets (9,10). Furthermore, syntaxin 2 and SNAP 23, which are members of the SNARE family, are essential for the dense-core granule secretion (10), and syntaxin 4 and SNAP 23 function in the ␣-granule secretion (11). Recently, we have identified small GTPase Rab4 as an essential regulator of ␣-granule secretion, but not of dense-core granule secretion, using the assay system used in this investigation (12).
In addition to membrane-associated factors such as SNARE proteins and Rab small GTPases, cytosolic factors also play critical roles in intracellular membrane docking/fusion (13). For example, Rab5-regulated membrane docking/fusion of early endosomes in vitro is cytosol-dependent and the ratelimiting factor in cytosol was demonstrated to be a protein complex composed of a Rab5 effector, Rabaptin-5, and a Rab5 GDP/GTP exchange factor, Rabex-5 (14). The Ca 2ϩ -induced norepinephrine release from neuroendocrine cells also requires cytosolic factors, including the phosphatidylinositol transfer protein (15), phosphatidylinositol-4-phosphate 5-kinase (16) and Ca 2ϩ -dependent activator protein for secretion (CAPS) (17,18), a mammalian homologue of Caenorhabditis elegans unc31. However, cytosolic factors involved in the platelet secretion have not been identified.
Protein kinase C (PKC) family members are important signaling molecules (19,20). Conventional PKCs have regulatory Ca 2ϩ -and phorbol ester-binding domains (19,20). The function of PKC as a signaling molecule has been analyzed mainly pharmacologically using small cell-permeable compounds of inhibitors and stimulators such as phorbol esters. Recently, Munc13-1, which contains a phorbol ester-binding domain, has been demonstrated to play an important role in membrane docking/fusion in exocytosis in synapses (21,22). Specifically, we do not understand whether the effects of phorbol ester are mediated by Munc13 and/or PKC. Therefore, the interpretation of the earlier reports with experiments using phorbol ester has become complicated especially concerning vesicle transport. Furthermore, since PKC increases intracellular Ca 2ϩ concen-tration through modulation of Ca 2ϩ channels in the plasma membrane in neurons (23,24), it remains a question whether PKC acts upstream and/or downstream of increase of Ca 2ϩ .
In platelets, PKC inhibitors have been shown to inhibit platelet aggregation and granule release, and phorbol 12-myristate 13-acetate, a phorbol ester, have been shown to stimulate both functions (25)(26)(27). Therefore, PKC has been considered to be an important signaling molecule in platelet activation (7). However, with reasons described above, further characterization of PKC function in platelet activation is required especially in the granule secretions. Here, we have demonstrated that the Ca 2ϩ -induced secretions of ␣and dense-core granules in platelets are ATP-and cytosol-dependent and that PKC␣ is an essential, but not sufficient, cytosolic factor for both secretions.

EXPERIMENTAL PROCEDURES
Assay for Secretion of ␣and Dense-core Granules-The standard assay method was essentially as described previously (12). Briefly, freshly obtained washed platelets (5 ϫ 10 7 platelets per assay, counted with Coulter counter) were incubated with [ 3 H]5-HT (Amersham Pharmacia Biotech) to allow uptake into dense-core granules. After washing the platelets, the platelet plasma membrane was permeabilized in 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, the calculated free calcium ion concentration was ϳ20 nM (28)) containing 4 mg/ml bovine serum albumin and 0.6 g/ml SLO as described (12). Approximately, more than 80% of the cytosolic lactate dehydrogenase was recovered in the media by the procedure, measured with lactate dehydrogenase-cytotoxic test kit (Wako Chemical, Osaka, Japan). Usually, the permeabilized platelets per assay contained ϳ20,000 cpm of [ 3 H]5-HT. Unless otherwise specified, the platelets were incubated in Buffer A containing 4 mg/ml bovine serum albumin with an ATP regeneration system containing 8 mM creatine phosphate, 50 g/ml creatine phosphokinase, and 1 mM ATP, cytosol, and/or others at 4°C for 60 min followed by further incubation at 30°C for 5 min. Finally, the platelets were stimulated with 20 M Ca 2ϩ (28) by addition of CaCl 2 at 30°C for 1 min. Then, after removing the platelets by centrifugation, aliquots of the supernatant containing any secreted [ 3 H]5-HT and vWF were measured for the presence of [ 3 H]5-HT by a liquid scintillation counter (Beckman) and of vWF by Western blot analysis after immunoprecipitation with anti-vWF antibody (Sigma) followed by quantification using densitometric detection and NIH Image software. Secretion levels of vWF and [ 3 H]5-HT were expressed as percentages of the total vWF and [ 3 H]5-HT, respectively, above those (3-8% of total vWF and 5-10% of total [ 3 H]5-HT) released before the final incubation. Unless otherwise specified, they were expressed as mean Ϯ S.E. of more than five independent experiments with similar results. For vWF secretion, representative Western blots were also shown in some figures.
Preparation of Cytosols-Fifty rats were anesthetized and sacrificed. The rat brains were homogenized in Buffer A using a Potter-type blender and centrifuged at 100,000 ϫ g for 30 min at 4°C. The supernatant was used as rat brain cytosol. The cytosols of Jurkat cells and HeLa cells were prepared in a similar way. The cytosol of human platelets isolated from blood of healthy volunteers was prepared in a similar way except using sonication instead of a Potter-type blender.
Purification and Identification of the Most Potent Factor in Cytosol for ␣and Dense-core Granule Secretion-The purification procedures were carried out at 4°C and column chromatography was performed using the fast protein liquid chromatography system (Amersham Phar-macia Biotech). The assay during the purification was carried out with rat brain cytosol at 0.5 mg of protein/ml, which supported 5-10% secretions over the controls (Fig. 7). In each purification procedure (Table I), the activity for both ␣and dense-core granule secretions was recovered in a single peak which supported both ␣and dense-core granule secretions. The rat brain cytosol was fractionated by ammonium sulfate precipitation. The precipitate was dialyzed against Buffer A without KCl. The peak was recovered in a 35-55% precipitate. The fraction was loaded onto a Mono-Q HR 10/10 column equilibrated with Buffer A without KCl, and eluted with a linear gradient of KCl. The peak of the activity was recovered at around 100 mM KCl. The collected fractions were dialyzed with Buffer A (pH 6.5) without KCl and loaded onto a Mono-S HR 5/5 equilibrated with buffer A (pH 6.5) without KCl, and eluted with a linear gradient of KCl. The peak of the activity was recovered at KCl concentration around 150 mM. The collected peak fractions were then concentrated with Centricon 10 (Amicon) and separated by Superdex 200 gel filtration column in Buffer A. The peak of the activity was recovered between the protein markers (Bio-Rad) of 44 and 158 kDa. The collected peak fractions were added with KCl to adjust the KCl concentration at 2 M. Then the sample was loaded onto a hydrophobic-interaction column, Resource PHE, equilibrated with Buffer A containing 2 M KCl, and eluted with a decreasing linear gradient of KCl. The peak of the activity was recovered at a KCl concentration around 250 mM. For the characterization of PKC␣, the samples purified by the last column Resource PHE were used as purified PKC␣. By using this procedure, a protein band at 80 kDa was purified to homogeneity in a Coomassie Blue-stained SDS-PAGE gel and the intensity was paralleled with the secretion activity. The band was excised and analyzed for partial amino acid sequencing by Edman's method (29) performed by Apro Science Co. (Tokushima, Japan). For the characterization of the human platelet cytosol, the cytosol was separated with the Superdex 200 gel filtration column in the same way as described above.
Assay for PKC Activation in the Platelets-The cytosolic domain of junctional adhesion molecule (JAM) has been demonstrated to be phosphorylated by PKC (30). The glutathione S-transferase (GST)-tagged partial cytosolic domain of JAM (JAM290) (20 nM) containing the PKC phosphorylation site (30) was incubated for 1 min at 30°C with 1 mM [␥-32 P]ATP (300 cpm/pmol), 1 M okadaic acid, and 20 M Ca 2ϩ in the presence or absence of permeabilized platelets, purified PKC␣, and phorbol 12-myristate 13-acetate. Then, the GST-JAM290 was collected with glutathione-Sepharose beads (Amersham Pharmacia Biotech). The phosphorylation levels of the JAM290 were analyzed by autoradiography after SDS-PAGE.
Antibodies, Materials, and Others-Anti-vWF rabbit polyclonal antibody was purchased from Sigma, anti-PKC␣ rabbit polyclonal antibody used for Western blot analysis was from Santa Cruz, anti-PKC␣ mouse monoclonal antibody used for immunodepletion was from Transduction Laboratories, and a control mouse IgG from Zymed Laboratories Inc. Horseradish peroxidase-labeled anti-mouse and anti-rabbit IgG monoclonal antibodies were from Amersham Pharmacia Biotech, which were used as secondary antibodies for Western blot analysis visualized by the enhanced chemiluminescence method (Amersham Pharmacia Biotech). Unless otherwise specified, all the chemicals were purchased from Sigma, except Gö6850 (bisindolylmaleimide 1) which was from Calbiochem Co. and streptolysin-O (SLO) from Dr. Bhakdi, Mainz University, Mainz, Germany. Protein concentrations were determined by Bradford's method (31) (Bio-Rad) or densitometric scanning of the Coomassie Blue-stained band of SDS-PAGE, using bovine serum albumin as a standard.

Ca 2ϩ -induced Secretions of ␣and Dense-core Granules in SLO-permeabilized Platelets
Were ATP-and Cytosol-dependent-We established an in vitro assay system using SLO-permeabilized platelets by monitoring secreted vWF stored in ␣-granules and [ 3 H]5-HT in dense-core granules (12). It is well known that agonists drive the granule secretions by increasing intracellular calcium ion concentrations in platelets (32). Since the calcium ion concentration inside the platelets would be the same as that of the outside following permeabilization, we used calcium chloride as a stimulus. Therefore, we analyzed the secretion mechanism triggered by increased Ca 2ϩ . In the assay, secretions of ␣and dense-core granules appeared physiological, since the characterization of the secretions revealed similar time course and Ca 2ϩ sensitivity to those in intact platelets (12).
Without addition of ATP, Ca 2ϩ did not induce ␣-granule secretion (vWF) or dense-core granule secretion ([ 3 H]5-HT) (data not shown) (33,34), indicating that ATP is essential for both secretions. The permeabilized platelets were extensively depleted of cytosol since more than 80% of lactate dehydrogenase, a cytosolic protein, was recovered in the media following permeabilization (data not shown). Under this condition, Ca 2ϩ did not induce ␣- (Fig. 1A) or dense-core (Fig. 1B) granule secretion without addition of exogenous cytosol, even in the presence of ATP. On the other hand, the Ca 2ϩ -induced secretions of both types of granules were efficiently reconstituted by addition of both ATP and rat brain cytosol (Fig. 1, A and B). The efficient secretions were also confirmed morphologically with electron microscopy, since most of the granules in the permeabilized platelets disappeared upon Ca 2ϩ stimulation in the presence of ATP and cytosol (data not shown), as shown previously (35). These results indicated that both ATP and cytosolic factor(s) are essential for the Ca 2ϩ -induced secretions in platelets.
The cytosol dependence indicated the presence of essential cytosolic factor(s) for the secretions. Cytosols of human fibroblast cell line, HeLa cells, and human T-cell cell line, Jurkat cells, also supported the Ca 2ϩ -induced secretions of both granules (data not shown), suggesting that the essential cytosolic factor(s) are expressed ubiquitously among various species and tissues.
Purification and Identification of PKC␣ as the Essential Cytosolic Factor in Rat Brain Cytosol-The cytosol dependence prompted us to identify the essential factor(s) in the cytosol. The Ca 2ϩ -induced secretions of both vWF and [ 3 H]5-HT were almost maximal in the presence of rat brain cytosol at 2-5 mg of protein/ml (Figs. 6 and 7). In order to maintain minimal secretory activity, the assays during the purification procedure were performed in the presence of rat brain cytosol at 0.5 mg of protein/ml, which supported ϳ5-10% of both secretions induced by Ca 2ϩ (Figs. 6 and 7).
We first separated rat brain cytosol by Superdex 200 gel filtration column ( Fig. 2A). We detected a potent peak of secretion supporting activity between markers 44 and 158 kDa (Bio-Rad) ( Fig. 2A). Notably, the peaks for ␣and dense-core granule secretions were almost completely overlapped, suggesting that the factor is common for both secretions.
Then, we scaled up the purification procedure. The purification was performed by ammonium sulfate precipitation and sequential column chromatography with Mono-Q, Mono-S, Superdex 200, and Resource PHE columns as described under "Experimental Procedures." During each procedure (Table I), the most potent activity for either ␣or dense-core granule secretion was recovered in a single peak. Then, the peaks of both ␣and dense-core granule secretions were almost completely overlapped in each column chromatography (data not shown). In Superdex 200 gel filtration column chromatography, the peak fraction emerged at a position between 44 and 158 kDa (data not shown), similar to the results in experiments of the direct separation of the cytosol by Superdex 200 (Fig. 2A).  2. Purification of an essential factor for ␣and dense-core granule secretions in the permeabilized platelets. A, rat brain cytosol (10 mg of protein/ml) was separated by Superdex 200 gel filtration column chromatography and each 0.4-ml fraction was collected. The permeabilized platelets were incubated with the fractions in the presence of 0.5 mg of protein/ml rat brain cytosol and the ATP regeneration system for 60 min at 4°C. The platelets were then stimulated with 20 M Ca 2ϩ for 1 min at 30°C and released vWF (open columns) and [ 3 H]5-HT (closed columns) were measured as described under "Experimental Procedures." The data shown were representative of three independent experiments with similar results. The inset immunoblot is one of the typical results for measuring the released vWF. B, the most potent activity for both secretions was purified from brain cytosol of 50 rats by sequential column chromatography as shown in Table I. Collected active fractions after each column chromatography for both secretions were analyzed with a Coomassie Blue-stained SDS-PAGE gel.
As shown in Fig. 2B, a band with apparent molecular mass at ϳ80 kDa was purified to homogeneity after the fifth step of Superdex 200 column chromatography (Table I). Furthermore, the active fractions were collected and fractionated by a hydrophobic interaction column, Resource PHE. The secretion-supporting activities for both granules again paralleled the 80-kDa band (data not shown). The molecular weight was consistent with the results in the Superdex 200 gel filtration column chromatography for direct separation of rat brain cytosol ( Fig.  2A). Taken together, the 80-kDa protein was most likely the essential common factor for both secretions.
The partial amino acid sequences of the purified 80-kDa protein excised from the SDS-PAGE gel after the Resource PHE column chromatography (Fig. 2B, lane 6) were determined by the Edman's method (29). The determined sequence IARFF, LIPMDPNGL, and IRSTLNPQW completely corresponded to the sequences of rat PKC␣ 38 -44, 182-191, and 214 -222, respectively (36). Since all three sequences corresponded to parts of the rat PKC␣ sequence, the 80-kDa protein was determined to be PKC␣ which was a conventional type of PKC composed of 672 amino acids with a calculated molecular mass of 76.6 kDa (36).
Accordingly, Gö6850, a PKC inhibitor, concentration dependently inhibited the Ca 2ϩ -induced secretions of both types of granules in the permeabilized platelets (Fig. 3, A and B). Gö6850 at 10 M almost completely inhibited both secretions (Fig. 3, A and B). Furthermore, when PKC␣ was immunodepleted from the purified samples with anti-PKC␣ monoclonal antibody-coated beads (Fig. 4A), the immunodepleted samples lost the secretion stimulatory activity for both types of granules (Fig. 4, B and C). Thus, PKC␣ was identified to be an essential factor in rat brain cytosol for the Ca 2ϩ -induced secretions of both ␣and dense-core granules in platelets.
We then examined whether the added PKC␣ was activated in the assay. The PKC activity in the assay was monitored by phosphorylation levels of GST-JAM290, which contained a PKCdependent phosphorylation site (30) of JAM, a cell-cell adhesion molecule (37,38). GST-JAM290 was hardly phosphorylated by the permeabilized platelets alone without the purified PKC␣ (Fig. 5, lane 3). On the other hand, the purified PKC␣ alone weakly phosphorylated GST-JAM290 (Fig. 5, lane 4). The phosphorylation level of GST-JAM290 was strongly enhanced in the presence of both purified PKC␣ and permeabilized platelets (Fig. 5, lane 5) to the similar level obtained in the experiments with the purified PKC␣, permeabilized platelets, and phorbol 12-myristate 13-acetate, a PKC activator (Fig. 5, lane  1). These results suggested that the exogenously added PKC␣ was indeed activated in the assay by Ca 2ϩ and the permeabilized platelets. Since PKC␣ is activated by Ca 2ϩ in the presence of activators such as phosphatidylserine and diacylglycerol (19,20), the permeabilized platelet membrane could substitute for the activators.
In another set of experiments, 10 M Gö6850 completely inhibited Ca 2ϩ -induced phosphorylation of GST-JAM290 in the presence of permeabilized platelets (data not shown). Since Gö6850 at 10 M also inhibited platelet secretion as shown in Fig. 3, catalytic activity of PKC could be required for secretion.
PKC␣ Alone Did Not Support the Ca 2ϩ -induced Secretions-We characterized the purified PKC␣ in the Ca 2ϩ -induced secretions in permeabilized platelets. First, we examined the amount of PKC remaining after permeabilization by Western blot analysis. Since we detected very little PKC associated with the permeabilized platelets after the procedure (data not shown), most of the PKC in the platelets appeared to be lost by diffusion through holes in the plasma membrane. Second, we carefully determined the PKC concentration in rat brain cy- tosol by Western blot analysis using purified PKC␣ as a control. The PKC concentration in the rat brain cytosol preparation used in the assay (10 mg of protein/ml) was 250 nM.
Then, we examined whether PKC␣ alone was enough to support the Ca 2ϩ -induced secretions. The Ca 2ϩ -induced secretions of both types of granules were examined with various concentrations of the purified PKC␣ and rat brain cytosol Fig.  6). The activities of the cytosol and the purified PKC␣ were compared based on concentrations of contained PKC (Fig. 6, B and C). Without addition of any cytosol or purified PKC␣ (see the points of PKC concentration at 0 nM in Fig. 6, B and C), Ca 2ϩ did not induce either vWF or [ 3 H]5-HT secretion over the levels obtained without Ca 2ϩ stimulation (Fig. 6, B and C). Under these conditions, rat brain cytosol efficiently supported the Ca 2ϩ -induced secretions of both granules in a concentrationdependent manner (Fig. 6, B and C). The cytosol was saturated at 4 mg of protein/ml (containing 100 nM PKC) for the vWF secretion and at 2 mg of protein/ml (containing 50 nM PKC) for the [ 3 H]5-HT secretion (Fig. 6, B and C). The concentrations of rat brain cytosol which supported half-maximal ␣and densecore granule secretion were 2.0 mg of protein/ml (containing 50 nM PKC) (Fig. 6B) and 1.2 mg of protein/ml (containing 30 nM PKC), respectively (Fig. 6C). On the other hand, addition of purified PKC␣ alone at comparable concentrations hardly stimulated either secretion (Fig. 6, B and C), indicating that PKC␣ is not a sufficient factor for the Ca 2ϩ -induced secretions in platelets.
Next, in another set of experiments, we examined the effects of the purified PKC␣ on the Ca 2ϩ -induced secretions of ␣- (Fig.  7, A and B) and dense-core granules (Fig. 7C) in the presence of various concentrations of rat brain cytosol. Rat brain cytosol alone (closed circles in Fig. 7, B and C) supported both secretions concentration dependently, as shown in Fig. 6. In the experiments with the cytosol plus 40 nM purified PKC␣ (open circles in Fig. 7, B and C), the purified PKC␣ strongly enhanced both secretions in the assay with low concentrations of the cytosol, indicating that PKC␣ is indeed a limiting factor for the secretions. Similar levels of the secretions with high concentrations of the cytosol in the absence or presence of 40 nM purified PKC␣ (Fig. 7, B and C) could reflect the saturated effects of PKC␣ in the assay.
From another point of view, although 40 nM purified PKC␣ alone did not support the Ca 2ϩ -induced secretions, addition of low concentrations (0.5-1 mg of protein/ml containing 12.5-25 nM PKC, respectively) of rat brain cytosol in addition to 40 nM purified PKC␣ strikingly enhanced the Ca 2ϩ -induced secretions of ␣- (Fig. 7B) and dense-core granules (Fig. 7C). These results suggested that other essential factor(s) in the cytosol are involved in the regulation of the secretions. If we could assume that factors other than PKC would become rate-limiting by addition of 40 nM PKC␣, the half-maximal concentration of rat brain cytosol for such factor(s) would be 0.5 mg of protein/ml for ␣-granule secretion and 0.7 mg of protein/ml for dense-core granule secretion (Fig. 7, B and C).
PKC Could be a Core Regulatory Factor Also in Human Platelet Cytosol for Dense-core Granule Secretion-In the last set of experiments, we analyzed whether PKC plays an important role in platelet secretion also in the human platelet cytosol. Since we detected large amounts of vWF in the human platelet cytosol, which was probably from broken ␣-granules during preparation of the platelet cytosol, we analyzed the effect of the platelet cytosol only on dense-core granule secretion ([ 3 H]5-HT). After fractionation of the platelet cytosol by Superdex 200 gel filtration column, the fractions were examined for the presence of PKC by Western blot and the activity for the Ca 2ϩ -induced secretion of dense-core granules. As shown in Fig. 8, we detected a peak activity for the dense-core granule secretion between 44 and 158 kDa marker, which paralleled with the concentrations of PKC. The weaker activity than that of rat brain cytosol ( Fig. 2A) could be due to limitation of the PKC concentrations brought into the assays, since the secretion assay for the peak fraction (fraction 25 in Fig. 8) was performed in the presence of 27.5 nM PKC (15 nM from the fraction plus 12.5 nM from rat brain cytosol added to the assay). The parallel emergence of the secretion supporting activity with PKC suggested that PKC is a core regulatory component also in platelet cytosol for the Ca 2ϩ -induced secretion of densecore granules. DISCUSSION We have established an assay system to analyze Ca 2ϩ -induced secretions of ␣and dense-core granules in SLO-permeabilized platelets (12). In the system, both secretions were ATP-and cytosol-dependent, indicating the presence of essential cytosolic factor(s). We purified an essential factor for both secretions in rat brain cytosol and identified it to be PKC␣. Although the purified PKC␣ stimulated both secretions in the presence of exogenously added cytosol, PKC␣ alone without minimal addition of cytosol did not support either secretion, indicating that PKC␣ is not a sufficient factor. We also showed that in human platelet cytosol PKC␣ could be such a factor. Thus, we have identified PKC␣ as an essential, but not sufficient, common factor in cytosol for both ␣and dense-core granule secretions in platelets.
PKC has been shown to play important roles in platelet activation mainly by pharmacological experiments with cellpermeable small compounds of inhibitors and stimulators such as phorbol esters (27). However, since phorbol ester-binding proteins, besides PKCs, such as Munc13-1 have been identified and demonstrated to be essential for vesicle fusion (21,22), re-evaluation of the effects of phorbol esters in the regulated exocytosis has been proposed (39,40). Recently, using the PC12 crackled-cell assay, Chen et al. (40) have identified the norepinephrine release-stimulatory factors in the EGTA extract of rat brain membrane fraction to be PKC␣ and calmodulin. They have shown that PKC␣ and calmodulin are sufficient factors since either of the factors alone stimulated the secretion (40). Then, they have shown that their activities were additive and that the contributions of PKC␣ and calmodulin were 33-44% and 13-22%, respectively, of the total activity of the cytosol (40). In the case of platelets, we have here demonstrated that PKC␣ is an essential factor in cytosol. However, PKC␣ alone supported very little secretion of either ␣or dense-core granules (Fig. 6, B and C), indicating that PKC␣ is not a sufficient factor in platelet secretions. The difference from the results of Chen et al. (40) could be due to a different mechanism of regulation used by platelets and PC12 cells, or due to different assay conditions. PKC has been demonstrated to increase intracellular Ca 2ϩ concentration through modulation of Ca 2ϩ channels in the plasma membrane in the case of neuronal cells (23,24). Therefore, another question remains: whether PKC acts downstream and/or upstream of increased concentration of Ca 2ϩ . Here, we could safely say that PKC␣ plays a critical role at least downstream of the increased Ca 2ϩ in the granule secretions in platelets, since Ca 2ϩ itself was the stimulus in our assay.
The function of PKC␣ is presumed to phosphorylate its substrate proteins involved in the exocytosis machinery. Recently, syntaxin 4, a SNARE protein, has been shown to be phosphorylated by PKC upon thrombin stimulation in platelets (41). Since syntaxin 4 has been implicated in ␣-granule secretion (11), a putative target molecule of PKC␣ could be syntaxin 4. Interestingly, syntaxin 4 has very recently been shown to interact directly with Rab4 in pancreatic ␤ cells (42), which we have shown to be an essential regulator for ␣-granule secretion in platelets (12). The function of syntaxin 4 could be finely regulated by both PKC␣ and Rab4 in ␣-granule secretion in platelets. Furthermore, Munc18-1, a syntaxin-binding protein essential for neurotransmitter release in the synapses (43)(44)(45), has also been shown to be phosphorylated by PKC (46). Since a homologue of Munc18-1 is present in platelets (47), it could also be a substrate. Further investigation is required to elucidate how PKC␣ regulates the Ca 2ϩ -induced secretions of both ␣and dense-core granules in platelets.
Besides membrane-associated components such as SNARE proteins and RabGTPases, cytosolic factors have also been demonstrated to play critical roles in membrane docking/fusion (13,15). However, only a few proteins have been identified so far in exocytosis. Ca 2ϩ -induced norepinephrine release has been analyzed in vitro in the crackled-cell assay using PC12 FIG. 8. PKC␣ paralleled the activity for the Ca 2؉ -induced secretion of dense-core granule in gel filtration column chromatography of human platelet cytosol. Human platelet cytosol (10 mg of protein/ml) was fractionated by Superdex 200 gel filtration column chromatography and each 0.6-ml fraction was collected. The permeabilized platelets were incubated with the fractions in the presence of 0.5 mg of protein/ml rat brain cytosol and the ATP regeneration system for 60 min at 4°C. The platelets were then stimulated with 20 M Ca 2ϩ for 1 min at 30°C and released [ 3 H]5-HT (closed columns) were measured as described under "Experimental Procedures." The presence of PKC in each fraction was examined by Western blot analysis. The data shown are representative of three independent experiments with similar results.
cells (17). Using the assay, Hay et al. (15) detected three peaks of secretion stimulating activities in rat brain cytosol in gel filtration column chromatography and two of these were identified to be phosphatidylinositol transfer protein and phosphatidylinositol-4-phosphate 5-kinase (16). Furthermore, they have found the Ca 2ϩ -dependent activator protein for secretion, a mammalian homologue of C. elegance unc31, as the cytosolic stimulatory factor in rat brain cytosol (17,18).
However, we detected only one peak of the activity obtained with PKC␣ ( Fig. 2A), although we have shown that other cytosolic factor(s) are also required for secretions. In the assay used in the purification procedure, rat brain cytosol at 0.5 mg of protein/ml was added since the activity would not be detected if the essential factors were multiple and separated by column chromatography. With the results shown in Fig. 7, B and C, we speculated that the concentrations of rat brain cytosol which supported half-maximal secretions of both granules were 1.2-2.0 mg of protein/ml for PKC␣, whereas it was 0.5-0.7 mg of protein/ml for other factor(s). Therefore, the rat brain cytosol at 0.5 mg of protein/ml might be limiting for PKC while it could contain relatively high concentrations of other factors, which could be essential, but not rate-limiting. This would not allow us to detect the stimulatory activity caused by the putative factors other than PKC␣. Identification of other cytosolic factors is absolutely required for elucidation of the regulatory mechanism taking place in the cytosol for the granule secretions in platelets.