Possible involvement of cytoskeleton in collagen-stimulated activation of phospholipases in human platelets.

The action of phospholipases A2 and C in the course of collagen-stimulated platelet activation and the effect of cytochalasins on the responses were studied. Stimulation of human platelets with collagen was accompanied by aggregation, Ca2+ mobilization, inositol phosphate formation, and arachidonic acid release. However, in the presence of a cyclooxygenase inhibitor or a thromboxane A2 (TXA2) receptor antagonist, collagen induced only weak arachidonic acid release and weak inositol phosphate formation. The TXA2 mimetic agonist U46619 induced all the responses except for arachidonic acid release, which was induced by synergistic action of collagen and U46619. The result that U46619 did not induce arachidonic acid release despite the activation of phospholipase C suggested that arachidonic acid was not released via phospholipase C but by phospholipase A2. These findings suggested that collagen initially induced weak activation of phospholipases A2 and C and that further activation of phospholipase C as well as Ca2+ mobilization and aggregation were induced by TXA2, whereas further activation of phospholipase A2 required the synergistic action of collagen and TXA2. Platelets pretreated with cytochalasins did not respond to collagen. Further analysis revealed that the initial activation of phospholipases A2 and C was specifically inhibited by cytochalasins, but the responses induced by U46619 or a synergistic action of collagen and U46619 were not inhibited. Therefore, we proposed that interaction of collagen receptor with actin filaments might have some roles in the collagen-induced initial activation of phospholipases.


Possible Involvement of Cytoskeleton in Collagen-stimulated Activation of Phospholipases in Human Platelets*
Tohru Nakano, Kohji Hanasaki, and Hitoshi AritaS From the Shionogi Research Laboratories, Shionogi & Co., Ltd., Fukushimu-ky Osaka,553,Japan The action of phospholipases AP and C in the course of collagen-stimulated platelet activation and the effect of cytochalasins on the responses were studied. Stimulation of human platelets with collagen was accompanied by aggregation, Ca2+ mobilization, inositol phosphate formation, and arachidonic acid release. However, in the presence of a cyclooxygenase inhibitor or a thromboxane A2 (TXA2) receptor antagonist, collagen induced only weak arachidonic acid release and weak inositol phosphate formation. The TXAz mimetic agonist U46619 induced all the responses except for arachidonic acid release, which was induced by synergistic action of collagen and U46619. The result that U46619 did not induce arachidonic acid release despite the activation of phospholipase C suggested that arachidonic acid was not released via phospholipase C but by phospholipase Az. These findings suggested that collagen initially induced weak activation of phospholipases AS and C and that further activation of phospholipase C as well as Ca2+ mobilization and aggregation were induced by TXA2, whereas further activation of phospholipase AZ required the synergistic action of collagen and TXA2. Platelets pretreated with cytochalasins did not respond to collagen. Further analysis revealed that the initial activation of phospholipases Az and C was specifically inhibited by cytochalasins, but the responses induced by U46619 or a synergistic action of collagen and U46619 were not inhibited. Therefore, we proposed that interaction of collagen receptor with actin filaments might have some roles in the collagen-induced initial activation of phospholipases.
Phospholipases AP and C play important roles in platelet activation. Phospholipase AB acts on phospholipids and liberates arachidonic acid (1-4). The cyclooxygenase pathway converts arachidonic acid into thromboxane A2 (TXAZ)' which is a potent inducer of platelet aggregation as well as vascular contraction (5-9). On the other hand, phospholipase C acts on inositol phospholipids (10). Degradation of phos-phatidylinositol4,5-bisphosphate by phospholipase C results in the formation of two second messengers, diacylglycerol and inositol 1,4,5-trisphosphate (11,12). Moreover, it is postulated that diacylglycerol is further hydrolyzed by diacylglycerol and 1 To whom correspondence should be addressed.
Collagen-induced platelet aggregation is essentially dependent on endogenously generated TXA2, since it is inhibited by cyclooxygenase inhibitors and TXAn/prostaglandin Hz . Using rat platelets, we have clarified that phospholipase Az action is an initial, critical event in collagen-induced platelet activation and that synergistic action of collagen and TXAz is necessary to induce phospholipase C action and further activation of phospholipase Az (20)(21)(22)(23). Therefore, collagen-stimulated platelets might be suitable for studying the regulation of phospholipases Az and C. However, detailed analysis of responses of collagen-stimulated human platelets had not been done. In this study, we separated the collagen-induced responses of human platelets into two phases, TXG-independent and TXAZ-dependent responses by means of a cyclooxygenase inhibitor and a TXA2/PGH2 receptor antagonist and investigated the actions of phospholipases Ai? and C in the respective phases.
To obtain clues to the regulation of these enzymes, we studied the effects of cytochalasins on collagen-induced responses. Cytochalasins prevent assembly of actin and actinbinding protein bundles in platelets (24-26). Actin filaments construct the platelet cytoskeleton together with other components such as actin-binding protein and microtubles (27). Recently, some glycoproteins (GPs) on platelet plasma membranes have been found to be linked to actin filaments (28, 29). A candidate for a receptor of collagen is involved in the GPs (30,31). However, the role of the linkage of the GPs with actin filaments is not yet known. The present study is the first to demonstrate that the actin filaments may be involved in receptor-mediated signal transduction in platelets.
Preparation of Human Platelets-Platelets were isolated from the blood of healthy human donors who had not taken medication for at least the previous 2 weeks. The blood was anticoagulated with 0.15 volumes of acid citrate dextrose (85 mM trisodium citrate, 70 mM citric acid, and 110 mM dextrose) and 0.5 pg/ml PGEI and centrifuged at 160 x g for 10 min. The platelet-rich plasma was then centrifuged for 15 min at 1,200 X g, and the platelets were resuspended in the appropriate volume of resuspension buffer as described below. The resuspension buffer, adjusted to pH 7.35, contained 137 mM NaC1, 2.7 mM KC1, 1.0 mM MgC12, 3.8 mM NaHZPO4, 3.8 mM Hepes, 5. mM dextrose, 0.035% bovine serum albumin. After labeling, platelets were sedimented onto 40% bovine serum albumin, isolated by gel filtration with a column of Sepharose 2B, and suspended in the resuspension buffer at 2.5 X 10' cells/ml. CaCI2 (1 mM) was added to the platelets 2 min before stimulation. Measurement of Intracellular Ca2+ Concentration-Platelets were suspended in the resuspension buffer containing 0.5 pg/ml PGE, at 1 X lo9 cells/ml and incubated with 1 p~ Fura2-AM for 30 min at room temperature. Platelets were isolated and resuspended as described above. The fluorescence ratio, obtained by dividing the fluorescence at 340 nm by that at 380 nm, was determined using a CAF-100 Caz+ analyzer (Japan Spectroscopic Co., Ltd., Japan) while stirring at 37°C. The emission wavelength was 500 nm. Changes in the fluorescence ratio were calibrated to changes in intracellular Ca2+ levels using the method of Grynkewicz et al. (32).
Measurement of Platelet Aggregation-Platelets were stimulated in an NKK HEMA Tracer 1 (Nikoh Bioscience Co., Ltd., Japan) and aggregation of platelets was followed continuously in it.
Measurement of Inositol Phospholipid Hydrolysis-Platelets were suspended in the resuspension buffer containing 0.5 pg/ml PGE, at 2 X lo9 cells/ml and incubated with 100 pCi/ml [3H]inositol for 2 h at room temperature. Platelets were isolated and resuspended as described above. LiCl (15 mM) was added 30 min before stimulation.
[3H]Inositol-labeled platelets were stimulated as indicated in the figures, and the reactions were stopped by addition of an equal volume of an 15% trichloroacetic acid. The acid-soluble inositol phosphates were separated by anion exchange chromatography as described by Berridge et al. (33).
Measurement of Arachidonic Acid Release-Platelets were suspended in the resuspension buffer containing 0.5 pg/ml PGE, at 2 X lo9 cells/ml and incubated with 20 pCi/ml [3H]arachidonic acid for 2 h at room temperature. Platelets were isolated and resuspended as described above and stimulated as indicated in the figures. The reactions were stopped by addition of 4 volumes of chloroform/ methanol (1:2, v/v), and lipids were extracted by the method of Bligh and Dyer (34). 3H-Labeled eicosanoids were separated by thin layer chromatography according to the method of Berteli. et al. (35). The areas corresponding to arachidonic acid, TXB2, hydroxyeicosatetraenoic acid, and hydroxyheptadecatrienoic acid were scraped off and the radioactivity measured. Arachidonic acid metabolites other than these eicosanoids were scarcely detected.
Measurement of TXAz Formation-Platelets were stimulated as indicated in the figure. The reactions were stopped by addition of an equal volume of ice-cold 10 mM EGTA and 50 p~ indomethacin. Platelets were precipitated and TXB2, a stable metabolite of TXA,, in the supernatant was measured by TXB2 radioimmunoassay kits.

TXA, Dependence of Collagen-induced Aggregation and Ca2+
Mobilization-As shown in Fig. 2 A , collagen (10 pg/ml) induced aggregation of human platelets after a lag period of about 1 min. The threshold concentration of collagen for the induction of full aggregation was 5-10 pg/ml. Indomethacin (10 p~) , an inhibitor of cyclooxygenase, inhibited the collagen-induced aggregation. Furthermore, 1 p~ SQ29,548 which is a specific antagonist of TXAz/PGH2 receptor (36) also completely blocked the platelet response to collagen.
Using Fura2-loaded platelets, the collagen-induced elevation of cytoplasmic Ca2+ concentration was observed (Fig.  2B). Ca2' mobilization was not detected during the lag period. Along with the shape change and aggregation, Ca2+ concentration increased and then decreased. The elevation of Ca2+ was also completely inhibited by indomethacin and SQ29,548. Vehicle ( a ) , 10 p~ indomethacin ( b ) , or 1 phi SQ29,548 (c) was added to platelets 3 min before stimulation. Washed human platelets (A) or Fura2-loaded platelets ( B ) were stimulated with 10 pg/ml collagen at the time indicated by the arrow. Increase and decrease of light transmission represent aggregation and shape change, respectively. Changes in the fluorescence ratio were monitored and calibrated into Ca2+ level as described in "Experimental Procedures." 3. Effect of indomethacin and SQ29.548 on collageninduced inositol phosphate formation. Vehicle (a), 10 M M indomethacin ( b ) , or 1 p~ SQ29,548 (c) was added to [3H]inositol-labeled platelets 3 min before stimulation. The platelets were stimulated for 3 min with 10 pg/ml collagen. Increases of inositol monophosphate (ZPI), inositol bisphosphate (ZP,), and inositol trisphosphate (IP3) over the resting level are represented. The resting levels of radioactivity in inositol monophosphate, inositol bisphosphate, and inositol trisphosphate in the absence of indomethacin and SQ29,548 were 360 f 41 dpm, 81 2 36 dpm, and 45 f 12 dpm, respectively. Indomethacin and SQ29,548 did not significantly affect the resting levels. Data are mean f S.E. (n = 3).
These results indicated that collagen-induced aggregation and Ca2+ mobilization were completely dependent on the action of endogenously generated TXA,.
TXA, Dependence of Collagen-stimulated Inositol Phosphate Formation-To assess phospholipase C action, stimulation of the inositol phosphate formation was investigated. Fig. 3 shows collagen-stimulated increase of inositol monophosphate, inositol bisphosphate, and inositol trisphosphate in [3H]inositol-labeled platelets. Collagen-stimulated inositol phosphate formation was also inhibited by indomethacin and SQ29,548. These compounds inhibited the response by about 80%. This finding indicated that 20% of inositol phosphate formation was induced by the action of collagen alone, which was independent of TXA2, and that 80% of the response was induced by the action of TXA,.
TXAz Dependence of Collagen-stimulated T U 2 Formation and Arachidonic Acid Release-These results indicated that TXAz was initially generated by the stimulation of platelets with collagen. As shown in Fig. 4A, collagen induced TXA2 formation. A cyclooxygenase inhibitor, indomethacin (10 p~) , completely blocked it. SQ29,548 (1 p~) also inhibited the TXA2 formation, although unlike indomethacin, it could not interfere with the pathway of TXAz synthesis (36). Similar results were obtained on arachidonic acid release as shown in Fig. 4B. Collagen-induced arachidonic acid release represented by [3H]eicosanoid formation was greatly reduced by inhibiting TXAz formation with indomethacin and also by a blockade of TXA, binding to the receptor with SQ29,548, although these compounds did not directly affect the arachidonic acid release (21, 36). These compounds inhibited the collagen-induced responses by about 80%, indicating that about 20% of arachidonic acid release and TXAz formation was induced by the action of collagen alone and that the action of TXAZ was necessary to induce further arachidonic acid release and TXA2 formation.
Responses of Human Platelets Stimulated with U46619--In order to find whether TXA, would induce the responses inhibited by indomethacin and SQ29,548, responses of platelets stimulated with U46619, an agonist of TXAZ/F'GH2 receptor (37), were studied. As shown in Fig. 5A, U46619 dose-  Indomethacin (10 p~) was added 3 min before stimulation. Washed human platelets ( A ) or Fura2-loaded platelets ( E ) were stimulated with 10, 100, or 1000 nM U46619 at the time indicated by the arrow. Aggregation and ea2' mobilization were monitored as described in Fig. 2. dependently induced aggregation of human platelets. Increase of intracellular Ca2+ concentration was also observed upon addition of U46619 (Fig. 5B). It reached maximum within 10 s and rapidly decreased. As shown in Fig. 6, U46619 induced inositol phosphate formation. U46619 a t 1000 nM induced almost the same level of inositol monophosphate formation as collagen and induced more than 60% increase of inositol bisphosphate and inositol trisphosphate. These results indicated that aggregation, ca*+ mobilization, and activation of phospholipase C could be induced by the action of TXA2. However, as shown in Fig. 7, U46619 did not induce arachidonic acid release. TXA, formation, as measured by radioimmunoassay, was also not observed by the stimulation with U46619 (data not shown).
Synergistic Action of Collagen and TXAZ on Platelet Responses-Since a synergistic action of collagen and TXAt had been revealed to be important for full activation of rat plate- The platelets were stimulated for 3 min with 10 pg/ml collagen (O), various concentrations of U46619 (0), or U46619 plus 10 pg/ml collagen (0). ['H]Inositol phosphate formation is represented as described in Fig. 3. The resting levels of radioactivity in inositol monophosphate, inositol bisphosphate, and inositol trisphosphate in the absence of indomethacin were 360 f 41 dpm, 81 f 36 dpm, and 45 f 12 dpm, respectively.  10 p~ indomethacin (0,O) was added to ['Hlarachidonic acid-labeled platelets 3 min before stimulation. The platelets were stimulated for 3 min with 10 pg/ml collagen (O), various concentrations of U46619 (0), or U46619 plus 10 pg/ml collagen (0). [3H]Eicosanoid formation is represented as described in Fig. 4. The resting level of radioactivity in eicosanoid in the absence of indomethacin was 5130 f 550 dpm. Indomethacin did not significantly change the resting level. lets (20-22), we supposed that similar synergistic action might be necessary for inducing full arachidonic acid release in human platelets. The synergistic action was studied by stimulating platelets with collagen plus U46619 in the presence of indomethacin (Fig. 7). Indomethacin was used to prevent endogenous TXA, generation. In the presence of indomethacin, collagen induced only a small amount of arachidonic acid release. However, further addition of U46619 induced arachidonic acid release in a dose-dependent manner, indicating that collagen and U46619 synergistically induced arachidonic acid release.
In addition to arachidonic acid release, the formation of inositol phosphates was also enhanced by the synergistic action (Fig. 6), although U46619 alone could induce the response. The effect of the synergistic action was remarkable on the formation of inositol bisphosphate and inositol trisphosphate but not on the formation of inositol monophosphate. Kaibuchi et al. (38) also reported the synergistic action of collagen and a TXAz analogue on phosphatidic acid formation.
Furthermore, the synergistic action was also observed in Ca" mobilization ( Fig. 8). Stimulation of platelets with collagen plus U46619 resulted in the appearance of a second Ca2+ peak after the rapid, first Ca2+ peak which could be induced by U46619 alone. The appearance of the second Ca2+ peak has also been observed in rat platelets by the synergistic action of collagen and U46619 (20, 21). Fig. 9, we found that 10 g~ cytochalasin B completely inhibited collagen-induced platelet aggregation and Ca2+ mobilization. Furthermore, cytochalasin B completely inhibited collagen-induced arachidonic acid release (Fig. lOA, I) and inositol phosphate formation (Fig. 11, Indomethacin (10 p~) was added to Fura2-loaded platelets 3 min before stimulation. The platelets were stimulated with 100 or 1000 nM U46619 plus 10 pg/ml collagen at the time indicated by the arrow. Ca2+ mobilization was monitored as described in Fig. 2. I 10oL FIG. 9. Effect of cytochalasin B on collagen-induced platelet aggregation (A) and Ca'+ mobilization (B). Vehicle (a) or 10 p~ cytochalasin B (b) was added to platelets 4 min before stimulation. Washed human platelets ( A ) or Fura2-loaded platelets ( B ) were stimulated with collagen at the time indicated by the arrow. Aggregation and Ca2' mobilization were monitored as described in Fig. 2.

Effect of Cytochalasin B on the Responses of Human Platelets Stimulated by U46619
and Thrombin-In order to find whether the inhibitory effect of cytochalasins would be common among different stimuli, the effect of cytochalasin B on thrombin-and U46619-induced responses of platelets was examined. Fig. 1OB shows the effect of cytochalasin B on thrombin-induced arachidonic acid release. Cytochalasin B did not inhibit the response. As shown in Fig. 11, III, U46619induced inositol phosphate formation was not inhibited. Moreover, U46619-induced CaZ+ mobilization was not affected by cytochalasin B (Fig. 12). However, 10 p~ cytochalasin B completely blocked the thrombin-and U46619-induced platelet shape change (data not shown), for which polymerization Fura2-loaded platelets 4 min before stimulation. Indomethacin (10 PM) was added 3 min before stimulation. The platelets were stimulated with 1 pM U46619 at the time indicated by the arrow. Ca2+ mobilization was monitored as described in Fig. 2. of actin was postulated to be important (25,26).
Analysis of the Site Affected by Cytochalasin B-To clarify what step of collagen-stimulated signal transduction was inhibited by cytochalasin B, more precise analysis of the effect of cytochalasin was carried out on collagen-stimulated arachidonic acid release and inositol phosphate formation. In the presence of indomethacin, the responses induced by collagen were those induced by the action of collagen alone. As shown in Fig. lOA, 11, and Fig. 11, 11, arachidonic acid release and inositol phosphate formation induced by the action of collagen alone were inhibited by cytochalasin B. When platelets were stimulated with collagen plus U46619 in the presence of indomethacin, the total response consisted oE 1) the responses induced by collagen alone, 2) the responses induced by U46619 alone, and 3) the responses induced by the synergistic action of collagen and U46619. As U46619 alone could not induce arachidonic acid release, the arachidonic acid release represented in Fig. 10, IV, was a combination of factors 1 and 3. Cytochalasin B slightly decreased the response, and the decrease was almost the same as that of the response for factor 1 shown in Fig. lOA, if. Inositol phosphate formation induced by collagen plus U46619 in the presence of indomethacin is represented in Fig. 11, ZV. The response was factor 1 plus factor 2 plus factor 3. Like arachidonic acid release, the decrease by cytochalasin B was almost the same as that of factor 1 which is represented in Fig. 11, II. The factor 2 response has been shown to not be inhibited by cytochalasin B. These findings indicated that the responses inhibited by cytochalasin B were only the responses induced by the action of collagen alone, which were independent of TXA2. The responses induced by U46619 alone and the responses induced by the synergistic action of collagen and U46619 were not inhibited, indicating that cytochalasin B did not impair the interaction of U46619 and collagen with platelets. The reason the responses induced by collagen in the absence of indomethacin were completely inhibited by cytochalasin B (Fig.   lOA, I and Fig. 11, I) can be explained well from our results that all subsequent reactions depended on the TXAZ initially generated by the action of collagen alone, which was susceptible to cytochalasin B.

DISCUSSION
Collagen-induced platelet activation is greatly dependent on endogenously generated TXA2. The responses of collagenstimulated platelets can be separated into two categories, TXAz-independent and TXAz-dependent responses. First, collagen induces the TXAz-independent responses including initial TXAz generation. Next, the TXA2-dependent responses are successively induced. We refer to the TXAzindependent process as the "first-phase" responses and the TXAz-dependent one as the "second-phase'' responses. The first-phase responses are observed when TXAz formation is inhibited by indomethacin or when binding of TXA2 to the receptor is prevented by a receptor antagonist SQ29,548. In the presence of these compounds, collagen induced release of arachidonic acid, formation of TXAz (only when SQ29,548 was used), and formation of inositol phosphates without Caz+ mobilization and aggregation. However, these responses were very weak in comparison to the responses observed in the absence of indomethacin and SQ29,548. Therefore, the greater part of the collagen-induced responses seems to be induced by the action of TXA2.
It has been postulated that arachidonic acid is liberated by a combination of phospholipase C (10) and diacylglycerol and monoacylglycerol lipase activities (13-15) and by phospholipase AP activity (1-4). However, the finding that activation of phospholipase C by U46619 was not accompanied by arachidonic acid release suggests that the former pathway does not actually function in human platelets, and arachidonic acid is mainly released by the latter pathway, i.e. phospholipase AP. We have obtained the same results with rabbit platelets.' The importance of phospholipase Az in arachidonic acid release is also suggested by the finding that an inhibitor of diacylglycerol lipase does not inhibit arachidonic acid (39). Degradation of several phospholipid classes by phospholipase A2 in collagen-stimulated human platelets has been confirmed by Takamura et al. (40) and Pollock et al. (41). Our results also suggest that phospholipase A2 and phospholipase C are independently controlled.
In order to study the synergistic action of collagen and TXAz in the second-phase responses, platelets were stimulated with collagen plus U46619 in the presence of indomethacin. For the second-phase arachidonic acid release, the synergistic action was essential as shown in Fig. 7. Moreover, U46619-induced inositol phosphate formation was also enhanced by the synergistic action of collagen and U46619. Enhanced formation of inositol trisphosphate may induce the second Ca'+ peak as shown in Fig. 8 by releasing Ca2+ from intracellular Ca2+ store (11).
From these findings, we summarized collagen-induced signal transduction as occurring as follows. In the first phase, collagen slightly activates phospholipases A2 and C. Activation of phospholipase A2 results in formation of a small amount of TXA2. Despite the activation of phospholipase C, Ca'+ mobilization is not detected, probably because of insufficient formation of inositol trisphosphate. In the second phase, phospholipase C is further activated by the action of TXAz. However, for the further activation of phospholipase AP, the synergistic action of TXA, and collagen is necessary.
Actin is a major protein in platelets and is one of the components of the platelet cytoskeleton (27). The relationship T. Nakano, K. Hanasaki, S. Matsumoto, and H. Arita, submitted for publication. between the cytoskeleton and platelet shape change is well accepted (56)(57)(58)(59). On the other hand, the interaction of cytoskeleton with platelet membrane has recently become a subject of attention. Fox and Solum et al. (28,29) have reported that GPIa and GPIb are linked to actin filaments. Actinbinding protein has been indicated to be involved in the association of the GPs with the actin filaments (28, 60). Moreover, GPIa and GPIb are proposed to be a collagen receptor and a receptor of thrombin and von Willebrand factor, respectively (30,31,(61)(62)(63). Although these findings suggest involvement of the linkage of GPs with actin filaments in signal transduction across plasma membrane, no evidence has been presented.
Cytochalasins inhibited all the collagen-induced responses. However, U46619-induced inositol phosphate formation and Caz+ mobilization, and thrombin-induced arachidonic acid release was not inhibited. These results indicate that phospholipases A, and C and CaZ+ mobilization are not directly inhibited by cytochalasins and that post receptor signal transduction of TXA, as well as thrombin receptors are not impaired. It has been reported that cytochalasin B does not inhibit ADP-induced platelet aggregation (64) and that cellular secretions (65-68), including dense body secretion in platelets (69,70), were rather enhanced by cytochalasins. These observations suggest that the site inhibited by cytochalasins is specific to collagen receptor. Based on these findings, we propose that the linkage of GPIa with actin filaments may play an important role in collagen receptormediated signal transduction and that cytochalasins probably impair the linkage. It is not known how actin filaments participate in activation of phospholipases A, and C. However, the linkage is suggestive of the relationship between phospholipase activity and mobility of the receptor on platelet membrane. Moreover, interaction of lipocortin (calpactin) with cytoskeleton (71) is also an interesting finding. Binding of collagen to GPIa may alter the interaction of lipocortin with cytoskeleton, which may affect the activity of phospholipase Az. Microtubles might not have been involved in the signal transduction, since Hashizume et al. (72) reported that a low dose of vinblastine, an inhibitor of polymerization of microtubles, did not inhibit collagen-induced platelet aggregation despite disorganization of the microtuble system.
Although activation of phospholipases At and C in the first phase was inhibited by cytochalasins, activation of the enzymes in the second phase was not inhibited. These results indicate that the action of collagen in the first phase is affected by cytochalasins, whereas that in the second phase is not. This difference may suggest the existence of two receptors of collagen, one acts in the first phase and the other in the second phase. In addition to GPIa, GPIIb and coagulation factor XI11 on the platelet membrane have been also proposed to be collagen receptors (73,74).
The activation mechanisms of the enzymes in the first phase and in the second phase have been suggested to be different. We discussed above that phospholipase A2 and phospholipase C are independently controlled. It also seems that the regulation mechanisms of the enzymes are different among agonists. Such a diversity would be necessary to strictly control cellular activity. In order to clarify the cellular regulation, more precise analyses of the diverse regulation mechanisms of the enzymes, including synergistic action of agonists such as collagen and TXA2, are needed. 14.