J Biol Chem, Vol. 275, Issue 11, 7527-7533, March 17, 2000

A Role for the Actin Cytoskeleton in the Initiation and Maintenance of Store-mediated Calcium Entry in Human Platelets
EVIDENCE FOR CONFORMATIONAL COUPLING^*

Juan A. Rosado, Susanne Jenner, and Stewart O. Sage^§

From the Department of Physiology, University of Cambridge, Downing Street, Cambridge CB2 3EG, United Kingdom

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The nature of the mechanism underlying store-mediated Ca²⁺ entry has been investigated in human platelets through a combination of cytoskeletal modifications. Inhibition of actin polymerization by cytochalasin D or latrunculin A had a biphasic time-dependent effect on Ca²⁺ entry, showing an initial potentiation followed by inhibition of Ca²⁺ entry. Moreover, addition of these agents after induction of store-mediated Ca²⁺ entry inhibited the Ca²⁺ influx mechanism. Jasplakinolide, which reorganizes actin filaments into a tight cortical layer adjacent to the plasma membrane, prevented activation of store-mediated Ca²⁺ entry but did not modify this process after its activation. In addition, jasplakinolide prevented cytochalasin D-induced inhibition of store-mediated Ca²⁺ entry. Calyculin A, an inhibitor of protein serine/threonine phosphatases 1 and 2 which activates translocation of existing F-actin to the cell periphery without inducing actin polymerization, also prevented activation of store-mediated Ca²⁺ entry. Finally, inhibition of vesicular transport with brefeldin A inhibited activation of store-mediated Ca²⁺ entry but did not alter this mechanism once initiated. These data suggest that store-mediated Ca²⁺ entry in platelets may be mediated by a reversible trafficking and coupling of the endoplasmic reticulum with the plasma membrane, which shows close parallels to the events mediating secretion.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In many cell types, including human platelets, depletion of the intracellular Ca²⁺ stores induces entry of Ca²⁺ across the plasma membrane (PM)¹ (1). However, the mechanism by which the filling state of stores is communicated to the PM remains unclear. Hypotheses have considered both indirect and direct coupling mechanisms. Indirect coupling assumes the existence of a diffusible messenger generated by the storage organelles such as cyclic GMP (2), cytochrome P-450 metabolites (3), tyrosine kinases (4, 5), or release of a calcium influx factor from depleted stores (6); however, no messenger molecule has been identified. Alternatively, direct coupling (conformational coupling) proposes a physical interaction between the endoplasmic reticulum (ER) and the PM and considers that some calcium sensitivity may reside on the InsP₃ receptor, which is thought to be responsible for transmitting information from the ER to the PM (7). Consistent with this model, some studies indicate that Ca²⁺ entry is closely localized to sites where InsP₃ evokes emptying of the Ca²⁺ stores and that Ca²⁺ entry is unlikely to be activated by a diffusible molecule (8-10). A different model suggests that Ca²⁺ channels or membrane- bound activator molecules are exocytotically incorporated into the PM upon store depletion (10). This hypothesis is compatible with the conformational coupling model only if it is assumed that the link between the ER and the PM is mechanically weak before store depletion and strong afterward.

Recently, a secretion-like coupling model has been proposed by Patterson et al. (11). This mechanism involves a physical but reversible interaction between the ER and the PM based on a physical trafficking of the ER toward the PM. In support of this hypothesis, small GTP-binding proteins, which modulate actin reorganization and vesicular trafficking, have been shown to be important for store-mediated Ca²⁺ entry (SMCE) in different cell types (e.g. 12, 13) including platelets (14). In addition, it has been shown that GTP, possibly through activation of a small GTP-binding protein, may activate a rapid and reversible interaction between membrane surfaces involving the formation of a prefusion pore through which Ca²⁺ ions can flow (15, 16). In the secretion-like coupling model, the actin cytoskeleton close to the PM plays a key regulatory role in Ca²⁺ entry as it does in secretion. Redistribution of actin filaments into a tight layer subjacent to the PM prevents Ca²⁺ entry by acting as a barrier that blocks the coupling between ER and PM. This effect is reversed by disassembly of the cortical actin layer, thus providing the reestablishment of the coupling and Ca²⁺ entry (11).

The model suggested by Patterson et al. (11) is compatible with the hypothesis of Yao et al. (10) because a model based on vesicular trafficking may support transport of the ER or Ca²⁺ channels toward the PM. In the present study we sought to expand our understanding of the mechanisms underlying SMCE in platelets. We have shown previously (14) that small GTP-binding proteins are required for the activation of Ca²⁺ entry after store depletion in platelets and that this mechanism may involve the actin cytoskeleton. We report here that modification of the actin cytoskeleton by stabilizing actin filaments prevents SMCE without affecting chemical coupling mediated by InsP₃; however this modification has no effect once Ca²⁺ entry has been activated by depletion of the stores. These results suggest that under these conditions actin filaments act as a physical barrier that prevents a close interaction between the ER and the PM. Moreover, cytochalasin D (Cyt D) and latrunculin A, two agents that disrupt the actin filament network, can reverse Ca²⁺ entry after activation. This suggests that a direct interaction between the ER and the PM, which may need mechanical support, is more likely than the incorporation of channels into the PM by vesicle fusion.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Fura-2 acetoxymethyl ester (fura-2/AM) was from Texas Fluorescence (Austin, TX). Apyrase (grade V), aspirin, bovine serum albumin, paraformaldehyde, Nonidet P-40, fluorescein isothiocyanate-labeled phalloidin and thapsigargin (TG) were from Sigma (Poole, Dorset, U. K.). Cyt D and brefeldin A were from Calbiochem (Nottingham, U. K.). Jasplakinolide and latrunculin A were from Molecular Probes (Leiden, The Netherlands). Calyculin A was from Alexis (Nottingham, U. K.). All other reagents were of analytical grade.

Platelet Preparation-- Fura-2-loaded platelets were prepared as described previously (5). Briefly, blood was obtained from healthy drug-free volunteers and mixed with one-sixth volume of acid/citrate dextrose anticoagulant containing (in mM) 85 sodium citrate, 78 citric acid and 111 D-glucose. Platelet-rich plasma was then prepared by centrifugation for 5 min at 700 × g and 100 µM aspirin and 40 µg/ml apyrase were added. Platelet-rich plasma was incubated at 37 °C with 2 µM fura-2/AM for 45 min. Cells were then collected by centrifugation at 350 × g for 20 min and resuspended in HEPES-buffered saline containing (in mM) 145 NaCl, 10 HEPES, 10 D-glucose, 5 KCl, 1 MgSO₄, pH 7.45, and supplemented with 0.1% (w/v) bovine serum albumin and 40 µg/ml apyrase.

Measurement of [Ca²⁺]_i-- Fluorescence was recorded from 1.5-ml aliquots of magnetically stirred platelet suspension (10⁸ cells/ml) at 37 °C using a Cairn Research Spectrophotometer (Cairn Research Ltd., Sittingbourne, Kent, U. K.) with excitation wavelengths of 340 and 380 nm and emission at 500 nm. Changes in [Ca²⁺]_i were monitored using the fura-2 340/380 fluorescence ratio and calibrated according to the method of Grynkiewicz et al. (17).

Determination of Ca²⁺ Entry-- Ca²⁺ influx in TG-induced store-depleted platelets was estimated using the integral of the rise in [Ca²⁺]_i for 2.5 min after the addition of CaCl₂. Thrombin-evoked Ca²⁺ influx was measured as the integral of the rise in [Ca²⁺]_i above basal for 1 min after the addition of thrombin in the presence of external Ca²⁺, corrected by subtraction of the integral over the same period for stimulation in the absence of external Ca²⁺ (with 1 mM EGTA).

Measurement of F-actin Content-- The F-actin content of resting and activated platelets was determined according to the modifications (18) of a previously published procedure (19). Briefly, washed platelets (2 × 10⁸ cells/ml) were activated in HEPES-buffered saline. Samples of platelet suspension were transferred to 200 µl of ice-cold 3% (w/v) formaldehyde in phosphate-buffered saline for 10 min. Fixed platelets were permeabilized by incubation for 10 min with 0.025% (v/v) Nonidet P-40 detergent dissolved in phosphate-buffered saline. Platelets were then incubated for 30 min with 1 µM fluorescein isothiocyanate-labeled phalloidin in phosphate-buffered saline supplemented with 0.5% (w/v) bovine serum albumin. After incubation the platelets were collected by centrifugation in an MSE Micro-Centaur Centrifuge (MSE Scientific Instruments, Crawley, Sussex, U. K.) for 90 s at 3,000 × g and resuspended in phosphate-buffered saline. Staining of 2 × 10⁷ cells/ml was measured using a Perkin-Elmer fluorescence spectrophotometer (Perkin-Elmer, Norwalk, CT). Samples were excited at 496 nm, and emission was at 516 nm.

Because jasplakinolide (JP) binds to the same site as phalloidin (20), fixed platelets were incubated with 10 µM fluorescein isothiocyanate-labeled phalloidin, and JP was added at the same time as phalloidin in the control samples for JP experiments.

Statistical Analysis-- Analysis of statistical significance was performed using Student's unpaired t test except for calyculin A Ca²⁺ entry data where Student's paired t test was used. For multiple comparison, one-way analysis of variance combined with the Dunnett tests was used.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Breakdown of the Actin Cytoskeleton Inhibits Store-mediated Ca²⁺ Entry in Human Platelets-- In the absence of extracellular Ca²⁺, the addition of 200 nM TG to fura-2-loaded human platelets in stirred cuvettes at 37 °C evoked a prolonged elevation in [Ca²⁺]_i because of the release of Ca²⁺ from internal stores. Subsequent addition of 300 µM Ca²⁺ to the external medium induced a sustained increase in [Ca²⁺]_i indicative of SMCE (Fig. 1, A and D).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Effect of Cyt D and latrunculin A on TG-evoked Ca2+ entry in human platelets. Fura-2-loaded human platelets were incubated at 37 °C either in the presence of 10 µM Cyt D for 1 min (B) and 40 min (C), 3 µM latrunculin A for 1 min (E) and 1 h (F), or the vehicles (A and D). At the time of experiment 100 µM EGTA was added. Cells were then stimulated with 200 nM TG, and 3 min later CaCl2 (final concentration 300 µM) was added to the medium to initiate Ca2+ entry. The traces are representative of five independent experiments.

Pretreatment of human platelets with 10 µM Cyt D, a widely utilized membrane-permeant inhibitor of actin polymerization which binds to the barbed end of actin filaments (26), prevents TG-evoked actin filament formation in a time-dependent manner, reaching complete inhibition after 40 min of treatment (Table I). On the other hand, Cyt D had no effect on the actin filament content of unstimulated platelets when treated for up to 40 min (Table I). Similar results were obtained with latrunculin A, an agent that inhibits actin polymerization by binding to actin monomers (27). Treatment of human platelets with 3 µM latrunculin A for 1 h at 37 °C abolished TG-induced actin filament formation without having any effect on the actin filament content of unstimulated cells (Table I). Treatment of platelets at 37 °C with 10 µM Cyt D for 40 min or 3 µM latrunculin A for 1 h reduced TG-evoked Ca²⁺ entry by 50 and 65%, respectively. In contrast, preincubation with these agents for 1 min increased TG-evoked SMCE (Fig. 1, A-F). Cyt D- or latrunculin A-treated platelets showed an identical release of Ca²⁺ from the intracellular stores upon stimulation with TG compared with untreated cells, indicating that accumulation of Ca²⁺ in the internal stores was unaffected by inhibition of actin polymerization (Fig. 1, A-F).

JP Inhibits Store-mediated Ca²⁺ Entry-- JP, a cell-permeant peptide isolated from Jaspis johnstoni which induces polymerization and stabilization of actin filaments (20), is a useful tool for studying further the role of the cortical actin cytoskeleton in SMCE. JP has been shown to elongate and organize actin filaments exclusively at the cell periphery, near the PM (11). Treatment of human platelets with 10 µM JP for 30 min at 37 °C resulted in a significant enhancement of the F-actin content in unstimulated cells (n = 4; p < 0.001; Table I). In addition, no further increase in F-actin content was observed after stimulation with TG in JP-treated cells (Table I).

Pretreatment of human platelets for 30 min at 37 °C with JP attenuates TG-evoked SMCE in a concentration-dependent manner (Fig. 2, A-C). Ca²⁺ entry was reduced significantly by 48.9 ± 4.0 and 77.7 ± 5.4% after treatment for 30 min with 5 or 10 µM JP, respectively (n = 6; p < 0.001). In contrast, JP had no effect on resting cytosolic Ca²⁺ levels nor any effect on TG-evoked release of Ca²⁺ from the intracellular stores.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Effect of JP on TG- and thrombin-evoked Ca2+ entry. Panels A-C, cells were preincubated for 30 min at 37 °C in the presence of 5 µM JP (B) or 10 µM JP (C) or the vehicle (methanol, A). At the time of experiment, 100 µM EGTA was added. Platelets were then stimulated with 200 nM TG, and 3 min later 300 µM CaCl2 was added to the medium. D-E, cells were pretreated for 30 min at 37 °C with 10 µM JP (traces b) or the vehicle (traces a). At the time of experiment, 1 mM CaCl2 (D) or 1 mM EGTA (E) was added. Platelets were then stimulated with 0.5 unit/ml thrombin at the time indicated. Traces shown are representative of six separate experiments.

As shown in Fig. 2D, treatment of platelets with 10 µM JP for 30 min at 37 °C resulted in a substantial inhibition of the elevation in [Ca²⁺]_i evoked by the physiological agonist thrombin (0.5 unit/ml) in medium containing 1 mM Ca²⁺. The initial peak [Ca²⁺]_i elevation above basal after agonist was reduced significantly from 1,000 ± 60 to 328 ± 22 nM (n = 6; p < 0.001). In the absence of external Ca²⁺ (1 mM EGTA added), JP was without effect on the thrombin-evoked rise in [Ca²⁺]_i. The initial peak elevation in [Ca²⁺]_i above basal after agonist stimulation was 233 ± 12 nM in control cells and 233 ± 19 nM in JP-treated cells (Fig. 2E; n = 6). If we consider the entry of Ca²⁺ stimulated by thrombin (see "Experimental Procedures"), JP significantly reduced thrombin-evoked Ca²⁺ entry by 67.1 ± 1.8% (p < 0.001).

Effect of Cyt D and JP on Preactivated Store-mediated Ca²⁺ Entry-- As for the secretory pathway, where actin filaments in the cell periphery act as a barrier preventing exocytosis, the cortical actin cytoskeleton might prevent the interaction between ER and the PM necessary for activation of Ca²⁺ entry. To investigate whether the actin filament network is required for the maintenance of Ca²⁺ entry we studied the effects of Cyt D and JP on Ca²⁺ influx in platelets after SMCE had been stimulated using TG.

Fig. 3A shows the effect of adding Cyt D to store-depleted platelets. 10 µM Cyt D or the vehicle was added 3 min after TG. As shown in Fig. 3A (control t = 3 min) Ca²⁺ entry was clearly stimulated at this point in time. Subsequent addition of Cyt D for 1 h blocked SMCE without having any effect in TG-evoked release of Ca²⁺ from the internal stores (Fig. 3A; n = 4). Under these conditions Cyt D reversed the F-actin filament formation stimulated by TG (Table II; n = 6). These results were confirmed using latrunculin A. The addition of latrunculin A after SMCE has been stimulated with TG abolished Ca²⁺ entry (data not shown) as well as completely reversing the F-actin polymerization induced by TG (Table II; n = 4).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Effects of Cyt D and JP on TG-evoked store-mediated Ca2+ entry. Fura-2-loaded human platelets were suspended in a Ca2+-free medium (100 µM EGTA was added) as described under "Experimental Procedures." Cells were then stimulated with 200 nM TG, and 3 min later 10 µM Cyt D (A), 10 µM JP (B), both (C), or the vehicles (control t = 63 min) were added as indicated by the arrows. CaCl2 (final concentration 300 µM) was added to the medium at the same time, as a control (control t = 3 min) or 1 h after Cyt D, JP, or both to initiate Ca2+ entry. Traces shown are representative of four independent experiments.

On the other hand, JP, which induces polymerization and stabilization of actin filaments (20), did not significantly modify Ca²⁺ entry in platelets once SMCE has been activated by adding TG (Fig. 3B; p = 0.085, n = 4). Finally, experiments were carried out to test for reversal of the action of Cyt D by JP because the binding affinity of JP for F-actin (K_D = 15 nM) (20) is greater than of Cyt D (K_D = 50 nM) (21). Simultaneous addition of both JP and Cyt D resulted in an increase in the F-actin content (Table II; n = 4) similar to that observed with JP alone (Table I; n = 4). In addition, JP prevented the inhibitory effect of Cyt D on activated SMCE (Fig. 3C; compare with Fig. 3A, n = 4).

Calyculin A Prevents Store-mediated Ca²⁺ Entry-- To study further the involvement of the cytoskeleton in SMCE we used an independent means to check whether the cortical actin cytoskeleton could interfere with the communication between the ER and the PM. Recent studies have shown that phosphorylation of the moesin, ezrin, and radixin family of proteins is required for mediation of actin cross-linking to the PM (22, 23). To induce phosphorylation-dependent association of the actin-binding proteins with the PM we treated human platelets with calyculin A, a serine/threonine phosphatase inhibitor that inhibits protein phosphatases 1 and 2 (11, 23). In many cells, including platelets, treatment with calyculin A results in a condensation of actin filaments at the PM (11, 24, 25). Treatment of human platelets for 2 min at 37 °C with calyculin A caused a concentration-dependent reduction of Ca²⁺ entry induced by TG (Fig. 4). Calyculin A significantly reduced store-regulated Ca²⁺ entry by 26.8 ± 6.9, 55.1 ± 3.9, 67.2 ± 3.1, 69.3 ± 4.0, and 72.3 ± 1.3% (n = 6) at concentrations of 3, 10, 30, 100, and 300 nM, respectively (p < 0.05) without having significant effects on Ca²⁺ release from the intracellular stores (data not shown). The effects of calyculin A on actin polymerization are shown in Table I. Exposure of platelets to 100 nM calyculin A for 2 min appeared to increase the F-actin content, but the difference was not statistically significant. However, pretreatment of platelets with 100 nM calyculin A significantly inhibited the increase of F-actin induced by thapsigargin (Table I; n = 6; p < 0.001).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effect of calyculin A on TG-evoked Ca2+ entry. Fura-2-loaded human platelets were incubated for 2 min at 37 °C in the presence of increasing concentrations of calyculin A or the vehicle (dimethyl sulfoxide). At the time of the experiment, 100 µM EGTA was added. Cells were then stimulated with 200 nM TG, and 3 min later CaCl2 (final concentration 300 µM) was added to the medium to initiate Ca2+ entry. Values represent the Ca2+ influx normalized to paired controls without calyculin A. Data shown are presented as means ± S.E. of six independent experiments.

Role of Intracellular Transport in Store-mediated Ca²⁺ Entry-- Brefeldin A, which specifically inhibits vesicular transport (28, 29), was used to study the involvement of intracellular transport in SMCE. The results shown in Fig. 5, A and B, indicate that treatment of human platelets for 1 h at 37 °C with 100 µM brefeldin A inhibited TG-stimulated Ca²⁺ entry by 78.2 ± 6.1% (n = 6; p < 0.01). Brefeldin A did not alter the resting cytosolic Ca²⁺ level, nor did it have any effect on TG-induced Ca²⁺ release from the intracellular stores, indicating that accumulation of Ca²⁺ in the internal stores was unaffected by preincubation with brefeldin A (Fig. 5, A and B). Because the effect of brefeldin A on the intracellular transport is reversible (30), we examined whether SMCE might be reactivated after removal of brefeldin A. Human platelets were preincubated with 100 µM brefeldin A for 1 h at 37 °C followed by an incubation for a further 1 h in the absence of brefeldin A. The results indicated that, as for secretion, SMCE had been reactivated after this period in the absence of brefeldin A (Fig. 5C; n = 6). Treatment of cells for 1 h with 100 µM brefeldin A after SMCE had been activated by the addition of TG did not modify the entry of extracellular Ca²⁺ (Fig. 5D; n = 4). This finding is compatible with a role for intracellular transport in the activation but not maintenance of SMCE and furthermore indicates that brefeldin A has no nonspecific effect as a Ca²⁺ chelator or a Ca²⁺ channel blocker.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Effect of brefeldin A on TG-evoked store-mediated Ca2+ entry. A-C, fura-2-loaded human platelets were incubated at 37 °C for 1 h in the presence of 100 µM brefeldin A (B) or the vehicle (methanol, A) or preincubated for 1 h in the presence of 100 µM brefeldin A followed by incubation for 1 h in the absence of brefeldin A (C). At the time of experiment, 100 µM EGTA was added. Cells were then stimulated with 200 nM TG, and 3 min later CaCl2 (final concentration 300 µM) was added to the medium to initiate Ca2+ entry. Traces shown are representative of six independent experiments. D, fura-2-loaded human platelets were suspended in a Ca2+-free medium (100 µM EGTA added) as described under "Experimental Procedures." Cells were then stimulated with 200 nM TG, and 3 min later 100 µM brefeldin A or the vehicle (methanol; control t = 63 min) was added as indicated by the arrow. CaCl2 (final concentration 300 µM) was added to the medium at the same time, as a control (control t = 3 min) or 1 h after brefeldin A or the vehicle to initiate Ca2+ entry. Traces are representative of four independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The activation of Ca²⁺ entry following the depletion of intracellular Ca²⁺ stores is a signaling process of great relevance. However, the mechanism by which the filling state of the Ca²⁺ stores activates such influx has remained unclear. Current hypotheses are centered on either a diffusible messenger released from the ER or a direct coupling mechanism involving physical interaction between the PM and the ER. Using different cytoskeletal modifications before and after activation of SMCE, our results provide new insights as to the nature of the coupling mechanism in human platelets.

Cyt D, a widely used membrane-permeant inhibitor of actin polymerization, prevented TG-induced actin polymerization in human platelets in a time-dependent manner without having any effect on basal F-actin content (Table I). Similar effects were observed after treatment with latrunculin A, which inhibits actin polymerization by binding to monomeric G-actin in a 1:1 complex (27). These findings showed, in agreement with previous studies (26), that "treadmilling" in resting platelets is extremely slow. Interestingly, treatment of platelets with Cyt D or latrunculin A for different times has a biphasic effect on Ca²⁺ entry. A short preincubation (1 min) with these agents increases SMCE while inhibiting TG-induced actin polymerization by about 20%. This is in agreement with an earlier report (31). On the other hand, when Cyt D and latrunculin A had completely inhibited actin polymerization Ca²⁺ entry was significantly reduced.

Our observations are in agreement with previous studies in vascular endothelial cells (32). In contrast, Pedrosa-Ribeiro et al. (33) and Patterson et al. (11) have obtained essentially the opposite result: Cyt D did not block TG-induced SMCE in NIH 3T3 cells or in smooth muscle cell lines. The reason for these discrepancies is not certain. In platelets, where actin accounts for about 20% of the total protein, and in vascular endothelial cells, actin filaments are located mainly in the cell periphery, and a dynamic actin polymerization is required upon activation (34, 35). In contrast, NIH 3T3 and smooth muscle cells have a more evenly distributed and organized cytoskeletal architecture (11, 33). Reorganization of the dense subplasmalemmal actin filament network found in platelets and endothelial cells may thus be essential in allowing coupling between the ER and PM in these cells but not others, so explaining the differences in the reported effects of Cyt-D on SMCE in different cell types.

Recently a secretion-like coupling model for SMCE has been proposed in cultured smooth muscle cell lines (11). It is well known that cortical F-actin prevents exocytosis and that disassembly of such actin filaments by Cyt D increases stimulated secretion by permitting the approach and docking of secretory granules with the PM (e.g. 36). This phenomenon could explain the potentiation of SMCE after a short preincubation of platelets with Cyt D or latrunculin A. Because both agents inhibit TG-induced actin polymerization in a time-dependent manner, it is likely that this cell-permeant agent initially affects the actin filaments near the PM, which could thus facilitate the coupling between the ER and the PM.

In a secretion-based coupling model cortical F-actin should play an inhibitory role. In agreement with this, JP, a cell-permeant peptide that induces polymerization and stabilization of cortical actin filaments (37), prevented TG-induced SMCE. It has been shown that JP, as well as its membrane-impermeant structural analog, phalloidin, causes elongation and reorganization of actin filaments exclusively near the PM (11, 37). Such an effect is consistent with the enhanced formation of actin filaments by JP observed in platelets. Treatment of human platelets with JP enhanced F-actin content to 200% of basal. Considering that recent investigations have established that almost half of the actin in resting platelets is unpolymerized (26, 34), our results indicate that JP induced a full actin polymerization in these cells, in agreement with previous studies in muscle cells (11). This finding allows us to assume that despite the difficulty of visualizing the effect of JP by confocal microscopy because of the small size of platelets, the action of JP in platelets must be similar to that observed by Patterson et al. (11) in smooth muscle cells. JP inhibits Ca²⁺ entry stimulated by TG or the physiological agonist, thrombin, without having any effect on the release of Ca²⁺ from the intracellular stores. This suggests that neither accumulation of Ca²⁺ in the stores nor the chemical coupling between the receptor and Ca²⁺ stores mediated by InsP₃ is modified by JP. JP, like phalloidin, has been reported to prevent secretion by formation of a cortical actin barrier at the PM, so excluding cytoplasmic organelles from this region and thus preventing close association between the PM and internal organelles (38).

A different means of investigating whether cortical actin could interfere with the coupling process is through the activation of the moesin, ezrin, radixin family of proteins. These proteins act as linkers between the PM and the actin cytoskeleton (22), and phosphorylation of specific threonine residues activates their function (39). A highly effective means to induce phosphorylation-dependent association of these proteins, and thus actin filaments with the PM, is by using serine/threonine phosphatase inhibitors like calyculin A, a very potent and highly specific inhibitor of protein phosphatases 1 and 2. As shown previously in many cell types, including platelets, calyculin A at concentrations  100 nM induced reorganization of the actin cytoskeleton (11, 24, 25). Actin filaments appear tightly condensed at the PM, without induction of polymerization, and thrombin-induced actin polymerization is prevented (25). In agreement with these observations, our results show that calyculin A did not alter the F-actin content of unstimulated platelets but did abolish TG-induced actin polymerization. Consistent with the results obtained using JP, calyculin A blocked TG-induced Ca²⁺ entry in a concentration-dependent manner. The effect of calyculin A on Ca²⁺ entry has been shown to be mediated exclusively by cytoskeletal modifications (11) and cannot be attributable to a modification of membrane potential or by action as a Ca²⁺ channel blocker (40).

Previous studies using primaquine in rat megakaryocytes have suggested that vesicular trafficking is required for SMCE activation (41). However, it has been reported that the action of primaquine may be a direct inhibition of Ca²⁺ channels (42). Our results with brefeldin A suggest that some form of intracellular transport is required for SMCE in human platelets. Brefeldin A inhibits intracellular transport by inactivating the small GTP-binding protein, Arf (43). Our present results are thus in agreement with a previous study (14) that reported the importance of small GTP-binding proteins in SMCE in human platelets, and consistent with the effect of brefeldin A on the secretory pathway, brefeldin A inhibition of SMCE was reversed by removal of the inhibitor.

The observations described above suggest that it is unlikely that a model based on a diffusible molecule released from the ER to gate a PM Ca²⁺ channel mediates SMCE in human platelets. Such a factor should easily be able to reach the PM after cytoskeletal modifications because InsP₃ was able to function normally to stimulate Ca²⁺ release channels in the ER under these conditions. In contrast, the message originating in the ER to activate PM Ca²⁺ channels was interrupted by the peripheral actin barrier induced by JP or calyculin A. Our results are in favor of a model based either on a physical coupling between the ER and the PM or on the exocytotic insertion of vesicle-carried channels into the PM.

To distinguish whether activation of SMCE is caused by a direct coupling between the ER and the PM or by the insertion of Ca²⁺ channels in the cell membrane we studied the effect of actin cytoskeletal modifications after activation of SMCE. In a "conformational coupling" hypothesis, one would postulate that contact between the ER and the PM would be easily prevented before store depletion and also interrupted after activation of SMCE by disruption of an actin-provided scaffold required to hold the ER near the PM. In contrast, a model based in the insertion of Ca²⁺ channels into the PM might be expected to be affected before store depletion but should become stable enough after activation of SMCE to survive disruption of the actin cytoskeleton. Addition of Cyt D or latrunculin A after activation of SMCE reversed both the actin polymerization and Ca²⁺ entry activated by TG. This suggests a direct and reversible coupling between the ER and the PM, which requires a physical support provided by the actin cytoskeleton (44), is the most likely model for SMCE in human platelets. Consistent with the above, JP, which stabilizes actin polymerization, did not alter preactivated SMCE and prevented the inhibitory effect of Cyt D on both actin polymerization and Ca²⁺ entry, a result in agreement with the higher affinity of JP for actin filaments (20, 21). Indeed, this observation also provides evidence that the action of Cyt D is specifically mediated by actin disassembly because it is prevented when an actin-polymerizing agent is present. In addition, our observations provide evidence that vesicular trafficking is not required for the maintenance of SMCE after activation in human platelets. This conclusion is strongly supported by the lack of effect of brefeldin A on Ca²⁺ entry when added after induction of SMCE, a result in agreement with previous studies in human leukemia cells (45).

As we have reported previously, the activity of small GTP-binding proteins is required for SMCE in human platelets (14). Our new data suggest that these proteins could both mediate ER transport toward the PM and facilitate reversible coupling between these structures as described previously in permeabilized DDT₁MF-2 smooth muscle cells (16), where ER membranes containing functional InsP₃ receptors have been shown to be subjected to a membrane-coupling process activated by GTP and dependent on GTP hydrolysis (16).

In conclusion, in support of the new model of Patterson and co-workers (11) we have provided evidence for a new model for SMCE in platelets based on a reversible trafficking and coupling of the ER with the PM. Because inhibition of actin polymerization using Cyt D and latrunculin A alters SMCE, this model, in contrast to that proposed for smooth muscle cells (11), includes a role for actin filaments in the initiation and maintenance of the coupling process.

	ACKNOWLEDGEMENT

We thank Duncan Towers for assistance in some experiments.

	FOOTNOTES

^* This work was supported in part by The Wellcome Trust Grant 051560.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by a grant from the Junta de Extremadura-Consejería de Educación y Juventud and Fondo Social Europeo, Spain.

^§ To whom correspondence should be addressed. Tel.: 44-1223-333-870; Fax: 44-1223-333-840; E-mail: sos10@cam.ac.uk.

	ABBREVIATIONS

The abbreviations used are: PM, plasma membrane; ER, endoplasmic reticulum; InsP₃, inositol 1,4,5-trisphosphate; SMCE, store-mediated calcium entry; Cyt D, cytochalasin D; TG, thapsigargin; JP, jasplakinolide.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES