SNAP-23 Is a Target for Calpain Cleavage in Activated Platelets*

The role of calpain in platelet function is generally associated with aggregation and clot retraction. In this report, data are presented to show that one component of the platelet secretory machinery, SNAP-23, is specifically cleaved by calpain in activated cells. Other proteins of the membrane fusion machinery,e.g. syntaxins 2 and 4 and α-SNAP, are not affected.In vitro studies, using permeabilized platelets, demonstrate that cleavage is time- and calcium-dependent. Analysis of SNAP-23 cleavage products suggests that the calpain cleavage site(s) is in the C-terminal third of the molecule potentially between the cysteine-rich acyl attachment sites and the C-terminal coiled-coil domain. The time course of cleavage is most consistent with late calpain-mediated events such as pp60c-src cleavage, but not early events such as protein-tyrosine phosphatase-1B activation. SNAP-23 cleavage is inhibited by calpeptin, calpastatin, calpain inhibitor IV, and E-64d, but not by caspase 3 inhibitor III or cathepsin inhibitor I. When tested for their effect on secretion, none of the calpain-specific inhibitors significantly affected release of soluble components from any of the three platelet granule storage pools. These results indicate that SNAP-23 cleavage occurs after granule release and therefore may play a role in affecting granule membrane exteriorization. This is consistent with the ultrastructural morphology of calpeptin-treated platelets after activation.

The role of calpain in platelet function is generally associated with aggregation and clot retraction. In this report, data are presented to show that one component of the platelet secretory machinery, SNAP-23, is specifically cleaved by calpain in activated cells. Other proteins of the membrane fusion machinery, e.g. syntaxins 2 and 4 and ␣-SNAP, are not affected. In vitro studies, using permeabilized platelets, demonstrate that cleavage is time-and calcium-dependent. Analysis of SNAP-23 cleavage products suggests that the calpain cleavage site(s) is in the C-terminal third of the molecule potentially between the cysteine-rich acyl attachment sites and the C-terminal coiled-coil domain. The time course of cleavage is most consistent with late calpain-mediated events such as pp60 c-src cleavage, but not early events such as protein-tyrosine phosphatase-1B activation. SNAP-23 cleavage is inhibited by calpeptin, calpastatin, calpain inhibitor IV, and E-64d, but not by caspase 3 inhibitor III or cathepsin inhibitor I. When tested for their effect on secretion, none of the calpainspecific inhibitors significantly affected release of soluble components from any of the three platelet granule storage pools. These results indicate that SNAP-23 cleavage occurs after granule release and therefore may play a role in affecting granule membrane exteriorization. This is consistent with the ultrastructural morphology of calpeptin-treated platelets after activation.
Platelets play a critical role in hemostasis by recognizing vascular lesions, binding to the damaged site, and then sealing the breach (1). Two general steps are essential to this function. First, platelets must adhere to the site of injury and secrete the components required for primary clot formation and wound healing (2). Adherence and initial activation is mediated by specific cell surface receptors. The secretory process is mediated by integral membrane proteins contributed from the granule, called v-SNAREs, 1 and from the plasma membrane, called t-SNAREs (3)(4)(5)(6)(7)(8). These proteins form a heterotrimeric complex that facilitates the fusion of granule membranes with the open canalicular system (9). In the second phase, there is a rapid and dramatic reorganization of the cytoskeleton and plasma membrane (2). The intracellular granule membranes, which are now fused with the peripheral plasma membrane, are exteriorized, thereby increasing the overall surface area of the cell, the platelet-platelet contacts, and the platelet to vascular wall contacts. Three central processes characterize this second phase of platelet activation: cell spreading, cytoskeletal rearrangements, and microvesicle blebbing. One key enzyme in these late phase events is the cysteine protease, calpain.
Calpains are heterodimers consisting of a conserved 30-kDa regulatory subunit and a variable, catalytic 80-kDa subunit (10). Calcium binding to the multiple EF hands in the 80-kDa subunit triggers translocation of calpain from the cytosol to the plasma membrane (17). At the plasma membrane, calpain maintains membrane association through calcium-dependent C2 domain-phospholipid binding (18). The membrane-bound calpain undergoes sequential autolysis of the catalytic domain, generating two distinct enzymes: the 78-and 76-kDa forms (17, 19 -22). These two calpain forms display distinct substrate specificities, and each can either enhance or diminish the activity of their substrate through cleavage. For example, the 78-kDa form cleaves protein-tyrosine phosphatase-1B (PTP-1B) early in platelet activation and stimulates phosphatase activity (23)(24)(25), whereas the fully autolyzed 76-kDa calpain is required for the cleavage and inactivation of pp60 c-src (25,26). Calpain is also responsible for the limited proteolysis of a wide spectrum of platelet proteins including talin, filamin, ␣-actinin, integrin ␣ IIb ␤ 3 , and protein kinase C, linking the function of calpain to a variety of processes including cell adhesion, cell motility, cell signaling, and cytoskeletal arrangements (27)(28)(29)(30)(31). However, the role of calpain in secretory processes is less defined. In experiments using intact platelets and a membrane permeant calpain inhibitor, Croce et al. (32) demonstrated that calpain activity is required for secretion from alpha granules. In contrast, mice deficient in -calpain display normal bleeding times and normal platelet granule secretion (33). In other cell types, such as mouse pancreatic islet cells, calpain inhibitors have been shown to actually increase insulin secretion (34). Still other reports indicate that elements of the secretory ma-chinery (e.g. syntaxin 1, VAMP 2, and SNAP-25) are cleaved by calpain in stimulated alveolar epithelial cells (35).
The experiments reported here seek to further probe the role of calpain in the secretory process. Initial experiments demonstrate that calpain cleaves one of the central elements of the platelet secretory machinery, SNAP-23, a t-SNARE required for all three platelet release events (3,(5)(6)(7). Calpain cleavage of SNAP-23 was dependent on platelet activation in both intact and permeabilized cells and was specific, because other components of the secretory machinery, specifically other t-SNAREs, were left intact. Calpain inhibitors, however, were without effect on secretion from dense core granules, alpha granules, or lysosomes. This data, together with the time course of SNAP-23 cleavage, suggests that it is a late event, potentially occurring after granule release. The role of calpain cleavage of SNAP-23 appears to be related to some post-secretion event, perhaps granule membrane mobilization. Consistently, calpeptin-treated platelets show an increase in intracellular membrane structures following stimulation, suggesting a defect in membrane exteriorization.
Calpain Activation in Intact Platelets-Washed platelets were resuspended in Hepes-buffered saline (20 mM Hepes/NaOH, pH 7.4, 154 mM NaCl, 2.7 mM KCl), and 100-l aliquots were incubated with 300 M calpeptin for 30 min at 37°C. Reactions were then recalcified with 2 mM CaCl 2 for 5 min and stimulated with 1 M A23187 for the indicated times. Reactions were solubilized in 5ϫ SDS sample buffer and analyzed by SDS-PAGE and Western blotting.
Calpain Activation in Permeabilized Platelets-Washed platelets were resuspended in Buffer A (20 mM Hepes/NaOH, pH 7.4, 120 mM sodium glutamate, 5 mM potassium glutamate, 2.5 mM EDTA, 2.5 mM EGTA, 3.15 mM MgCl 2 , 1 mM dithiothreitol) and 100-l platelet aliquots were incubated with 0.8 units/ml SLO, 4 mM ATP in Buffer A, and calpain inhibitors (as indicated) for 10 min at 25°C. Platelets were activated with CaCl 2 (at given concentrations) for the specified times. Reactions were solubilized in 5ϫ SDS sample buffer and analyzed by SDS-PAGE and Western blotting.
Permeabilized Platelet Assay for the Detection of Dense Core, Lysosome, and Alpha Granule Secretion-Platelets were assayed for dense core release of [ 3 H]5-HT, lysosomal release of ␤-hexosaminidase, and alpha granule release of PF4, as described previously (5)(6)(7), with the following modifications. Prior to isolation of platelets, the platelet-rich plasma was incubated with 0.4 Ci/ml [ 3 H]5-HT for 45 min at 37°C with gentle shaking. Washed platelets were resuspended in Buffer A and 100-l platelet aliquots were incubated with 0.8 units/ml SLO, 4 mM ATP in Buffer A, and calpain inhibitors (as indicated) for 10 min at 25°C. Platelets were then stimulated to secrete granule contents with 100 M CaCl 2 for 5 min. The reactions were centrifuged at 13,000 rpm for 1 min and the supernatants were retained. The remaining platelet pellet was solubilized in an equal volume of 0.5% Triton X-100 in Buffer A. Equal volumes of both supernatant and pellet fractions were then assayed for all granule exocytosis events. The data were tabulated as the percentage of marker released compared with the total marker present in each platelet reaction for lysosome and dense core granule assays, whereas alpha granule secretion was monitored as described previously (7).
Electrophoresis and Western Blotting-Visualization of calpain cleavage of talin (235 kDa) and filamin (280 kDa) required the use of 7.5% SDS-PAGE gels with a ratio of 29.6% acrylamide to 0.4% bisacrylamide (Fisher Scientific, Pittsburgh, PA). All other gels had a composition of 29% acrylamide:1% bisacrylamide. For all Western blotting experiments, the enhanced chemiluminescence (ECL) detection system (Pierce) was used with secondary anti-mouse and anti-rabbit IgG, peroxidase-conjugated antibodies purchased from Sigma.

Analysis of SNAP-23 Cleavage Fragments by MALDI-TOF Mass Spectrometry (MS)-SNAP-23-encoding
DNA was inserted into the pQE-9 vector (Qiagen, Chatsworth, CA) as described previously (5). Recombinant His 6 -tagged SNAP-23 was prepared from Escherichia coli and purified by Ni 2ϩ -NTA affinity chromatography (Qiagen) (41). The His 6 tag and linker sequence are represented in the MS data and for ease of interpretation have the following numerical nomenclature, where Met 1 represents the natural start residue of SNAP-23: Met Ϫ12 -Arg-Gly-Ser-His-His-His-His-His-His-Gly-Ser-Met 1 -Asp 2 . . . Reactions (50 l) containing 20 g of SNAP-23 alone, 2 g of -calpain alone, or 20 g of SNAP-23 and 2 g of -calpain were incubated in the presence of 100 M CaCl 2 in calpain cleavage buffer (20 mM Tris-HCl, pH 7.5, 50 mM NaCl, 5 mM 2-mercaptoethanol) for 30 min at 37°C. Reactions containing 50 M calpeptin, 20 g of SNAP-23, and 2 g of -calpain prevented SNAP-23 degradation (data not shown). All reactions were stopped simultaneously with the addition of 5ϫ SDS-sample buffer. Entire reactions were loaded on 15% SDS-PAGE gels, and the separated bands were excised for analysis by MALDI-TOF MS. Digestion with trypsin and MALDI-TOF MS were performed Biomolecular Mass Spectrometry Core Laboratory at the University of Louisville (www. louisville.edu/ϳwmpier01/biomassspec.htm) on a fee-for-service basis.
Electron Microscopy-Permeabilized platelets were preincubated with 300 M calpeptin (as indicated) and were either left resting or stimulated for 10 min with 500 M CaCl 2 . The platelets were then prepared for ultrastructural analysis by electron microscopy as described previously (6,7), with the following modifications. An equal volume of 6% glutaraldehyde (Electron Microscopy Sciences, Ft. Washington, PA), 80 mM lysine (Sigma) in 0.1 M Sorenson's phosphate buffer (8.1 mg/ml KH 2 PO 4 , 1.88 mg/ml Na 2 HPO 4 ) was added to the reactions, and the platelets were incubated at 4°C for 1.5 h. The platelets were washed three times for 15 min with 0.1 M Sorenson's phosphate buffer and osmicated with 1% OsO 4 in 0.1 M Sorenson's for 30 min on ice. Following two brief washes in ice-cold H 2 O, 4% aqueous uranyl acetate was added to the platelets and they were incubated 1 h at 4°C on a rotator. The platelets were washed with 0.1 M Sorenson's and dehydrated in a series of ethyl alcohols for 5 min. The platelets were rinsed with three changes of propylene oxide and infiltrated overnight in a 1:1 mixture of propylene oxide and Spurr's resin. Samples were embedded in Spurr's resin at 50°C for 48 h. Polymerized blocks were sectioned, mounted on copper grids, and examined by transmission electron microscopy as described previously (6,7).

RESULTS
Degradation of SNAP-23 in Activated Platelets-Initial experiments indicated that the t-SNARE, SNAP-23, was de-graded in activated platelets (Fig. 1). Stimulation of intact platelets with 1 M A23187 in the presence of 2 mM CaCl 2 caused a time-dependent degradation of talin and filamin (ABP-280), which was maximal at 5 min. Talin and filamin cleavage was inhibited by 300 M calpeptin, consistent with the fact that these two proteins are known calpain substrates (27). The proteins of the platelet secretory machinery were then examined by Western blotting (Fig. 1), and it was found that SNAP-23 was also degraded in a time-dependent manner in intact cells. Other components of the platelet secretory machinery, such as the t-SNARE syntaxin 4, which is known to associate with SNAP-23 (3,(5)(6)(7), and the adaptor protein ␣-SNAP, were not degraded under these same conditions. As with talin and filamin cleavage, SNAP-23 degradation was inhibited by calpeptin, suggesting a role for calpain.
The permeabilized platelet assay system (5-7), used to probe platelet exocytosis, was next tested. In Fig. 2, concentrations of free calcium greater than 10 M were required to induce SNAP-23 degradation in SLO-permeabilized platelets. A similar free calcium concentration is required to induce secretion from permeabilized platelets (5-7). In the same reactions, talin and filamin were also degraded in response to greater than 10 M calcium. Because the permeabilized platelet system recapitulates the cleavage events initially observed in intact cells (Fig. 1), it was used for subsequent studies. Fig. 3 shows that SNAP-23 degradation in permeabilized platelets was timedependent ( Fig. 3A) and reached a maximum after 5 min of stimulation at 37°C. At each point, SNAP-23 cleavage was inhibited by 400 M E-64d (E). SNAP-23 degradation was delayed relative to PTP-1B, which was significantly cleaved by 45 s. (Fig. 3B). The antibody used for Western blotting does not recognize the cleavage product of PTP-1B. PTP-1B is indicative of an "early" calpain substrate because it is cleaved rapidly after activation (25). As seen in Fig. 3C, the time course of SNAP-23 cleavage was more similar to "late" substrates, such as pp60 c-src , which is cleaved by calpain later in the platelet activation process (25). Interestingly, the production of the lower molecular weight form of pp60 c-src was less efficient than that seen for cleavage of SNAP-23.
SNAP-23 Degradation Is Specific and Calpain-dependent-The data to this point indicate that SNAP-23 is specifically degraded by calpain in platelets upon activation. To further address these points, permeabilized platelets were either held in a resting state (ϾpCa ϭ 9, R), stimulated with CaCl 2 (S), or stimulated in the presence of calpeptin (C). To confirm activation of calpain, the reactions were analyzed by SDS-PAGE to demonstrate that both talin and filamin were cleaved (data not shown). Inhibition of talin and filamin cleavage confirmed that the calpeptin treatment was effective (data not shown). These reactions were then probed by Western blotting using a number of antibodies to proteins known to be part of the secretory machinery in platelets. SNAP-23 was cleaved in the stimulated cells (Fig. 4, S), but the six other proteins tested were not cleaved. Of specific note, syntaxins 2 and 4, which are both known to complex with SNAP-23 in platelets (3, 5-7) and would represent equally accessible substrates, are not cleaved. Although this is not a complete selection of platelet secretory machinery proteins, it does represent a significant number of  proteins that could associate with SNAP-23 in a platelet during the secretion process.
SNAP-23 cleavage was also sensitive to other calpain inhibitors. In permeabilized cells, calcium-stimulated cleavage of SNAP-23 was inhibited by calpastatin, calpain inhibitor IV, as well as by E-64d and calpeptin (Fig. 5). In the same reactions, each inhibitor blocked cleavage of talin and filamin (data not shown). As shown, the carrier solvents, Me 2 SO and ethanol, had no effect (Fig. 5A). The effect of E-64d is particularly notable, given that this inhibitor blocks conversion of the partially activated 78-kDa calpain to the fully activated 76-kDa form (25). The inhibitory effect of E-64d implies that fully activated calpain is required for SNAP-23 cleavage. Platelets contain other proteases, which also do not appear to play a role in SNAP-23 degradation. A broad spectrum inhibitor of caspase, caspase 3 inhibitor III (42,43), had no effect on SNAP-23 cleavage, nor did the broad spectrum cathepsin inhibitor I (44,45). This indicates that neither caspases nor cathepsins are involved in SNAP-23 cleavage. The effect of the four different calpain inhibitors confirms that calpain is indeed cleaving SNAP-23 upon Ca 2ϩ -mediated activation of the permeabilized platelets. Further, when intact platelets were treated with dibucaine, a known calpain activator (46), SNAP-23 degradation was dose-dependent (Fig. 5C). Dibucaine-induced cleavage of SNAP-23 was detectable as early as 5 min but was fully complete by 10 min, comparable with other calpain substrates, e.g. talin and filamin (data not shown).
To further characterize SNAP-23 cleavage, recombinant SNAP-23 was incubated with purified calpain. In Fig. 6, calpain completely cleaved recombinant SNAP-23 during a 30min incubation (lane I). Unfortunately, the anti-SNAP-23 antibody used for Western blotting was unable to recognize any of the cleavage products that were detected by Coomassie staining (data not shown). The starting preparation of recombinant SNAP-23 (lane II) contained one degradation product (band 1) that could not be easily purified from the full-length protein. This protein appears to represent a cleavage of the most C-terminal end (see below). Like intact SNAP-23, this protein was also a substrate for calpain and disappeared upon incubation with calpain in the presence of calcium. In the complete reaction (lane I), several lower molecular weight bands (bands 2-5) were detected. The appearance of each band was dependent on calpain digestion and did not appear in either the SNAP-23 alone (lane II) or the calpain alone reactions (lane III). Bands 2-5 (from lane I), as well as band 1 from the recombinant protein preparation (lane II), were excised from the SDS-PAGE gel and digested with trypsin, and the tryptic peptides were analyzed by MALDI-TOF MS. Examination of the spectra showed that each band (bands 1-5) was derived from recombinant SNAP-23. Twelve peptides from band 1 (A-L, Fig. 6B) were identified based on molecular weight identities with those predicted from a trypsin digestion of recombinant His 6 -SNAP-23. These 12 peptides extend over the length of the protein, but none of the potential peptides corresponding to the immediate C terminus were detected. The four internal regions not represented in the MS spectrum correspond to sequences high in both lysine and arginine, which would have yielded peptides too small (Ͻ600 m/z) for accurate detection and interpretation. For each of the four calpain cleavage products (bands 2-5), the peptides identified were derived from the N-terminal region of the molecule (A-I). This was confirmed by Western blotting using INDIA HisProbe-HRP (Pierce) to detect the N-terminal His 6 tag (data not shown). The most C-terminal of the peptides for each fragment corresponded to amino acids Thr 101 -Lys 117 . None of the large peptides (Ͼ600 m/z) predicted to be derived from the C terminus were detected. Based on this analysis, the cleavage site is unlikely to be N-terminal of amino acid Lys 117 as most of those peptides were accounted for. The lower molecular weight of the calpain cleavage products, together with the lack of C-terminal-derived peptides and the fact that the C-terminal derived peptides could be detected from band 1, suggests that the C-terminal region of SNAP-23 must contain the calpain cleavage site. Because there was heterogeneity in the cleavage product size (lane I), it is possible that the C terminus contains multiple calpain cleavage sites. Although these data do not allow for an exact determination of the cleavage site(s), it does suggest that the site is somewhere C-terminal of amino acid Lys 117 , possibly between Lys 117 and Ile 143 . The site may lie between the cysteine-rich region (Fig.  6B, black bar; Cys 79 -Cys 87 ), which serves as a palmitoylation site for membrane anchorage (47), and the C-terminal coiledcoil domain (Fig. 6B, gray bar; Val 122 -Asp 201 ), which is necessary for SNARE complex formation (48). Given that the consensus site for calpain cleavage is rather ill defined (49 -51), further analysis will be required to more accurately characterize the calpain recognition and cleavage site in SNAP-23.
Role of SNAP-23 Cleavage by Calpain-The role of calpain in platelet exocytosis has been controversial. Some calpain inhibitors appear to block release (32), whereas platelets from -calpain knockout mice show normal secretion, but reduced clot retraction (33). Given that SNAP-23 is involved in all three of the platelet secretion events (3,(5)(6)(7) and that it is a calpain substrate, the next experiments focused on the effect of calpain inhibitors in platelet secretion. Permeabilized platelets were incubated with increasing concentrations of calpeptin, and then stimulated to secrete by increasing the free calcium concentration (Fig. 7A). Calpeptin at all concentrations inhibited talin and filamin cleavage (data not shown), but had no effect on secretion from dense core granules, from alpha granules, or from lysosomes. If there was any consistent effect, secretion was slightly augmented (10 -50%) by calpeptin treatment (Fig.  7 and data not shown). Three other calpain inhibitors were also tested (Fig. 7B), and, although each inhibited talin and filamin cleavage (data not shown), there was no effect on secretion from any of the intraplatelet stores. These experiments indicate that calpain activity is not required for secretion from any of the three platelet granule stores. Although perhaps less signifi- cant, these data imply that calpain activation may actually be inhibitory to platelet secretion.
To gain further insight into the role that calpain cleavage of SNAP-23 may play, the morphology of the permeabilized platelets was examined. Permeabilized platelets were either treated or not with 300 M calpeptin and then stimulated to secrete by increasing free calcium. The platelets were subsequently fixed and prepared for examination by electron microscopy. Morphometric analysis of permeabilized platelets is difficult because of an inability to precisely maintain subcellular structures during fixation. However, granules and mitochondria can easily be distinguished based on the presence of a double membrane and internal cristae in the mitochondria. In Fig. 8A, resting platelets show normal distribution of granules (G) and sub-platelet organelles (e.g. mitochondria (M)). Upon stimulation for 10 min, the permeabilized platelets appeared devoid of granules and generally lacked significant intracellular membrane structures other than mitochondria (Fig. 8, D-F). This is consistent with granule exocytosis and incorporation of the granule membranes into the peripheral plasma membrane. When the platelets were preincubated with 300 M calpeptin and subsequently stimulated, there was a consistent decrease in cargo-containing granules as expected based on the secretion data discussed above (Fig. 8, G-I), but there was a significant increase in the amount of intracellular membrane structures (Fig. 8G, arrowheads), relative to the untreated controls. These large membranous structures frequently appeared to contain diffuse protein complexes consistent with a releasing granule that had fused to the open canalicular system. This phenotype might be expected if the granule membranes failed to extrude to the peripheral plasma membrane once secretion was completed. To further document this phenotype, calpeptin-treated, stimulated platelets were classified based on the presence of internal membrane accumulation as seen in Fig. 8 (G-I). Only platelets with visible mitochondria were counted to assure that comparable sections were being examined. Of the 211 cells examined, 198 (94%) had visible internal membrane accumulation as depicted in Fig. 8 (G-I). DISCUSSION In this paper, we demonstrate that SNAP-23, which is used for all three platelet secretion events (5)(6)(7)52), is a substrate for calpain cleavage. The degradation of SNAP-23 is specific in that only calpain inhibitors block cleavage and that other proteins involved in secretion are not degraded. The time course for SNAP-23 cleavage is consistent with its being a "late" substrate, and inhibition by E-64d suggests that full activation of calpain is required. This late cleavage event is also consistent with the lack of a significant effect of calpain inhibitors on platelet secretion. Secretion is a relatively rapid step, even in the permeabilized assay, reaching completion before 5 min (5). It would appear that SNAP-23 cleavage occurs after secretion is complete.
A consensus calpain cleavage site has been difficult to determine, based on the wide range of substrates reported (49 -51). The exact position of the calpain cleavage site in SNAP-23 cannot be determined based on the data presented here; however, the data do indicate where SNAP-23 is not cleaved and suggest where cleavage could occur. Based on the recovery of the peptides from the calpain cleavage products, it is possible to conclude that the site is not N-terminal of amino acid Lys 117 . If the cleavage site is close to this residue, then it falls in a region between the cysteine-rich, acylated domain and the coiled-coil domain (see Fig. 6B). This is a particularly intriguing site as it is predicted to be readily accessible even when SNAP-23 is in complex with other SNAREs. Based on the crystal structure of SNAP-25 (a SNAP-23 homologue) in complex with syntaxin 1 and VAMP-2 (48), this potential cleavage site is in a membrane distal region at the end of the SNARE complex where it would be exposed and easily accessible. Such accessibility of both complexed and free pools may, in part, explain the efficiency by which calpain cleaves platelet SNAP-23.
Croce et al. (32) showed that the calpain inhibitor, calpastat-ALA, inhibits secretion from alpha granules. The assay system used relied on the surface exposure of the granule membrane protein, P-selectin. This scheme differs from the assay reported here which relies on release of the soluble protein, PF4. One explanation that could rectify these two observations would be that calpain activation is in some way important for surface mobilization of granule membranes after fusion. SNAP-23-containing SNARE complexes or other cytoskeletal elements could restrict granule membrane protein mobility and lessen P-selectin exposure, but not PF4 release. Although this explanation is difficult to rigorously test, one might expect that calpeptintreated platelets would show an increase in internal membranes in stimulated platelets when compared with untreated controls. In Fig. 8, resting platelets (panels A-C) show a normal distribution of granules and organelles (e.g. mitochondria). The granules are not present in the stimulated platelets (panels D-F), nor are there any significant membrane structures other than the normal organelles such as mitochondria. This is the result of normal membrane extrusion seen in activated platelets and is consistent with exposure of granule membrane proteins such as P-selectin on the peripheral plasma membrane. After calpeptin treatment, the majority (Ͼ90%) of stimulated platelets show a consistent increase in internal membrane structures (panels G-I) that was not seen in untreated cells. These structures were consistently seen in all fields of platelets examined. The striking feature is that the membrane structures frequently appeared to contain protein complexes that may represent secreted granule contents. The connection between this phenotype and the calpain-mediated cleavage of SNAP-23 is tenuous at best; however, a role for calpain in membrane mobilization events may explain why calpain inhibitors affect P-selectin exposure, but not the release of PF4.
The importance of SNAP-23 cleavage may also lie in the product produced. Given the potential SNAP-23 cleavage site suggested by this report, it is possible that calpain releases the C-terminal coiled-coil domain. This domain is an effective inhibitor of secretion (5); therefore, its release into the platelet cytosol could limit further secretory steps. Because a similar cleavage of SNAP-25 by botulinum toxin E also eliminates secretion (53), cleavage of the intact SNARE and generation of an inhibitor peptide would be a doubly effective method to prevent platelet exocytosis. This scenario, although possible, seems unlikely because secretion appears to already be complete before SNAP-23 is cleaved. However, it should be noted that inhibition of calpain activation did slightly augment release from all three granular stores (Fig. 7A). Further analysis will be required to understand the physiological significance of SNAP-23 cleavage by calpain.