Increased Association of Synaptosome-associated Protein of 25 kDa with Syntaxin and Vesicle-associated Membrane Protein following Acrosomal Exocytosis of Sea Urchin Sperm*

Synaptosomal-associated protein of 25 kDa (SNAP-25) is a palmitoylated integral membrane protein expressed almost exclusively in neuronal and neuroendocrine tissues. This protein forms a ternary complex with vesicle-associated membrane protein (VAMP) and syntaxin, which is thought to regulate the fusion of plasma and vesicle membranes during exocytosis. We report the identification of SNAP-25 expressed in sea urchin sperm. Sea urchin SNAP-25 shares greater identity with mammalian SNAP-25 than with mammalian SNAP-23, a ubiquitously expressed homologue believed to regulate membrane fusion in non-neuronal tissues. Sea urchin sperm contain a single exocytotic vesicle, the acrosomal vesicle, whose contents are exposed during the acrosome reaction. Fusion of the plasma membrane with the acrosomal vesicle membrane at multiple points (vesiculation) results in the release of SNAP-25 with the shed acrosome reaction vesicles. A complex containing SNAP-25, syntaxin, and VAMP is present in sperm, as detected by affinity chromatography and immunoprecipitation. Although this complex is present prior to the acrosome reaction, the amount of complex increases over 4-fold following acrosomal exocytosis. These findings support the involvement of SNAP-25 in the invertebrate sperm acrosome reaction, possibly through increased association with VAMP and syntaxin driving the fusion of plasma and acrosomal membranes.

During fertilization in many animals, including sea urchins and mammals, outer investments of the egg trigger the exocytosis of a single exocytotic vesicle in sperm, the acrosomal vesicle (1)(2)(3). This process, the acrosome reaction, exposes the sperm membrane that will fuse with the egg plasma membrane to initiate development. Exocytosis of the acrosomal vesicle during the acrosome reaction is unique in that the plasma and acrosomal membranes fuse at multiple points resulting in formation of acrosome reaction vesicles (ARVs), 1 which are subsequently shed from sperm (4 -6). ARVs can be collected and the associated proteins studied (7).
The sea urchin sperm acrosome reaction is triggered by interaction of a plasma membrane receptor (receptor for egg jelly) with a sulfated fucan in the egg jelly coat (8). The activated receptor regulates ion channels, resulting in the influx of Ca 2ϩ and exocytosis (9). This results in the exposure of bindin and the elongation of the acrosomal process. Acrosomal exocytosis can be induced by ionophores (10), or by the addition of Ca 2ϩ to digitonin-permeabilized sperm (11), making sperm an interesting model for studying Ca 2ϩ -triggered exocytosis.
To understand the mechanism of acrosomal exocytosis, we have identified homologues in sea urchin sperm of proteins believed to be key regulators of membrane fusion during exocytosis. Syntaxin, an intracellular protein integral to the plasma membrane (12,13), and VAMP (synaptobrevin; Refs. 14 and 15), which is integral to the vesicle membrane, are expressed in sea urchin sperm and are shed with the ARVs during the acrosome reaction. Previous work demonstrated that sperm syntaxin and VAMP increase their association following acrosomal exocytosis (7). In neurons, these proteins form a ternary complex with SNAP-25 (16), which is postulated to regulate membrane fusion (17). Here, we describe a sea urchin SNAP-25 homologue expressed in a non-neuronal cell type. Sperm SNAP-25 is found in a complex with syntaxin and VAMP and is also shed with the ARVs. The amount of complex of these three proteins increases following the acrosome reaction. This increase may represent the post-exocytotic state of these proteins in the absence of an endocytic membrane retrieval cycle common to most cells.
Preparation of Proteins-Sea urchins (S. purpuratus) were spawned by intracoelomic injection of 0.5 M KCl. Isolation of sperm heads and * This work was supported by National Institutes of Health Grant HD-12986 (to V. D. V.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF036902.
Immunoblots and Immunoprecipitations-Hen egg yolk antibodies (IgY) were generated against the His-tagged sea urchin SNAP-25 fusion protein. Antibodies were purified from egg yolks as described (21). Briefly, egg yolks were resuspended in 30 ml of 0.01 M potassium phosphate (pH 7.2) and 0.1 M NaCl per yolk. To the resuspended sample was added 30 ml per yolk 7% (w/v) PEG 8000 (Sigma) in the same buffer. The sample was centrifuged at 14,000 ϫ g for 10 min, and the supernatant was brought to 12% PEG 8000 and centrifuged at 14,000 ϫ g for 10 min. The pellet was resuspended in the above buffer (20 ml/yolk) and precipitated by addition of an equal volume of buffer containing 24% PEG and centrifugation at 14,000 ϫ g for 10 min. Pellets were resuspended and dialyzed against the same buffer. Antisera to recombinant sea urchin syntaxin and VAMP were kindly provided by G. M. Wessel and S. Conner (7,22). A polyclonal antibody directed against mammalian SNAP-25 was obtained from Alomone Laboratories. Antibodies were used at a dilution of 1:1000 for immunoblots on Immobilon P membranes (Millipore). Immunoblot signals were detected with Supersignal CL-HRP reagents (Pierce). Peroxidase-conjugated antibodies against hen egg and rabbit antibodies were obtained from Jackson Immunoresearch Laboratories, Inc. Immunoprecipitations using syntaxin antiserum and SNAP-25 antibody coupled beads (see below) were performed as described (7). Consistent loading of the immunoprecipitation samples was confirmed by india ink staining of the immunoblots following exposure. SNAP-25 affinity chromatography was performed by coupling anti-SNAP-25 IgY antibodies to an Affi-Gel Hz support (Bio-Rad) following the manufacturer's instructions. Solubilized sperm proteins were loaded on the column and washed in solubilization buffer (150 mM NaCl, 50 mM Tris, pH 7.4, 1 mM EDTA, 0.4% Nonidet P-40) followed by washing in solubilization buffer containing 0.5 M NaCl. The bound proteins were then eluted with 100 mM citrate, pH 2.8, and neutralized immediately with 0.5 volume of 1 M Tris, pH 8.0.
Sucrose Density Gradients of Solubilized Proteins-Protein samples solubilized in 0.4% Nonidet P-40 were layered on 7.5-25% sucrose gradients (4.4 ml), prepared in solubilization buffer, and centrifuged at 200,000 ϫ g for 16 h at 4°C in a SW60 rotor (Beckman Instruments). Fractions of 190 l were collected, and 50 l of each separated on SDS-polyacrylamide gel electrophoresis and immunoblotted. To estimate the fold increase in ternary complex following acrosomal exocytosis, immunoblots of SNAP-25 present in syntaxin immunoprecipitates of gradient fractions ( Fig. 5; UnR, ARV fraction 6) were normalized for the total amount of SNAP-25 present on gradients of solubilized sperm proteins prior to the acrosome reaction and ARVs. A 1:1:1 stoichiometry of the ternary complex was assumed (30). These immunoblots contained recombinant SNAP-25 dilution standards, and the resulting autoradiograms were scanned and analyzed using the NIH Image program. 2 To quantitate the amounts and molar ratios of SNAP-25 and syntaxin in ARVs and Nonidet P-40 sperm protein extracts, a dilution series for each of the recombinant proteins was created and used on immunoblots with the sperm protein samples. Linear regressions for the syntaxin and SNAP-25 standards were used to calculate the molar ratios of SNAP-25 and syntaxin present in sperm sources from the same exposure used to calculate the regression as described above.

Identification of SNAP-25 cDNA in Sea Urchin
Testis-Multiple overlapping PCR products were amplified from a sea urchin testis cDNA library that encoded a single isoform of SNAP-25. Based on the sequences of these products, primers were designed to amplify the entire sequence encoding sea urchin testis SNAP-25. The sequence encodes a 212-amino acid protein and, when aligned to other SNAP-25 family members, is conserved throughout the sequence (Fig. 1). The length of the protein is conserved with other invertebrate SNAP-25 sequences (Fig. 1, e.g. Leech 25). Sea urchin SNAP-25 is 59% identical to electric ray SNAP-25 (Ray 25, Table I) and 58% identical to human SNAP-25. It is 51% identical to human SNAP-23 (23). Neighbor-joining distance analysis was performed with SNAP-25 protein sequences (Fig. 2). Sea urchin SNAP-25 falls between vertebrate and invertebrate sequences and is an outgroup to a SNAP-25/-23 clade consisting exclusively of vertebrate members.
Identification of SNAP-25 in Sea Urchin Sperm-Other invertebrate SNAP-25 homologues have been identified in neurons at  either the mRNA or protein level (24,25). To determine if sea urchin SNAP-25 is present in sperm (a non-neuronal cell type), antibodies generated against sea urchin SNAP-25 were used to identify SNAP-25 in sperm. These antibodies react with the expressed protein with the His tag removed by thrombin digestion (32 kDa; Fig. 3A, lane 1). SNAP-25 (32 kDa) is present in the ARVs released from sperm during the acrosome reaction (Fig. 3A,  lane 3). Sea urchin and leech (Hirudo medicinalis) SNAP-25 (25) have apparent molecular masses larger than mammalian SNAP-25 homologues. The signals for recombinant and ARV SNAP-25 are blocked (Fig. 3A, lanes 2 and 4) by pretreatment of the antibody with the SNAP-25 fusion protein.
Knowing that SNAP-25 is present in ARVs, we investigated the fate of SNAP-25 in sperm that had been acrosome-reacted by ionophore. SNAP-25 is present in sperm prior to the acrosome reaction (Fig. 3B, lane 2), but is qualitatively lost from acrosome-reacted sperm (Fig. 3B, lane 3). This loss is comparable to the loss of the exocytosis regulatory proteins syntaxin and VAMP from acrosome-reacted sperm with shed ARVs (7).
Isolation of SNAP-25 in a Complex with VAMP and Syntaxin-As SNAP-25 forms a ternary complex with syntaxin and VAMP in neurons, and sea urchin SNAP-25 is present in ARVs isolated from acrosome-reacted sperm, we wished to determine if SNAP-25 is associated with syntaxin and VAMP in sperm. SNAP-25 was affinity-purified from solubilized sperm proteins using an anti-SNAP-25 IgY affinity column (Fig. 4A, lane 3) and compared with the eluate of a control column containing normal IgY (Fig. 4A, lane 2). Multiple bands co-eluted specifically with SNAP-25. To confirm the identity of the co-eluting proteins (Fig. 4A, lane 3), the eluate was immunoblotted with SNAP-25 antibodies (Fig. 4B, lane 1) and syntaxin and VAMP antibodies (Fig. 4B, lane 3). Syntaxin and VAMP co-elute from the affinity column with SNAP-25 and are greatly enriched (Fig. 4A, lane 3). SNAP-25 antibodies reacted with the fulllength protein (32 kDa) as well as a breakdown product (28 kDa) present in the starting material (Fig. 4B, lane 2). The identity of the breakdown product was confirmed by N-terminal protein sequencing, corresponding to a truncation of the N terminus commencing at residue 19 (Gln) of the SNAP-25 sequence. Multiple proteins co-eluted from the SNAP-25 affinity column in addition to syntaxin and VAMP (Fig. 4A, compare  lanes 2 and 3, asterisks), suggesting that SNAP-25 is involved in multiple protein interactions (either directly or indirectly) prior to the acrosome reaction.
Following acrosomal exocytosis, there is a shift in the sedimentation patterns of syntaxin and VAMP on sucrose density gradients to denser fractions having an estimated sedimentation coefficient of 6.2 S (Ref. 7; Fig. 5). To compare this with SNAP-25, detergent-solubilized proteins from unreacted sperm and ARVs were analyzed by sucrose gradient velocity sedimentation. The sedimentation patterns of the proteins from both sources are consistent with each other, and the sedimentation patterns of the major proteins present in both unreacted sperm and ARV protein gradients are similar (data not shown). A proteolytic degradation (7) product co-sediments with fulllength syntaxin following the acrosome reaction (Fig. 5, ARV/ syn), suggesting that this breakdown product is able to associate with the same proteins as is full-length syntaxin. SNAP-25 solubilized from ARVs does not undergo a complete shift; only a portion of SNAP-25 sediments to denser fractions following the acrosome reaction (Fig. 7, ARV/S25), suggesting that  (19). Numbers at the nodes indicate bootstrap values from 1000 replicates. The sequences indicated by tVERT 25a and b represent the terrestrial vertebrate isoforms from human, mouse, and chicken, which are identical to each other at the amino acid level. The tVERT 25a and b isoforms arise from alternative splicing of a duplicated exon (33,34). SNAP-25 is not limiting. To test this, the amount of SNAP-25 and syntaxin present in ARVs and a detergent extract of sperm were calculated (see "Experimental Procedures"). SNAP-25 is present at 30.5 ng/10 g of ARV protein and 11.1 ng/10 g of protein in the sperm detergent extract. Syntaxin is present at 4.9 ng/10 g of ARV protein and 1.8 ng/10 g of protein in the sperm detergent extract. The molar ratios of SNAP-25 to syntaxin were calculated for both ARVs and the detergent extract of sperm. SNAP-25 is 6.9-fold more abundant on a molar basis than syntaxin in both sources, suggesting that SNAP-25 may have multiple binding partners in addition to syntaxin and VAMP. Using syntaxin antibodies, SNAP-25 was co-immunoprecipitated from sperm proteins prior to (Fig. 6A, UnR), but not following (Fig. 6A, R) the acrosome reaction. The preimmune antibodies did not precipitate SNAP-25 (Fig. 6A, PI). To test whether the changes in sedimentation patterns for all three proteins correlates with the formation of a ternary complex, gradient fractions were used as a source to immunoprecipitate the complex. Anti-syntaxin antibodies were used to co-immunoprecipitate VAMP and SNAP-25 from fractions 6 and 9 (Fig. 6B). SNAP-25 does not co-precipitate with syntaxin from fraction 9 (Fig. 6B, UnR/S25) prior to the acrosome reaction despite the abundance of SNAP-25 (Fig. 5) in these samples. VAMP is also absent from syntaxin immunoprecipitates of fraction 9 samples prior to the acrosome reaction (Fig.  6B, UnR/V). However, SNAP-25 and VAMP are co-precipitated with syntaxin from fraction 6 (Fig. 6B, ARV/V, S25) samples following the acrosome reaction (ARVs), correlating with the change in syntaxin's sedimentation pattern. SNAP-25 and VAMP are also co-precipitated to a lesser extent with syntaxin from fraction 6 samples prior to the acrosome reaction (Fig. 6B). Similarly, anti-SNAP-25 antibodies co-precipitated syntaxin and VAMP from gradient fraction 6 samples both prior to and following the acrosome reaction (Fig. 6C) The syntaxin breakdown product present following the acrosome reaction was also co-precipitated from fraction 6 ( Fig. 6C) in agreement with its co-sedimenting with full-length syntaxin on sucrose gradients. While the ternary complex (syntaxin, VAMP, and SNAP-25) is present prior to the acrosome reaction, it appears to increase over 4-fold in abundance following the acrosome reaction based on the amount of SNAP-25 present in syntaxin immunoprecipitates of the fraction 6 (6.2 S) gradient samples (see "Experimental Procedures"). Similarly, VAMP increases in abundance in syntaxin immunoprecipitates following the acrosome reaction (Fig. 6B), correlating with the observed shifts in the syntaxin and VAMP sedimentation patterns following acrosomal exocytosis (7). DISCUSSION SNAP-25 is an axonally transported protein in mammalian neurons where it localizes to nerve terminals (26,27). At the presynaptic membrane, SNAP-25 regulates neurotransmission by forming a ternary complex with syntaxin and VAMP (28 -31). In this report, we have identified a sea urchin homologue of SNAP-25. This is the first report of SNAP-25 being expressed in the sperm of any animal. Phylogenetic analysis of SNAP-25 protein sequences suggests that the mammalian SNAP-25 homologue, SNAP-23 (23,32), arose by a gene duplication in the vertebrate lineage to function in non-neuronal cell types. In agreement with this idea, only a single isoform of SNAP-25 was amplified from a sea urchin testis cDNA library despite the use FIG. 6. Immunoprecipitation of sucrose gradient fractions with syntaxin and SNAP-25 antibodies. A, co-immunoprecipitation of SNAP-25 with syntaxin antibodies. SNAP-25 (S25) was detected on immunoblots of syntaxin immunoprecipitates from solubilized sperm proteins (UnR) but absent from immunoprecipitates with syntaxin preimmune antibodies (PI). SNAP-25 was also absent from syntaxin immunoprecipitates of solubilized proteins from acrosome-reacted sperm (R). B, immunoprecipitation with syntaxin antibodies of VAMP (V) and SNAP-25 (S25) from the sucrose gradient fractions 6 and 9 (shown in Fig. 5). VAMP and SNAP-25 are present in syntaxin immunoprecipitates of fraction 6 from gradients containing solubilized sperm proteins (unreacted, UnR) and ARV proteins, but are absent from fraction 9 despite the abundance of syntaxin, SNAP-25, and VAMP (7) in these fractions. When controlled for the amount of SNAP-25 present on gradients of solubilized sperm proteins and ARVs (Fig. 5), the amount of ternary complex increases 4.3-fold (see "Experimental Procedures"). C, syntaxin was co-immunoprecipitated with SNAP-25 from solubilized sperm proteins (UnR) using SNAP-25 antibodies coupled to Affi-Gel Hz beads (see "Experimental Procedures" and Fig. 4) but blocked by pretreatment of the antibody beads with 20 g of expressed SNAP-25 (Block). D, immunoprecipitation of syntaxin (syn) and VAMP (V) with the SNAP-25 support-coupled antibodies from fraction 6 of the gradients containing solubilized sperm proteins (unreacted, UnR) and ARV proteins. A syntaxin degradation product (34 kDa) previously noted (7) is also present in SNAP-25 immunoprecipitates, in agreement with the co-sedimentation of this breakdown product with full-length syntaxin (36 kDa) on sucrose gradients of solubilized ARVs (Fig. 5). of degenerate primers designed to amplify both SNAP-25 and -23. This suggests that the sea urchin testis does not express a SNAP-23 homologue. The identification of SNAP-23 in mouse testis (32) raises the possibility that SNAP-23 has evolved to function in non-neuronal cell types in vertebrates, in agreement with the phylogenetic sequence analysis presented in Fig. 2.
Sea urchin sperm SNAP-25 shares many properties with syntaxin and VAMP. All three proteins are shed from sperm with the ARVs during the acrosome reaction. These proteins can be isolated as a complex by immunoprecipitation and affinity chromatography prior to the acrosome reaction. Following acrosomal exocytosis, there is greater than a 4-fold increase in the amount of ternary complex present, as demonstrated by the changes in the sedimentation patterns of syntaxin, VAMP, and SNAP-25 and by the immunoprecipitation of gradient fractions. Interestingly, only a portion of SNAP-25 sediments into denser fractions following the acrosome reaction. The 6.9-fold molar excess of SNAP-25 over syntaxin in sperm extracts and ARVs suggests that SNAP-25 is not limiting for ternary complex formation. However, all the syntaxin present in ARVs completely shifts to denser fractions after the acrosome reaction. The estimated sedimentation coefficient for the syntaxin peak fractions following the acrosome reaction is 6.2 S (7), which correlates with the size of the ternary complex based on the sedimentation of molecular weight standards.
The presence of the ternary complex in ARVs may represent the post-exocytotic state of these proteins in the same lipid bilayer, and the increase in complex formation following the acrosome reaction may be an exocytotic intermediate formed in the absence a subsequent endocytic cycle.
As the amount of ternary complex increases following acrosomal exocytosis, it will be of interest to determine what regulates complex formation prior to the acrosome reaction and how this correlates with the influx of Ca 2ϩ , which triggers acrosomal exocytosis.