Conserved N-terminal Cysteine Motif Is Essential for Homo- and Heterodimer Formation of Synaptotagmins III, V, VI, and X*

The synaptotagmins now constitute a large family of membrane proteins characterized by one transmembrane region and two C2 domains. Dimerization of synaptotagmin (Syt) I, a putative low affinity Ca2+ sensor for neurotransmitter release, is thought to be important for expression of function during exocytosis of synaptic vesicles. However, little is known about the self-dimerization properties of other isoforms. In this study, we demonstrate that a subclass of synaptotagmins (III, V, VI, and X) (Ibata, K., Fukuda, M., and Mikoshiba, K. (1998) J. Biol. Chem. 273, 12267–12273) forms β-mercaptoethanol-sensitive homodimers and identify three evolutionarily conserved cysteine residues at the N terminus (N-terminal cysteine motif, at amino acids 10, 21, and 33 of mouse Syt III) that are not conserved in other isoforms. Site-directed mutagenesis of these cysteine residues and co-immunoprecipitation experiments clearly indicate that the first cysteine residue is essential for the stable homodimer formation of Syt III, V, or VI, and heterodimer formation between Syts III, V, VI, and X. We also show that native Syt III from mouse brain forms a β-mercaptoethanol-sensitive homodimer. Our results suggest that the cysteine-based heterodimerization between Syt III and Syt V, VI, or X, which have different biochemical properties, may modulate the proposed function of Syt III as a putative high affinity Ca2+ sensor for neurotransmitter release.

Recent genetic and biochemical evidence has indicated that the proteins involved in vesicle traffic are evolutionarily conserved and form a large family, each member of which is thought to regulate constitutive and/or regulated vesicle traffic (1)(2)(3). Synaptotagmin is one such protein family and is characterized by a short amino terminus, a single transmembrane region, and two C2 domains (known as the C2A and C2B domains) homologous to the C2 regulatory region of protein kinase C (reviewed in Refs. 4 -7). To date, 12 members (synaptotagmins I-XI and Srg1) have been described in rat or mouse (8 -19). With the exception of synaptotagmin (Syt) 1 I, which functions in the Ca 2ϩ -regulated exocytosis of secretory vesicles from neurons or some endocrine cells (Refs. 20 -24 and reviewed in Refs. 4 -7), the exact localization and function of the other synaptotagmins remain largely unknown.
Genetic analysis of Drosophila synaptotagmin mutants (25) and in vitro biochemical studies (26 -30) have suggested that self-oligomerization is important for the Ca 2ϩ -sensing function of Syt I. Self-dimerization of Syt I occurs through a Ca 2ϩindependent multimerization, which is probably mediated by a region just downstream of the transmembrane region (31), or through a Ca 2ϩ -dependent dimerization mediated by the C2B domain (26 -30). Although only the latter interaction has been well studied, the former one is also important for bridging the two molecules before Ca 2ϩ ions enter through Ca 2ϩ channels to achieve efficient Ca 2ϩ -dependent dimerization via the C2B domain within a very short period of synaptic vesicle exocytosis (Ͻ300 s), the fusion of synaptic vesicle to presynaptic plasma membrane. However, it remains undetermined whether other isoforms (Syts III-XI and Srg1) also show Ca 2ϩ -independent interactions and whether they are important for the function of Syts III-XI and Srg1.
In this study, we cloned cDNAs covering the full coding region of mouse Syts I-XI and examined their Ca 2ϩ -independent dimerization properties by expressing T7-tagged Syts I-XI in COS-7 cells. We identified a novel conserved cysteine motif in the N-terminal domain present only in the previously described subclass of synaptotagmins (Syts III, V, VI, and X), which is characterized by a lack of inositol 1,3,4,5-tetrakisphosphate binding activity (32). This cysteine motif was shown to be essential for the stable homo-and heterodimer formation through disulfide bonds of Syts III, V, VI, and X by site-directed mutagenesis. On the basis of these results, we discuss the role of cysteine-based heterodimerization properties of Syts III, V, VI, and X.

EXPERIMENTAL PROCEDURES
Materials-ExTaq and AmpliTaq DNA polymerases were obtained from Takara Biomedicals and Perkin-Elmer, respectively. All other chemicals were commercial products of reagent grade. Solutions were made in deionized water.
Molecular Cloning of Mouse Synaptotagmin Isoforms-cDNAs encoding the open reading frame of synaptotagmin isoforms (Syts IV, VI, VII, IX, and XI) were amplified by reverse transcriptase-polymerase chain reaction (PCR) from mouse cerebellum cDNAs (10) using the following sets of primers designed on the basis of mouse or rat sequences previously reported (12, 14 -16, 18): Syt IV, 5Ј-ACATGGCTCCTATCACCAC-C-3Ј (sense; amino acids 1-6) and 5Ј-AGCTAACCATCACAGAGCAT-3Ј (antisense; amino acids 421-425); Syt VI, 5Ј-GCATGAGCGGAGTTTG-G-3Ј (sense; amino acids 1-5), 5Ј-CGAATTCAGTAGCGTACTGGATG-TCCT-3Ј (antisense; amino acids 351-357), 5Ј-CGGATCCGCCGCCAA-* This work was supported in part by grants from the Science and Technology Agency of Japan. 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) AB026802-AB026808.
The putative initiation codon of the mouse Syt V cDNA (GGGC-GATG) closely corresponded to the Kozak sequence (CC(A/G)C-CATG(G)) (33). Although we could not obtain the in-frame stop codon preceding the putative initiation codon, the first methionine presented in Fig. 3 is most likely to be the initiation methionine because this position matches well those of Syts III, VI, and X, the same subclass of synaptotagmins.
Sequence analysis of the 5Ј-RACE products of Syt VIII showed that there are two forms of cDNA. The major, longer form contains a putative retained intron just upstream of the transmembrane region (data not shown) because it was identical to the genome sequence previously reported (16) and did not contain any apparent open reading frame. In the second, shorter form, there are three methionine residues just after the 5Ј-in-frame stop codon, each of which did not correspond well to the Kozak sequence (33). However, expression of Syt VIII proteins from 1, 100, or 118 to 1291 base pairs in COS-7 cells and comparison of their size indicated that the first methionine presented in Fig. 3 is the initiation methionine (data not shown).
Construction of T7-and FLAG-tagged Synaptotagmin Isoforms-Full-length synaptotagmin isoforms (Syts V-VIII and X) were constructed using the appropriate restriction enzyme sites on the pGEM-T Easy vector. Addition of the FLAG tag to the N terminus of each synaptotagmin isoform was done by PCR with primers encoding the FLAG tag sequence (in italics below) following the Kozak sequence (underlined below). In the case of FLAG-tagged Syt IX, for example, the FLAG-Syt IX cDNA fragment was amplified by primer FLAG-Syt IX (5Ј-CCACCATGGACTACAAGGACGATGACGACAAGGGATCCATGT-TCCCGGAACCC-3Ј) and the SP6 primer using pGEM-T-Syt IX as a template. Purified PCR products were inserted into the pGEM-T Easy vector and verified by DNA sequencing as described above. The fulllength synaptotagmin inserts with FLAG tags were then excised from the pGEM-T Easy vector by NotI digestion and subcloned into the NotI site of pEF-BOS (10, 34) (pEF-FLAG-Syts I-XI). Construction of pEF-T7-Syts I-XI was also performed by the same procedures using a primer with a T7 tag sequence (in italics; 5Ј-CCACCATGGCTAGCATGACTG-GTGGACAGCAAATGGGTCGCGGATCC-3Ј) in the 5Ј-flanking region.

FIG. 1. Phylogenetic analysis of the mouse synaptotagmin family (Syts I-XI).
The phylogenetic tree was depicted as described under "Experimental Procedures." At least three distinct subclasses of synaptotagmins were observed: Syts IV and XI, which are characterized by a lack of Ca 2ϩ -dependent phospholipid (phosphatidylserine (PS)/phosphatidylcholine (PC) (1:2.5, w/w) liposome) binding activity (18,42); Syts III, V, VI, and X, which are deficient in inositol 1,3,4,5-tetrakisphosphate (IP 4 ) binding activity (32) and showed cysteine-based heterodimerization in this study; and Syts I, II, and IX. Syts I and II are thought to be low affinity Ca 2ϩ sensors for neurotransmitter release (4 -7). The sequences of mouse Syts I and II are from Ref. 10; that of mouse Syt III is from Ref. 13; and those of mouse Syts IV-XI are from this study.
DNA Transfection-Five micrograms of pEF-T7-Syt (or pEF-BOS as a control) and pEF-FLAG-Syt were cotransfected into COS-7 cells (5 ϫ 10 5 cells/10-cm dish) by the DEAE-dextran method. Cells were harvested 72 h after transfection and homogenized in 1 ml of buffer containing 50 mM HEPES-KOH, pH 7.2, 250 mM NaCl, 0.1 mM phenylmethylsulfonyl fluoride, 10 M leupeptin, and 10 M pepstatin A. After solubilization with 1% Triton X-100 at 4°C for 1 h, the supernatants (400 l) were obtained by centrifugation at 15,000 rpm for 10 min. After incubation with anti-T7 tag antibody-conjugated agarose (wet volume 10 l; Novagen) at 4°C for 1 h, the beads were washed five times with 1 ml of 50 mM HEPES-KOH, pH 7.2, 250 mM NaCl, 0.2% Triton X-100, 0.1 mM phenylmethylsulfonyl fluoride, 10 M leupeptin, and 10 M pepstatin A and then resuspended in SDS sample buffer. Immunoprecipitates were analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) and immunoblotting as described below.
Polyclonal Antibody Production and Purification-Balb/c mice were immunized with 50 g of the purified Syt III C2A domain fused to glutathione S-transferase (13) by intraperitoneal injection with the RIBI adjuvant at intervals of 21 days. Antisera were collected after the fourth booster injection. After preabsorption by the glutathione S-transferase-Syt VI C2A domain fusion protein to remove the cross-reactive component, anti-Syt III IgG was affinity-purified by protein A-Sepharose (Amersham Pharmacia Biotech). The specificity of the anti-Syt III antibody was checked by probing extracts from COS-7 cells transfected with individual synaptotagmins (Syts I-XI) (data not shown).
Immunoblotting-Proteins were subjected to SDS-PAGE and transferred to polyvinylidene difluoride membrane (Millipore Corp.) by electroblotting. Blots were blocked with 1% skim milk and 0.1% Tween 20 in phosphate-buffered saline and then incubated with primary antibody. After the primary antibody incubation, the membrane was washed with phosphate-buffered saline containing 0.1% Tween 20 and then incubated with peroxidase-conjugated goat anti-rabbit or goat anti-mouse IgG. Immunoreactive bands were detected by enhanced chemiluminescence (Amersham Pharmacia Biotech). Polyclonal and monoclonal antibodies (M2) against the FLAG peptide (DYKDDDDK) were obtained from Zymed Laboratories Inc. and Sigma, respectively. Horseradish peroxidase-conjugated anti-T7 tag (MASMTGGQQMG) antibody and monoclonal antibody against Syt I (SYA148) were from Novagen and Stressgen Biotech Corp., respectively.
Sequence Analyses-Multiple sequence alignment and phylogenetic analysis of the mouse synaptotagmin family were performed using the PILEUP program of the GCG software package (Version 8.1) using default parameters (window size ϭ 12 and gap length weight ϭ 4).

Phylogenetic Analysis of Mouse Synaptotagmins I-XI-To
understand the phylogenetic relationship between synaptotagmin family proteins, we cloned the mouse synaptotagmin I-XI cDNAs by PCR on the basis of previously described rat sequences. Since only the partial amino acid sequences of Syts V, VIII, and X have been reported, we first determined the nucleotide sequences of the N termini of Syts V and VIII or the C terminus of Syt X using the 5Ј-or 3Ј-RACE reaction, respectively (see "Experimental Procedures" for details). The predicted amino acid sequences of mouse Syts V, VIII, and X showed that these synaptotagmins consist of 491, 398, and 523 amino acids, respectively (data not shown). Although the sequence similarities among the synaptotagmin family proteins were quite low especially at the N-terminal domain, Syts V and X were highly homologous to Syts III and VI throughout the entire protein, including the N terminus, transmembrane region, C2 domains, and C terminus. Syt V had 43.4, 55.5, and 55.1% identities to Syts III, VI, and X, respectively; and Syt X had 44.7, 55.1, and 60% identities to Syts III, V, and VI at the amino acid level, respectively, indicating that these synaptotagmins constitute a subfamily. This finding was make clearer when a phylogenetic tree of mouse synaptotagmin was constructed ( Fig. 1). According to this phylogenetic tree, the synaptotagmin family can be classified into at least four distinct groups: Syts IV and IX; Syts III, V, VI, and X; Syts I, II, and IX; and others (Syt VII or VIII).

FIG. 2. SDS-insensitive homodimerization properties of synaptotagmins I-XI in the presence (A) or absence (B) of ␤-mercaptoethanol.
T7-tagged Syt I-XI were expressed in COS-7 cells as described under "Experimental Procedures." Cells were homogenized in 1% SDS using a 27-gauge syringe. The solubilized proteins were boiled for 3 min with or without ␤-mercaptoethanol (␤-Me), subjected to 10% SDS-PAGE, and transferred to a polyvinylidene difluoride membrane. Expressed T7-tagged synaptotagmins were detected by horseradish peroxidaseconjugated anti-T7 tag antibody (1:1000 dilution). Except for Syts IV, IX, and XI, the SDS-resistant dimer of synaptotagmins can be easily observed; but probably due to the relatively high concentration of polyacrylamide gel, the dimer band of T7-Syt III was not detected in A. Note that in the absence of ␤-mercaptoethanol, the immunoreactive bands of Syts III, V, VI, and X were mostly shifted to higher molecular weights, mins I-XI-To examine whether the SDS-resistant homodimerization properties observed in Syt I or II on SDS-PAGE (10,31,35) are also retained in other isoforms, we expressed T7-tagged Syts I-XI in COS-7 cells and analyzed them by SDS-PAGE in the presence or absence of ␤-mercaptoethanol (Fig. 2). Except for Syts IV, IX, and XI, we could easily detect the minor immunoreactive bands corresponding to the dimer positions predicted from the cDNA sequences in addition to the major monomer bands (Fig. 2A). The dimer band of Syt III was difficult to see on 10% polyacrylamide gel, but could be detected on 7.5% gel (data not shown). Surprisingly, in the absence of ␤-mercaptoethanol, the majority of the Syt III, V, VI, and X immunoreactive bands were shifted to dimer positions (Fig. 2B, lanes 3, 5, 6, and 10), whereas ␤-mercaptoethanol had almost no effect on other isoforms' dimerization properties. This result indicates that certain cysteine residues are involved in the homodimerization of Syt III, V, VI, or X through a disulfide bond. Sequence alignment of Syts III, V, VI, and X revealed seven conserved cysteine residues among these isoforms: three cysteine residues in the N-terminal domain, one cysteine residue within the transmembrane region, one cysteine residue just downstream of the transmembrane region, and two cysteine residues within the C2 domains ( Fig. 3 and data not shown). Among these cysteine residues, we initially focused on the N-terminal cysteine residues that probably reside in the intravesicular region (4 -7) because there are no cysteine residues in the N-terminal domains of the other isoforms (Fig. 3, asterisks), and some of the cysteine residues around the transmembrane region are thought to be palmitoylated (36,37) and are therefore unlikely to participate in disulfide bond formation. Two cysteine residues in the cytoplasmic domain are also unlikely to be involved in disulfide bond formation because the cytoplasmic domain of Syts III and VI did not show ␤-mercaptoethanol-sensitive dimers (data not shown). Furthermore, these N-terminal cysteine residues are also conserved in electric ray synaptotagmin (p65-c), a homologue of mouse Syts III, V, VI, and X, suggesting that these are at least retained in vertebrate evolution (38).
Mutational Analysis of the N-terminal Cysteine Residues of Synaptotagmin III-To identify the cysteine residues respon-  3. Alignment of the amino-terminal domains of the mouse synaptotagmin family (Syts I-XI). According to the phylogenetic tree in Fig. 1, synaptotagmins were divided into four groups (Syts I, II, and IX; Syts III, V, VI, and X; Syts IV and XI; and others) and compared within the same group. Residues that were conserved in half of the sequences within the same subclass are boxed. Cysteine residues are shown in boldface. Asterisks indicate the conserved cysteine residues only in Syts III, V, VI, and X. The number sign indicates the N-glycosylation sites of Syts I and II (10,31). The transmembrane region (TM) is indicated by a box. Amino acid numbers are indicated on the right. sible for homodimerization of Syt III, N-terminal cysteine residues (at positions 10, 21, and 33) were replaced individually or in combination with alanine residues (Fig. 4A). The C10A single, C10A/C21A double, and C10A/C21A/C33A triple substitutions abolished the ␤-mercaptoethanol-sensitive homodimerization, whereas the C21A and C33A single and C21A/C33A and C10A/C33A double mutations had almost no effect (Fig.  4B). This result indicates that the first cysteine at position 10 is essential for the intermolecular disulfide bridge to form a stable homodimer of Syt III, but that the C10A/C33A mutant was artificially modified probably at Cys-21 due to the replacement of the first and third cysteines with alanines. Since the cysteine at position 10 of Syt III is also conserved in Syts V, VI, and X, as described above, it will be interesting to see whether the corresponding cysteine residues are involved in heterodimerization of Syts III, V, VI, and X.
Cysteine-based Heterodimerization Properties of Synaptotagmins III, V, VI, and X-To further explore the processes of heterodimerization of Syts III, V, VI, and X, we introduced two different tags, the T7 and FLAG tags, into the N terminus of each synaptotagmin and evaluated their association by immunoprecipitation (see "Experimental Procedures" for details). Briefly, T7-and FLAG-tagged synaptotagmins were coexpressed in COS-7 cells, and T7-tagged synaptotagmins were immunoprecipitated by anti-T7 tag antibody-conjugated agarose. Then, the coprecipitated FLAG-tagged synaptotagmins were detected by immunoblotting using anti-FLAG antibody.
As shown in Fig. 5A (lanes 6 -10), FLAG-Syt III was efficiently co-immunoprecipitated with T7-Syts V, VI, and X as well as T7-Syt III. In contrast, when FLAG-Syt III and T7-Syt III, V, VI, or X were separately expressed in COS-7 cells and mixed 1 h before immunoprecipitation, we could not detect the association of FLAG-Syt III with any of the T7-tagged synaptotagmins (Fig. 5A, lanes 1-5). These results indicate that the cysteine-based homo-and heterodimerization only occurred in the living cells.
In the next set of experiments, we examined the possible involvement of cysteine residues in the heterodimerization of Syts III, V, VI, and X using mutant FLAG-Syt III carrying cysteine-to-alanine mutations. Substitution of the first cysteine residue (C10A, C10A/C21A, C10A/C33A, or C10A/C21A/C33A) almost completely abolished the stable dimerization with Syt III, V, VI, or X (Fig. 5A (lanes 11-15) and data not shown), indicating that the disulfide bond mediated by the cysteine at position 10 of Syt III is crucial for the stable homo-and heterodimerization of Syt III and Syt V, VI, or X. Since a small amount of Syts V and X still associated with Syt III C10A (Fig.  5A, lanes 12 and 14), the second and third cysteine residues may also partly contribute to the heterodimerization.
To further determine whether the N-terminal cysteine motif (especially the first cysteine residue) of Syt V, VI, or X is also responsible for homo-and heterodimerization, we produced mutant Syts V (C9A), VI (C12A), and X (C13A) by site-directed mutagenesis. Except for the homodimerization of Syt X, these FIG. 5. N-terminal cysteine motif of synaptotagmin III is essential for homo-and heterodimer formation between synaptotagmins III, V, VI, and X. pEF-T7-Syt III, V, VI, or X or vector control and pEF-FLAG-Syt III, V, VI, or X were cotransfected into COS-7 cells. Expressed proteins were solubilized with 1% Triton X-100 and immunoprecipitated by anti-T7 tag antibody-conjugated agarose as described under "Experimental Procedures." Co-immunoprecipitated FLAG-tagged synaptotagmins were first detected by anti-FLAG rabbit antibody (2.7 g/ml; upper panels). Then, the same blot was stripped (43) and reprobed with horseradish peroxidase-conjugated anti-T7 tag antibody to ensure that the same amounts of T7-tagged proteins were loaded (lower panels). A, mutational analysis of Syt III. Lanes 1-5, FLAG-Syt III and T7-Syt III, V, VI, or X or a mock control, respectively, were separately expressed in COS-7 cells; mixed (1:1, v/v) after 1% Triton X-100 solubilization; and preincubated for 1 h at 4°C with gentle agitation before adding anti-T7 tag-conjugated agarose beads. Lanes 6 -10, FLAG-Syt III and T7-Syt III, V, VI, or X or a control, respectively, were coexpressed. Lanes 11-15, FLAG-Syt III C10A and T7-Syt III, V, VI, or X or a control, respectively, were coexpressed. B-D, mutational analysis of Syts V, VI and X, respectively. Lane 1, coexpression of wild-type FLAG-Syt V, VI, or X and T7-Syt V, VI, or X, respectively; lanes 2-6, coexpression of mutant FLAG-Syt V C9A (B), mutant FLAG-Syt VI C12A (C), or mutant FLAG-Syt X C13A (D) and wild-type T7-Syt III, V, VI, or X or a mock control, respectively. The positions of M r markers (ϫ10 Ϫ3 ) are shown on the left. mutants did not essentially interact with wild-type Syt III, V, VI, or X, indicating that the first cysteine residue of Syts III, V, VI, and X is the site of the major intermolecular disulfide bond, but that the second and third cysteine residues of Syt X also partly participate in intermolecular disulfide bond formation.
␤-Mercaptoethanol-sensitive Homodimerization of Native Synaptotagmin III from Mouse Brain-Finally, we examined the ␤-mercaptoethanol-sensitive homodimerization of native Syt III to rule out the possibility that the cysteine-based dimerization of Syt III observed in COS-7 cells is an artifact due to the ectopic or overexpression of Syt III. Using a specific antibody against Syt I or III, we confirmed that native Syt III from mouse adult brain also forms a ␤-mercaptoethanol-sensitive homodimer, whereas Syt I does not show such dimerization properties (Fig. 6).

DISCUSSION
In a previous study, we identified a subclass of synaptotagmins (Syts III, V, VI, and X) based on the inositol 1,3,4,5tetrakisphosphate binding properties of the C2B domain (32). In the present study, we first found that this class of synaptotagmins form ␤-mercaptoethanol-sensitive homo-and heterodimers through the evolutionarily conserved N-terminal cysteine residues (at amino acids 10, 21, and 33 of Syt III), whereas the homodimerization of other isoforms is independent of the presence of ␤-mercaptoethanol (Figs. 2B and 6). The disulfide bond-based dimerization properties of Syts III, V, VI and X are apparently different from the multimerization of Syt I, which is probably mediated by the predicted amphipathic ␣-helix region just downstream of the transmembrane region (31). A phylogenetic tree of synaptotagmins I-XI also demonstrated that Syts III, V, VI, and X form a small branch (Fig. 1). In addition, genomic sequences around the C2A domains of Syts I-III and VIII indicate that the genomic organization of Syt III (positions of exon-intron boundaries) is different from that of Syts I, II, and VIII (16,39). 2 These results also support the idea that Syts III, V, VI, and X form a distinct class of synaptotagmins.
Systematic substitutions of these cysteine residues with alanines in Syt III indicated that the first cysteine residue is crucial for homodimerization and is also important for heterodimerization with Syt V, VI, or X. The second and third cysteine residues may also be involved in heterodimerization with Syt V or X. Therefore, Syts III, V, VI, and X can form hetero-oligomers of various combinations through these cysteine residues by disulfide bonds. These disulfide bonds facilitate easy dimerization of Syts III, V, VI, and X within the cell (Fig. 5), but probably restrict the stable heterodimer with other isoforms.
What is the function of cysteine-based dimerization of Syts III, V, VI, and X in brain? In the accompanying paper (44), we show that one alternatively spliced variant of Syt VI (named Syt VI⌬TM) that lacks the conserved N-terminal domain did not interact with full-length Syt VI even when both proteins were coexpressed in COS-7 cells. Therefore, the main function of the cysteine-based dimerization is to tether the two molecules by the N terminus in the resting state and to allow them to efficiently bind each other at the C2B domain (26 -30) or the presynaptic protein syntaxin (16) in response to Ca 2ϩ . The second function of the cysteine-based dimerization is to produce a variety of sets of synaptotagmins of different biochemical natures, which may confer more diverse functions from a limited number of synaptotagmin isoforms. In previous reports, the C2 domains of Syts III, V, VI, and X were shown to have some distinct biochemical properties, including Ca 2ϩ -dependent phospholipid or syntaxin binding (16), divalent cation selectivity (40,41), and inositol 1,3,4,5-tetrakisphosphate binding (32). Among Syts III, V, VI, and X, Syt III is suggested to function as a high affinity Ca 2ϩ sensor for neurotransmitter release because it has a relatively higher affinity for Sr 2ϩ (41). Since all these synaptotagmins are expressed in brain (Syts III, V, and X are neuronal types, and Syt VI is ubiquitous) (11,16,17), these isoforms may hetero-oligomerize via disulfide bonding, and Syts V, VI, and X may modulate the Ca 2ϩ -sensing function of Syt III. In addition, the expression of Syt X mRNA in adult brain is known to be regulated by depolarization (17). Thus, it is also possible that the heterodimer between Syts III and X occurs only after some kind of synaptic plasticity.
In conclusion, we first identified, by sequence comparison and mutational analysis, the evolutionarily conserved cysteine residues responsible for intermolecular disulfide bonding at the N terminus of Syts III, V, VI, and X. Our findings suggest that hetero-oligomerization of Syts III, V, VI, and X may produce a variety of Ca 2ϩ sensors that function in neurotransmitter release with different biochemical natures.