(Received for publication, September 24, 1996)
From the Max-Planck-Institut für experimentelle
Medizin, Abteilung Molekulare Neurobiologie, Hermann-Rein-Strasse 3, D-37075 Göttingen, Federal Republic of Germany and the
§ Howard Hughes Medical Institute, University of Texas
Southwestern Medical Center, Dallas, Texas 75235
unc-13 mutants in Caenorhabditis elegans are characterized by a severe deficit in neurotransmitter release. Their phenotype is similar to that of the C. elegans unc-18 mutation, which is thought to affect synaptic vesicle docking to the active zone. This suggests a crucial role for the unc-13 gene product in the mediation or regulation of synaptic vesicle exocytosis. Munc13-1 is one of three closely related rat homologues of unc-13. Based on the high degree of similarity between unc-13 and Munc13 proteins, it is thought that their essential function has been conserved from C. elegans to mammals. Munc13-1 is a brain-specific peripheral membrane protein with multiple regulatory domains that may mediate diacylglycerol, phospholipid, and calcium binding. In the present study, we demonstrate by three independent methods that the C terminus of Munc13-1 interacts directly with a putative coiled coil domain in the N-terminal part of syntaxin. Syntaxin is a component of the exocytotic synaptic core complex, a heterotrimeric protein complex with an essential role in transmitter release. Through this interaction, Munc13-1 binds to a subpopulation of the exocytotic core complex containing synaptobrevin, SNAP25 (synaptosomal-associated protein of 25 kDa), and syntaxin, but to no other tested syntaxin-interacting or core complex-interacting protein. The site of interaction in syntaxin is similar to the binding site for the unc-18 homologue Munc18, but different from that of all other known syntaxin interactors. These data indicate that unc-13-related proteins may indeed be involved in the mediation or regulation of synaptic vesicle exocytosis by modulating or regulating core complex formation. The similarity between the unc-13 and unc-18 phenotypes is paralleled by the coincidence of the binding sites for Munc13-1 and Munc18 in syntaxin. It is possible that the phenotype of unc-13 and unc-18 mutations is caused by the inability of the respective mutated gene products to bind to syntaxin.
Nerve cells store neurotransmitters in synaptic vesicles. These vesicles dock to a specialized region of the synaptic plasma membrane, the active zone, where they undergo a maturation or priming process. Upon depolarization and a consequential rise in the intracellular calcium concentration, primed vesicles release their content by exocytosis. Following release, vesicular membrane and protein components are retrieved by endocytosis and recycled through an early endosomal compartment. From there, synaptic vesicles bud off for a new round of regulated exocytosis (1, 2).
During the last decade, the molecular mechanisms underlying synaptic vesicle exocytosis and recycling have been the focus of numerous studies. In particular, the components of the exocytotic synaptic core complex, the synaptic vesicle protein synaptobrevin and the plasma membrane proteins syntaxin and SNAP25,1 have been analyzed in detail. These proteins are absolutely essential for neurotransmitter release, and their complex formation was originally thought to mediate the vesicle docking process (3). A modification of the core complex was interpreted as the structural basis of the maturation process that leads to a metastable hemifusion state, rendering docked vesicles fusion-competent (1-4). However, syntaxin mutants in Drosophila exhibit a normal synaptic ultrastructure and normal numbers of docked vesicles but a general block of secretion, indicating that syntaxin is essential for transmitter release and other fusion reactions but may not be directly involved in vesicle targeting or docking. Rather, it is likely that syntaxin exerts its critical role at a later stage in the vesicle fusion process. Likewise, the formation of the synaptic core complex presumably occurs upon vesicle docking to the active zone but probably does not cause the docking process itself (5).
These and other observations demonstrate that the exact functional role of the majority of synaptic vesicle components and other presynaptic proteins involved in synaptic exocytosis remains unclear. In addition, it is likely that novel, as yet unidentified proteins will have to be included in the current molecular models of neurotransmitter release. The discovery of novel candidate proteins with a constitutive or regulatory function in neurotransmitter release was greatly facilitated by mutagenesis studies in Caenorhabditis elegans. Pioneering experiments by Brenner (6) identified a collection of partially or completely paralyzed C. elegans mutants. A subgroup of these so called unc mutants is further characterized by high levels of acetylcholine and resistance to acetylcholine esterase inhibitors, but normal acetylcholine esterase, choline acetyltransferase, and postsynaptic receptor activities (7, 8), indicating impaired neurotransmitter release. A detailed molecular characterization demonstrated that several members of this subgroup of unc genes do indeed encode proteins that are directly involved in synaptic vesicle function (e.g. unc-17, the vesicular acetylcholine transporter (9), and unc-18, a protein whose mammalian homologue, Munc18, interacts with the core complex component syntaxin, and whose Drosophila homologue, rop, is an essential regulator of transmitter release (10-14)).
Of the known unc mutants with impaired transmitter release, unc-13 is of particular interest because of its severe phenotype (8). Its interesting domain structure with a phorbol ester-binding C1 domain and two C2 domains homologous to the calcium/phospholipid-binding domain of protein kinase C (15-17) suggested a calcium- and diacylglycerol-dependent regulatory role for unc-13 in transmitter release (15, 18). In the rat, three distinct, brain-specific homologues of unc-13 are expressed (Munc13-1, Munc13-2, and Munc13-3) (18). Like unc-13, Munc13 isoforms are very large proteins (195-225 kDa). They are highly homologous to each other and to unc-13 in the C-terminal two-thirds (50% identity), whereas the N-terminal parts are dissimilar. The most abundant isoform, Munc13-1, is a peripheral membrane protein that is presumably attached to the cytoplasmic face of the synaptic plasma membrane (18).
Based on the similarity between the phenotypes of unc-13 and unc-18, it has been proposed that unc-13, like unc-18, may be involved in synaptic vesicle docking (19). In addition, the structural similarities between unc-13 and Munc13 isoforms suggest that the essential function of unc-13 in C. elegans is conserved in mammals. Supporting the notion of an essential role for Munc13 homologues in neurotransmitter release, we demonstrate in the present study that Munc13-1 interacts with the synaptic core complex by binding directly to syntaxin. Using yeast two-hybrid technology, immunoprecipitation analyses, and cosedimentation assays with recombinant protein fragments, we show that a conserved domain of 165 amino acids between the second and third C2 domains of Munc13-1 is sufficient for syntaxin binding. As in the case of Munc18, but in contrast to all other syntaxin-binding proteins, the binding of Munc13-1 to syntaxins requires the syntaxin N terminus. The second of two predicted coiled coil domains in the syntaxin N terminus appears to be necessary for Munc13-1 binding. We postulate that the very similar phenotypes of unc-13 and unc-18 mutations may be caused by the inability of the respective mutated gene products to bind to syntaxin.
A bait vector (pLexN-N7N9) encoding residues 1181-1736
of Munc13-1 (Fig. 1) was constructed by amplifying the
corresponding DNA fragment from pCMV-Munc13-1 (18) using engineered
oligonucleotide primers and subcloning the resulting 1.65-kb fragment
into the EcoRI-SalI sites of pLexN, a
modification of pBTM116 containing an SV40 large T-antigen nuclear
localization signal N-terminal to the polylinker (20). This results in
a vector expressing a LexA fusion protein with Munc13-1 starting at
residue 1181. A cDNA library was constructed in the
EcoRI site of pVP16-3, a modification of pVP16 (20) with an
extended polylinker, from poly(A)+-enriched embryonic day
18 rat brain RNA. In brief, total RNA was prepared from E18 rat brains
by acid guanidinium thiocyanate-phenol-chloroform extraction with RNA
Stat-60 (Tel-Test, Friendswood, TX), and mRNA was purified using an
mRNA purification kit (Pharmacia Biotech, Inc., Freiburg, Federal
Republic of Germany). The Superscript Choice System (Life Technologies,
Inc., Eggenstein, Federal Republic of Germany) was used for production
of cDNA from 10 µg of mRNA. After ligation of
EcoRI adaptors, the cDNA was size fractionated (>2 kb)
by column chromatography. The fractionated cDNA was then ligated
into the EcoRI site of pVP16-3. DH10B Escherichia
coli was transformed with this ligation, resulting in a library
size of 8 × 106 independent clones. YTH screens were
performed essentially as described (20-23). 100 million yeast
transformants were screened, and 84 clones positive upon
retransformation were isolated and sequenced using the dideoxy chain
termination method with dye terminators on an Applied Biosystems 373 DNA sequencer (Applied Biosystems, Weiterstadt, Federal Republic of
Germany).
An additional YTH bait vector covering residues 444-1182 of Munc13-1 was cloned using restriction sites in cDNA clones (pLexN-M13Xho2.2) and used as a negative control in the analysis of positive prey clones. Prey vectors in pVP16 encoding SNAP25A (pVP16-SNAP25, full-length SNAP25A), synaptobrevin 2 (pVP16-Synaptobrevin 2, residues 1-96), or synaptotagmin 1 (pVP16-Synaptotagmin 1, residues 120-422) fusion proteins with VP16 were obtained from Drs. Y. Hata, T. C. Südhof, and S. Sugita (Dallas, TX) (22).
-Galactosidase assays were performed according to Rose et
al. (24) on extracts from yeast strain L40 cotransformed with the
indicated bait and prey constructs.
Recombinant GST-Munc13-1 fusion proteins covering the entire conserved part of Munc13-1 were synthesized from the following expression plasmids in the pGEX-KG vector (25): pGEX-Munc13-1A (residues 556-808), pGEX-Munc13-1B (residues 708-1032), pGEX-Munc13-1C (residues 1032-1345), pGEX-Munc13-1D (residues 1181-1345), pGEX-Munc13-1E (residues 1399-1622), and pGEX-Munc13-1F (residues 1399-1736). Recombinant proteins were purified on glutathione-agarose (Sigma, Deisenhofen, Federal Republic of Germany) and used, immobilized on the resin, for cosedimentation assays.
Subcellular FractionationSubcellular fractions were prepared essentially as described by Jones and Matus (26). They were designated as follows: H, homogenate; P1, nuclear pellet; P2, crude synaptosomal pellet; P3, light membrane pellet; S3, cytosolic fraction; LP1, lysed synaptosomal membranes; LP2, crude synaptic vesicle fraction; LS2, cytosolic synaptosomal fraction; SPM, synaptic plasma membranes.
Cosedimentation AssaysCrude synaptosomes from rat brain were solubilized at a protein concentration of 2 mg/ml in 100 mM NaCl, 25 mM HEPES/KOH, pH 7.4, 1 mM EGTA, 1% Triton X-100, 0.2 mM phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 0.5 µg/ml leupeptin (solubilization buffer A). After being stirred on ice for 30 min, insoluble material was removed by centrifugation (90 min at 250,000 × gmax and 4 °C). The equivalent of 15 mg of total protein was then incubated with 20 µg of immobilized GST-fusion protein in the presence or absence of 2 mM CaCl2 for 60 min at 4 °C. Beads were then washed 5 times with solubilization buffer A containing 0.1% Triton X-100 with or without 2 mM CaCl2, resuspended in SDS-PAGE sample buffer, and analyzed by SDS-PAGE and immunoblotting.
ImmunoprecipitationsA novel polyclonal antibody directed
against the Munc13-1 C terminus was generated using the pGEX-Munc13-1F
GST fusion protein as antigen. The affinity-purified antibodies
(N395AP) recognize a single band of 206 kDa in Western blots of brain
membranes and HEK293 cells transfected with pcDNA3-Munc13-1 (which
expresses full-length Munc13-1), but not in mock-transfected cells
(data not shown). N395AP specifically immunoprecipitates a 206 kDa
protein that is recognized by a previously characterized antibody
directed against Munc13-1 (I475AP) (18) (see Fig. 6 and data not
shown). N395AP antibodies were used for Munc13-1 immunoprecipitations. For this purpose, purified synaptosomes from rat brain were solubilized at a protein concentration of 2 mg/ml in 100 mM NaCl, 25 mM HEPES/KOH, pH 7.4, 1 mM EGTA, 1% sodium
cholate (Wako, Neuss, Federal Republic of Germany), 0.2 mM
phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 0.5 µg/ml
leupeptin (solubilization buffer B). Cholate was used instead of Triton
X-100 because it was more potent in solubilizing Munc13-1, and no
qualitative differences in immunoprecipitations were observed between
cholate- and Triton X-100-extracted material. After being stirred on
ice for 30 min, insoluble material was removed by centrifugation (90 min at 250,000 × gmax, 4 °C). Aliquots of 10 ml were incubated with 5 µg of affinity-purified antibody N395AP (2-12 h at 4 °C). After addition of 15 µl bed volume of protein G-Sepharose (Pharmacia) and incubation for 2 h at 4 °C, the beads containing the immune complexes were washed three times with
solubilization buffer B, resuspended in SDS-PAGE sample buffer, and
analyzed by SDS-PAGE and immunoblotting.
Miscellaneous
All chemicals were of highest available
purity and purchased from standard sources. The following antibodies
were used: monoclonal antibodies to syntaxin 1A/B (27), SNAP25 (Cl
71.2) (28), synaptobrevin 2 (Cl 69.1) (29), synaptotagmin 1 (Cl 41.1)
(30), and N-methyl-D-aspartate R1 (Cl 54.1)
(31); polyclonal antibodies to Munc13-1 (18), Munc18-1 (10), complexin
1 and 2 (32), rabphilin 3A and rab 3A (33), synaptophysin 1 (34), NSF,
and SNAP (32, 35).
SDS-PAGE and immunoblotting were performed using standard procedures (36, 37).
We screened 100 million yeast colonies transformed with
a rat E18 brain cDNA library and a bait construct encoding a fusion protein consisting of LexA and residues 1181-1736 of Munc13-1. 84 positive prey clones were isolated, retested for interaction with the
bait construct, and sequenced. Apart from 17 single isolates, the
sequenced clones fell into several large groups of identical or
overlapping clones. Interestingly, one family of 10 isolated clones
encoded residues 1-79 of syntaxin 1B (pPrey 103 and others), while
another family of 4 clones encoded a fragment of syntaxin 2 starting
with residue 53 and ending with residue 227 (pPrey 128 and others; Fig.
1). Both types of prey clones interacted strongly with the Munc13-1
bait construct pLexN-N7N9 but not with the control construct
pLexN-M13Xho2.2 in YTH -galactosidase assays (Fig.
2B), demonstrating a direct and specific
interaction of the Munc13-1 C-terminal residues 1181-1736 with the
core complex component syntaxin. The interaction of the Munc13-1 C
terminus (residues 1181-1736) is specific for syntaxin, as the bait
construct pLexN-N7N9 selectively interacted with syntaxin prey vectors
pPrey 103 and pPrey 128 in YTH
-galactosidase assays, but not with prey vectors encoding SNAP25, synaptobrevin 2 or synaptotagmin (Fig.
2C). Similarly, the control vector pLexN-M13Xho2.2 did not interact with prey vectors encoding SNAP25, synaptobrevin 2, or synaptotagmin (data not shown), indicating that syntaxin is the only
core complex component that binds to Munc13-1. The two groups of
syntaxin prey clones overlap in a region that covers one of two
predicted coiled coil domains in the syntaxin N terminus (Fig. 1),
indicating that this part of syntaxin is necessary for Munc13-1 binding. However, truncation of pPrey 103 resulting in a prey construct
that only covers the second coiled coil domain of syntaxin 1B (residues
47-79) resulted in a fusion protein that was incapable of binding to
the Munc13-1 bait construct, presumably caused by misfolding of the
isolated interacting domain (data not shown). These data show that
apart from Munc18, an essential regulator of synaptic exocytosis,
Munc13-1 is the only known protein that binds to the syntaxin N
terminus. All other known interactors (synaptobrevin, SNAP25,
SNAP,
N-type calcium channel, complexins, and synaptotagmin) bind to a
C-terminal region just upstream of the transmembrane domain (Fig.
3).
Mapping of the Syntaxin Binding Site on Munc13-1
In order to
determine the region in Munc13-1 that is responsible for the
interaction with syntaxin, we used recombinant protein fragments of
Munc13-1 in syntaxin binding assays. For this purpose, several
GST-Munc13-1 fusion constructs were generated, covering the entire
conserved part of Munc13-1 (Fig. 4A). These
constructs were used in cosedimentation assays with rat brain extract,
and bound material was assayed for syntaxin 1 by immunoblotting (see "Materials and Methods"). Of the six fusion proteins tested, only pGEX-Munc13-1C and pGEX-Munc13-1D were capable of binding to syntaxin 1 (Fig. 4B). The two constructs overlap in a 165-amino acid
region (residues 1181-1345) between the central and C-terminal C2
domains of Munc13-1. These data demonstrate that residues 1181-1345
(or a part of this domain) of Munc13-1 are necessary for syntaxin binding.
Binding of Munc13-1 to the Synaptic Core Complex
The
formation of a heterotrimeric complex between the vesicle protein
synaptobrevin and the plasma membrane proteins syntaxin and SNAP25, the
synaptic core complex, is thought to represent a critical step in
synaptic vesicle exocytosis. Therefore, we examined whether the complex
between Munc13-1 and syntaxin also includes SNAP25 and synaptobrevin.
In cosedimentation assays, the syntaxin-interacting fusion protein
pGEX-Munc13-1C bound not only syntaxin 1A/B from brain extracts (as
shown in Fig. 4B) but also SNAP25 and synaptobrevin 2 (Fig.
5). In order to examine Munc13-interacting synaptic
proteins in more detail, we performed Munc13-1 immunoprecipitations and
analyzed the precipitated material with a number of antibodies directed
against synaptic proteins involved in the formation and regulation of
the core complex. Fig. 6 demonstrates that syntaxin
1A/B, SNAP25, and synaptobrevin 2 clearly coprecipitated with Munc13-1.
Some coprecipitation was also observed for synaptotagmin 1. However,
Munc18-1, complexin 1 and 2, NSF, Rabphilin 3A, rab 3A, synaptophysin
1, and SNAP did not interact with Munc13-1. Interestingly, only a
fraction of the core complex components was coprecipitated with
Munc13-1. These data indicate that Munc13-1 interacts directly with
syntaxin 1 and associates with the synaptic core complex by binding to syntaxin. In agreement with this is the observation that in YTH experiments none of the available Munc13-1 bait constructs interacted with prey constructs encoding VP16 fusion proteins with SNAP25A, synaptobrevin 2, synaptotagmin 1, and Munc18-1 (Fig. 2C and
data not shown).
Colocalization of Munc13-1 and Core Complex Components
So
far, the exact subcellular localization of Munc13-1 is not known. None
of the available antibodies give reliable signals in
immunocytochemistry. As a consequence, the only data concerning colocalization of Munc13-1 with components of the synaptic core complex
were obtained in subcellular fractionation experiments. Here, most of
the Munc13-1 copurified with synaptic markers (Fig. 7)
and was highly enriched in synaptic plasma membranes but not in
synaptic vesicle fractions. These data suggest a localization of
Munc13-1 to the presynaptic plasma membrane, coinciding well with the
localization of syntaxin and SNAP25.
unc-13 belongs to a group of C. elegans mutations with deficits in neurotransmitter release. The unc-13 gene product is characterized by a unique domain structure. It contains a diacylglycerol-binding C1 domain that represents a single copy of a zinc finger-like region. In addition, it contains two C2 domains that were originally described in protein kinase C as calcium- and phospholipid-binding domains (15). The first of the two C2 domains, which is localized in the middle of the protein, just C-terminal of the C1 domain, is very similar to other calcium-binding C2 domains and contains all structural features that are thought to be necessary to form the calcium binding pocket (18, 38, 39). These structural features, together with the phenotype of the mutation, led to the suggestion that the unc-13 gene product may be involved in the calcium- and diacylglycerol-dependent regulation of synaptic exocytosis. Interestingly, the phenotype of unc-13 is very similar to that of unc-18 (19). unc-18, its Drosophila homologue rop, and its mammalian homologue Munc18 are syntaxin binding proteins (10, 12, 13). This interaction has been evolutionarily conserved because even the respective yeast homologues appear to interact in a functional manner. Null mutations of rop in Drosophila are lethal and electrophysiological analyses indicate a block in synaptic vesicle release (11). Likewise, secretion from other cells appears to be affected by the mutation, indicating an essential secretory function of rop (11). These data, together with the observation of an accumulation of secretory vesicles in yeast mutants of the unc-18 homologue sec1 (40), suggest a facilitatory role of rop in release. However, rop overexpression studies in Drosophila indicate additional inhibitory functions for the protein. Like deletion, overexpression of rop also leads to a general reduction in secretory processes (14). The bimodal action of rop can be explained by a model in which the protein associates with syntaxin to prevent indiscriminate formation of the core complex outside the active zone. Upon association of the core complex at the active zone, which would require the previous syntaxin/rop association as well as another "catalyzing" protein, rop is displaced from syntaxin, and fusion can occur (14). Based on the similarity of phenotypes in C. elegans, it was therefore hypothesized that unc-13, like unc-18, may be involved in the regulation of core complex formation and, more specifically, vesicle docking (19).
Munc13-1 is a rat homologue of unc-13. The high degree of similarity and brain-specific localization indicate that it may also represent a true functional homologue (18, 41). The current hypothesis for unc-13 function was therefore extended to Munc13-1, and a regulatory role for Munc13-1 in synaptic vesicle docking and exocytosis was postulated (41). The present study supports this hypothesis by demonstrating a direct interaction of Munc13-1 with the core complex component syntaxin.
Multiple documented protein interactions indicate a central role for
syntaxin in the exocytotic machinery. It is also a substrate for
botulinum neurotoxin C1, and its cleavage by the toxin is paralleled by
a block in neurotransmitter release, demonstrating a critical role for
syntaxin in the release process (42). In addition to the core complex
components synaptobrevin and SNAP25, syntaxin interacts with SNAP, a
modulator of the core complex and a recruiting protein for NSF in
fusion reactions; with complexin, another modulator of core complex
formation; with N-type calcium channels; with the exocytotic calcium
sensor synaptotagmin; and with Munc18 (2, 43). The syntaxin domain
necessary for most of these interactions is localized to a putative
coiled coil domain of the protein just upstream of its C-terminal
transmembrane domain (43-46) (Fig. 3). In contrast to this, binding of
Munc13-1 requires the syntaxin N terminus (Figs. 1, 2, 3). Interestingly,
the only other known syntaxin-interacting protein requiring the
syntaxin N terminus for binding is Munc18 (46), highlighting another parallel between Munc13 and Munc18 and supporting the idea of similar
roles of the two proteins in synaptic exocytosis.
A detailed analysis of Munc13-1 immunoprecipitation experiments and cosedimentation assays showed that in addition to syntaxin, Munc13-1 also binds synaptobrevin, SNAP25, and to a lesser extent, synaptotagmin (Figs. 4, 5, 6). Synaptobrevin and SNAP25 are constitutive components of the synaptic core complex, while synaptotagmin is thought to associate with a subpopulation of the heterotrimeric core complex with unknown function (2, 4). Because neither synaptobrevin nor SNAP25 binds directly to Munc13-1 in YTH experiments (Fig. 2, B and C), it is likely that they associate with Munc13-1 as part of the core complex via syntaxin.
None of the other tested proteins interacted with Munc13-1 or the core
complex subfraction that associates with it (Fig. 6). These proteins
included all other known syntaxin interactors with the exception of
N-type calcium channels, which were shown in separate experiments using
radiolabeled -conotoxin GVIA not to bind to Munc13-1 (data not
shown). This implies that the syntaxin complexes containing Munc18,
complexin,
SNAP, or N-type calcium channels do not bind
simultaneously to Munc13-1, and the complexes that do associate with
Munc13-1 represent a distinct intermediate. The latter is supported by
the fact that despite almost complete immunoprecipitation of Munc13-1,
only a fraction of total syntaxin, synaptobrevin, and SNAP25 is
coprecipitated.
One possible interpretation of our findings is that Munc13-1 modulates or stabilizes an intermediate form of the synaptic core complex. To test this hypothesis, we used the spontaneous association of recombinant core complex components as an assay to study modulatory, stabilizing, or destabilizing effects of recombinant Munc13-1 fragments. In addition, we studied the effects of exogenous recombinant protein fragments on the composition of Munc13-1 immunoprecipitates. However, so far no clear indication of a regulatory effect of Munc13-1 on core complex formation was obtained (data not shown). Likewise, no evidence for a calcium or phorbol ester dependence of the Munc13-1 binding to syntaxin was observed (Fig. 4 and data not shown).
Our study presents evidence for an involvement of Munc13-1 in synaptic vesicle trafficking. The observed binding of Munc13-1 to syntaxin and to the core complex is compatible with previous hypotheses that were based on the similarity between unc-13 and unc-18 phenotypes in C. elegans and that postulated an involvement of unc-13 and Munc13 in synaptic vesicle docking. The coincidence of Munc13 and Munc18 binding sites on syntaxin lends further support to this view and provides an explanation for the similarities between the phenotypes of mutations in the respective genes of C. elegans. In fact, disruption of the syntaxin binding by the mutated unc-13 and unc-18 gene products may represent the mechanistic basis of the observed phenotypes.
We thank Drs. Y. Hata, R. Jahn, T. C. Südhof, and S. Sugita for the kind gift of antibodies and plasmids, Drs. S. Hollenberg, P. Bartel, and S. Fields for YTH plasmids, J. Ficner for artwork, Dr. F. Schmitz and O. Schlüter for comments on the manuscript, and Dr. T. C. Südhof for support and advice.