In addition, we have used this in vitro assay to examine the biochemical effects of several mutations in Sec9 that result in pronounced growth defects in vivo. As expected, a temperature-sensitive mutation in the region most highly conserved between Sec9 and SNAP-25 is severely diminished in its ability to bind Sso1 and Snc1 in vitro. In contrast, a temperature-sensitive mutation near the C terminus of Sec9 shows no defect in SNARE binding in vitro. Similarly, a deletion of the Cterminal 17 residues, which is lethal in vivo, also binds Sso1 and Snc1 normally in vitro. Interestingly, we find that these same two C-terminal mutants, but not mutants that show SNARE assembly defects in vitro, act as potent dominant negative alleles when expressed behind a strong regulated promoter. Taken together these results suggest that the C-terminal domain of Sec9 is specifically required for a novel interaction that is required at a step following SNARE assembly.In recent years, studies on neuronal exocytosis, biochemical analysis of in vitro transport systems, and yeast genetic analysis have converged on a set of structurally related proteins known as SNARE proteins, as critical for the process of vesicle targeting and fusion in eukaryotic cells (1). This has led to the SNARE hypothesis, which suggests that SNARE proteins on the surface of vesicles (v-SNAREs) 1 can interact specifically with SNARE proteins on the target membrane (t-SNAREs) to form a complex that recruits factors required for fusion of the two membrane bilayers (2). In neurons the vesicular protein, synaptobrevin, associates with the two plasma membrane proteins syntaxin and soluble SNAP-25 to form a complex that acts as a receptor for ␣-SNAP and NSF. Hydrolysis of ATP by NSF leads to SNARE complex disassembly and is thought to be linked to membrane fusion, although precisely how NSF is involved in the membrane fusion event is still unclear (3).Binding studies with recombinant neuronal proteins have shown that although syntaxin and SNAP-25 can bind to one another with high affinity, they can also individually bind to the vesicle protein, synaptobrevin, with lower affinities (4). The presence of both t-SNAREs, however, potentiates the interaction of syntaxin with synaptobrevin about 10-fold and that of SNAP-25 with synaptobrevin approximately 2-fold. This suggests that the formation of a highly stable ternary complex drives the interaction between these proteins and consequently aids in determining the overall specificity of synaptic vesicle docking (4).The regions of each neuronal SNARE protein that mediate these binding activities have been extensively characterized. Within the cytoplasmic domain of syntaxin 1, for example, a small 73-residue juxtamembrane region, predicted to form coiled-coils, can mediate both binding to synaptobrevin (5, 6) and 7,8). The amino-terminal region of syntaxin is also capable of binding the carboxyl-terminal domain of the protein, and this interaction appears to be somewhat inhibitory to synaptobrevin binding (5). The amino-termi...