Unraveling the Role of SNARE Interactions in Neurotransmitter Release

The release of neurotransmitters by Ca2+-triggered synaptic vesicle exocytosis is tightly controlled by an intricate protein machinery. Essential components of this machinery are the synaptic vesicle protein synaptobrevin and the plasma membrane proteins syntaxin 1 and SNAP-25, which are collectively known as SNAREs and form a tight complex (the core complex). The assembly of the core complex may mediate membrane fusion. Complexin is a highly conserved cytoplasmic protein that binds tightly to the SNARE complex. Analysis of the interaction between complexin and the SNARE complex showed that complexin binds to the groove between the synaptobrevin and syntaxin helices, and the binding stabilizes the syntaxin/synaptobrevin interface. These results led to a model whereby complexin stabilizes the fully assembled SNARE complex, which is critical for the fast Ca2+-triggered neurotransmitter release. The N-terminal domain of syntaxin 1 folds back and forms a 'closed' conformation, which interacts with munc18-1, an essential protein in the neurotransmitter release. It has been proposed that the binding of munc18-1 might change the closed conformation. To test this model, I solved the solution structure of the N-terminal domain within the closed conformation of syntaxin 1 and structure comparisons showed that the N-terminal domain adopts the same conformation whether it is isolated, bound to Munc18-1, or within the closed conformation. Analysis of the Ca2+-binding properties of the core complex revealed that it contains several low affinity Ca2+ binding sites and most of them are nonspecific for Ca2+. A SNAP-25 mutation that causes a change in the Ca2+-dependence of secretion in chromaffin cells has no effect on the SNARE/synaptotagmin 1 interactions, but has a conspicuous effect on core complex assembly. Thus, the SNAREs are unlikely to directly act as Ca2+ sensors, but SNARE complex assembly is tightly coupled to Ca2+ sensing in neurotransmitter release. To directly test SNARE function, I reconstituted v- and t-SNAREs into separate liposomes and carefully characterized the proteoliposomes containing v- and t-SNAREs. Fusion between the v- and t-SNARE proteoliposomes was then monitored with a lipid mixing assay. Interestingly, little fusion was observed. The results suggest that the SNAREs alone are not sufficient to mediate membrane fusion.