A tripartite interface that regulates vesicle fusion

In neurons, SNARE proteins mediate the fusion of neurotransmitter-containing vesicles with the plasma membrane, leading to the release of neurotransmitters upon action potential. To prime fusion, SNARE proteins in the vesicle membrane (synaptobrevin-2) and in the target membrane (SNAP-25 and syntaxin-1A) form a complex characterized by a core four-helix bundle, called the SNARE complex (pictured; synaptobrevin-2 is dark blue, SNAP-25 is green and syntaxin-1A is red).

However, SNARE proteins do not work alone. In fact, vesicle fusion is stringently regulated, both to prevent spontaneous release and to ensure that neurotransmitter release occurs in a fast and synchronous manner upon stimuli. Ca²⁺ influx, induced by action potential, triggers the fusion of vesicles docked to the cytoplasmic membrane. The Ca²⁺ sensor in this process is synaptotagmin-1 (Syt1), which has two Ca²⁺-binding domains, C2A and C2B. Another regulator is complexin (Cpx), which is required to suppress spontaneous fusion.

Both Syt1 and Cpx interact with the SNARE complex and previous structural work has revealed details of those binary interactions, but the interplay between the two factors remained unclear. Now, Zhou et al. (Nature http://dx.doi.org/10.1038/nature23484) have determined the structure of the prefusion complex containing SNAREs, Syt1 and Cpx. Along with extensive analysis in neuronal cells, the work uncovers previously unknown interactions and provides major insights into the complex regulation of neuronal vesicle fusion.

Unexpectedly, the new structural data reveal two Syt1 molecules interacting with the SNARE complex via different interfaces, with their C2B domains bound on opposite sides of the four-helix bundle. The first C2B domain (gray) binds the SNARE complex via the previously known, or primary, interface. The second C2B domain (orange) interacts with the SNARE complex and with Cpx (light blue), in what the authors call a 'tripartite interface'. This interface (marked by a dashed box in the figure) involves the conserved HA α-helix in C2B, which extends the Cpx α-helix that interacts with the SNARE complex.

Mutagenesis analyses of the tripartite interface show that it is required for Ca²⁺-triggered synchronous release by neurons, as is the primary interface. Thus, both binding interfaces are essential to reach the primed prefusion state captured in the new structure. However, the tripartite interface also serves to 'lock' the complex in a primed state; thus, Syt1 and Cpx cooperate to prevent calcium-independent vesicle fusion. The tripartite interface would be unlocked upon Ca²⁺ binding to the C2 domains, which would likely result in release of the Syt1 C2B domain involved in the tripartite complex and conformational rearrangements of the remaining complex. This proposed unlocking mechanism remains to be confirmed, but the present findings can already explain the phenotypes of Cpx or Syt1 knockouts and of dominant-negative Syt1 mutations.