Exocytosis is greatly reduced but not eliminated at synapses lacking the SNARE fusion-promoting protein synaptobrevin/VAMP. The group found that endocytosis was delayed in these neurons even if amounts of exocytosis were first equalized at wild-type and mutant synapses by using different stimulation protocols. Dyes provided a further suggestion that mutant synapses had selectively lost fast endocytosis, as mutant synapses released both fast- and slow-diffusing styryl dyes and wild-type synapses released only the fast-diffusing dye.
Hints of an endocytic SNARE role have been seen previously in yeast and flies. But in neither has the link been made to the special case of fast endocytosis. This process is essential for neurons, but the responsible molecules, unlike the clathrin used during slow endocytosis, remain obscure. Fast endocytosis may involve partial retention of vesicle structure (“kiss and run”) between fusion and endocytosis.
What synaptobrevin is actually doing to promote endocytosis remains unclear. It may either act as a nucleator of endocytic proteins or, the authors suggest, be required to take the exocytic vesicle into a pathway that is primed for fast endocytosis. For example, synaptobrevin on the vesicle might attach to plasma membrane SNAREs to set up a hemi-fusion state, but then release the plasma membrane SNAREs before full fusion. Thus disengaged, the hemi-fusion structure could, when the pro-fusion calcium signal arrived, move into full fusion mode and back out to endocytosis without having to wait for SNARE disentanglement.