Xue Han, Meyer Jackson and colleagues, reporting in Science, now show that the fusion pore that forms in Ca2+-regulated exocytosis has a protein lining. Until now, the molecular composition of this important intermediate was unknown and the possibility that it was composed of lipids could not be ruled out.

The authors reasoned that if the fusion pore is composed of protein, neurotransmitter flux through this pore should be affected by mutations that alter the size of amino-acid side chains lining the pore. They focused their study on syntaxin for several reasons: for example, it is essential for exocytosis and it is the only essential plasma-membrane protein that contains a transmembrane segment.

By mutating residues in this segment, in turn, to tryptophan and taking measurements of neurotransmitter flux through the fusion pore, Han et al. could elucidate whether each mutation obstructed the fusion pore. They identified three mutations — I269W, G276W and I283W — that significantly reduced flux through the fusion pore: these residues all lie along one face of the syntaxin transmembrane α-helix.

To further examine this effect, the authors created a series of mutants such that, at each of the positions mentioned above, the amino-acid side chain gradually increased in size. As expected, increasing the side-chain size at these positions produced a graded reduction in neurotransmitter flux through the pore. Mutating non-pore-lining residues in this way had no effect on flux.

Han et al. therefore propose that the plasma-membrane-spanning part of the fusion pore is a barrel that, from their model, is composed of 5–8 copies of the syntaxin transmembrane segment (although they acknowledge that further proteins and/or lipids might also constitute the pore). Furthermore, they suggest that this syntaxin-lined pore is linked to a complementary protein (synaptobrevin)-lined pore in the vesicle membrane by SNARE complexes.