A well-established approach to study a cellular process is to reconstitute the system in vitro. Researchers in several laboratories, including that of Axel Brunger at Stanford University, have been using this strategy to understand neurotransmitter release. In a recently published paper, Brunger and colleagues describe an in vitro system that qualitatively recapitulates several of the fundamental characteristics of the process as it occurs in neurons.

The researchers had to begin with very stable preparations of proteoliposomes, incorporating full-length, functional components of the known neuronal fusion machinery, the SNAREs, at physiological protein levels. For most experiments, this minimal reconstituted system included the calcium-sensing protein synaptotagmin and the synaptic protein complexin.

Critically, they included dyes in the donor vesicle preparations that would report on both lipid mixing (DiD) and content mixing (sulforhodamine B) during the vesicle fusion process, a unique feature of their system. The dyes are quenched at concentrations at which they are incorporated into the proteoliposomes and become dequenched upon fusion. But some of the dyes themselves destabilize the liposomes, explains Brunger, so optimizing the system to minimize leakage was very important early on.

Equally importantly, the researchers set up their system so that they could begin with a very well-defined set of conditions. They immobilized acceptor vesicles onto a surface and then allowed labeled donor vesicles to bind in the absence of calcium, a known trigger for neuronal fusion. They monitored the system with very sensitive cameras, using total internal reflection microscopy to detect interacting vesicles via the weak signal from the quenched lipid dye in the donor. After extensive washing and a sufficiently long incubation to allow calcium-independent processes to reach a plateau, Brunger and colleagues had in hand a meta-stable preparation of interacting vesicles mimicking the ready-releasable pool of synaptic vesicles in the neuron. “When they are docked,” explains Minjoung Kyoung, first author on the paper describing this work, “the vesicles don't burst; they don't leak; they stay very well together, just interacting; and then when we add calcium, they fuse.”

As the system includes both lipid- and content-mixing reporters, and because the researchers monitor single vesicles, different events have characteristic fluorescent signals and can be distinguished on this basis. “We can distinguish interacting vesicles from just lipid mixing and from full fusion, so we know the nature of the events,” says Brunger. Also, in contrast to ensemble measurements, in which single vesicles are not observed and which is how reconstituted systems have been monitored in the past, this allows one to effectively distinguish between fusing vesicles and those that may burst or leak accidentally.

Notably, the reconstituted system showed rapid full vesicle fusion upon injection of calcium and could recapitulate known in vivo effects of mutant synaptotagmin and complexin. As in vivo, the system behaves cooperatively as calcium concentration is increased, though the levels of calcium that trigger fusion are one to two orders of magnitude higher than those in vivo. The minimal system is most probably lacking components needed to perfectly mimic the situation in the cell. It is, however, undoubtedly an excellent starting point to quantitatively study the functions of additional players in neurotransmitter release.

Brunger and colleagues hope to reach a complete mechanistic understanding of synaptic vesicle fusion at the single-molecule level. “We want,” says Brunger, “to make movies of this process, and this system is a major stepping stone to this goal.”