Single-domain/bound calcium hypothesis of transmitter release and facilitation.

1. We describe a model of transmitter release that is based on the finding that release can be gated during the opening of individual Ca2+ channels, suggesting that the release site can be activated by the Ca2+ domain under a single channel. In this model each release site contains four independent Ca2+ binding sites or gates with unbinding kinetics graded from slow to fast and affinities ranging from high to low. All four gates must be bound for release to occur. Thus synaptic dynamics are governed by the kinetics of Ca2+ binding and unbinding from release sites, not Ca2+ diffusion. 2. Fast facilitation occurs when an action potential invades a terminal with one or more ions remaining bound to the release sites. Residual free Ca2+ is not necessary for facilitation with this mechanism, but if present it would enhance facilitation by binding to high-affinity gates between pulses. 3. This model can account for key features of release. These include fourth-power cooperativity with regard to external Ca2+; a release time course that is virtually independent of an increase in quantal content; an inverse relation between external Ca2+ and the degree of facilitation; and a steplike increase in facilitation with increasing stimulus frequency, with each step corresponding to a unitary decline in the Ca2+ cooperativity. 4. Facilitation of single-channel-based secretion is shown to be robust even if channel opening is stochastic. Spontaneous release of transmitter, assumed to be due in part to spontaneous Ca2+ channel openings, is shown to be elevated during and after a train of impulses. 5. An extension of the model to include multiple Ca2+ channels per release site demonstrates that one role of overlapping Ca2+ domains may be to accentuate depolarization-evoked release relative to spontaneous release.