Are transmitter release statistics meaningful?

Abstract

A BASIC premise of synaptic physiology is that neurotransmitter is released from motor nerve terminals in multimolecular packets called quanta^1,2. The rate of release of quanta is low at rest, but when an action potential invades the nerve terminal there is a large increase in the mean rate of release leading to the generation of an endplate potential. It has been found that the distribution of the number of quanta released by an action potential is accurately described by a binomial distribution^3–8. The assumption which underlies the binomial process is that n independent units (quanta) each have the same probability of success (release), p, on a single trial (stimulation). The parameters n and p can be estimated from the mean and variance of the binomially distributed variable. Although it is not clear what the physical correlates of n and p might be, it has been suggested that n may refer to the number of releasable quanta¹, or to the number of release sites which can be activated on stimulation⁶. The changes in the apparent values of n and p which occur during synaptic processes such as facilitation^6,8 and depression⁷, have been interpreted as providing evidence for particular mechanisms which may underly those processes.