Spontaneous quantal transmitter release: a statistical analysis and some implications.

1. Miniature end-plate potentials (m.e.p.p.s) were intra- and extracellularly recorded from neuromuscular junctions in rat phrenic nerve-diaphragm preparations in vitro.2. Statistical analysis of the intervals between m.e.p.p.s showed that when the mean number of events in time t was plotted as a function of the variance of the events in time t there was a significant deviation from the straight line relationship expected for a Poisson process. Computer simulation showed that this deviation is explicable if release was generated by the random phasing of the activity of a number of releasing sites.3. There was no indication that release of one quantum influences the probability of release of remaining quanta (drag, clustering). It is suggested that m.e.p.p.s whose amplitude is larger than the mode result from the release of the contents of vesicles whose volume is also supramodal.4. The effects of depolarization of nerve terminals upon the variance-mean curve suggest an increase in the activity of sites rather than an increase in their number.5. Statistical analysis indicated at least 200 +/- 100 (mean +/- 1 S.E.) releasing sites. This number is of the same order as the number of sites of vesicle aggregation and presynaptic membrane density seen in electron micrographs of nerve terminals of this preparation.