Is the quantum of transmitter release composed of subunits? A critical analysis in the mouse and frog.

1. Miniature end-plate potentials (m.e.p.p.s) were recorded from mouse diaphragm and frog cutaneous pectoris muscles and miniature end-plate currents (m.e.p.c.s) were recorded from frog cutaneous pectoris to investigate the proposal that the m.e.p.p. is built up of one to thirty subunits. Evidence for this hypothesis is drawn mainly from the observations that there is a class of m.e.p.p.s smaller than the classical m.e.p.p.s, and that histograms of m.e.p.p. amplitudes display multiple peaks which often appear to be regularly spaced and which extend throughout the histograms; in terms of the subunit hypothesis each successive peak results from an increasing integral number of subunits per m.e.p.p. (Kriebel, Llados & Matteson, 1976; Wernig & Stirner, 1977).2. Histograms of m.e.p.p. amplitudes and m.e.p.c. areas confirmed the existence of two classes of m.e.p.p.s as reported by Kriebel & Gross (1974): a larger class (well described by a Gaussian curve) which consists of the classical m.e.p.p.s, and a smaller class with amplitudes considerable less than the classical m.e.p.p.s.3. Histograms of m.e.p.p. amplitudes and m.e.p.c. areas showed multiple peaks that extended throughout the histograms.4. Autocorrelations of the histograms, an unbiased method used to test for regularity in data, showed that the multiple peaks were not regularly spaced, as required by the subunit hypothesis.5. A series of computer simulations demonstrated that, for expected levels of base-line noise in the recording system, multiple peaks that extend throughout histograms of m.e.p.p. amplitudes could arise from subunits only if the standard deviation of the subunit amplitude were less than 2-5% of the mean subunit amplitude and the standard deviation of the variability in post-synaptic sensitivity were less than 2% of the mean post-synaptic sensitivity. It seems unlikely that the variability in post-synaptic sensitivity and in proposed subunit amplitude would be as small as this.6. Taking more realistic estimates for the standard deviation of the subunit amplitude of 12% of the mean subunit amplitude and standard deviation of the variation in post-synaptic sensitivity of 4% of the mean sensitivity, it was found that at most three to four regularly spaced peaks would be apparent in m.e.p.p. amplitude histograms due to subunits.7. On the basis of these theoretical considerations it seems doubtful that the multiple peaks observed to extend throughout histograms of m.e.p.p. amplitudes could arise from subunits; therefore, the experimental data that have been used to support the subunit hypothesis are unlikely to have arisen from a subunit mechanism.8. We suggest that there are few, if any, data that directly support the subunit hypothesis. The multiple peaks observed to extend throughout m.e.p.p. amplitude histograms most likely arise from random variation in the data, although additional factors cannot be ruled out.