Transmission at a central inhibitory synapse. II. Quantal description of release, with a physical correlate for binomial n.

Publisher Website
Google Scholar
Abstract
Simultaneous intracellular redordings were obtained from the goldfish Mauthner cell (M-cell) and adjacent identifiable inhibitory interneurons. Amplitude fluctuations of unitary inhibitory postsynaptic potentials (IPSP) produced by directly evoked presynaptic impulses were subjected to quantal analysis. The release parameters were correlated with histological features of the presynaptic cells, which were systematically iontophoretically injected with horseradish peroxidase (HRP) and reconstructed. A computational procedure was developed that provided the probability density function (PDF) for a given set of responses and, by minimizing the effects of background noise through a deconvolution process, allowed optimal fits of the data according to the predictions of both Poisson and binomial equations. These 2 models were further compared on the basis of their likelihood criteria. The binomial relation always gave a more adequate description of the PDF than did the Poisson, this conclusion being substantiated by statistical tests. A striking equivalence was found between the binomial term n and the number of stained presynaptic boutons. This identity was verified for 18 cells that met the mathematical requirement for the analysis (namely, a mean quantal content < 7), with binomial n ranging from 3-28. The probability of release parameter, p, was consistnetly high, averaging 0.37, which further favors the adoption of the binomial model. There was a possible tendency for p to be inversely related to n. In the analysis of data from 6 additional cells for which an elevated np product required that the binomial n be fixed to the value of its histological counterpart, the average p was similar, suggesting that the validity of the model extends to cells with an even larger number of terminal boutons. The binomial quantal size, q, was compared for all cells studied by expressing its respective computed amplitudes relative to that of the full-sized collateral IPSP, which is itself a constant fraction of the driving force. In this manner, q was remarkably constant at .apprx. 1.1% of the full-sized response. Calculations taking into account the value of q, the resting membrane conductance and the conductance increase during a unitary IPSP, provided support for the postulate that the action of a single quantum (and, therefore, of 1 releasing temrinal) is functionally similar to that of the contents of a single synaptic vesicle. The method used for this study was applied to the synaptic depression observed during high-frequency stimulation. Despite a limited sample, evidence was obtained that even under those conditions every synaptic bouton continues to function as an independent all-or-none releasing unit; the reduction in IPSP amplitudes could be solely attributed to a lower probability of release, p.