Evaluation of quantal neurosecretion from evoked and miniature postsynaptic responses by deconvolution method

A new deconvolution algorithm has been developed for evaluation of quantal content and its variability at high-output synapses. The algorithm derives the distribution of the number of neurosecretory quanta released in a trial (M) from the measured sizes of evoked postsynaptic responses. The deconvolution employs the distribution of quantal sizes obtained by measuring sizes of miniature postsynaptic responses. The distribution of quantal content M is derived by ridge regression method from the distributions of sizes of the responses and of quantal sizes. The deconvolution method was applied to postsynaptic responses from the excitory innervation of lobster dactyl opener muscle obtained by focal extracellular recordings. The obtained solution (distribution of M) had six to eight components and was stable. The method was tested by the analysis of simulated multiquantal responses. For the simulated responses, the ridge regression solution reproduced the imposed distribution of M within the limits of the calculated confidence intervals. To further test the algorithm, the distribution of M at a low-output synapse was obtained both by deconvolution method and by the method of direct quantal counts. The results of these two methods were found to be in a very good agreement.