Transmission at a central inhibitory synapse. IV. Quantal structure of synaptic noise.

1. Continuous segments of synaptic noise were recorded, in 12 experiments, from the voltage-clamped goldfish Mauthner (M-)-cell soma during depolarizations beyond the Cl- equilibrium potential so that spontaneous inhibitory postsynaptic currents (IPSCs) were outward going, i.e., opposite in polarity to excitatory transients. Inhibitory components constituting the noise were detected with a program based on their expected waveforms, and their amplitude distributions were analyzed according to a quantal model to determine whether these responses, presumably because of asynchronous impulses in a population of presynaptic cells, were integral multiples of the same minimal unit. 2. Synaptic noise in the M-cell is predominantly inhibitory, as it is abolished by the glycine antagonist strychnine. IPSC amplitudes varied by 20-fold or more and about one-third were grouped in bursts, with one or several components often occurring during the falling phase of a preceding one. To measure individual amplitudes, a base-line correction procedure was designed to extrapolate the decay of the leading IPSC, on the assumption that it was exponential with a time constant, tau, equal to that obtained by fitting averaged waveforms of distinct IPSCs recorded in quiet periods. Currents were expressed in nanoamperes and, for comparison between experiments, as percentages of the full-sized IPSC evoked by activation of the collateral network. 3. The probability distribution functions (PDFs) of inhibitory events generated with such an analysis had from 5 to 12 clear and equally spaced peaks, larger transients not being included because of their sparsity. These PDFs were fit with a sum of Gaussians, on the assumption that all peaks were integral multiples of the smallest unit and had the same standard deviation, sigma. The fits were statistically satisfactory, according to a chi 2 test, and mean quantal size was 0.63 +/- 0.17% (SD; range, 0.42-1.0%) of the collateral IPSC. 4. In eight experiments, tetrodotoxin (TTX) was subsequently applied topically to block presynaptic impulses and isolate "miniature" responses because of single exocytotic events. It had no apparent nonspecific effects, as indicated by the constancy of M-cell input conductance (8.01 +/- 3.11 microS vs. 7.90 +/- 2.89 microS). Amplitude distributions of residual IPSCs were unimodal and Gaussian, with their mean size (0.60 +/- 0.19%) and average rise time (0.548 +/- 0.128 ms) the same as those of the first peaks of control histograms. The latter parameter was also comparable with that of unitary IPSCs studied previously. These results thus confirmed that spontaneous and evoked events are due to transmitter release from the same afferent population.(ABSTRACT TRUNCATED AT 400 WORDS)