LARGE-AMPLITUDE VARIABILITY OF GABAERGIC IPSCS IN MELANOTROPES FROM XENOPUS-LAEVIS - EVIDENCE THAT QUANTAL SIZE DIFFERS BETWEEN SYNAPSES

1. We made in situ whole-cell recordings from melanotropes in the inte rmediate lobe of the pituitary gland of Xenopua laevis. Melanotropes r eceived spontaneous synaptic inputs that had a fast rise time and a mu ch slower decay. These inputs were GABAergic inhibitory postsynaptic c urrents(IPSCs): they followed the reversal potential for chloride ions and they were blocked by the gamma-aminobutyric acid-A (GABA,) recept or antagonist bicuculline. 2. Because of the very low baseline noise i t was possible to see discrete levels in the tails of IPSCs that corre sponded to the opening of one or more synaptic GABA(A) receptor channe ls. ''All-points'' histograms of the IPSCs showed that the chord condu ctance of the channels in the tails of the IPSCs was 21.6 +/- 0.6 pS ( mean +/- SE, n = 6). 3. The amplitudes of the spontaneous IPSCs were v ery variable, ranging from 3 to 390 pA at a holding potential of -80 m V. The average of the median amplitudes was -67.5 +/- 5.9 pA (n = 28). The amplitude distributions of the IPSCs were well described by the s um of two lognormal distributions with large SDs. The average of the m eans of the first lognormal distribution was 27.8 +/- 5.3 pA(n = 10); the average of the SDs was 24.7 +/- 8.1 pA. For the second lognormal d istribution these values were 87.0 +/- 13.4 and 33.7 +/- 7.4 pA. An av erage of 41.8 +/- 6.9% ofthe IPSCs originated from the first lognormal distribution. 4. The large variability in the amplitudes of spontaneo us IPSCs was not the result of presynaptic action potentials because i t was not reduced by tetrodotoxin (TTX), Ca2+-free extracellular solut ion, or the combined application of TTX and Mn2+ 5. The time course of the IPSCs from the first and the second lognormal distributions were very similar: averages of the median 20- to 80% rise times were 585 +/ - 64 and 488 +/- 28 mu s, respectively (n = X), whereas the decays wer e well described by the sum of two exponential functions, with fast ti me constants of 8.9 +/- 1.1 (n = 7) and 9.3 +/- 3.3 ms and slow time c onstants of 29.5 +/- 3.3 and 3 1.7 +/- 2.6 ms, respectively 6. The dec ay of the IPSCs was voltage dependent; it was similar to 3 times slowe r at a holding potential of +40 mV than at -80 mV (n = 5). Desensitiza tion of the GABA(A) receptor does not seem to play a role in the decay of the IPSCs because in outside-out patches this desensitization was much slower (1.3 +/- 0.3 s, n = 6) than the decay of the IPSCs and not dependent on voltage. 7. Melanotropes from animals that had been adap ted to a black background for greater than or equal to 3 wk had larger whole-cell capacitances (8.6 +/- 1.1 pF, n = 36) than cells from whit e-adapted animals (4.7 +/- 0.4 pF, n = 23), which indicates a more tha n twofold increase in cell volume. During most experiments the frequen cy of the spontaneous IPSCs remained constant. Average intervals were 9.8 +/- 2.9 s (n = 17) in cells from black-adapted animals and 2.8 +/- 0.9 s(n 11) in cells from white-adapted animals; however, this differ ence was not significant. No clear differences were observed between t he amplitudes and time courses of the IPSCs in white-adapted versus bl ack-adapted animals. 8. Stimulus-evoked IPSCs were blocked by TTX (1 m u M) and by bicuculIine (20 mu M), a GABA, receptor antagonist, which means that they were evoked by action potentials in the afferent GABAe rgic fibers. The stimulus-evoked IPSCs had sharp thresholds, their ave rage amplitudes did not increase with higher stimulation intensities, and their delays were well described by a single Gaussian distribution , suggesting that a single fiber was stimulated. 9. The probability of release was dependent on the extracellular Ca2+ concen