Properties of the early phases of crotoxin poisoning at frog neuromuscular junctions.

Addition of crotoxin to the frog neuromuscular junction in either 0.9 mM Ca2+ plus tubocurarine or 0.5 mM Ca2+ Ringer solution produced a triphasic change in the amplitude of nerve-evoked endplate potentials (e.p.p.s) and, with 0.5 mM Ca2+, a biphasic change in miniature endplate potential (m.e.p.p.) frequency. The secondary phase of rising e.p.p. amplitude was associated with an increase in facilitation of e.p.p. amplitude with closely spaced twin impulses; the increase in spontaneous release lagged behind that of evoked release. When a calcium chelator, BAPTA, was loaded into presynaptic nerve terminals to buffer cytosolic free Ca2+, both e.p.p. amplitude and twin-impulse facilitation were increased by crotoxin to a similar extent relative to that in the control without BAPTA. The duration of the increase in twin-impulse facilitation was reduced but the duration of the increase in e.p.p. amplitude was unaffected by BAPTA loading. The presence of BAPTA did not alter the characteristic changes in spontaneous release in response to crotoxin. These results suggest that the augmentation of evoked and spontaneous transmitter release by crotoxin is not primarily due to changes in cytosolic Ca2+. The response time between stimulus and e.p.p. peak was lengthened in all phases due to prolongation of the interval between stimulus and e.p.p. onset and, in the secondary and tertiary phases, slowing of e.p.p. rise-time. The protein kinase C inhibitor, H-7, produced complex changes in e.p.p.s. under control conditions but did not alter the triphasic response characteristic of intoxication. These results suggest that crotoxin initiates a primary disturbance in the phasic release process leading to a series of time-gated changes which include transient facilitation then uncoupling of phasic release and generalized acceleration of spontaneous release.