Abstract
1. Facilitation of endplate potentials (EPPs) and frequency of miniature endplate potentials (MEPPs) were studied, in the presence of low Ca2+/raised Mg2+, in isolated mouse hemidiaphragm, using pseudo-random sequences of nerve stimulation and automated (computer) counting of MEPPs and quantal components of EPPs. 2. The facilitation in quantal content of EPPs (m) produced by one or more antecedent stimuli was accompanied by facilitation of MEPP frequency (fm) that was similar in magnitude and substantially less than expected if facilitation reflects persistent (residual) intraterminal Ca2+. The time course of 'phasic' quantal release, associated with the EPP, was little if at all altered with facilitation. 3. The magnitude and time course of facilitation was consistent with two distinct presynaptic processes, each manifest both in m and fm, (i) an effect to multiply transmitter release, and (ii) residual Ca2+ which adds to Ca2+ brought in by nerve impulses. These have distinct time courses. 4. After loading nerve terminals with bis (O-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA), facilitation of m and fm became very small. 5. At sufficiently low Ca2+/raised Mg2+ facilitation of m and fm became very small although latency histograms showed clear EPPs. However, the multiplicative component of facilitation became maximal at Ca2+/Mg2+ concentrations giving an average m value less than 0.1, corresponding to about 5% of normal Ca2+ entry per pulse. At lower Ca2+, facilitation was restored when EPPs were made larger using 4-aminopyridine. 6. With EPPs elicited by brief 'direct' nerve terminal depolarizations, facilitation was graded with pulse intensity (and m) and could be much less than with EPPs with similar m evoked by nerve stimuli at lower Ca2+ and/or higher Mg2+. 7. It was concluded that fast facilitation is primarily multiplicative and reflects activity within the nerve terminal of a Ca(2+)-sensitive process distinct from that generating Ca(2+)-dependent release.