Quantal mechanism of long-term synaptic potentiation ( synaptic memory / neuronal plasticity / crustacean neuromuscular synapses / opener muscle

D. A. Baxter,George D. BITTNERt,T. H. Brown

Abstract:Intracellular recordings were used to demonstrate the occurrence and to analyze the microphysiology of long-term synaptic potentiation (LTP) in the crayfish opener neuromuscular synapse. Brief stimulation of the single excitor motor axon enhanced the amplitudes of subsequent postsynaptic potentials for several hours. Three methods of quantal analysis were used to evaluate the mechanism responsible for LTP. The results of all three methods supported predictions of the hypothesis that LTP results from a presynaptic mechanism that increases the average of neurotransmitter quanta evoked by nerve impulses in the excitor axon. Long-term potentiation (LTP) is a use-dependent form of enhanced synaptic efficacy that can persist for hours and can be induced by activation of the synapses for only a few seconds or less (1-3). This great asymmetry between the duration of the synaptic activity and the duration of the subsequent synaptic change is the defining characteristic of LTP-a property that makes this phenomenon an interesting possible mechanism for long-term control of information transmission. Originally thought to be unique to the hippocampal formation, LTP is now known to occur in many vertebrate and invertebrate synapses (3-7). Both preand postsynaptic mechanisms have been proposed to be responsible for LTP (2, 3, 7-11). In testing preversus postsynaptic hypotheses regarding changes in synaptic efficacy, the technique of quantal analysis has proven historically to be valuable (12-20). Unfortunately, the conventional methods of quantal analysis cannot be universally applied, because the results can be subject to interpretational errors if certain key assumptions or conditions are unsatisfied (refs. 20-23; cf. refs. 24 and 25). For our first quantal analysis ofLTP, we therefore decided to select a synaptic preparation that would be most immune to such interpretational errors. An obvious choice was the crayfish opener-excitor neuromuscular synapse. This classical preparation has wellknown advantages for quantal analysis of changes in synaptic efficacy (12-14, 17) and our preliminary results (5) demonstrated that these synapses display LTP. We report here that three conventional methods of quantal analysis all supported predictions of the hypothesis that LTP in these synapses results from a presynaptic mechanism that increases the average number of neurotransmitter quanta evoked by nerve impulses in the axon. MATERIALS AND METHODS Experimental Preparation. All experiments were performed on cheliped opener muscles of crayfish (Procambarus clarkii). These muscles are innervated by a single excitor and a single inhibitor motor axon. Small animals (body length of 1-2 cm) were selected for the quantal analysis because the input resistance of their muscle fibers is sufficiently large that spontaneous miniature excitatory postsynaptic potential (MEPSP) amplitudes can be accurately measured by using intracellular recording techniques (cf. refs. 26-28). MEPSP amplitudes are known to be difficult or impossible to measure in crustacean muscle fibers, whose input resistance is much less than 200 kil (cf. refs. 12, 26-29). Isolated claws were maintained in physiological saline (205 mM NaCl/5.4 mM KCl/14 mM CaCl2/2.6 mM MgCl2, pH 7.4) at room temperature. The dorsal surface of the opener muscle was exposed in the propodite and the nerve bundle containing the single opener-excitor motor axon was isolated and stimulated in the meropodite by using conventional bipolar electrodes. This nerve bundle also contains some efferent innervation to the closer muscle as well as several afferent axons, but in the meropodite it is believed to contain only one efferent to the opener muscle-the excitor axon. Synaptic inhibition was blocked by adding 50 gM picrotoxin to the saline and severing the opener-inhibitor axon. Electrophysiological Methods. Because of the need to maintain stable measurements for long durations, intracellular recordings seemed preferable to the focal extracellular technique (28) that is more commonly used to study the microphysiology of these synapses. Evoked excitatory postsynaptic potentials (EPSPs) and spontaneous MEPSPs were recorded differentially. The intracellular and extracellular electrode tips, which were placed within <500 ,um of each other, were filled with 3 M KCl and had matching resistances of 2-5 MI. The filtered noise level was usually -<20 uV peak-to-peak. Experiments were terminated and data were excluded from analysis if the electrode noise noticeably increased, if the resting potential changed by >5 mV, or if the membrane potential became less negative than -70 mV. Single suprathreshold electrical stimuli were delivered to the nerve bundle containing the opener-excitor motor axon at 0.333 Hz for 20 min to 2 hr before and for 25 min to 9 hr after a brief stimulus train was delivered to the excitor axon. The stimulus trains used in the six most carefully studied claw preparations are given in Table 1. The amplified postsynaptic potentials were recorded on magnetic tape and later digitized (5 kHz) for computer analysis. Quantal Analysis. The quantum hypothesis (20, 21, 28, 30) holds that neurotransmitter substances are discharged from nerve terminals in the form of integral numbers of multimolecular packets or quanta. The mean size of an evoked EPSP V is a function of the average number of quanta released (mean quantal content m) and the mean amplitude of the postsynaptic potential produced by individual quantal releases (mean quantal size q). Specifically,