Simulation of synaptic depression, posttetanic potentiation, and presynaptic facilitation of synaptic potentials from sensory neurons mediating gill-withdrawal reflex in Aplysia

Abstract

The defensive gill-withdrawal reflex in Aplysia has proven to be an attractive system for analyzing the neural mechanisms underlying simple forms of learning such as habituation, sensitization, and classic conditioning. Previous studies have shown that habituation is associated with synaptic depression and sensitization with presynaptic facilitation of transmitter release from sensory neurons mediating the reflex. The synaptic depression, in turn, is associated with a decrease in Ca2+ currents in the sensory neurons, whereas presynaptic facilitation with increased Ca2+ currents produced indirectly by a decrease in a novel serotonergic sensitive K+ current. The present work represents an initial quantitative examination of the extent to which these mechanisms account for each of these types of synaptic plasticity. To address these issues a lumped parameter mathematical model of the sensory neuron release process was constructed. Major components of this model include Ca2+-channel inactivation, Ca2+-mediated neurotransmitter release and mobilization, and readily releasable and upstream feeding pools of neurotransmitter. In the model, release of neurotransmitter has a linear function of Ca2+ concentration and is not affected directly by residual Ca2+. The model not only simulates the data of synaptic depression and recovery from depression, but also qualitatively predicts other features of neurotransmitter release that it was not designed to fit. These include features of synaptic depression with high and low levels of transmitter release, posttetanic potentiation, a steep relationship between action potential duration and transmitter release, enhanced release produced by broadening the sensory neuron action potential (presynaptic facilitation), and dramatic synaptic depression with two closely spaced tetraethylammonium (TEA) spikes. The model cannot account fully for synaptic depression with empirically observed somatic Ca2+-current kinetics. Rather a large component of synaptic depression is due to reduction to the pools of releasable neurotransmitter (depletion). In the model when spike durations are greater than 15-20 ms, spike broadening produces little facilitation. However, when spike durations are more physiological, spike broadening leads to enhanced transmitter release.