JOURNAL~F NEUROPHYSIOLOGY Vol. 74, No. 2, August 1995. Printed in U.S.A. Depression of Synaptic Connections Between Identified Motor Neurons in the Locust DAVID PARKER Department of Zoology, University of Cambridge, Cambridge CB2 3EJ, United Kingdom SUMMARY AND CONCLUSIONS 1. The fast extensor tibiae motor neuron makes direct excitatory central connections with the posterior group of flexor tibiae motor neurons in the locust metathoracic ganglion. The flexor group has a slow, a fast, and an intermediate motor neuron. The motor neu- rons are involved in the motor program for kicking and jumping, the defensive and escape behaviors of the locust. An antidromic action potential in fast extensor tibiae motor neuron (FETi) results in a monosynaptic, glutamatergic excitatory postsynaptic potential (EPSP) in each of the flexor motor neurons. 2. A train of 10 antidromic spikes in FETi at frequencies of l- 20 Hz resulted in depression of the amplitude of the EPSP in each of the flexor motor neurons. The depression was not significantly different in the different flexor motor neurons. The depression was greater with higher frequency stimulation and was reduced in low calcium saline. 3. After stimulation at 20 Hz, the EPSP amplitude was de- pressed by -80%. This did not change when the number of stimuli was increased to 20, when stimulation was done in high calcium saline, or when the frequency of stimulation was increased to 50 or 100 Hz. The recovery from depression was greater after 20-Hz stimulation than at lower frequencies, although the recovery was reduced when the number of stimuli was increased, and also in high calcium saline. 4. In normal saline the depression of the EPSP amplitude was associated with a reduction of the presynaptic spike amplitude at frequencies of 25 Hz. In tetraethylammonium (TEA) saline the width of a TEA-broadened spike was also reduced. The reduction in spike amplitude and spike width correlated with the depression of the EPSP. 5. Certain of these results are consistent with a depletion model of synaptic depression, whereas others are not consistent with this model. The depression may be partly due to an initial depletion of transmitter stores, and partly to modulation of the presynaptic ac- tion potential that reduces calcium entry, and therefore transmitter release. The significance of the depression on the motor program for kicking and jumping is discussed. INTRODUCTION The properties of connections between neurons are not static but can change depending on the history of the syn- apse. This plasticity allows for alterations in the output of neural circuits, due to an increase or decrease in synaptic strength. This paper analyzes synaptic depressionof connec- tions between identified motor neurons in the locust. In the classical model of synaptic depression, repeated stimulation depletes the releasable transmitter store, re- sulting in a reduction of transmitter release and therefore of the postsynaptic response (Liley and North 1953; Wachtel and Kandel 1967). Elmqvist and Quastel ( 1965) suggested that each action potential released a fraction of the releasable transmitter store and that this store was incompletely replen- ished at high frequencies, resulting in a successivereduction in the amount of transmitter released. Support for this hy- pothesis has been obtained from the squid giant synapse, where there is a linear relationship between depression and the amount of transmitter released (Kusano and Landau 1975), but no effect on the presynaptic action-potential shape(Takeuchi and Takeuchi 1962) or presynaptic calcium entry (Charlton et al. 1982). Rosenthal ( 1969) also sug- gested that transmitter depletion was involved in synaptic depression at frog end-plate synapses. The junction poten- tials were initially depressed after 50-Hz stimulation, the depressionbeing followed by potentiation. In low calcium/ high magnesium saline, there was no depression.The reduc- tion in calcium concentration was suggestedto prevent the depletion of the transmitter store by reducing transmitter release, thus reducing the depression. A morphological basis for depletion has also been shown in the terminals of Aplysia sensory neurons, where the number of vesicles adjacent to active zones has been shown to be reduced at depressed synapses compared with control (Bailey and Chen 1988). Depletion of releasable transmitter stores, however, does not explain depression in all cases. In the crayfish, depression at the abdominal fast flexor neuromuscular junction was found to be greater after a single impulse than for impulses in a train (Zucker and Brunner 1977). In addition, the de- pression was not affected by high magnesium solutions, which would be expected to reduce transmitter release and thus depletion. Byrne (1982) found that in Aplysia sensory neurons the depression was not a monotonic function of spike interval, and the recovery from depressionwas greater for shorter interstimulus intervals (where depression was greater), than for longer interstimulus intervals. These re- sults, therefore, do not fit with a depletion model of synaptic depression but suggest that other mechanisms are involved. At sensory neuron synapses in Aplysia, depression was correlated with a reduction in presynaptic calcium entry, presumably reducing transmitter release and hence postsyn- aptic response (Klein et al. 1980). The depressionwas there- fore not due to a reduction in the amount of transmitter available for release, but to a reduction in probability of release.Similarly, the depressionof inhibitory synapses onto the goldfish Mauthner cell is due to a reduction in the proba- bility of release, P, and not to a reduction in the amount of transmitter available for release, n, as the depletion hypothe- sis predicts (Korn et al. 1984). However, modeling of the sensorimotor synapse with the use of the data of Klein et 0022-3077/95 $3.00 Copyright 0 1995 The American Physiological Society 529