A correlative physiological and morphological analysis of monosynaptically connected propriospinal axon–motoneuron pairs in the lumbar spinal cord of frogs

Intracellular stimulation of single propriospinal axons evoked excitatory postsynaptic potentials (EPSPs) in lumbar motoneurons. Mean EPSP amplitudes differed by two orders of magnitude when measured in different connections. After analyzing the distribution of mean amplitudes of 47 single-fiber EPSPs, two populations of responses could be defined: (1) those with mean amplitudes between 0.1 and 1.2 mV (mean±S.D.: 0.48±0.30 mV, 34 pairs), which is in the range of values typical for single-fiber EPSPs evoked by stimulation of supraspinal fibers and primary muscle afferents, (2) those with mean amplitudes between 1.6 and 8 mV (4.2±2.0 mV, 13 pairs). Both populations of responses had similarly short latencies and rise times and responded similarly to paired-pulse stimulation, consistent with monosynaptic transmission. However, the high-efficacy connections had significantly smaller coefficients of variation of EPSPs, as well as increased quantal content and quantal size. Tetanic stimulation gradually depressed the amplitude of large EPSPs by 81–86%, but did not affect small EPSPs. Recovery of large EPSPs was exponential with a time constant of 3–5.6 min. During post-tetanic depression the amplitude ratio between the test and conditioned EPSPs evoked by paired-pulse stimulation was not changed but the coefficient of variation was increased, suggesting that the depression was due to depletion of synaptic vesicles available for release. Intracellular labeling of seven electrophysiologically studied propriospinal axon–motoneuron pairs revealed that the number of axon varicosities establishing close appositions with dendrites of the labeled motoneuron was higher for connections where large-amplitude EPSPs were recorded. These varicosities were more often located on proximal dendrites of motoneurons than those of low-efficacy connections. In addition, the number of boutons in highly effective connections was several times lower than the maximal number of available quanta estimated from physiological data, implying that the large EPSPs may be generated by multivesicular release from presynaptic boutons. We conclude that the efficacy and related mode of use-dependent modulation of propriospinal connections is determined by a number of factors, including the number and position of synaptic contacts and the number of active zones or vesicles available for release.