Presynaptic inhibition of primary afferent transmitter release by 5- hydroxytryptamine at a mechanosensory synapse in the vertebrate spinal cord

The effects of the neuromodulatory monoamine 5-HT (serotonin) on a cutaneous mechanosensory (Rohon-Beard, R-B neuron) pathway in the spinal cord of postembryonic Xenopus laevis tadpoles have been examined. In paralyzed animals, exogenous 5-HT at 1-10 μM reversibly inhibits (within 1-2 min) the activation of fictive swimming in response to electrical stimulation of R-B free nerve endings in the skin. At threshold stimulus intensities for swimming under control conditions, intracellularly recorded EPSPs in contralateral motoneurons are completely abolished by 5-HT without any obvious change in neuronal conductance or membrane potential. However, increasing the stimulus voltage can activate swimming with enhanced motor burst discharge on each cycle (Sillar et al., 1992). This suggested that 5- HT inhibits the swim-initiating pathway rather than the motor rhythm- generating circuitry itself. Extracellular recordings from the central projections of R-B neurons indicated that the amine does not impair the generation of mechanoafferent impulses or their propagation into the spinal cord. However, 5-HT application blocks impulse activity in dorsolaterally positioned sensory interneurons (DLis) that are contacted by R-B neurons, suggesting that 5-HT acts at R-B to DLi synapses in the dorsal cord. By recording with microelectrodes from DLis, we find that skin stimulus-evoked EPSPs at this first-order synapse in the swim-initiating pathway are reversibly suppressed by 5-HT. No obvious change in DLi membrane potential or conductance could be detected during the inhibition, suggesting a presynaptic site of action for 5-HT. To investigate this suggestion further, the effects of 5-HT on the spontaneous release of R-B sensory transmitter (excitatory amino acid, EAA) were examined, again by recording postsynaptically from DLis. In quiescent preparations, DLis receive spontaneous glycinergic, GABAergic, and EAA receptor-mediated PSPs. The inhibitory potentials are abolished by strychnine and curare, respectively. The excitatory potentials that remain are not blocked by application of the calcium channel blocker cadmium chloride at 1 mM, but are suppressed by the EAA receptor antagonist kynurenic acid. They therefore resemble the TTX-resistant EPSPs described previously in Xenopus DLis (Sillar and Roberts, 1991), which are presumed to arise from the spontaneous liberation of EAA transmitter from R-B terminals. Bath application of 5-HT dramatically reduces the rate of occurrence of these spontaneous EPSPs consistent with a presynaptic locus for the inhibitory effects of 5-HT. Moreover, the reduction in the rate of spontaneous release and the suppression of evoked excitation in response to skin stimulation are comparable in time course and magnitude, implying that the two amine effects are causally related. In contrast, there is no apparent change in the distribution of spontaneous EPSP amplitudes, suggesting that the inhibition of evoked excitation by 5-HT does not additionally rely on a change in presynaptic quantal content or efficacy of postsynaptic EAA receptors. We conclude that one major site at which 5-HT inhibits R-B sensory pathway function is via a direct activation of 5-HT receptors located on the R-B afferent terminals, which leads to a reduction in the probability of EAA transmitter release onto DLis.