The stopping response of Xenopus laevis embryos: pharmacology and intracellular physiology of rhythmic spinal neurones and hindbrain neurones.

1. Xenopus laevis embryos stop swimming in response to pressure on the cement gland. This behaviour and 'fictive' stopping are blocked by bicuculline (10 mumol 1(-1)), tubocurarine (110 mumol 1(-1)) and kynurenic acid (0.5 mmol 1(-1)). 2. Intracellular recordings from spinal neurones active during swimming have shown that pressure on the cement gland evokes compound, chloride-dependent inhibitory postsynaptic potentials (IPSPs). These are blocked by bicuculline, tubocurarine and kynurenic acid, but are unaffected by strychnine (2 mumol 1(-1)). 3. When the cement gland is pressed, trigeminal ganglion activity precedes both the IPSPs and the termination of 'fictive' swimming activity recorded in rhythmic spinal neurones. The trigeminal discharge is unaffected by the antagonists bicuculline, tubocurarine, kynurenic acid and strychnine. 4. Intracellular recordings from the hindbrain have revealed neurones that are normally silent, but rhythmically inhibited during 'fictive' swimming. In these neurones pressure on the cement gland evokes depolarising potentials, often with one or more spikes. 5. We propose that the stopping response depends on the excitation of pressure-sensitive trigeminal receptors which innervate the cement gland. These release an excitatory amino acid to excite brainstem GABAergic reticulospinal neurones, which inhibit spinal neurones to turn off the central pattern generator for swimming. There may also be a less direct pathway.