Positive feedback as a general mechanism for sustaining rhythmic and non-rhythmic activity.

Our aim is to reassess our proposal that various states of motor output may be sustained by positive feedback generated within the premotor neural circuitry. The evidence for this proposal came from the Xenopus embryo which when touched can swim for many seconds even after all movement has been prevented by a neuromuscular blocking agent. Experiments showed that even the spinal cord could sustain its own swimming activity for a few seconds after stimulation. We proposed that this was the result of the glutamatergic excitatory spinal interneurons synapsing with each other. Because this excitation is of long duration compared to the swimming cycle period it can sum from cycle to cycle to sustain swimming by a form of positive feedback. We have tested the plausibility of these ideas by making realistic computer simulations of the spinal networks and have shown that positive feedback can sustain stable swimming activity. Pharmacological evidence recently suggested that acetylcholine contributes to the excitation underlying swimming in spinal embryos so we investigated the central synapses made by motoneurons. Recordings from pairs of synergistic motoneurons then showed: a) cholinergic chemical synapses from more rostral motoneurons activate nicotinic receptors and produce excitation; and b) local intrasegmental electrical synapses also lead to mutual excitation. The presence of central motoneuron synapses suggested that they could contribute to excitation during swimming. We therefore used local drug applications to see if spinal neurons received cholinergic or electrical excitation during fictive swimming. The results show that motoneurons received both types of excitation while interneurons received only cholinergic excitation. This evidence suggests that when motoneurons are active during swimming they contribute positive feedback excitation not only to themselves but also to the premotor interneurons of the spinal rhythm generating network. This excitation would sum with that from 'glutamatergic' excitatory interneurons. We conclude that in addition to our original proposal of feedback between excitatory interneurons, there are other forms of positive feedback during swimming in the Xenopus embryo spinal cord. Motoneurons feed excitation back to each other. They may also contribute cholinergic excitation to premotor interneurons which could sum with the excitation from 'glutamatergic' interneurons and help to sustain swimming. If they do this, motoneurons may be a component part of the central pattern generator for swimming. Since central motoneuron synapses are a feature of most vertebrate groups, these results suggest a reevaluation of such synapses in these groups also.