Rhythmic patterns in the thoracic nerve cord of the stick insect induced by pilocarpine.

Bath application of the muscarinic agonist pilocarpine onto the deafferented stick insect thoracic nerve cord induced long-lasting rhythmic activity in leg motoneurones. Rhythmicity was induced at concentrations as low as 1x10(-4) mol l-1 pilocarpine. The most stable rhythms were reliably elicited at concentrations from 2x10(-3) mol l-1 to 5x10(-3) mol l-1. Rhythmicity could be completely abolished by application of atropine. The rhythm in antagonistic motoneurone pools of the three proximal leg joints, the subcoxal, the coxo-trochanteral (CT) and the femoro-tibial (FT), was strictly alternating. In the subcoxal motoneurones, the rhythm was characterised by the retractor burst duration being correlated with cycle period, whereas the protractor burst duration was almost independent of it. The cycle periods of the rhythms in the subcoxal and CT motoneurone pools were in a similar range for a given preparation. In contrast, the rhythm exhibited by motoneurones supplying the FT joint often had about half the duration. The pilocarpine-induced rhythm was generated independently in each hemiganglion. There was no strict intersegmental coupling, although the protractor motoneurone pools of the three thoracic ganglia tended to be active in phase. There was no stereotyped cycle-to-cycle coupling in the activities of the motoneurone pools of the subcoxal joint, the CT joint and the FT joint in an isolated mesothoracic ganglion. However, three distinct 'spontaneous, recurrent patterns' (SRPs) of motoneuronal activity were reliably generated. Within each pattern, there was strong coupling of the activity of the motoneurone pools. The SRPs resembled the motor output during step-phase transitions in walking: for example, the most often generated SRP (SRP1) was exclusively exhibited coincident with a burst of the fast depressor trochanteris motoneurone. During this burst, there was a switch from subcoxal protractor to retractor activity after a constant latency. The activity of the FT joint extensor motoneurones was strongly decreased during SRP1. SRP1 thus qualitatively resembled the motoneuronal activity during the transition from swing to stance of the middle legs in forward walking. Hence, we refer to SRPs as 'fictive step-phase transitions'. In intact, restrained animals, application of pilocarpine also induced alternating activity in antagonistic motoneurone pools supplying the proximal leg joints. However, there were marked differences from the deafferented preparation. For example, SRP1 was not generated in the latter situation. However, if the ipsilateral main leg nerve was cut, SRP1s reliably occurred. Our results on the rhythmicity in leg motoneurone pools of deafferented preparations demonstrate central coupling in the activity of the leg motoneurones that might be incorporated into the generation of locomotion in vivo.