Pattern generators for muscles crossing more than one joint.

The activation pattern of many hindlimb flexor and extensor motoneurons (MNs) in walking can be described as simple: a single burst of spikes in the appropriate phase of walking. However, in muscles which cross more than one joint (here called bifunctional muscles), the activation pattern is complex, consisting of activity during both the flexion and extension phases. The neural origin of the complex activation pattern is unknown. Perret and Cabelguen presented a circuit diagram which they hypothesized would give the observed patterns. Their model suggested that the bifunctional MNs received output from both flexor and extensor portions of the central pattern generator (CPG), and included effects of sensory feedback on the CPG. Jordan et al. showed a strict reciprocity in locomotor drive potentials between flexors and extensors, and hypothesized a distribution network composed of excitatory and reciprocally inhibitory interneurons (INs). We have collected fictive locomotion data on simple and bifunctional MNs and have begun to test hypotheses about how the brief burst at onset of flexion is produced in the bifunctional MNs. Here we report our initial experiments, using large-scale simulation, on minimal circuits which might generate such patterns.