Contributions of the motor cortex to the control of the hindlimbs during locomotion in the cat.

Although the corticospinal tract is not essential for the production of the basic locomotor rhythm in cats, it does contribute to the regulation of locomotion, particularly in situations in which there is a requirement for precise control over paw placement or limb trajectory. Lesions of the dorsolateral funiculi at the low thoracic level (T(13)) that completely interrupted both the cortico- and rubrospinal pathways produced long-term deficits in locomotion on a level surface. These deficits included a paw-drag that was probably caused both by a loss of cortico- and rubrospinal input to motoneurones controlling distal muscles as well as by a change in the relative timing of muscles acting around the hip and knee. Smaller lesions produced similar deficits from which the cats recovered relatively quickly. Cats with the largest lesions of the dorsolateral funiculi were unable to modify their gait sufficiently to step over obstacles attached to the treadmill belt even 3-5 months postlesion. These results imply that the medial pathways, the reticulo- and vestibulospinal pathways, are unable to fully compensate for damage to the lateral pathways. Single unit recordings from identified pyramidal tract neurones (PTNs) within the hindlimb representation of the primary motor cortex (area 4) showed that a substantial proportion of neurones (67%) significantly increased their discharge frequency when the cats modified their gait to step over obstacles attached to the treadmill belt. Of those PTNs that showed increased activity during the swing phase, populations of neurones were activated at different times. A large proportion of PTNS discharged early in swing, in phase with knee flexors such as the semitendinosus. Others discharged slightly later, in phase with the activity of ankle flexors, such as tibialis anterior, while still others discharged at the end of swing, in phase with digit dorsiflexors, such as the extensor digitorum brevis. We suggest that different populations of cortical neurones may specifically modify the activity of selected groups of close synergistic muscles during different parts of the swing phase. We further suggest that these modifications are mediated, in part, by groups of interneurones that are involved in determining the base locomotor rhythm. This provides a means by which the changes specified by the descending signal from the motor cortex may be smoothly, and appropriately, incorporated into the locomotor cycle.