Spatiotemporal activation of lumbosacral motoneurons in the locomotor step cycle

The aim of this study was to produce a dynamic model of the spatiotemporal activation of ensembles of alpha motoneurons (MNs) in the cat lumbosacral spinal cord during the locomotor step cycle. The coordinates of MNs of 27 hindlimb muscles of the cat were digitized from transverse sections of spinal cord spanning the entire lumbosacral enlargement from the caudal part of L(4) to the rostral part of S(1) segments. Outlines of the spinal cord gray matter were also digitized. Models of the spinal cord were generated from these digitized data and displayed on a computer screen as three-dimensional (3-D) images. We compiled a chart of electromyographic (EMG) profiles of the same 27 muscles during the cat step cycle from previous studies and used these to modulate the number of active MNs in the 3-D images. The step cycle was divided into 100 equal intervals corresponding to about 7 ms each for gait of moderate speed. For each of these 100 intervals, the level of EMG of each muscle was used to scale the number of dots displayed randomly within the volume of the corresponding MN pool in the digital model. One hundred images of the spinal cord were thereby generated, and these could be played in sequence as a continuous-loop movie representing rhythmical stepping. A rostrocaudal oscillation of activity in hindlimb MN pools emerged. This was confirmed by computing the locus of the center of activation of the MNs in the 100 consecutive frames of the movie. The caudal third of the lumbosacral enlargement showed intense MN activity during the stance phase of locomotion. During the swing phase, the focus of activation shifted abruptly to the rostral part of the enlargement. At the stance-swing transition, a transient focus of activity formed in the most caudal part of the lumbosacral enlargement. This was associated with activation of gracilis, posterior biceps, posterior semimembranosus, and semitendinosus muscles. These muscles move the foot back and up to clear the ground during locomotion, a role that could be described as retraction. The spatiotemporal distribution of neuronal activity in the spinal cord during normal locomotion with descending control and sensory inputs intact has not been visualized before. The model can be used in the future to characterize spatiotemporal activity of spinal MNs in the absence of descending and sensory inputs and to compare these to spatiotemporal patterns in spinal MNs in normal locomotion.