Rhythmic motor activities including walking, running and swimming are controlled by spinal circuits known as central pattern generators (CPGs). These circuits integrate descending command signals from the brain and ascending feedback signals indicating muscle length and force. The basic motor pattern for walking is generated by a CPG located within the lumbar spinal cord. Although the descending commands are compromised in spinal cord injury, afferent sensory signals retain CPG access and are thought to be recruited in gait rehabilitation therapies including treadmill training and epidural stimulation. More effective engagement of CPG elements via afferent signaling may improve therapeutic outcomes. Extensive data from the cat and the rat have identified hip extension and ankle load as strong modulators of stance and swing phase timing. Activation of these afferents perturb the gait cycle in ways that may or may not persist as a phase angle shift in subsequent cycles. These results have suggested a two-layer CPG structure in which a rhythm-generating (RG) layer acts as the metronome and directs a pattern-forming (PF) layer that recruits motoneurons. Sensory inputs to the RG neurons are expected to have the most profound effects on locomotor function. However, the organization of sensory feedback within the CPG is poorly understood, in part because the cellular identity and connectivity of CPG circuit elements have only begun to be described within the last decade. Recently, a population of neurons that are Shox2+/Chx10- (Shox2RG) has been putatively identified as part of the RG layer. The goal of the proposed project is to use a combination of biological and computational approaches to explore how sensory afferent inputs are organized within the RG layer of the CPG and contribute to stance/swing phase duration and onset. We hypothesize that evoked polysynaptic inputs to Shox2RG neurons underlie the afferent modulation of gait timing. We have preliminary data showing that stimulation of specific proprioceptive afferents results in postsynaptic responses in many Shox2 neurons. We will test integration of information from specific afferents in Shox2RG neurons by separately stimulating nerves innervating flexor and extensor muscles at the hip and ankle during visually-guided patch clamp of Shox2RG neurons in isolated spinal cord preparations. Intermediate neurons in these polysynaptic pathways will be tested to see whether convergence of afferent input is mediated at the level of Shox2RG neurons or before. The pattern of input measured in the quiescent preparation will be tested and compared to that obtained during fictive locomotion. These data will be integrated into a computational model of the CPG to test for the sufficiency of the experimental data to account for observed sensory modulation of CPG timing and activation. Results are expected to provide new insights into the fundamental operation of feedback systems affecting RG circuits which could be used in the design of effective therapeutics for the recruitment of locomotor CPGs in injury and disease.
The basic motor pattern for rhythmic activities such as walking and running are generated by spinal neural circuits known as central pattern generators (CPGs). The locomotor CPG integrates descending information from the brain and ascending sensory information to produce fluid walking movements. Understanding the organization of sensory inputs modulating the CPG is expected to have therapeutic relevance to spinal cord injury and stroke, in which descending controls are impaired but ascending controls remain intact.
