Proprioceptive Coding in the Cuneate Nucleus of Awake MonkeysPublic
Proprioception, or the sense of one’s body in space, provides critical feedback that the brain usesto generate controlled movements. When proprioceptive feedback is lost, people find it difficult to perform even basic motor tasks. Despite its importance, proprioceptive coding of single neurons in the cuneate nucleus (CN), the most peripheral somatosensory nucleus of the brain, had never been studied in awake animals. During my doctoral work, I developed methods to record single neurons in CN of awakeanimals for the first time ever. I examined two fundamental properties of CN neurons, 1) how their sensitivity to proprioceptive information changes across contexts and 2) how many muscles typically compose their receptive fields (RFs). I recorded from CN of monkeys trained to perform reaching tasks and to tolerate bumps appliedto their hand. I found that in contrast to the typical “sensory gating” of tactile signals, the responses of proprioceptive CN neurons to movement are, on average, modestly potentiated during reach compared to rest. I propose that CN modulates sensitivity to enhance relevant information and attenuate irrelevant information. I also found that CN neurons with muscle-like RFs have properties that resemble those ofmuscle spindle afferents and don’t typically include signals from more than a single muscle, evidence for limited spatial convergence in CN. Looking for signs of processing along the neuraxis, I compared proprioceptive responses in CN to previous recordings from somatosensory cortex and found that many features of cortical proprioceptive neurons are already evident in CN, perhaps inherited from muscle receptors themselves. These results suggest that although CN relays proprioceptive signals that resemble muscle receptors, it does so in a context-dependent manner that allows for flexible representation of the sensory input, potentially to build “smart” brainstem and transcortical reflexes or to improve proprioceptive acuity when required by the task. My experiments, conducted as part of a research group focused primarily on motor control,sought to understand how proprioceptive coding in the medulla contributes to the generation of motor behaviors; however, a prevailing theory of motor cortical activity, neural dynamical systems (NDS), doesn’t typically take proprioceptive inputs into consideration. To address this shortcoming, I developed a model of motor cortex that combines the field of Optimal Feedback Control with NDS, in which feedback controllers in motor cortex are built using the intrinsic dynamics of sensory and motor cortices. In this dissertation, I lay out the features of this model and propose experiments that could validate or falsify its key predictions.