Adaptive Cerebellar Processing at Cellular Resolution in Flexible Behavior

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT The cerebellum integrates sensory, motor, and internal information to rapidly guide and fine-tune action. This process has been investigated most extensively for movement control, but the cerebellum is also involved in the updating of internal states such as reward and working memory. Previous work from this laboratory shows that the cerebellar region crus I is required for evidence accumulation and decision-making. These findings, along with preliminary data, led to the hypothesis that cerebellar processing of sensory and internal information evolves over the course of learning to exert moment-to-moment predictive influence and shape flexible behavior. The proposed experiments will determine, with quantitative rigor, how cognitive regions of the cerebellum contribute to neural coding, predictive learning, and forebrain target activity. Past studies of cerebellar contributions to cognition have been hampered by the coarseness with which neuronal activity could be monitored and perturbed, pathways traced, and behavior measured. This proposal will overcome these limitations by using advanced tools, including two-photon calcium imaging, whole-brain transsynaptic viral tracing, high-density silicon probe recording, and optogenetic perturbation. Aim 1 will determine how predictive information in cerebellar activity influences working memory. In an evidence-accumulation decision task that distinguishes neural activity related to evidence accumulation, information retention, and decisions, preliminary data show that optogenetic inactivation of crus I removes the dependence of decisions on previous evidence, indicating a necessary role in evidence integration. This aim will examine the main cerebellar pathway with optogenetics, two-photon imaging, and many-electrode recording to probe learned cerebellar contributions to sensory processing, working memory, decisions, and motor output with subsecond time resolution. Aim 2 will characterize learning and transfer of working memory-related neural dynamics. This aim will examine how task representations evolve during learning in Purkinje cells and deep-nuclear neurons to test the idea that intrinsic cerebellar signals involved in movement preparation provide a foundation for learning neural responses that accumulate sensory evidence over time. Aim 3 will evaluate how cerebellar areas involved in cognition shape activity in connected forebrain areas. This aim will use transsynaptic viral tracing to identify pathways from crus I through midbrain and thalamus to their targets in the neocortex, and then specifically perturb and monitor these pathways to identify their contribution to task performance. The long-term goal of this project is to build a quantitative explanatory framework for cerebellar function in complex behavior. The results are expected to inform computational models that predict and explain the impact of detailed cerebellum-forebrain interactions. Together, these studies will significantly advance basic neuroscience of the cerebellum and contribute to understanding of syndromes marked by cerebellar dysfunction, including attention-deficit hyperactivity disorder and autism spectrum disorder.