Cerebellar Circuits - Abstract The cerebellar cortex transforms mossy fiber inputs to granule cells, into predictions in the form of Purkinje cell outputs, and facilitates learning through synaptic refinement to correct inaccurate predictions. This scheme was originally advanced to account for motor control and motor learning, but we now know that the cerebellum is also critically involved in social and emotional behaviors, balance, cognition, and learning in reward tasks. The traditional view is that the same simple circuit comprised of a small number of cell types is repeated throughout the cerebellar cortex, with the only difference being the source of the inputs and the targets of the outputs. It is unlikely that this simple circuit captures the full complexity of cerebellar processing, particularly in light of our recent study that revealed many subtypes of cells: 9 Purkinje cells, 2 molecular layer interneurons, 3 Purkinje layer interneurons, 2 Golgi cells, 5 granule cells, and a continuum of unipolar brush cells. This tremendous complexity indicates that the basic cerebellar circuit is incomplete. The primary goal of my laboratory during the next eight years is to answer the following question: How do these different subtypes of cerebellar neurons contribute to signal processing and behavior? Clarifying the roles of these newly identified cell types and subtypes promises to lead to a new understanding of cerebellar processing. We also found that many of these cell types are differentially distributed within the cerebellar cortex. How is cerebellar circuitry regionally specialized to contribute to diverse behaviors? To answer these questions, we generated and acquired intersectional mouse genetic tools to precisely target cell types and subtypes, we combined electrophysiological and optogenetic methods to assess connectivity, we generated a large serial EM data set to reconstruct cerebellar circuits, we established in vivo multielectrode recording techniques, we established a battery of behavioral tests to assess cerebellar function, and we developed a highly constrained model of the cerebellar cortex. In the past three years, we began to study all types and subtypes of neurons in the cerebellar cortex and established a solid foundation for future investigations. Achieving the goals of this R35 promises to revolutionize our understanding of how different types and subtypes of neurons contribute to cerebellar processing. This will provide crucial insights for understanding how cerebellar dysfunction leads to neurological disorders.
