Publication: Synaptic and Circuit Specializations for Temporal Processing in the Cerebellar Cortex
Open/View Files
Date
Authors
Published Version
Published Version
Journal Title
Journal ISSN
Volume Title
Publisher
Citation
Abstract
Many behaviorally salient computations require the ability to represent and manipulate information in time. To do so, neurons in many brain areas evolved to generate time-varying activities via either cell-intrinsic or network mechanisms. The cerebellum supports temporal processing in many sensorimotor and cognitive behaviors across a variety of time scales, but how cerebellar circuits generate time-varying activities is not well understood. It is generally accepted that the same underlying circuit is repeated throughout the cerebellar cortex and that different behaviors are represented in different regions of the cerebellar cortex. Here we provide anatomical, physiological, and behavior results supporting regionally specialized synaptic and circuit mechanisms that could generate time-varying activities. Unipolar brush cells (UBCs) are excitatory interneurons in the input layer that are enriched in select lobules. By examining their synaptic responses to mossy fiber inputs, we found that the UBC population provided a rich temporal basis set comprised of shifted Log-Gaussian response curves spanning two orders of magnitude in time. Interestingly, the synaptic responses are mediated by excitatory and inhibitory metabotropic glutamate receptors, which showed diverse and continuously varying kinetics across cells. We also found that in specialized areas of the cerebellar cortex, Purkinje cells (PCs) have extensive collaterals that directly inhibit granule cells (GrCs) and a subset of UBCs within the input layer. PC provides both GABAA receptor-mediated phasic inhibition and long-lasting inhibition to both cell types. Long-lasting inhibition is mediated primarily by high-affinity δ subunit-containing GABAA receptor in GrCs, and by metabotropic GABAB receptors in UBCs. These results suggest that multiple interacting neural mechanisms may work in synergy to support temporal processing in the cerebellum. Furthermore, the circuit is regionally specialized to accommodate divergent computational demands across multiple behaviors.