ELSC Seminar Series

In striking contrast to the rest of the brain, evolutionary expansion of the cerebellum has kept pace with that of neocortex, and these two structures contain ~99% of all human neurons. Neocortex and cerebellum are also reciprocally connected by some of the brain’s densest long-range projections that are universal across mammals. This strongly hints at a general and conserved algorithm at the heart of cortico-cerebellar circuitry. Our work aims to understand these basic computations by observing and manipulating cortico-cerebellar circuit interactions while animals acquire novel skills, via multisite two-photon calcium imaging, optogenetics, and viral-genetic synaptic connectivity mapping. Our two-photon calcium imaging of cerebellar granule cells unexpectedly revealed reward-related signals reflecting prominent cognitive input, likely from neocortex. Simultaneous two photon calcium imaging of neocortical output and downstream cerebellar granule cells over weeks of learning showed that these structures gradually converge onto common task encoding strategies. Most recently, we published chronic two photon mescoscope recordings of climbing fiber activity originating in the inferior olive. Learning drove the emergence of sharp olivary synchronization transitions, which were causally dependent on active cerebellar regulation. Kuramoto models suggest that olivary state transitions likely reflect basic organizing principles of coupled phase oscillator networks, which could be used to flexibly coordinate large-scale ensemble dynamics across the cerebellum to guide learning and behavior. In ongoing work, we are integrating studies of all of these circuit components to more completely understand the evolution of cortico-cerebellar transmission during novel skill acquisition.