A cerebellar synaptic mechanism for temporally precise behaviors

PROJECT SUMMARY/ABSTRACT An important feature of the brain is its ability to combine information about the external environment with internal state in order to generate precisely timed actions. These processes occur on varying timescales and involve different parts of the brain, but the cerebellum is central to shaping sub-second behaviors. Damage to the cerebellum can affect the timing of a conditioned eyeblink and can cause diminished movement endpoint accuracy and oscillatory movements around a target. While considerable work has focused on how cerebellar output can fine-tune movements, it is not known how the cerebellum generates precise temporal representations. This proposal leverages the well-defined structure of the cerebellum and novel optical tools to examine how the synaptic properties of the cerebellar circuit might underlie precisely timed behavioral outputs. Cerebellar granule cells integrate mossy fiber inputs that carry information from different cortical and subcortical streams and that have diverse short-term dynamics. To investigate whether this synaptic diversity generates sub-second temporal representations, genetically encoded calcium, glutamate, and voltage indicators will be used to measure mossy fiber and granule cell activity while mice are performing a skilled reaching behavior that requires temporal precision. Mossy fiber and granule cell activity will be compared to mossy fiber glutamate release to test the hypothesis that mossy fiber activity is transformed at the synaptic level to generate a set of sparse, reliable, and distinct granule cell activity patterns, i.e. a temporal basis set. To test the hypothesis that short-term plasticity dynamics at mossy fiber-granule cell synapses underlie temporal patterns of granule cell activity, a light-activated G-protein-coupled receptor will be used to lower mossy fiber release probabilities. The synaptic consequence of this will be measured using ex vivo whole-cell recordings of granule cells. Further, the effect of altering mossy fiber release probabilities on in vivo granule cell temporal dynamics and on reach endpoint accuracy will be assessed. These findings will elucidate how synaptic diversity can subserve temporal computations in the cerebellum and may provide a more generalizable principle for how synaptic diversity can drive patterns of activity.