IntTempComp

The cerebellum is critical for fine-tuning behaviors under specific sensory contexts. How its microcircuitry generates temporally precise and context-dependent neural activity remains an open question. We hypothesize that the specific properties of soma- and dendrite-targeting interneurons (STIs and DTIs), which inhibit the sole output of the cerebellar cortex (Purkinje cells, PCs), confer distinct but critical computational abilities within the cerebellar cortex. We hypothesize that electrical feed-forward inhibition between STIs will synchronize PCs, while sublinear synaptic integration of SCs will enhance pattern separation by PCs. To test these hypotheses, we will combine 1) dual-patch clamp and neurotransmitter uncaging in brain slices, 2) mathematical modeling, and 3) state-of-the-art 2-photon imaging and high-density electrical recordings of network activity in awake mice. Recent literature suggests that despite the inhibitory nature of PC outputs, their target cerebellar nuclear neurons are more efficiently driven by PC when their activity is synchronized. But how rapid and transient PC synchrony emerges despite heterogeneity in the source and timing of synaptic input it is not well understood. STIs are well-positioned to resolve this conundrum, but they are subject to the same heterogeneity in excitatory synaptic drive as PCs. Thus a major question remains: how does the cerebellar cortical circuit achieve precise transient synchrony from less precise synaptic inputs? Partner 1 has recently shown that STIs in the cerebellum can generate rapid feed-forward synchronization of their firing by transmitting filtered action potentials, known as spikelets, and subthreshold EPSPs. Partner 3 has previously shown in live animals that STIs powerfully regulate PCs. We, therefore, hypothesize that the precise timing of cerebellar output neurons requires two network stages of synchrony: in the first stage, the synchronized firing of cerebellar interneurons is achieved through electrical synapses, and in the second stage, the precise STI inhibition narrows the integration window, thus ensuring synchronous PC firing. Furthermore, an electrically coupled network of interneurons can generate a spatially extended, transiently synchronized inhibition that ensures the parasagittal recruitment of many PCs converging on common cerebellar nuclear neurons. Such a multi-stage synchrony mechanism would have transformative implications for the understanding of how PCs control downstream targets. Another critical function of PCs is their ability to discriminate contexts (internal and external sensory information) and transform specific contextual information into output firing patterns that adapt temporally precise behaviors. The expansion and recoding of information within the granule cell layers is thought to decorrelate input activity patterns that project to PCs to allow their discrimination. Partners 1 and 2 have shown previously that cerebellar DTIs can enhance their computational power through sublinear dendritic integration. Therefore, we propose that DTIs are ideally suited to further enhance this discrimination and transformation into an appropriate output for modifying behavior. We propose that the diverse properties of cerebellar interneurons are leveraged for the precise control of PC output to precisely adapt motor and cognitive behaviors for different sensory contexts. Since STI and DTI subnetworks are ubiquitous throughout the brain, we propose that they provide a canonical strategy for generating precisely timed neural activity essential for adapting and adjusting behaviors.