Synaptic Shaping of Cortical Dynamics at Species Specific Scales

Abstract

Events in the world are structured both spatially and temporally: while brains vary greatly in size across species, they must coordinate neuronal populations in response to shared environmental timescales. How are spatial and temporal structure scalably encoded in cortical dynamics? Individually, synapses show preferential responses to particular temporal patterns of activity enabled by transient changes in connection strength known as short-term plasticity. At a group level, one idea is that sensory stimulation evokes cortical travelling waves which share information between spatially distant neurons and act as a record of the recent past. However, individual synaptic dynamics display a large degree of heterogeneity that is informally separated into classes despite the precise number and properties of such classes being unclear. Moreover, species with larger brains have more neurons distributed over larger spatial volumes: for cortical waves to integrate information from topologically organized receptive fields in different species, they must propagate via mechanisms that scale proportionately. To make sense of this complexity, this thesis leverages modelling spanning from the scale of synapses to networks, as well as machine learning tools, a large open synaptic physiology dataset, and novel experiments performed by collaborating scientists. We present a picture of the cortical microcircuit where synaptic dynamics belong to a nuanced dictionary of functional subtypes and where spatially structured synaptic connectivity encodes local perturbations into cortical travelling waves. In rodent data, we find five functional clusters of synaptic dynamics that partially converge with transgenically defined subtypes. Strikingly, the application of the same clustering method in human data infers a highly similar number of subtypes, supportive of stable clustering. At the network level, our modelling elucidates how cortical synaptic connectivity may selectively encode localized perturbations into transient travelling waves, a prediction consistent with new experimental results from electrical microstimulation of rodent cortical tissue. Crucially, our modelling finds these rodent data consistent with two rival connectivity schemes: one in which perturbations of both human and mouse tissue produce waves of similar scale and a second in which evoked waves propagate more than twice as far in models of human cortex, consistent with a role in species specific scaling of spatiotemporal integration. Taken together, this thesis provides a cross-species account of how the structure of cortical connectivity shapes dynamics across networks including a nuanced dictionary of specific functional subtypes of cortical connections and a falsifiable mechanistic prediction about how neuronal populations of species specific scale may be coordinated in response to shared environmental scales.