UC San Diego

A distinguishing feature of the mammalian cerebral cortex is its laminar architecture, each layer containing a unique composition of neuronal types with distinct morphologies, molecular markers, and electrophysiological properties. These neurons form precise, specific synaptic connections with one another to form complex microcircuits that underlie sensory information processing. By compartmentalizing computation into layers, the cortex can efficiently channel and transform information to represent and interact with the external world. Therefore, deciphering the precise input and output connectivity structure of different neuronal types in the context of their respective layers is necessary to fully appreciate their unique functional roles in the representation and manipulation of sensory information. This dissertation builds on the traditional idea of a canonical interlaminar circuit by characterizing fundamental intracortical connections between excitatory and inhibitory cell types. Chapter 1 explores the relative functional input distributions from 5 layer-specific excitatory subpopulations to 4 cell types in mouse primary visual cortex (V1). By optogenetically activating these excitatory subpopulations and recording from targeted excitatory and inhibitory subtypes across cortical layers 2/3-6, I elucidate a complex interlaminar network that provides a novel framework for visual information processing. In Chapter 2, I approach the interlaminar connectivity of mouse V1 from a transcriptomic perspective using our newly developed method Single Transcriptome Assisted Rabies Tracing (START). By combining rabies tracing using glycoprotein (G)-deleted rabies virus (RVdG) with snRNAseq, we identify, and transcriptomic ally characterize cells projecting to the same layer-specific subpopulations as in Chapter 1. We find that START generates results consistent with established circuit models validating the utility of START as a circuit tracing tool. More importantly, with the increased cell type granularity achieved with transcriptomic characterization of inputs, we were able to uncover specific subtypes of somatostatin and parvalbumin interneurons that provide input to excitatory cells across layers. Taken together, findings from Chapters 1 and 2 demonstrate layer and cell type specificity in cortical circuit structure, indicating that a cell’s laminar position and synaptic connectivity are deeply intertwined with its functional role. Understanding cell type diversity in the context of circuit architecture forms the foundation of a novel framework for cortical information processing.