Establishing a Ground Truth for Inter-Areal Neuronal Communication

PROJECT SUMMARY/ABSTRACT This proposal focuses on the development of a powerful neural recording approach that leverages recent technical and conceptual innovations in neuroscience— and could take them to the next level for answering how neurons perform “computations” upon their inputs. Specifically, we propose to develop techniques for performing dual in vivo whole-cell and large-scale extracellular Neuropixels recordings in the awake marmoset. We will record a target cell using the powerful whole-cell technique, providing continuous, high signal-to-noise access to the subthreshold excitatory and inhibitory inputs of the neuron (as well as its output spikes). By pairing that recording with simultaneous extracellular measurement of 100s of neurons using a Neuropixels probe, we will have access to an appropriately large number of possible input neurons to the target cell. We propose to conduct our investigation in the visual cortex, leveraging retinotopic organization and a foundational understanding of feedforward computations to perform comprehensive and physiological tests of how information is transformed. Critically, we will conduct these experiments in the marmoset, establishing these techniques in an exciting model system that allows for both technical and conceptual translation from rodent model systems in one direction, and to humans in the other. This project will establish technical feasibility and refinement that would support a larger-scale, hypothesis driven project at the R01 scale, focusing on how visual signals are transformed across the visual cortical hierarchy and how they interact with non visual contributions coming from elsewhere in the brain (e.g., eye movements, body movements, changes in behavioral state, etc). Because of the smooth nature of the cortex in both marmosets and rodents, the approach we develop in this project could then used by our labs and the larger scientific community to assess many questions across a large fraction of cortical and subcortical areas. We further propose to share the resulting data publicly, with no delays or constraints on its use. Our dataset will be the first of its kind, producing new, much-needed measurements of functional synaptic connectivity between physiologically-defined neurons (i.e., mapped spatial receptive fields and visually-driven responses). It will thus serve as a `gold-standard', `ground-truth' dataset to leverage for inferring neural connectivity from purely extracellular recordings.