UNTANGLING CORTICAL CIRCUITRY WITH HOLOGRAPHIC OPTOGENETICS: HOW THE CORTEX SHAPES INPUT AND HOW INPUT SHAPES THE CORTEX

Abstract

To survey and navigate the world, animals must learn to efficiently transform and filter vast amounts of sensory information from their environment. How the brain coordinates across millions of neurons to process, or compute with, this sensory information is thus a central question in systems neuroscience. Over the past several decades, research has described how sensory stimuli might be processed across a hierarchy of brain regions, where simpler features of stimuli computed in lower order cortical areas combine to form more complex representations downstream. However, this view ignores a large piece of the puzzle: that cortical neurons receive a majority of their connections from their immediate neighbors, not from long-range projections. The role of this extensive recurrent connectivity within a given area remains unclear. This ongoing gap in our understanding is partly due to limitations in methodology that make challenging either investigations in vivo, where ongoing network activity critically influences network output, or perturbations on a biological scale. In the present thesis, we bridge this gap by developing and leveraging state-of-the-art optogenetic approaches, including two-photon imaging and holographic stimulation, to examine key building blocks of neural computation in the awake, intact cortex. We investigate how cortical neurons transform their aggregate inputs into spiking output in the awake brain, and how cortical networks change as a function of learning specific patterns of input. By employing holographic stimulation methods to deliver precise inputs to individual patterns of neurons, we begin to untangle how the cortex relies on its local recurrent connections to shape its output, and set a foundation for further studies to probe how the cortex and downstream regions interact and are reshaped to efficiently produce sensation to guide animals through the world.