Topographic organization of feedback projections to mouse primary visual cortex
Context dependent top-down modulation in visual processing has been a topic of wide interest. Recent findings on context dependent modulation, combined with the tools available to investigate network mechanisms in the mouse, make the mouse primary visual cortex an ideal system to investigate context-dependent modulation. However, the distribution of inputs to V1 from across the brain is still relatively unknown. In this study, we investigate inputs to V1 by injecting cholera toxin B subunit (CTB), a retrograde tracer, across the extent of V1. To identify CTB labelled cell bodies and quantify their distribution across various brain regions, we developed a software pipeline that maps each labelled cell body to its corresponding brain region. We found over fourteen brain regions that provided inputs to V1. Higher visual areas (HVAs) provided the most inputs to V1, followed by the retrosplenial, cingulate, and other sensory cortices. As our injections spanned a range of coordinates along the mediolateral axis of V1, we asked if there was any topographic organisation of inputs to V1: do particular areas project preferentially to specific regions of V1. Based on the distribution of inputs from different HVAs, injection sites broadly clustered into two groups, consistent with a retinotopic separation into sites within the central visual field and the peripheral visual field. Furthermore, the number of cells detected in HVAs was correlated to the azimuthal retinotopic location of each injection site. This topographic organization of feedback projections along the medio-lateral axis of V1 suggests that V1 cells representing peripheral vs central visual fields are differentially modulated by HVAs, which may have an ethological relevance for a navigating animal.