A major challenge in neuroscience is to understand how neural circuits in the visual cortex compute the response properties of cortical neurons, and how these contribute to visual perception. To study cortical computations, we first need to identify the circuits themselves, and to understand how they are organized with respect to the functional architecture of the visual cortex. The output pathways from the primary visual cortex (V1) to the secondary visual area (V2) are a good model to study cortical computations, because they show a highly specialized organization. It was first thought that three parallel processing pathways to V2 (for the processing of color, form and stereopsis, respectively) leave V1, and segregate into distinct V2 subregions (thick, thin and pale cytochrome-oxidase -CO- stripes). Recent studies have revised this model and proposed that only two pathways to V2 leave V1, one to the thin stripes, the other to both pale and thick stripes. With pale and thick stripes receiving a common message from V1, models of parallel processing in the visual system are being challenged. However, thick and pale stripes segregate their outputs to cortical areas having different functional specialization. This observation, and a critical evaluation of the recent anatomical data, suggests that the pale and thick stripes instead may receive segregated inputs from V1. We propose to use retrograde tracer injections targeted to functionally identified (using optical imaging - OI) specific V2 stripes, and quantitative anatomical methods, to test the hypothesis that more than just two segregated pathways to V2 leave from V1. Using a novel retrograde viral tracer (a genetically modified GFP- expressing rabies virus) to label dendritic and axonal arbors of single V1 output cells, we further propose to examine, at the single cell level, the degree of specialization of the V1 output pathways to different V2 stripes. In particular, we will test the hypothesis that V1 projections to different V2 stripes arise from distinct cell populations. Information on how V1 output pathways are organized with respect to the cortical maps of visual stimulus features in V1 and V2 is necessary to understand their computational role. The response properties of V2 neurons in different CO stripes suggest that thin strips are involved in surface processing, and thick and pale stripes in contour processing. Thick and pale stripes may be further specialized in processing different aspects of object contours. To determine what and how V1 contributes to the response properties of V2 cells, we propose to examine, at the neuronal population and single cell level, how the V1 output pathways to different V2 stripes are organized with respect to the retinotopic maps, and maps of visual stimulus orientation and spatial frequency in V1 and V2. Retrograde tracers and the rabies-GFP virus will be co-injected into specific orientation or spatial frequency domains within specific V2 stripes, identified by OI. The distribution of resulting labeled cells and boutons on the V1 feature maps will be quantitatively analyzed. These studies will provide insight into parallel information processing in the early visual system, and into the kinds of computations that are performed by the V1 output pathways to V2.
Normal vision depends on the orderly development of circuits in the visual cortex and on their intact function. Our studies of the normal circuitry between cortical areas V1 and V2, which are at the early stages of visual processing, will also provide greater insight into the causes and effects of central vision defects when these circuits are damaged by stroke or other insult.
Showing the most recent 10 out of 22 publications
