Preferential facilitation of V1 excitatory neuron responses by V2 feedback in mouse

PAFUNDO, DE; KUHLMAN, SK

Ashburn, VA

Conferencia; Signal Transforms in the Early Visual System; 2014

Howard Hughes Medical Institute

Feedback connections are a prominent feature of all sensory cortices. In visual cortex, top-down signals from association areas (V2) aid in context-dependent perception of visual scenes. Feedback from V2 to primary cortex (V1) targets both excitatory (EX) and parvalbumin positive (PV) inhibitory neurons within V1. Local interactions between these two cell types are thought to be a key component of cortical network computations, including gain control and gamma oscillations. It is unknown whether during visual stimulation, in vivo, V2 equally modulates the two cell types or whether there is preferential feedback to a specific cell type, which could thus alter excitation-inhibition balance and V1 computations. To address this, we used 2-photon guided electrophysiological recording of PV and EX response properties in vivo, in V1, while reversibly silencing V2 by optogentically releasing GABA from PV neurons in a localized area (600µm-diameter cortical column). V1 recording and V2 silencing sites were retinotopically matched, using intrinsic signal optical imaging. We found that the evoked firing rate of EX neurons at their preferred stimulus orientation was reduced 17± 7% by V2 silencing (n=13, p=0.012) in urethane-anesthetized mice. In contrast, PV neurons were unaffected (n=10). As expected, layer 4 EX neurons, which lack strong direct V2 input, were unaffected by V2 silencing (n=9). These results indicate that feedback from V2 preferentially modulates the activity of EX neurons in layer 2/3, and that the net impact of V2 feedback is to facilitate responses of V1 EX neurons at their preferred orientation. Interestingly, in primates arousal can invert the impact of V2 to net suppressive. Therefore, we will test the hypothesis that the impact of V2 feedback is dependent on cholinergic signaling, and that cholinergic modulation is required for a phenomenon described by Chen (Neuron 82:682 2014) during contour integration, V1 neurons with RFs on the background pattern are suppressed while V1 neurons with RFs along the contour are facilitated. Dr. Pafundo?s published work includes (1) in primates, defining the relationship between development of inhibitory synapses and production of gamma ex-vivo, and (2) quantifying cholinergic modulation of functional connectivity in prefrontal cortex circuits. Dr. Pafundo is now applying this expertise to understand how cholinergic activation can modulate the effects of V2 feedback to mediate scene integration.