Stability and plasticity of visual representations in the mouse cortex

In the mammalian brain, sensory representations emerge from the activity of interconnected neurons in large cortical networks. Learning novel representations requires modifications in the
functional connections between neurons across these networks. How these modifications are coordinated between cells for updating a representation while also preserving coexisting ones is largely unknown. I will show that optogenetically pairing the activation of just a single neuron to a surrounding population reorganizes the orientation selectivity of the network in the visual cortex of awake, adult mice. Within minutes, a sub-population of weak visually-responsive cells changed their preferred orientation to 90 degrees with respect to that of the single neuron. In contrast, strong visually-responsive neurons did not change their orientation and preserved the overall network’s ability to encode orientations. These results demonstrate that functional connectivity changes coordinated by a single neuron can rapidly reorganize network-level representations with precise feature specificity. The findings also reveal a dependence of the reorganization on sub-populations defined by their level of visual responsiveness, enabling concurrent stability and flexibility in the encoding of multiple sensory representations.