Representational Stability in the Mouse Visual System
1 Institute of Experimental Epileptology and Cognition Research (IEECR), Faculty of Medicine, University of Bonn, Bonn, Germany
Our brains are in constant flux, with neuronal circuits, cells, and even subcellular structures undergoing continuous changes. Despite this dynamic nature, our visual perception appears remarkably stable over time, raising questions about how unstable neural circuits can produce reliable cognition. Here, we use chronic two-photon Ca2+ imaging to examine the circuit mechanisms of representational stability in the mouse dorsal lateral geniculate nucleus (dLGN) and primary visual cortex (V1). Continual changes in neuronal representation of the same input – termed representational drift — have been observed to varying degrees in different brain areas, including the visual system. The factors influencing this drift, whether related to changes in behavior, brain states, or cellular and circuit mechanisms, remain largely unexplored. Based on recent theoretical and experimental work, we hypothesize that feedback from higher-order neuronal populations and synaptic plasticity within lower-order circuits contribute to stabilizing visual representations. To test this hypothesis, we use (chemo)genetic tools to disrupt feedback from V1 to dLGN and alter synaptic plasticity within the dLGN. In addition, we develop and employ machine learning models to disentangle the influences of behavior, latent states, and sensory input on neuronal responses. Combining these approaches, we aim to elucidate the interplay between stability and plasticity in the brain, shedding light on the mechanisms underlying stable perception within dynamic neural networks.