| dc.description.abstracteng | Rodents possess an array of facial whiskers on either side of their snout, which they use to collect tactile information about their surroundings. In layer IV of the somatosensory cortex, individual whiskers are represented by dense clusters of cells called barrels. Together, the barrels form the barrel field, a somatotopic representation of the whiskers on the snout. Layer 4 barrels receive direct sensory input from the thalamus, which is then distributed to all other layers of the cortex along the so-called canonical microcircuit, resulting in widespread cortical activation. In the reeler mouse, the loss of function of the reelin protein causes abnormal development of the cortex leading to the loss of cortical lamination. In this chaotic cortex, neurons that ought to be grouped together in layers are found in ectopic positions, scattered across the cortical depth. Although solid evidence exists that barrel equivalent structures may form in reeler, the functional organization and connectivity of its somatosensory cortex remains obscure. Here, using in vivo intrinsic signal optical imaging, we demonstrate that sensory input reaches and activates the barrel cortex of the mutant mouse normally, and that somatotopy persists in spite of laminar disorganization. Furthermore, using in vitro whole cell recordings and optogenetics, we demonstrate that thalamic input to layer IV equivalent neurons is direct in reeler. Direct thalamic input is weakened, however, while network activity increases with respect to the normal cortex. These results reveal an unexpected paradox in the reeler brain, one that opposes a weakened thalamic input to a normal activation of cortical networks. We propose a model of cortical network changes that solves this paradox, whereby a weakened thalamic input can be rescued by an increase in the gain of its intracortical amplification and an adaptive shift in the excitation-inhibition balance. These results question the hitherto prevailing view that both connectivity and function are left unchanged in the reeler mouse, and may contribute to a reappraisal of its relevance as a model of adaptive circuit plasticity. | de |