Weak Coupling Between Spontaneous Local Cortical Activity State Switches Under Anesthesia Leads to Strongly Correlated Global Cortical States

Abstract

Under anesthesia, neural dynamics deviate dramatically from those seen during wakefulness. During recovery from this perturbation, thalamocortical activity abruptly switches among a small set of metastable intermediate states. These metastable states and structured transitions among them form a scaffold that guides the brain back to the waking state. Here, we investigate the mechanisms that constrain cortical activity to discrete states and give rise to abrupt transitions among them. If state transitions were imposed onto the thalamocortical system by changes in the subcortical modulation, different cortical sites should exhibit near-synchronous state transitions. To test this hypothesis, we quantified state synchrony at different cortical sites in anesthetized rats. States were defined by compressing spectra of layer-specific local field potentials (LFPs) in visual and motor cortices. Transition synchrony, mutual information, and canonical correlations all demonstrate that most state transitions in the cortex are local and that coupling between sites is weak. Fluctuations in the LFP in the thalamic input layer 4 were particularly dissimilar from those in supra- and infra-granular layers. Thus, our results suggest that the discrete global cortical states are not imposed by the ascending modulatory pathways but emerge from the multitude of weak pairwise interactions within the cortex.