Resting-state Dynamics as a Cortical Signature of Anesthesia in Monkeys

Background: The mechanism by which anesthetics induce a loss of consciousness remains a puzzling problem. We hypothesized that a cortical signature of anesthesia could be found in an increase in similarity between the matrix of resting-state functional correlations and the anatomical connectivity matrix of the brain, resulting in an increased function-structure similarity. Methods: We acquired resting-state functional magnetic resonance images in macaque monkeys during wakefulness (n = 3) or anesthesia with propofol (n = 3), ketamine (n = 3), or sevoflurane (n = 3). We used the k-means algorithm to cluster dynamic resting-state data into independent functional brain states. For each condition, we performed a regression analysis to quantify function-structure similarity and the repertoire of functional brain states. Results: Seven functional brain states were clustered and ranked according to their similarity to structural connectivity, with higher ranks corresponding to higher function-structure similarity and lower ranks corresponding to lower correlation between brain function and brain anatomy. Anesthesia shifted the brain state composition from a low rank (rounded rank [mean SD]) in the awake condition (awake rank = 4 [3.58 +/- 1.03]) to high ranks in the different anesthetic conditions (ketamine rank = 6 [6.10 +/- 0.32]; moderate propofol rank = 6 [6.15 +/- 0.76]; deep propofol rank = 6 [6.16 +/- 0.46]; moderate sevoflurane rank = 5 [5.10 +/- 0.81]; deep sevoflurane rank = 6 [5.81 +/- 1.11]; P < 0.0001). Conclusions: Whatever the molecular mechanism, anesthesia led to a massive reconfiguration of the repertoire of functional brain states that became predominantly shaped by brain anatomy (high function-structure similarity), giving rise to a well-defined cortical signature of anesthesia-induced loss of consciousness.