Computational and Physiological Approaches to the Primate Anxiety Connectome
A fundamental goal of psychiatry is to guide and control emotion and anxiety when these become debilitating. However, we have not yet elucidated how the brain causally generates mood and anxiety or how these become dysfunctional; thus strategies for treatment are limited. Recent human fMRI studies reported several large-scale brain networks with coordinated activity, suggesting that coactivation in distant regions plays an essential role in interareal interactions in the human brain. Studies have also reported a large-scale network in the human brain specifically activated in the pessimistic conditions involved in anxiety disorders. However, it remains unclear whether or how those networks are causally involved in anxiety-like behaviours. Distant interareal interactions are particularly rich in the primate brain, and we have identified multiple areas in primates that are causally involved in the generation of anxiety-like states. This suggests the existence of a “primate anxiety connectome,” as a homologue of the human network. This homologue provides an unparalleled opportunity to examine the causal structure of the generation of anxiety. To do this, we will combine four experimental approaches: electrical microstimulation (EMS), behavioral analysis, computational modeling, and functional magnetic resonance imaging (fMRI). First, we will perform the EMS of the multiple limbic areas when macaque monkeys perform a task which is known to be useful for characterizing anxiety-related behavior: the fixed-ratio conflict. To quantify the subject’s internal states relevant to emotion and anxiety, we will model their behaviour using sophisticated forms of reinforcement learning (a key point of collaboration between Germany and Japan). This will provide us with the means of identifying the regions where the EMS can causally produce a change in the behaviour, specifically related to anxiety. Then, we will perform primate whole-brain fMRI under anaesthetic to examine how the EMS influences widespread activity. Our approach will provide the first evidence that the activation of the connected network in the primate brain is important to produce an anxiety state.
Professor Peter Dayan
Max-Planck-Institut für biologische Kybernetik