Context-invariant socioemotional encoding by prefrontal ensembles.

Abstract

The prefrontal cortex plays a key role in social interactions, anxiety-related avoidance, and flexible context-dependent behaviors, raising the question: How do prefrontal neurons represent socioemotional information across different environments? Are contextual and socioemotional representations segregated or intermixed, and does this cause socioemotional encoding to remap or generalize across environments? To address this, we imaged neuronal activity in the ventromedial prefrontal cortex of mice engaged in social interactions or anxiety-related avoidance within different environments. Neuronal ensembles representing context and social interaction overlapped more than expected by chance while nevertheless remaining largely orthogonal. Anxiety-related representations similarly generalized across environments while remaining largely orthogonal to contextual information. This shows how the ventromedial prefrontal cortex multiplexes parallel information streams using overlapping ensembles rather than transmitting information through largely distinct populations, thereby achieving context-invariant encoding alongside the context-specific reorganization of population-level activity.