CAUSAL ROLE OF INSULAR CORTEX NEURONAL ACTIVITY PATTERNS IN APPETITIVE AND AVERSIVE LEARNING

The insular cortex (InsCtx) integrates external and internal sensory information to guide diverse behaviors including feeding, drinking, anxiety, and addiction. We previously demonstrated that InsCtx neuronal population activity patterns form stereotyped manifolds across mice and physiological states like thirst, with dynamics predicting fulfillment of physiological needs. However, the causal contributions of these patterns to behavior and their distinct roles in appetitive versus aversive learning remained unclear.
Using longitudinal two-photon imaging and holography to track identical InsCtx ensembles, we found that learning reshapes manifold structure in state-dependent ways, with unique patterns emerging based on stimulus sensory modality and valence. Fiber photometry of Nucleus Accumbens dopamine signals revealed dynamic shifts in appetitive and aversive values across learning and physiological states. To establish causality, we employed two-photon holographic optogenetics targeting need- and stimulus-specific ensembles. Recreating natural activity patterns ("on-manifold" stimulation) revealed their causal role in network dynamics and behavior in a physiologically state-dependent manner, while artificial patterns ("off-manifold" stimulation) induced long-term InsCtx plasticity. Critically, preliminary findings show that "on-manifold" stimulation directly modulates behavioral satiation rate, demonstrating that InsCtx activity patterns regulate internal states and associated behaviors. These results establish InsCtx population activity patterns as fundamental computational units that bidirectionally encode and generate motivational states, offering unprecedented opportunities for future precision targeting of maladaptive feeding and addiction behaviors.