Unstable nucleus accumbens social representations in models of social behavioral dysfunction.

Abstract: The social motivation/reward theory of autism spectrum disorder (ASD) outlines that individuals with autism tend not to pursue, engage or maintain social interactions because they find social interactions less rewarding than individuals without ASD. A major hypothesis is that traditional reward circuits play an important role in social motivation and reward and may be dysfunctional in autism; yet, mechanistic proof for this intriguing hypothesis has been sparse. We have recently discovered that a significant proportion of nucleus accumbens (NAc) D1R neurons, a key node in the reward circuit, are activated during social interaction and that a subset of these neurons maintains a neural population representation of social interaction across days. We have also found that both short- and long-term NAc population representations of social interaction become destabilized in the Cntnap2 -/- mutant animals with social behavioral dysfunction. In addition, we have developed a social reward task where animals are rewarded for specific actions with social contact. VTA dopaminergic projections to NAc are activated and dopamine is released during the performance of this social reward task, again highlighting the importance of NAc for social reward. Furthermore, using Neuropixels recordings in freely behaving animals, we show reciprocal coordination of the prefrontal cortex and NAc during social interaction, suggesting that multiregional synchronization may play a key role in modulating social reward. The Golshani and Hong labs have a strong history of productive collaboration. Together, we will test the overarching hypothesis that unstable long-term NAc D1R social representations, resulting from altered inputs to NAc, drive abnormal social behaviors in models with social behavioral dysfunction. In Aim 1, using calcium imaging with wire-free miniaturized microscopes and selective labeling of D1R-MSN and D2R-MSNs in NAc, followed by decoding analysis, we will test the hypothesis that D1R-MSN neural representations of social interaction and social reward are degraded and less stable across days in the Cntnap2 and Shank3b mutant animals with social behavioral dysfunction. In Aim 2, using retroAAV-mediated labeling of specific VTA and mPFC inputs to NAc, miniaturized microscopy and multi-fiber photometry, we will test the hypothesis that the mPFC and dopaminergic VTA projections to NAc show degraded, unstable and uncoordinated social representations in the two mutant animals with abnormal social behavior. We will also probe fine-scale synchrony between mPFC and NAc using Neuropixels recordings in freely behaving animals. In Aim 3, we will use activity-dependent ChR2 labeling of NAc neurons activated by social interaction and closed-loop reactivation of these neurons to test whether stabilization of activity during social interaction can prolong and increase the probability of social interactions in both mutant animals with abnormal social behavior. These experiments will validate targets for future neuromodulation and find convergent causes for social dysfunction across different genetic models with social behavioral dysfunction.