PROPOSAL SUMMARY Deficits in social interaction are characteristic of several neuropsychiatric disorders, including Autism Spectrum Disorder (ASD). The cellular, molecular, and circuit mechanisms of social deficits in ASD are largely unknown and warrants further research. One of the most consistent etiological findings in ASD is a complete deletion of the SHANK3 gene, which encodes a postsynaptic scaffold protein in neurons. Our lab developed the first Shank3 complete knockout model by deleting exons 4-22 (Shank3∆e4-22). Shank3∆e4-22 mice show decreased social and reward-seeking behavior and blunted response of the Nucleus Accumbens (NAc) to social cues. The NAc is a well-established regulator of social behavior. Yet, the characteristics of neurons in the NAc that are active during social behavior, or social ensembles, are not well defined. ~95% of neurons in the NAc are medium-spiny neurons (MSNs) that express either dopamine receptor D1 (D1+), which encode reward reinforcement, or D2 (D2+), which encode aversive responses. NAc MSNs express high levels of SHANK3, and Shank3 deletion induces profound changes in D2+ MSN function. Taken together, this indicates a potential mechanism of action for social behavior deficits in Shank3∆e4-22 mice. SHANK3 scaffolds HOMER1b/c and metabotropic glutamate receptor 5 (mGluR5) to the postsynaptic density (PSD). HOMER1b/c and mGluR5 function in the NAc is crucial for regulating social and reward-seeking behaviors in WT mice. To study the role of SHANK3-HOMER1b/c interaction, our lab generated the first SHANK3-HOMER1b/c mutant mouse, Shank3PL. Since joining the Jiang lab, I have collected pilot data showing that Shank3PL mice have significantly decreased social and reward-seeking behavior. Pilot data also indicate Shank3PL mice have decreased HOMER1b/c expression in the PSD of the NAc. The specific objective of this proposal is to delineate how SHANK3 deficiency causes cellular and molecular malfunctions that underlie abnormal circuit and social behavior using our two novel mouse models: Shank3∆e4-22 and Shank3PL. First, I hypothesize that NAc social ensembles in Shank3∆e4-22 mice are primarily composed of D2+ MSNs and encode aversion and negative sociability; in contrast, I predict WT social ensembles are predominantly D1+ MSNs and encode reinforcement and positive social behavior. Second, I hypothesize that SHANK3-HOMER1b/c scaffolds are crucial for social and reward behavior. We will test these hypotheses using comprehensive methodologies. This study will be the first to characterize social ensembles in a well-validated genetic ASD model and the first to investigate the role of SHANK3-HOMER1b/c scaffolds on behaviors and NAc activity. Importantly, these studies may lead to the identification of novel therapeutic targets for treating social and reward deficits in ASD and other neuropsychiatric disorders.