Project Summary Opioid overdoses kill more than 130 Americans every day, yielding over 47,500 deaths per year, and the same number of deaths in the 9/11 attacks every 23 days. The best treatments for opioid use disorder (OUD) are medication-assisted therapies such as methadone, buprenorphine, or naltrexone, and are inadequate to buttress the epidemic. Failure to treat OUD is sourced in not understanding the molecular pathophysiology of OUD, including its clinical manifestations as a biological, psychological, and behavioral disease, the neurobiological mechanisms that reinforce its consumption, and how repeated consumption affects other neurocognitive processes. The Hurd lab recently discovered a genomic locus associated with OUD, downstream of the protein tyrosine-kinase FYN, in striatal neurons from human heroin abusers. Additionally, heroin increased Fyn expression in the striatum of rats that undergo heroin self-administration (SA), and in primary striatal neuronal cultures. FYN phosphorylates tau, and hyperphosphorylation of tau is known to precede neurofibrillary tangles, a hallmark of neurocognitive disease. This proposal leverages high-throughput electronic medical records (EMR) from the Mount Sinai Health System to examine how chronic opioid exposure affects the odds of neurocognitive disease (NCD) on a population level, and EMR from two other hospitals at Columbia University and the University of San Francisco to replicate these effects, using the same analysis pipeline. Additionally, I propose mechanistic studies of Fyn overexpression in the orbitofrontal cortex (OFC), a region known to be affected in NCD and OUD, on heroin SA and cognitive behaviors using animal models. Further, I used a machine learning approach with RNA-seq data obtained from human post-mortem OFC tissue to classify subjects as either belonging to the heroin overdose or control group based on the expression of ~10 genes. Three feature (gene) importance metrics highlighted a transcript, SHISA7, that was reduced in the human OFC, and this finding was replicated in the OFC of rats that underwent heroin self-administration. SHISA7 and FYN both regulate the synaptic function of AMPA receptors, and are components of the post-synaptic density of glutamatergic neurons. I will examine the impact of overexpressing SHISA7, identified using this machine learning approach, in the OFC on heroin SA and cognitive behaviors in animal models. Finally, I will conduct RNA-sequencing experiments on the OFC from Fyn- and Shisa7-manipulated animals, and in downstream bioinformatics analyses, parse key drivers of gene networks to identify potential mechanisms that account for the contribution of these genes to heroin SA and cognitive behaviors. High-throughput clinical and gene expression data from the human can inform translational, mechanistic studies using animal models that will enable a far-reaching understanding of the molecular disturbances underlying the effects of opioids on cogni...