Abstract
Cocaine use presents a worldwide public health problem with high socioeconomic cost. No current pharmacologic treatments are available for cocaine use disorder (CUD) or cocaine toxicity. To explore pharmaceutical treatments for tthis disorder and its sequelae we analyzed gene expression data from post-mortem brain tissue of individuals with CUD who died from cocaine-related causes with matched cocaine-free controls (n = 71, Mage = 39.9, 100% male, 49% with CUD, 3 samples/brain regions). To match molecular signatures from brain pathology with potential therapeutics, we leveraged the L1000 database honing in on neuronal mRNA profiles of 825 repurposable compounds (e.g., FDA approved). We identified 16 compounds that were negatively associated with CUD gene expression patterns across all brain regions (padj < 0.05), all of which outperformed current targets undergoing clinical trials for CUD (all padj > 0.05). An additional 43 compounds were positively associated with CUD expression. We performed an in silico follow-up potential therapeutics using independent transcriptome-wide in vitro (neuronal cocaine exposure; n = 18) and in vivo (mouse cocaine self-administration; n = 12-15) datasets to prioritize candidates for experimental validation. Among these medications, ibrutinib was consistently linked with the molecular profiles of both neuronal cocaine exposure and mouse cocaine self-administration. We assessed the therapeutic efficacy of ibrutinib using the Drosophila melanogaster model. Ibrutinib reduced cocaine-induced startle response and cocaine-induced seizures (n = 61-142 per group; sex: 51% female), despite increasing cocaine consumption. Our results suggest that ibrutinib could be used for the treatment of cocaine use disorder.