Consideration of genetic and sex effects in mice enhances consilience with human addiction studies

Concerns about translation of findings across species and external validity of rodent models are often based on results from narrow investigations of populations with limited diversity. Sources of individual variation – including genetics and sex – are not often directly encompassed in model organism studies. Explicit inclusion of individual differences in rodent research has the potential to reveal conserved phenotypes and molecular systems relevant to human addiction. As with most complex diseases, risk for cocaine use disorder is subject to considerable inter-individual variation. We surveyed cocaine-related behavioral traits in both males and females of eight inbred mouse strains whose genomes collectively capture 90% of the genetic diversity of the mouse species. Across these strains, individual differences explained a substantial proportion of variance in cocaine-responsive or cocaine response-predictive behavioral and physiological phenotypes. The inclusion of wild-derived mouse strains often extended the phenotypic ranges of these behaviors beyond the range observed in conventional domestic laboratory mouse strains. Striatum transcriptional responses to cocaine were also highly dependent upon strain and sex differences, with most cocaine-responsive genes being differentially expressed in a manner moderated by strain, sex, or their combination. We compared the strain- and sex-mediated transcriptional responses to cocaine in mice to transcriptomic analysis of people with cocaine use disorder, finding that mouse similarity to humans was highly dependent upon mouse genetic background and sex. Specifically, male WSB/EiJ mice and female NOD/ShiLtJ mice exhibited the greatest extent of neural transcriptional consilience with humans with cocaine use disorder. Model organism diversity thus represents a crucial source of biological information that can substantially improve external validity of research into addiction risk mechanisms.