Mate choice in the brain: Species differ in how male traits ‘turn on’ gene expression in female brains

Mate choice plays a fundamental role in speciation, yet we know little about the molecular mechanisms that underpin this crucial decision-making process. Stickleback fish differentially adapted to limnetic and benthic habitats are reproductively isolated and females of each species use different male traits to evaluate prospective partners and reject heterospecific males. Here, we integrate behavioral data from a mate choice experiment with gene expression profiles from the brains of females actively deciding whether to mate. We find substantial gene expression variation between limnetic and benthic females, regardless of behavioral context, suggesting general divergence in constitutive gene expression patterns, corresponding to their genetic differentiation. Intriguingly, female gene co-expression modules covary with male display traits but in opposing directions for sympatric populations of the two species, suggesting male displays elicit a dynamic neurogenomic response that reflects known differences in female preferences. Furthermore, we confirm the role of numerous candidate genes previously implicated in female mate choice in other species, suggesting that evolutionary tinkering with these conserved molecular processes underlies divergent mate preferences and sexual isolation. Taken together, our study adds important new insights to our understanding of the molecular processes underlying female decision-making critical for generating sexual isolation and speciation. Benthic and limnetic threespine sticklebacks were collected from two lakes on Texada Island, British Columbia and transferred to Michigan State University. Each female subject experienced one of three trial conditions: 1) courtship by a male from their lake and species (conspecific courtship), 2) courtship by a male from their lake and the opposite species (heterospecific courtship), or 3) a control in which female subjects were with another female from her same lake and species (social control). Brains were extracted after 20 min or when spawning occured, whichever occured first. Sample size for each treatment for Paxton limnetic, Paxton benthic, and Priest limnetic populations was n = 5-6; Priest benthic fish brains were not sequenced because of low sample siazes and we expected large indiviual variation in transcriptomes.