Comparing decoupling across development to decoupling between the sexes
Beyond metamorphosis, the logic of the ADH applies to any situation in which pervasive antagonistic selection favors the evolution of reduced genetic correlations. Although theories explaining sexual dimorphism are conceptually similar to the ADH, the evolutionary trajectories of sexually antagonistic selection may differ predictably from those under ontogenetically antagonistic selection. However, few studies have directly compared patterns of sex-biased and stage-biased gene expression. In one notable exception, Perry et al. (2014) evaluated how patterns of sex-biased expression in gonadal tissue change between three developmental stages of Drosophila melanogaster . In contrast to our finding that sex-biased expression was modest compared to stage-biased expression (Figure 2), they reported greater decoupling between the sexes than between larvae and prepupae. This apparent discrepancy could be explained by minimal antagonistic selection between dispersing Drosophila larvae and prepupae, the lack of a distinct molt between these observed stages, or tissue specificity. By contrast—and consistent with our findings—a comprehensive analysis of gene expression in different developmental stages, sexes, and castes of two ant species revealed that developmental stage had the largest impact on gene-expression profiles (Ometto et al., 2011). Given that the intensity of sexually antagonistic selection is likely to be highly variable across taxa, traits, and tissues (Connallon & Clark, 2013; Connallon, Débarre, & Li, 2018; Pennell & Morrow, 2013), studies in diverse taxa will be needed to evaluate if decoupling of stage-biased expression consistently exceeds sex-biased expression.
A final layer of complexity is that if the genetic targets and/or intensity of sexually antagonistic selection vary across an organism’s life cycle, patterns of sex-biased gene expression may also be decoupled over ontogeny. In support of this prediction, Perry et al. (2014) reported that 18-30% of sex-biased genes exhibited stage-specific sex-biased expression, with 4.5% of genes having opposing sex bias at different stages. Importantly, sex- and stage-specificity impacted evolutionary rates with male-biased genes evolving most rapidly when expressed throughout development while female-biased genes evolved most rapidly when expressed in larvae only. Overall, these results are consistent with the view that sexually antagonistic and stage-specific selection can lead to predictable differences in patterns of gene expression and molecular evolution.