Genomic architecture of the developing forelegs of <i>Drosophila prolongata</i> ; an exaggerated weapon and ornament

Elaborate secondary sexual traits are among the most striking phenotypes in nature. Functional genetic studies of these phenotypes generally focus on within-species analyses of candidate signalling pathways. These provide insight into mechanisms of trait expression, but not necessarily genetic differences contributing to evolutionary change. Drosophila prolongata offers an exceptional model to explore the evolution of trait exaggeration, as the only known species with male-specific foreleg size exaggeration in the melanogaster species group. We performed sex-specific RNA-seq in developing fore- and midleg tissues (imaginal discs) sampled at both the onset, and just after initiation of, sexually dimorphic growth in D. prolongata, D. carrolli (∼4MYA divergence) and D. melanogaster (∼35MYA). We observed a positive relationship between number, but not magnitude of differential expressed, sex-biased genes, with the extent of adult phenotypic dimorphism. Using genes identified from our differential expression analysis, we used RNAi knockdown in D. melanogaster to test their potential role in growth in the leg. Knockdown of several genes had quantitative impacts, whole one gene, grain, caused D. prolongata-like leg phenotypes. We also observed modest changes in magnitude and direction of expression differences, in signalling pathways previously implicated in sexually dimorphic evolution in other species. Our results are consistent with the evolution of sex-specific phenotypes being jointly mediated by few genes with large, and many more genes with modest impacts on sex-biased gene expression.