RNA-seq samples from male sex organs in Drosophila

Phenotypic evolution can result from gains and losses of genes, mutations in coding sequences, or regulatory mutations affecting gene expression. While the relative importance of these mechanisms is debated, regulatory evolution is recognized as a key driver of phenotypic diversity. In this study, we applied a phylogenetic model to discretized gene expression states (active or inactive) to investigate the evolutionary turnover of organ-specific transcriptomes, which we define as instances where gene expression is activated or deactivated in a particular organ. We focus on transcriptome turnover in two male reproductive organs in 11 species of the Drosophila melanogaster species group. Using the Bayesian inference method zigzag (Thompson et al. 2020), we estimate that testes express a higher proportion of the genome (65--75% of genes, depending on the species) compared to accessory glands (46--64%), with background expression noise producing less than 1% of transcripts in both organs. We find that many conserved genes have gained or lost expression in testes and accessory glands. Our model of joint transcriptome evolution, applied to 8,660 genes conserved as single copy families in all species (singleton genes), revealed similar turnover rates but distinct evolutionary trajectories in the two organs. We estimate that the genes in our data set transition between active and inactive expression states at rates on the order of 10^-9 yr^-1. The two organs experienced accelerated transcriptome turnover on different branches of the Drosophila phylogeny. The accessory glands exhibit greater variation in turnover rates among lineages, suggesting a lower baseline rate with bursts of rapid evolution in a subset of branches on the phylogeny. We do not observe significant differences in turnover rates between X-linked and autosomal genes. Genes that encode transcription factors transition between active and inactive states slower than non-TF genes in the accessory gland, but slightly faster in the testis. The results are robust to the choice of probability cut-offs used to discretize gene expression states, and there is good agreement between the estimates of expression states from zigzag and the inferences from the phylogenetic model. Overall, our study highlights the complex dynamics of transcriptome evolution in male reproductive organs. We discuss the benefits and challenges associated with investigating the evolution of gene expression as a binary trait and suggest potentially fruitful avenues for further methodological development. Overall design: To investigate the evolution of gene expression in drosophila male sex organs, we purified mRNA from the accessory glands and testis of pooled samples of about 25 adult male flies from two biological replicates from each of two different strains in 11 species. Libraries with insufficient read (< about 30 million) were augmented with additional sequencing runs. Relative expression was quantified and expression state inferred using latent mixture models. An evolutionay model and phylogenetic tree were fit to the comparitive data to infer evolutionary rates and history.