Taxonomic context and genomic architecture jointly shape expression divergence across animals.

Gene expression divergence is a major source of phenotypic variation, yet the factors that shift regulatory optima remain incompletely understood. In particular, it is unclear how broad taxonomic differences and local genomic architecture interact to shape the tempo and mode of expression evolution. Here, we analyze single-copy orthologs between species pairs in Drosophila and mammals to quantify taxon-specific rates and magnitudes of expression divergence, assess the influence of genome architecture, and evaluate tissue-specific and functional patterns. Using a computational framework that predicts expression divergence and species-specific expression optima, we identify markedly higher divergence rates in Drosophila than in mammals, consistent with theoretical expectations and prior empirical work. Contrary to expectations, genes located in nested structures, in which one gene lies entirely within the intron of another, were no more likely to diverge than unnested genes in either taxon. However, when divergence occurred in Drosophila, nested genes exhibited larger shifts in expression optima, with the strongest effects among internal genes. Divergence rates were also higher among young than old nested genes in both taxa, although magnitudes of expression shifts were indistinguishable, suggesting that nesting contributes to early regulatory instability but does not typically trigger large regulatory changes. Tissue-level and functional analyses further revealed taxon- and architecture-specific signatures of expression divergence, including contrasting patterns across reproductive and neural tissues, as well as enrichment of core regulatory processes among unnested genes and enrichment of drug and xenobiotic metabolism among nested genes in Drosophila. Collectively, these findings demonstrate that taxonomic context and genome architecture shape expression evolution in distinct and measurable ways, giving rise to contrasting patterns of regulatory divergence.