Analyses of nuclearly encoded mitochondrial genes suggest gene duplication as a mechanism for resolving intralocus sexually antagonistic conflict in Drosophila

Abstract

Gene duplication is probably the most important mechanism for generating new gene functions. However, gene duplication has been overlooked as a potentially effective way to resolve genetic conflicts. Here, we analyze the entire set of Drosophila melanogaster nuclearly encoded mitochondrial duplicate genes and show that both RNA- and DNA-mediated mitochondrial gene duplications exhibit an unexpectedly high rate of relocation (change in location between parental and duplicated gene) as well as an extreme tendency to avoid the X chromosome. These trends are likely related to our observation that relocated genes tend to have testis-specific expression. We also infer that these trends hold across the entire Drosophila genus. Importantly, analyses of gene ontology and functional interaction networks show that there is an overrepresentation of energy production-related functions in these mitochondrial duplicates. We discuss different hypotheses to explain our results and conclude that our findings substantiate the hypothesis that gene duplication for male germline function is likely a mechanism to resolve intralocus sexually antagonistic conflicts that we propose are common in testis. In the case of nuclearly encoded mitochondrial duplicates, our hypothesis is that past sexually antagonistic conflict related to mitochondrial energy function in Drosophila was resolved by gene duplication.

FIG. 1.—

Proportion of duplication events inferred to happen in a particular branch (nuclearly encoded mitochondrial genes/nuclear genes). Scale approximately reflects the evolutionary time measured in million years, based on Tamura et al. (2004).

FIG. 2.—

Averaged level of expression (±standard error of the mean) of parental (black bars) and duplicated (gray bars) genes in testis (T), ovary (O), and whole D. melanogaster body (W). The data were also divided in families with (right) and without (left) genes with testis-specific expression. Dotted gray bars indicate whole genome average level of expression in different tissues. Asterisk: P < 0.0001.

FIG. 3.—

Average expression level (±standard error of the mean) of duplicates in the mitotic (Mit), meiotic (Mei), and postmeiotic (Post Mei) cells of the D. melanogaster testis. The data set was divided in families with (squares) and without (diamonds) testis-specific genes. Black, duplicated genes; Gray, parental genes.

FIG. 4.—

Average evolutionary rates (±standard error of the mean) for duplicate (D) and parental (P) lineages. In (A), we show the results compiled for all families for which we could establish the parental lineage. In (B), the same data set is divided in families with and without testis-specific genes to show that only duplicated lineages that evolved testis-specific expression have higher evolutionary rates than their respective parental lineages.