Interpreting faster-X evolution in light of expression breadth and adaptation. Richard Meisel. Cornell University.

   Accelerated rates of molecular evolution can be driven by increased mutation rate, relaxed selective constraints, or higher rate of adaptive substitutions. Comparative and population genomics studies often measure the total divergence between genes (k), and the fraction of amino acid substitutions fixed by positive selection (). This has revealed that narrowly expressed genes are more divergent than genes broadly expressed across many tissues (k_narrow > k_broad), and X-linked genes tend to be more divergent than autosomal genes (k_X > k_A). The faster evolution of narrowly expressed genes can be attributed to relaxed constraints permitting more neutral fixations and/or more adaptive substitutions. The faster evolution of X-linked genes (the faster-X effect) has been hypothesized to be a result of the exposure to selection of X-linked recessive beneficial mutations in hemizygous males, which leads to more adaptive substitutions on the X chromosome. Comparisons of polymorphism and divergence between X-linked and autosomal genes (the McDonald-Kreitman test and its derivatives) have indeed revealed a higher frequency of substitutions fixed by positive selection in X-linked genes relative to autosomal genes (_X > _A). To simultaneously address the effects of X-linkage and expression breadth on divergence and adaptation, we tested for a relationship between the faster-X effect and gene expression profiles in Drosophila melanogaster within a McDonald-Kreitman framework. We find that, while faster-X divergence (k_X > k_A) is only observed amongst narrowly expressed genes, faster-X adaptation (_X > _A) is limited to broadly expressed genes. The faster-X adaptation in broadly expressed does not translate to faster-X divergence because the total number of substitutions in these genes is small. This low substitution rate can be attributed to increased constraints on broadly expressed genes, which has been shown to increase by decreasing the rate of neutral divergence. We therefore conclude that faster-X divergence is driven by relaxed selective constraints, and the specific type of faster-X evolution (divergence versus adaptation) depends on the constraints on the gene set under consideration.