Stronger positive and purifying selection on X-linked genes in a species with paternal genome elimination

Abstract

The faster-X effect predicts that X chromosomes may evolve more rapidly than autosomes due to selection on expressed recessive alleles. Quantifying the effect empirically usually requires numerous assumptions due to the impact of diplodiploid transmission dynamics, and so the relative contribution of dominance effects to sex chromosome evolution remains unclear. Springtails present a novel opportunity to ameliorate this. Male globular springtails have hemizygous X chromosomes, diploid autosomes, and eliminate their paternal genome during spermatogenesis. This leads to the autosomes matching the X chromosomes haplodiploid transmission. Hence, theoretically the X chromosomes and autosomes only differ in the fixation probability of male-beneficial alleles. We generated whole-genome resequencing and RNA-Seq data for the globular springtail Allacma fusca, to quantify the faster-X effect, leveraging this natural experiment. We find that after accounting for the deleterious distribution of fitness effects, and despite surprisingly depleted neutral X-linked diversity, the sex chromosomes show signals of both stronger purifying and positive selection, supporting the faster-X effect. Furthermore, we provide the first direct evidence for the loss of DNA methylation in Collembola. These results, obtained from a system where many of the confounding sex-autosome differences that typically obscure faster-X patterns are absent, support hemizygosity as a genuine driver of sex chromosome evolution and suggest that sex-linked genes could contribute disproportionately to adaptive potential.