Evidence for widespread positive and purifying selection across the European rabbit (Oryctolagus cuniculus) genome

Abstract

The nearly neutral theory of molecular evolution predicts that the efficacy of both positive and purifying selection is a function of the long-term effective population size (N(e)) of a species. Under this theory, the efficacy of natural selection should increase with N(e). Here, we tested this simple prediction by surveying ~1.5 to 1.8 Mb of protein coding sequence in the two subspecies of the European rabbit (Oryctolagus cuniculus algirus and O. c. cuniculus), a mammal species characterized by high levels of nucleotide diversity and N(e) estimates for each subspecies on the order of 1 × 10(6). When the segregation of slightly deleterious mutations and demographic effects were taken into account, we inferred that >60% of amino acid substitutions on the autosomes were driven to fixation by positive selection. Moreover, we inferred that a small fraction of new amino acid mutations (<4%) are effectively neutral (defined as 0 < N(e)s < 1) and that this fraction was negatively correlated with a gene's expression level. Consistent with models of recurrent adaptive evolution, we detected a negative correlation between levels of synonymous site polymorphism and the rate of protein evolution, although the correlation was weak and nonsignificant. No systematic X chromosome-autosome difference was found in the efficacy of selection. For example, the proportion of adaptive substitutions was significantly higher on the X chromosome compared with the autosomes in O. c. algirus but not in O. c. cuniculus. Our findings support widespread positive and purifying selection in rabbits and add to a growing list of examples suggesting that differences in N(e) among taxa play a substantial role in determining rates and patterns of protein evolution.

Fig. 1.

Unfolded site frequency spectrum of nonsynonymous (black bars) and synonymous (gray bars) mutations for autosomes and X chromosome in both subspecies of the European rabbit. White bars represent the expected site frequency spectrum. We used data on hares to infer the ancestral state of each polymorphism.

Fig. 2.

Proportion of amino acid substitutions driven to fixation by positive selection (α) in the autosomes (white) and X chromosome (gray) estimated using three different methods. FWW01: method of Fay et al. (2001); BEW04: method of Bierne and Eyre-Walker (2004); EWK09: method of Eyre-Walker and Keightley (2009).

Fig. 3.

(A) Synonymous site divergence (D_xy Syn) versus nonsynonymous site divergence (D_xy Nonsyn) for each gene and (B) ratio of synonymous site diversity (θ_w) to synonymous site divergence (D_xy Syn) versus nonsynonymous site divergence (D_xy Nonsyn) in Oryctolagus cuniculus algirus and O. c. cuniculus. Only genes with more than 300 synonymous sites were considered. Spearman rank correlation coefficients (R) and associated P-values are indicated.