Excess of Deleterious Mutations around HLA Genes Reveals Evolutionary Cost of Balancing Selection

Abstract

Deleterious mutations are expected to evolve under negative selection and are usually purged from the population. However, deleterious alleles segregate in the human population and some disease-associated variants are maintained at considerable frequencies. Here, we test the hypothesis that balancing selection may counteract purifying selection in neighboring regions and thus maintain deleterious variants at higher frequency than expected from their detrimental fitness effect. We first show in realistic simulations that balancing selection reduces the density of polymorphic sites surrounding a locus under balancing selection, but at the same time markedly increases the population frequency of the remaining variants, including even substantially deleterious alleles. To test the predictions of our simulations empirically, we then use whole-exome sequencing data from 6,500 human individuals and focus on the most established example for balancing selection in the human genome, the major histocompatibility complex (MHC). Our analysis shows an elevated frequency of putatively deleterious coding variants in nonhuman leukocyte antigen (non-HLA) genes localized in the MHC region. The mean frequency of these variants declined with physical distance from the classical HLA genes, indicating dependency on genetic linkage. These results reveal an indirect cost of the genetic diversity maintained by balancing selection, which has hitherto been perceived as mostly advantageous, and have implications both for the evolution of recombination and also for the epidemiology of various MHC-associated diseases.

Keywords: MHC/HLA.; balancing selection; deleterious variation; exome; mutation load; simulations.

Fig. 1.

Number of significant GWAS associations along the genome. The chromosomal location of significant trait associations from GWAS (N = 18,682) are shown for all autosomes. Data from NHGRI GWAS catalog.

Fig. 2.

Simulated polymorphism in regions surrounding the HLA gene under different selection scenarios. Polymorphic sites (SNPs) along the regions around an HLA gene are derived from simulations with three different selection scenarios on the HLA gene (white: no selection on HLA, blue: balancing selection, gray: recurrent sweeps of positive selection). Standard box plots show (a) the median number of all SNPs, (b) of only SNPs with derived allele frequency < 0.01, and (c) the genetic diversity (π) across all sites. Variants in surrounding regions evolved neutrally (s_nb= 0) or under codominant purifying selection with s_nb= −0.0001, s_nb= −0.001, or s_nb= −0.01, respectively. Nonoverlapping notches between box plots indicate significant difference. Note the different y axis scales.

Fig. 3.

Simulated site frequency spectrum of variants surrounding the HLA gene under different selection scenarios. Derived allele frequencies along the regions around an HLA gene are derived from simulations with three different selection scenarios on the HLA gene (white: no selection on HLA, blue: balancing selection, gray: recurrent sweeps of positive selection). Variants in neighboring regions evolved (a) neutrally (s_nb= 0) or under codominant purifying selection with (b) s_nb= −0.0001, (c) s_nb  = −0.001, or (d) s_nb= −0.01, respectively. Note the different y axis scales in the zoomed insets of panels (c) and (d) for better visualization.

Fig. 4.

Observed average number of polymorphic sites per gene. The median number of polymorphic sites per gene is averaged over the 113 genes represented in the MHC region (blue arrow, excluding classical HLA genes). Also shown is the distribution of equivalent values for 1,000 Monte Carlo-sampled sets of 113 random genes from the entire exome. Represented are (a) all SNPs and (b) only SNPs with derived allele frequency < 0.01.

Fig. 5.

Observed distribution of derived allele frequencies. Derived alleles are defined as the nonreference allele at polymorphic sites in the Exome Sequencing Project data. (a) The median frequency of derived alleles per gene is averaged over the 113 genes represented in the MHC region (blue arrow, excluding classical HLA genes). The yellow density curve shows the distribution of equivalent values for 1,000 Monte Carlo-sampled sets of 113 random genes from the entire exome. (b) Derived allele frequency at polymorphic sites (N = 4,175) in the MHC region is shown in relation to distance to the nearest classical HLA locus. Solid and dashed lines indicate linear fit and 95% confidence intervals, respectively. Color shading indicates local data point density for improved visualization of the extent of overlapping points in the plot, ranging from blue (low density) via green and yellow to red (high density).

Fig. 6.

Observed site frequency spectrum (SFS) of deleterious variants. The SFS of (a) probably damaging and (b) loss-of-function variants are shown for the entire exome (white bars) and the MHC region only (blue bars, excluding the classical HLA loci).