Ancient and recent selective pressures shaped genetic diversity at AIM2-like nucleic acid sensors

Abstract

AIM2-like receptors (ALRs) are a family of nucleic acid sensors essential for innate immune responses against viruses and bacteria. We performed an evolutionary analysis of ALR genes (MNDA, PYHIN1, IFI16, and AIM2) by analyzing inter- and intraspecies diversity. Maximum-likelihood analyses indicated that IFI16 and AIM2 evolved adaptively in primates, with branch-specific selection at the catarrhini lineage for IFI16. Application of a population genetics-phylogenetics approach also allowed identification of positive selection events in the human lineage. Positive selection in primates targeted sites located at the DNA-binding interface in both IFI16 and AIM2. In IFI16, several sites positively selected in primates and in the human lineage were located in the PYD domain, which is involved in protein-protein interaction and is bound by a human cytomegalovirus immune evasion protein. Finally, positive selection was found to target nuclear localization signals in IFI16 and the spacer region separating the two HIN domains. Population genetic analysis in humans revealed that an IFI16 genic region has been a target of long-standing balancing selection, possibly acting on two nonsynonymous polymorphisms located in the spacer region. Data herein indicate that ALRs have been repeatedly targeted by natural selection. The balancing selection region in IFI16 carries a variant with opposite risk effect for distinct autoimmune diseases, suggesting antagonistic pleiotropy. We propose that the underlying scenario is the result of an ancestral and still ongoing host-pathogen arms race and that the maintenance of susceptibility alleles for autoimmune diseases at IFI16 represents an evolutionary trade-off.

Keywords: AIM2-like receptors; IFI16; long-standing balancing selection; positive selection.

Fig. 1.—

ALR gene domain representation and analysis of positively selected sites. (A) Domain structure of the four ALR genes. Positively selected sites in whole phylogeny (identified through both BEB and MEME) are shown in red; residues subject to positive selection in the human (green) and in the catarrhini (black: BEB sites; magenta: BEB and MEME sites) lineages are also indicated. (B) Three-dimensional (3D) structure of the AIM2 HIN:DNA complex (pdb ID 3RN2). Positively selected sites are shown in red, and yellow indicates key residues at the HIN:DNA interface. Positions refer to the human sequence. (C) Alignment of the IFI16 PYD domain. Positively selected sites in whole phylogeny are shown in red, whereas human-specific positively selected sites are marked by a green asterisks. A lysine residue conserved among PYD domains from different proteins is shown in violet. The extension of the six α-helices, inferred from structure analysis of human PYDs, is shown. (D) 3D structure of the IFI16 HIN:DNA complex (pdb ID 3RNU). The 615T positively selected site (red) is located at the DNA-binding interface formed by the 4 IFI16 monomers.

Fig. 2.—

Lineage-specific selection and DFE analysis. (A) Branch-site analysis of positive selection in IFI16. Branch lengths are scaled to the expected number of substitutions per nucleotide, and branch colors indicate the strength of selection (ω). Red, positive selection (ω > 5); blue, purifying selection (ω = 0); gray, neutral evolution (ω = 1). The proportion of each color represents the fraction of the sequence undergoing the corresponding class of selection. Thick branches indicate statistical support for evolution under episodic diversifying selection as determined by BS-REL. (B) Violin plot of selection coefficients (γ) for ALR genes (median, white dot; interquartile range, black bar). Selection coefficients are classified as strongly beneficial (100, 50), moderately beneficial (10, 5), weakly beneficial (1), neutral (0), weakly deleterious (−1), moderately deleterious (−5, −10), strongly deleterious (−50, −100), and inviable (−500).

Fig. 3.—

F_ST analysis of the ALR gene cluster. Data from the 1000 Genomes Pilot Project were used to calculate F_ST in sliding windows of 20 SNPs moving along the ALR gene cluster (NCBI/hg18, chr1:157063927–157317926) with a step of three SNPs (upper panel). Color codes refer to population comparisons: red, YRI/CEU; blue, YRI/AS; and green, CEU/AS. Horizontal dashed lines represent the 95th percentile in the distribution of F_ST calculated for sliding windows deriving from 2,000 randomly selected human genes. SNPs genotyped in the HGDP-CEPH panel are represented as gray circles (no unusual F_ST value among continental groups) or black circles (F_ST outliers); a SNP associated to rheumatoid arthritis and celiac disease is reported in red. The resequenced IFI16 region is shaded in gray. The blue and cyan boxes represent the segmental duplication of exon 7. In the bottom panel, LD analysis for the IFI16 resequenced region (5 kb) is shown. LD analysis was performed with the Haploview software using resequencing data, and blocks were identified through the implemented confidence interval algorithm (see Materials and Methods). Variants within the first LD block were used for Network and GENETREE analyses.

Fig. 4.—

Haplotype analysis of IFI16. (A) Genealogy of haplotypes in the IFI16 LD region (1.7-kb region (NCBI/hg18, chr1:157267850–157269530, see text) reconstructed through a median-joining network. Each node represents a different haplotype, with the size of the circle proportional to frequency. Nucleotide differences between haplotypes are indicated on the branches of the network. Color codes are as follows: YRI, green; CEU, blue; and AS, red. The most recent common ancestor (MRCA) is also shown. SNPs mentioned in the text are reported. (B) GENETREE for the LD subregion of IFI16. Variants are represented as black dots; the absolute frequency of each haplotype is reported.

Fig. 5.—

TMRCA estimates for the IFI16-5 kb region and for reference windows. Probability density plot of TMRCAs from 5-kb windows deriving from autosomal NIEHS genes (solid line); the upper (TMRCA+SD) estimates are also shown as hatched lines. The TMRCA estimate for the IFI16-5 kb region is indicated with upper- and lower bounds (gray shading).