Positive selection in the chromosome 16 VKORC1 genomic region has contributed to the variability of anticoagulant response in humans

Abstract

VKORC1 (vitamin K epoxide reductase complex subunit 1, 16p11.2) is the main genetic determinant of human response to oral anticoagulants of antivitamin K type (AVK). This gene was recently suggested to be a putative target of positive selection in East Asian populations. In this study, we genotyped the HGDP-CEPH Panel for six VKORC1 SNPs and downloaded chromosome 16 genotypes from the HGDP-CEPH database in order to characterize the geographic distribution of footprints of positive selection within and around this locus. A unique VKORC1 haplotype carrying the promoter mutation associated with AVK sensitivity showed especially high frequencies in all the 17 HGDP-CEPH East Asian population samples. VKORC1 and 24 neighboring genes were found to lie in a 505 kb region of strong linkage disequilibrium in these populations. Patterns of allele frequency differentiation and haplotype structure suggest that this genomic region has been submitted to a near complete selective sweep in all East Asian populations and only in this geographic area. The most extreme scores of the different selection tests are found within a smaller 45 kb region that contains VKORC1 and three other genes (BCKDK, MYST1 (KAT8), and PRSS8) with different functions. Because of the strong linkage disequilibrium, it is not possible to determine if VKORC1 or one of the three other genes is the target of this strong positive selection that could explain present-day differences among human populations in AVK dose requirement. Our results show that the extended region surrounding a presumable single target of positive selection should be analyzed for genetic variation in a wide range of genetically diverse populations in order to account for other neighboring and confounding selective events and the hitchhiking effect.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Results of the VKORC1 haplotype study.

(A) Position of the seven SNPs along the VKORC1 gene.VKORC1 is a 4.1 kb gene (GenBank accession number AY587020) located at 16p11.2. The three exons of the gene are represented as boxes, with 5′UTR and 3′UTR regions colored in grey and coding regions in black. Flanking and intronic regions are represented as thin and thick lines, respectively. The seven studied SNPs are shown in their sequential order along the VKORC1 gene. The functional polymorphism rs9923231 located in the promoter, is highlighted in red and the SNP already present in the Illumina 650K chip in blue. Physical position along chromosome 16 is indicated in kb below. (B) Distribution of VKORC1 haplotypes at the global and regional level. For each haplotype, SNPs are listed in the same sequential order than in Figure 1A. Ancestral and derived alleles are shown in blue and orange, respectively. Haplotype labels H1 to H7 were given according to the global haplotype frequency. AF, sub-Saharan Africa; ME, Middle East; EUR, Europe; CSA, Central South Asia; EA, East Asia; OCE, Oceania; AM, America. (C) Median-joining network of the inferred VKORC1 haplotypes at the global level. Circles areas are proportional to the global haplotype frequency and branch lengths to the number of mutations separating haplotypes. Labels of haplotypes are indicated in corresponding circles, and labels of mutations on the network branches. The haplotype carrying the -1639A allele conferring the AVK sensitivity phenotype (H1) is shown in red and the ancestral haplotype (H6) in black.

Figure 2. Atypical patterns of genetic differentiation observed for VKORC1 SNPs.

Genome-wide empirical distributions of FST values were constructed from 644,143 SNPs having a MAF ≥0.001 at the global level. Individual values of FST calculated for each of the seven VKORC1 SNPs are plotted against their global MAF. The functional rs9923231 SNP is shown in red. The 50th, 95th and 99th percentiles are indicated as dotted, dashed and full red lines, respectively.

Figure 3. Distribution of –log10 (p-values) for four selection tests across a 2 Mb region centered on VKORC1.

A black vertical line indicates the physical position of VKORC1 on chromosome 16. Horizontal red dotted and dashed lines show 0.05 and 0.01 chromosome-wide significance levels, respectively. The selection tests (inter-regional FST, XP-CLR, XP-EHH and iHS, respectively) were separately applied in each of the seven geographic regions.

Figure 4. Detailed analysis of a 1.1 Mb genomic region surrounding the VKORC1 gene locus in East Asia.

The boundaries of the region displayed (chr16:30,271,572-31,391,123; UCSC human genome build hg18) were chosen so as to include the three clusters of significant scores detected in East Asia by the selection tests in the 2 Mb region centered on VKORC1 (Figure 3). (A) Name and location of genes. Exons are displayed as blue boxes and the transcribed strand is indicated with an arrow. Genes located in the block of strong LD encompassing VKORC1 and including the SNPs in the red box shown in Figure 4C, are highlighted in the grey area. (B) XP-EHH results in East Asia. The significance of the XP-EHH scores (−log10 empirical p-value) are shown for individual SNPs with a MAF ≥0.01 in East Asia. Horizontal dashed lines indicate 0.05 and 0.01 chromosome-wide significance levels. Recombination hotspots detected in HapMap Phase II data are indicated by red vertical dotted lines. The data and methods used to derive these hotspots are available from the HapMap website (http://www.hapmap.org/) , . (C) LD plot. Pairwise LD values, depicted as D’, are shown for SNPs with a MAF ≥0.01 in East Asia. D’ values are displayed in different colors from yellow to red for D’ = 0 to D’ = 1, respectively. The red box highlights SNPs included in the LD block encompassing VKORC1. The plot was produced using the snp.plotter R package .