Functional EPAS1/ HIF2A missense variant is associated with hematocrit in Andean highlanders

Abstract

Hypoxia-inducible factor pathway genes are linked to adaptation in both human and nonhuman highland species. EPAS1, a notable target of hypoxia adaptation, is associated with relatively lower hemoglobin concentration in Tibetans. We provide evidence for an association between an adaptive EPAS1 variant (rs570553380) and the same phenotype of relatively low hematocrit in Andean highlanders. This Andean-specific missense variant is present at a modest frequency in Andeans and absent in other human populations and vertebrate species except the coelacanth. CRISPR-base-edited human cells with this variant exhibit shifts in hypoxia-regulated gene expression, while metabolomic analyses reveal both genotype and phenotype associations and validation in a lowland population. Although this genocopy of relatively lower hematocrit in Andean highlanders parallels well-replicated findings in Tibetans, it likely involves distinct pathway responses based on a protein-coding versus noncoding variants, respectively. These findings illuminate how unique variants at EPAS1 contribute to the same phenotype in Tibetans and a subset of Andean highlanders despite distinct evolutionary trajectories.

Fig. 1.. EPAS1 variant under positive selection in Quechua Andean highlanders.

(A) iSAFE results from the 1-Mb window surrounding rs570553380. Regional plots of iSAFE scores (34) generated from the EPAS1 locus in the Andean whole genomes and visualized using Locuszoom. The left y axis shows the iSAFE score ranging from 0 to 1. Markers with iSAFE scores equal to or greater than 0.10 are considered favored by natural selection. The right y axis denotes the recombination rate in hundred crossovers per million base pairs (cM/Mb). The gray dotted line indicates the position of EPAS1 rs570553380 variant (unscored), and the purple diamond denotes the marker in the region with the highest iSAFE score (rs7559484). SNVs in the region are denoted in circles, color-coded based on linkage disequilibrium (LD) with the top iSAFE marker. (B) Haplotype-frequency plot of haplotypes in 1000 Genomes PEL subpopulation (n = 170 haplotypes) and the Andean whole genomes (WGS) (n = 80 haplotypes) containing rs570553380 and top iSAFE markers (rs7559484 and rs7556828). The black arrow indicates the haplotypes carrying the putatively adaptive “G” allele of rs570553380, in which the “C” and “A” alleles of respective iSAFE markers rs7559484 and rs7556828 are nearly always present (D′ > 0.99). (C) EHH plot around rs570553380 (A>G) in the 40 Quechuan Andean whole genomes.

Fig. 2.. Positively selected EPAS1 variant is associated with lower hematocrit, higher oxygen saturation, and lower incidence of EE within Andean males.

In Andean males, the EPAS1 missense variant rs570553380 (A>G, p.[His194Arg]) is associated with (A) hematocrit (%) (n = 139), (B) oxygen saturation (%SpO2) at $F_{I{O}_2}$ = 0.10 [subset of the 139 Andeans, n = 55; data from (40)], and (C) incidence of EE (n = 139). EE criteria: Hct ≥ 63% in men ([Hb] ≥ 21 g/dl) and ≥ 57% in women ([Hb] ≥ 19 g/dl) (63) (see fig. S4 for females). *P < 0.05, **P < 0.01, and ***P < 0.001. NS, not significant.

Fig. 3.. EPAS1 variant is highly conserved and results in altered expression of HIF-2α targets under hypoxia.

(A) Conservation of rs570553380 (A>G) in a subset of 100 Species Vertebrate Multiz Alignment & Conservation (fig. S6). Conservation scores (phyloP) of nucleotides across 100 vertebrates showed significant conservation at rs570553380 [−log(P) = 6.02; positive scores indicate slower evolutionary change while negative scores indicate accelerated evolution, greater than expected due to neutral drift (42)]. (B) Location of rs570553380 within the ribbon diagram and crystal structure of HIF-2α (white) as bound to aryl hydrocarbon receptor nuclear translocator (ARNT) (teal) and the hypoxia response element (HRE, orange) (Protein Data Bank ID: 4ZPK). His¹⁹⁴ (purple), shown in sphere representation, extends into one of the protein-protein interfaces between HIF-2α and ARNT. (C) Relative expression of HIF-2α target genes Adrenomedullin (ADM), Basic Helix-loop-helix Family Member e40 (BHLE40), Solute Carrier Family 2 Member 1 (GLUT1), Hypoxia Inducible Lipid Droplet Associated (HILPDA), Inhibitor of Growth Family Member 4 (ING4), and Vascular Endothelial Growth Factor A (VEGF) (table S5) in HEK293T cells nontransfected (WTNT, n = 3) and transfected wild-type (WTT, n = 3), heterozygous (HET, n = 3), and homozygous (HOM, n = 1) for rs570553380 (A>G) exposed to hypoxia (1% O2) relative to normoxia (21% O2) for 12 and 24 hours. Gene expression was determined by real-time reverse transcription quantitative polymerase chain reaction (RT-qPCR). Two-way analysis of variance (ANOVA) shows a significant effect of rs570553380 on gene expression of ADM, BHLE40, HILPDA, ING4, and VEGF after 24-hour exposure to hypoxia, as well as a transfection effect on BHLE40 gene expression. The post hoc generalized linear model shows that increased expression of ADM, BHLE40, HILPDA, ING4, and VEGF from 12- to 24-hour exposure to hypoxia is significantly lower in HET compared to WTT. *P < 0.05 and **P < 0.01.

Fig. 4.. Metabolite associations with hematocrit and/or rs570553380 (A>G, p.[His194Arg]).

(A) Hematocrit in the Andean cohort was associated with a previously identified metabolite (ID: M163) classified as endogenous cannabinoid arachidonoyl ethanolamide. Metabolites with putative IDs M934 (B), M1070 (C), and M2823 (D) were associated both with hematocrit and the EPAS1 rs570553380 variant in the Andean cohort. These metabolites were found in a large-cohort population study (FINRISK) and are yet to be classified. Male and female data points are indicated with closed and open circles, respectively.