Archaic Adaptive Introgression in TBX15/WARS2

Abstract

A recent study conducted the first genome-wide scan for selection in Inuit from Greenland using single nucleotide polymorphism chip data. Here, we report that selection in the region with the second most extreme signal of positive selection in Greenlandic Inuit favored a deeply divergent haplotype that is closely related to the sequence in the Denisovan genome, and was likely introgressed from an archaic population. The region contains two genes, WARS2 and TBX15, and has previously been associated with adipose tissue differentiation and body-fat distribution in humans. We show that the adaptively introgressed allele has been under selection in a much larger geographic region than just Greenland. Furthermore, it is associated with changes in expression of WARS2 and TBX15 in multiple tissues including the adrenal gland and subcutaneous adipose tissue, and with regional DNA methylation changes in TBX15.

Keywords: Denisova; Native Americans; adaptive introgression; admixture.; methylation; positive selection.

Fig. 1

(A) Geographic distribution of rs2298080 in different 1,000 Genomes populations. The color blue corresponds to the archaic allele in this SNP. For comparison, the allele frequency in Greenlandic Inuit is also shown. This figure was made using the Geography of Genetic Variants browser v.0.1, by J. Novembre and J.H. Marcus: http://popgen.uchicago.edu/ggv/. (B). Archaic allele frequencies of rs2298080 in the continental panels of the Simons Genome Diversity Project.

Fig. 2

(A) Genome-wide histogram of the number of uniquely shared sites where the Denisovan allele is at less than 1% frequency in Africans (AFR, excluding admixed African-Americans) and at more than 20% frequency in Eurasians (EUR + SAS + EAS). The counts were computed in non-overlapping 40 kb regions of the genome. The y-axis is truncated, as the vast majority of regions have zero uniquely shared sites. TBX15 and WARS2 are among the few regions that have three and four uniquely shared sites, respectively. (B) In each of the same 40 kb windows, we also computed the 95% quantile of Eurasian derived allele frequencies of all SNPs that are homozygous derived in Denisova and less than 1% derived in Africans. The figure shows a histogram of this score for all windows. The 95% quantile scores for TBX15 and WARS2 are 0.27 and 0.26, respectively.

Fig. 3

Introgression tracts inferred from the CRF method (Sankararaman et al. 2016) in the different continental panels of the SGDP data, using Denisova as the source population. The tract (denoted by the purple dashed line) coincides with the peak of PBS scores obtained in the GI scan for positive selection (red dots). For comparison, we also show the boundaries of the tract as inferred by the HMM method (green dashed line). The Even, Yakut and Naxi contain considerably longer versions of the tract. The frequency of the introgressed haplotype is highest among Native Americans, but is also at intermediate frequencies in Central and East Asians, and Siberians.

Fig. 4

We counted the differences between the Denisovan genome and all haplotypes in each 1,000G continental panel (AFR, EUR, SAS, EAS, and AMR). For each panel, we plot the cumulative number of haplotypes that have as many or less differences to the Denisovan genome than specified in the x-axis. Below each of these plots, we also plot the haplotype structure for each 1,000G panel in the introgressed tract, ordering haplotypes by decreasing similarity to the Denisovan genome (red arrow, at the top of each panel). The color codes at the bottom refer to 1,000G sub-populations to which the different haplotypes belong, as indicated by the colors in the left column in each panel.

Fig. 5

We computed F_ST in different non-African populations (CEU, CHB, PEL, MXL) against African Yoruba (YRI) in the TBX15/WARS2 region. We find a significant (P < 0.01) local increase in F_ST in American populations (PEL, MXL) exactly where the introgressed haplotype is inferred to be (CRF-tract: vertical dashed lines, HMM-tract: vertical dotted lines). In contrast, we found no significant increase in F_ST when comparing Europeans (CEU) and East Asians (CHB) against YRI (although there is an observable but non-significant peak in CHB). 99th percentile values of the F_ST empirical distribution for all comparisons are represented as horizontal dotted lines and color-coded according to the corresponding population.

Fig. 6

Network of archaic haplotypes and 20 most common present-day human haplotypes from the 1,000 Genomes Project in the TBX15/WARS2 introgressed region as inferred by the HMM. Each pie chart is a haplotype, and the dots along each line represent the number of differences between each haplotype. The size of each pie chart is proportional to the log base 2 of the number of individuals in which that haplotype appears, and the colors refer to the proportion of those individuals that come from different continental populations. AFR: Africans. AMR: Americans. EAS: East Asians. EUR: Europeans. SAS: South Asians.

Fig. 7

The regulatory effects of the introgressed haplotype on TBX15 and WARS2. (A) Methylation maps of the introgressed tract and its downstream region. The bottom panels show the reconstructed methylation maps of the Denisovan and the Neanderthal genomes, as well as two modern methylation maps from osteoblasts and fibroblasts. The maps are color coded from green (unmethylated) to red (methylated). The modern maps include fewer positions as they were produced using a reduced representation bisulfite sequencing (RRBS) protocol. Above the methylation maps are the CpG positions whose methylation levels are significantly associated with the introgressed haplotype. The top panels show TBX15 and WARS2, the introgressed tract (as defined by the CRF), and the previously identified DMRs, where the Denisovan genome differs from present-day humans. (B) Individuals carrying the introgressed haplotype (marked by black boxes) show lower levels of DNA methylation in both regions, compared with individuals who do not carry the introgressed haplotype (marked by orange boxes), in both the 16-CpG region around the TSS of TBX15 (left), and a CpG position within the introgressed region (right). (C) Expression of TBX15 in individuals carrying the introgressed haplotype versus individuals who do not carry it. TBX15 is expressed on average at 22% higher levels in individuals carrying the haplotype. Analogous boxplots but partitioned by genotype rather than presence/absence of haplotype are shown in figure S14.