The 88th Annual Meeting of the American Association of Physical Anthropologists (2019)


High Altitude Adaptation and the Evolutionary History of EPAS1 Gene in Tibetan Population

XINJUN ZHANG1, KELSEY WITT1, XIN JIN3, ASAN ASAN3, RASMUS NIELSEN4 and EMILIA HUERTA-SANCHEZ2.

1School of Natural Sciences, University of California, Merced, 2Department of Ecology and Evolutionary Biology, Brown University, 3Shenzhen Headquater, BGI, 4Department of Integrative Biology, University of California, Berkeley

March 28, 2019 5:00, CC Room 26 C Add to calendar

Modern humans have settled in some extreme environments since migrating out of Africa. One example is Tibetans, who have adapted to the hypoxic environment of high altitudes. The EPAS1 gene, which regulates physiological responses to the variation in oxygen supply, was previously identified as being under strong positive selection in Tibet. This beneficial haplotype also matched closely with the EPAS1 haplotype in Denisovans, an archaic human group from 40,000 years ago, suggesting that this haplotype was introduced through adaptive introgression. However, questions remain regarding the actual source population of the EPAS1 haplotype, as well as the timing of admixture.

In this study, DNA sequences (30x depth) including the EPAS1 region from 148 Tibetan individuals were obtained, with altitude ranging from 3000 to 4800 meters above sea level. We identified a positive correlation between altitude difference and EPAS1 mean Fst between groups within Tibet, which may indicate a direct correlation between the selection on EPAS1 and change in oxygen level. To further clarify the history of EPAS1, we applied SPrime method and identified two introgressed segments in EPAS1 region. By comparing the genetic divergence between Neanderthals and Tibetan EPAS1 haplotype, we confidently suggested that an archaic population highly alike the Denisovan, not Neanderthals, admixed with East Asian populations including Tibetans, and introduced the EPAS1 beneficial haplotype. In addition, we fit to the observed distribution of introgressed haplotype lengths, and estimated an approximate time of admixture.