Mitochondrial DNA Footprints from Western Eurasia in Modern Mongolia

TL;DR

This study analyzes mtDNA control-region sequences from 2,420 present-day Mongolians and 147 complete mitogenomes to characterize eastern and western Eurasian maternal ancestry and demographic history. It finds strong regional differentiation, with eastern Asian lineages predominating overall and western Eurasian lineages concentrated in western Mongolia, and links some western Eurasian lineages to post-glacial expansions and later Silk Route-era movements.

Key findings

• Of 2,420 Mongolian mtDNAs, 1,987 (82.1%) belong to eastern Asian haplogroups, with C at 19.6% and D4 at 19.8%; western Eurasian haplogroups reach 21.7% in western Mongolia and decrease eastwards, with H the most represented western Eurasian haplogroup at 6.5%.
• The 2,420 mtDNAs were classified into 413 lineages grouped into 18 macro-haplogroups; PCA separates the westernmost regions Bayan-Ölgii, Khovd, and Uvs due to JT, N2, R0, and X, while northern regions cluster together and southern regions are separated by eastern Asian lineages.
• Complete mitogenomes (N = 147; 121 western Eurasian and 26 eastern Asian) revealed two novel sub-branches, HV13b1 and U5b1c2; haplogroup H1 was present at about 3%, three times higher than reported for Inner Asia, with sub-clades H1b, H1c, and H1j.
• Bayesian analyses reconstructed two major increases in effective population size: one beginning in the late glacial period around 18–15 kya and a steeper early Holocene increase; some lineages coalesced around 5 kya during the Bronze Age without a significant demographic impact, and six sub-lineages matched the timeframe and geographic path of the Silk Route.

Methods

• Collected 2,420 unrelated subjects with a Mongolian terminal maternal ancestor from 20 of 21 provinces using mouthwash; sequenced the mtDNA control region (2,133 full control-region haplotypes and 2,335 HVS1 haplotypes), classified haplogroups with PhyloTree build 17, and analyzed diversity, PCA, MDS, and haplotype sharing.
• Sequenced 147 complete mitogenomes using the Ion Torrent Personal Genome Machine, built a most parsimonious tree and Bayesian BEAST analyses with an L3 outgroup, and compared results with 693 published ancient mitogenomes, including 25 excavated in Mongolia.

Limitations

• Stated: provinces or macro-areas with fewer than 30 individuals were excluded from intra-Mongolia analyses; sample bias in the Khangai Mountains (N = 12) cannot be excluded; it was not completely possible to identify specific lineages that evolved in situ; and recent events younger than 3 kya cannot be accurately described through coalescent estimates because mitogenomes accumulate about one mutation in more than two thousand years on average.
• Evident from the text: sampling is uneven across provinces, and 348 individuals lacked province information and were assigned to an “unspecified” group, which may affect regional comparisons; control-region-only analyses also have lower resolution than complete mitogenomes.