Mitochondrial DNA analysis of human remains from the Yuansha site in Xinjiang, China

Science in China Series C: Life Sciences

© 2008 SCIENCE IN CHINA PRESS

Springer

Sci China Ser C-Life Sci | Mar. 2008 | vol. 51 | no. 3 | 205-213

www.scichina.com life.scichina.com

www.springerlink.com

Mitochondrial DNA analysis of human remains from the Yuansha site in Xinjiang, China

GAO ShiZhu1,3, CUI YinQiu1,2, YANG YiDai1, DUAN RanHui1, Idelisi ABUDURESULE4, Victor H. MAIR5, ZHU Hong1 & ZHOU Hui1,2† 1 Laboratory of Ancient DNA, Research Center for Chinese Frontier Archaeology, Jilin University, Changchun 130012, China; 2 College of Life Science, Jilin University, Changchun 130023, China; 3 School of Pharmaceutical Sciences, Jilin University, Changchun 130021, China; 4 Xinjiang Cultural Relics and Archaeology Institute, Urumqi 830011, China; 5 Chinese Language and Literature Department of East Asian Languages and Civilizations, University of Pennsylvania, USA

The Yuansha site is located in the center of the Taklimakan Desert of Xinjiang, in the southern Silk Road region. MtDNA was extracted from fifteen human remains excavated from the Yuansha site, dating back 2,000―2,500 years. Analysis of the phylogenetic tree and the multidimensional scaling (MDS) reveals that the Yuansha population has relatively close relationships with the modern populations of South Central Asia and Indus Valley, as well as with the ancient population of Chawuhu.

Yuansha site, ancient DNA, mtDNA, Xinjiang, Central Asia

Central Asia is a vast geographic area connecting East and West, comprising Uzbekistan, Tajikistan, Turkmeni-stan, Kirghizstan, and part of Kazakhstan, an area extending from the Pamir Plateau and the Hindu Kush, far into the region of Xinjiang, China. Population migra-tions took place continually in Central Asia throughout its history, and the inhabitants of this region were com-posed of admixed populations of various physical types, cultures, and languages[1]. MtDNA analysis shows that modern populations in Central Asia occupy an interme-diate phylogenetic position between European and Asian lineages, due to the admixture of already differentiated Western and Eastern matrilines[2−4]. However, the ge-netic traces left by earlier migration events sometimes are erased by later ones. If analysis is based only on modern data, it is hard to decipher the migration and admixture history of Central Asian populations. Ancient DNA study of Central Asia provides a powerful tool with which to reconstruct migration history and to un-derstand the complex admixture in this region[5−7]. MtDNA, because of its multiple copies, maternal in-

heritance, lack of recombinants, and relatively high rate of evolution, has been widely used to study population history, migration, and evolution, and especially to re-search the genetic structure of ancient human remains.

Xinjiang, located on the eastern edge of Central Asia, and traversed by the Silk Road, is an important area for the culture contact of East and West. Various ancient civilizations encountered each other in this region, in-cluding those of the Chinese, Indian, Persian, and Greek. Recently, the origin of the Tarim Basin population has come to be of great interest. At present, the archaeo-logical evidence that supports the indigenous origin hy-pothesis is insufficient. It is widely acknowledged that three cultures from different regions contributed to the origin and development of the Tarim Basin civilization. One is nomadic culture from the Russo-Kazakh steppe, Received September 8, 2006; accepted November 29, 2007 doi: 10.1007/s11427-008-0034-8 †Corresponding author (email: [email protected]) Supported by the National Natural Science Foundation of China (Grant No. J0530184) and the “985” Foundation of Jilin University

206 GAO ShiZhu et al. Sci China Ser C-Life Sci | Mar. 2008 | vol. 51 | no. 3 | 205-213

north of the Tarim Basin, including the Bronze Age Afanasievo Culture, the Andronovo Culture, etc. Current archaeological data show that these cultures have played an important role in the earlier development of the Tarim Basin civilization[8]. The second influence is Eastern culture, such as the Bronze Age Qiang Culture from the eastern part of Qinghai and Gansu, the Han Culture after the Silk Road was establishied, as well as, later, the cul-tures from Northeast and North Asia[9] (for example, the westward movements of Xiongnu and Mongolia in the third and thirteenth centuries, AD, respectively). The third influence oases culture is from west of the Tarim Basin and north Central Asia in the Bronze Age, repre-sented by Bactrian-Margiana Culture[10]. However, it is generally believed by archaeologists that the contribu-tion of the oases culture to the early development of the Tarim Basin civilization is relatively small. In 200 BC or earlier, due to the expansion of the Aryans and its effect on the population migrations, the oases culture bearers might have crossed the Pamir Plateau directly into the Tarim Basin, with a resulting impact on the agricultural civilization of that region[11]. More archaeological evi-dence is needed to support this hypothesis. After the Silk Road was establishied, the cultural contacts between the Tarim Basin and western regions became more frequent. Based on physical anthropological data, Han[12] contends that the population migrations in the Tarim Basin can be summarized into three periods. At present, the earliest

inhabitants (Gumugou, 2300―1430 BC) discovered in the Tarim Basin are classified as "Proto-European type", and attributed to Afanasievo and Andronovo populations. The second immigration period involved an "Eastern Mediterranean type" (or"Indo-Afghan type") population, which became the major residents of that region in the Eastern Han Dynasty. Han suggested that in 200-300BC or even earlier, a branch of the Mediterranean popula-tions, traveling over the Pamir Plateau, and along the southern margin of the Tarim Basin, entered Xinjiang, penetrating far into the Lop Nur region. The third period occurred after the Silk Road had been established and involved a westward movement of the "Mongoloid type” from East and North Asia.

The Yuansha site is the oldest that has been discov-ered in the Taklimakan Desert to the present. The aim of this study is to investigate the genetic relationships be-tween the Yuansha population and the present and an-cient Eurasian populations, and to explore the origin and migration history of populations in the Tarim Basin as well as Central Asia.

1 Materials and methods

1.1 Sample collection

The Yuansha site is located in the center of the Takli-makan Desert, in the downstream of the Keriyan River, Xinjiang Province, northwestern China (Figure 1). Its

Figure 1 Location of the Yuansha site.

GAO ShiZhu et al. Sci China Ser C-Life Sci | Mar. 2008 | vol. 51 | no. 3 | 205-213 207

coordinates are 38°52′ N, 81°35′ E. According to carbon 14 (14C) testing, the Yuansha site is dated back about (2135±50) years. Archaeological studies show that the period of the Yuansha site is not later than the Western Han Dynasty. Since it was from the Han Dynasty onwar, the East began its contact with the West through the Silk Road. The Yuansha site is in the southern part of the Silk Road regions.

Sixteen samples collected from the Yuansha site were well preserved because of its dry and cold climate. Of these sixteen samples, fourteen were bones, and the other two were teeth. Two samples (No. 6 and 9) were from the same ancient individual.

1.2 Ancient DNA extraction, amplification, and se-quencing

The outer surface of the bones was removed to about 2―3 mm of depth with a dental drill to eliminate sur-face contamination. Then bone fragments were ground to powder in Waring Blandor (Waring, USA). Teeth were first soaked in 10% bleach for 20 min, rinsed with ethanol and distilled water, and then UV-irradiated for 30 min on each side (I. Ovchinnikov et al., 1999). Each tooth was powdered under liquid nitrogen in a 6750 Fre- ezer Mill (SPEX, USA).

DNA was carefully extracted from 500 mg bone or tooth powder using the Geneclean for Ancient DNA kit (BIO 101, Vista, Calif.), according to the manufacturer’s protocol. Four sets of overlapping primers (Table 1)

were used to amplify the mtDNA hypervariable region I (HVI) between positions 16035―16378. PCR amplifi-cation was performed in 25 µL of a reaction mixture containing 67 mmol/L Tris-HCl (pH8.8), 2 mmol/L MgCl2, 1 mmol/L each of dNTP, 0.5 µmol/L of each primer, and 1 U of Taq polymerase (Promega), and 1.3 mg/mL BSA. PCR was carried out as follows: predena-turation at 94℃ for 10 min, followed by 33 cycles with 94℃ for 1 min, 45 s at 52℃ for HVIa and HVIb or 45 s at 55℃ for HVIc and HVId, elongation for 1 min at 72℃, final extension at 72℃ for 10 min and storage at 4℃. To verify the reliability of the experiments, nega-tive controls containing all of the reagents but without remains DNA were included in each DNA extrac-tion/PCR run.

Amplification products were checked on a 2% aga-rose gel and purified with QIA quick Gel Extraction Kit (Qiagen). Sequence PCR was carried out using the ABI 310 Terminator Sequencing kit (PE Applied Biosystems) according to the manufacturer’s protocol and analyzed on the ABI PRISM 3100 automatic sequencer (PE Ap-plied Biosystems).

1.3 RFLP Typing

Seven sets of primers (Table 1) were used to amplify the mtDNA coding sequences for RFLP analysis. PCR am-plification was performed in 20 µL of a reaction mixture, and the amplifying conditions and the cycling proce-

Table 1 Primers for sequencing and RFLP

HVI segment /Haplogroup Primer Sequencing/RFLP Length (bp) Tm (℃)

HVI-a L16052 5′-GGAAGCAGATTTGGGTAC-3′ H16142 5′-ATGTACTACAGGTGGTCAAG-3′

sequencing 124 52

HVI-b L16131 5′-CACCATGAATATTGTACGGT-3′ H16218 5′-TGTGTGATAGTTGAGGGTTG-3′

sequencing 126 52

HVI-c L16185 5′-ACCCAATCCACATCAAAACC-3′ H16286 5′-TGTACTGTTAAGGGTGGGTAGG-3′

sequencing 142 55

HVI-d L16281 5′-CCTCACCCACTAGGATACCAA-3′ H16379 5′-CAAGGGACCCCTATCTGAGG-3′

sequencing 138 55

M L10304 5′-ATGAGCCCTACAAACAACT-3′ H10423 5′-GAGTCGAAATCATTCGTTT-3′

+AluI 10397 156 50

G L4701 5′-CTATCCTCTTCAACAATATACTCT-3′ H4848 5′-ATGTGAGAAGAAGCAGGC-3′

+HhaI 4831 182 52

H L6987 5′-GGCATTGTATTAGCAAACTCAT-3′ H7098 5′-TAGGGTGTAGCCTGAGAATAG-3′ −AluI7025 152 50

I L9959 5′-CTGTATGTCTCCATCTATTGC-3′ H10108 5′-TAGTAGTAAGGCTAGGAGGG-3′

+AluI 10032 150 47

D L5159 5′-GCACCACGACCCTACTACTA-3′ H5287 5′-GGGATGATGAGGCTATTGT-3′ −AluI 5176 166 48

A L648 5′-CCCACATCACCCCATAAACAA-3′ H745 5′-GCTTGATGCTTGTCCCTTT-3′

+HaeIII 663 136 50

T L13279 5′-TTACAATCGGCATCAACC-3′ H13424 5′-GAGTAGTCCTCCTATTTTTCG-3′

+BamHI 13366 146 51


dures were the same as HVI. Restriction digestions were carried out following the manufacturer’s recommenda-tions (New England Biolabs, Beverly, MA). Fragment length was analyzed using 2% agarose gel. For each sample, the characteristic site at 10397 of macrohap-logroup M/N was screened first. Then, on the basis of the HVI motifs and the mtDNA phylogenetic tree[13], other characteristic sites of the coding region were used to further define the haplogroup.

1.4 Precautions against contamination

Standard contamination precautions cited by O’Rourke et al.[14] were followed as closely as possible. Pre- and post-PCR experimental areas were separated. Facemasks, gloves, lab coats, and pipettes with aerosol-resistant tips were used. Negative extraction controls and negative PCR controls were employed. The equipment and benches were treated with DNA contamination removal solution (DNA-QUANT OFFTM Q-BIO gene) and ultra-violet irradiation at 254 nm. In order to detect possible exogenous sources of contamination, the mtDNA of the individuals involved in processing the samples were typed and compared with the results. At least three ex-tractions and three amplifications of each extract were carried out for each sample to assess the reproducibility and the authenticity of the results.

1.5 DNA analysis

MtDNA sequences were analyzed using CLUSTALX

1.83, excluding insertions and deletions[15]. Pairwise population distances (dA) were computed in MEGA ver-sion 2.0. A neighbor-joining tree (NJ tree) was con-structed using PHYLIP version 3 based on pairwise population distances. Multidimensional scaling (MDS) was performed using SPSS version 11 on dA values.

2 Results

2.1 Mitochondrial DNA polymorphisms

Reproducible sequences were obtained from a total of 16 samples. The 326-base pair (bp) fragments (16053－16378) of the mtDNA HVI control region were com-pared with the revised Cambridge Reference Se-quence[17]. A total of 15 different sequences defined by 22 variable nucleotide positions were obtained (Table 2), including one deletion, 21 transitions, and no transver-sion. Only the samples (No. 6 and 9) from the same in-dividual shared one mtDNA type.

To investigate the distribution of the Yuansha se-quences among modern populations, HVI motif searches were performed against mtDNA haplotypes in GenBank (Table 3). Of the 15 mtDNA types, a full sequence match was found in GenBank for nine types, and a one-step match was found for four types. The remaining two types had no shared sequences. The haplotypes of the ancient individuals matched are widely distributed in Eurasian populations.

Table 2 HVI polymorphic sites and haplogroups of mtDNA

Samples HVI polymorphic site RFLP Haplogroup 1 16126 16294 10397 AluI−, 13366 BamHI+ T 2 16223 16227 16278 16362 10397 AluI+, 4831HhaI+ G2a 3 CRS 10397 AluI−, 7025 AluI− H 4 16126 16163 16186 16189 16294 10397 AluI−, 13366 BamHI+ T 5 16129 16223 10397 AluI−, 10032 AluI+ I 6 16184 16223 16362 10397 AluI+, 5176 AluI− D 7 16184 16223 10397 AluI+, M 8 16126 10397 AluI, 7025 AluI− H 9 16184 16223 16362 10397 AluI+, 5176 AluI− D 10 16168 16223 10397 AluI N 11 16126 16179 16216 10397 AluI N 12 16223 16256 10397 AluI+ M 13 16311 16354 10397 AluI, 7025 AluI− H 14 16223 16270 (deletion) 10397 AluI N 15 16129 16223 16290 16319 10397 AluI, 663HaeIII+ A 16 16223 16253 10397 AluI+ M


Table 3 Distribution of HVI sequences in GenBank Sample Sharing sequence (number of individuals) a) 1 Continental European (18), Russian (west Siberia) (4), Greek (1) 2 Russian (Buyat) (5), Mongolian (1), Chinese Han (1), Chinese Uigur (1)

3 Continental European (364), Balkan (33), Russian (west Siberia) (30), Turk (6), Greek (31), Palestinian (6), Syrian (1), Nigerien (2), Australian (2), Chinese Han (1), Chinese Dai(1), Chinese Uigur (1)

4 Continental European (26), Balkan (3), Russian (5), Syrian (2), Turk(2), Indian (1), Chinese Han (1) 5 Continental European (42), Russian (1), Indian (3), Kenyan (1)

6,9 b) Russian (2), Georgian (3), Palestinian (1), Slovakian (1), Chinese Han (5), Chinese Mongolian (1), Chinese Dawuer (1), Chinese Ti-betan (1), Chinese Bai (2), Chinese Tujia (3), Chinese Yi (2), Chinese Korean (2)

7 b,c) 1. Romanian (1), Portuguese (1), Greek (1), Georgian (1), Caucasian (1), Aboriginal Australian (2) 2.c) Greek (1), Russian (2), Portuguese (1), Scottish (4), Chinese Tibetan (1), Chinese Miao (1), Chinese Han (1), Chinese Ani(1), In-donesian (6), Bolivian (1), African (3), Aboriginal Australian (2)

8 Norwegian (19), Scotch (48), Portuguese (4), Russian (3), Romanian (1), German (2), Hungarian (ancient individual) (1), Icelander (15), Adygeis (2), Chinese Ewenke(1), Chinese Dawuer (1)

10 Greek (1) 11 Unique 12 Indian (8) 13 Scottish (1) 14 b) The same as the second part of sample 7 15 Unique 16 b) The same as the second part of sample 7 a) All the shared sequences are from GenBank database of NCBI. b) Only one-step neighbor of the samples was found. c) The shared sequences in the

second part of sample 7 only show the mutation at site 16223.

2.2 Haplogroup distributions

Haplogroup assignment was attempted based on HVI sequences (from 16053 to 16378) in combination with RFLP data (Table 2). Samples 1 and 4 are assigned to haplogroup T. Haplogroup T is one of the European- specific haplogroups, with the frequency ranging from 6% to 14%[18]. Sample 2 is classified as G2a, a subhap-logroup of G. G is widely distributed in East Asian, North Asian, and Central Asian populations. It should be pointed out that only G2 of haplogroup G is found in Central Asian populations, and it may be distributed in the populations attributed to an earlier migration[3]. Samples 3, 8, and 13 are linked to haplogroup H, which is frequent in European and Near Eastern populations. Haplogroup H is also present in North Africa and Cen-tral Asia, but is rare or absent in East Asia. Sample 5 is assigned to haplogroup I, one of the West Eurasian line-ages. Haplogroup I is mainly distributed in the popula-tions of North and West Europe. The samples from the same ancient individual, 6 and 9, are classified as hap-logroup D. D is a common haplogroup in East Asia, Northeast Asia, and North Asia. Sample 5 is linked to haplogroup A, which has the same distribution trend as D. The haplogroups of samples 7, 12, and 16 are diffi-cult to define due to their unusual HVI motifs; they are assigned only to macrohaplogroup M. Similarly, samples 10, 11, and 14 are assigned to macrohaplogroup N.

2.3 Phylogenetic analysis

Thirty-six modern Eurasian populations and several an-cient populations, including the Chawuhu population in southern Xinjiang, the Jiaohe, Yanghai, and Subeixi populations in the Turfan Basin of Xinjiang, the Xiongnu population in Mongolia, as well as the Jiang-jialiang population in Hebei Province, were chosen for the phylogeneticanalysis (Table 4). The geographic dis-tributions of the modern populations can be grouped as East Asia, North Asia, Central Asia, Indus Valley, India, Iranian Plateau, Anatolia/Caucasus, and Europe.

In the NJ tree (Figure 3), modern populations seem to form several large branches according to their geographic distributions. All of the populations in Central Asia, both modern and ancient, occupy positions between the Asian and European lineages. The branches of the Cen-tral Asian populations are not only linked with the European branch through that of populations in Turkey and Anatolia, but are also linked with the East, North, and South Asian branches. This reflects the important role of Central Asia in the history of human expansion and migration. The branch of the Yuansha population is closest to the branch of the ancient Chawuhu population, and has a relatively short distance from the Tajikistan population (SHU) branch. Moreover, these three branches cluster with the branch of the Indus Valley populations. The Yuansha population branch, located between the


Table 4 The code, number and location of the modern and ancient populations Population Code Number Location EAST ASIA NORTH Northern Han[19] HAN 100 Qingdao and Liaoning, China Korean[20] KOR 64 Korea Japanese[20] JAP 50 Aboriginal Japanese SOUTH Southern Han[19] HAS 115 Guangzhou, Yunan , Wuhan in China

Southern ethnic groups[21] ETH 153 Dai (Yunnan, China), Zhuang(Guangxi, China), Thai (in Northern Thailand)

NORTH ASIA Buryat[22] BUR 61 Buryat Mongolian[23] MOG 103 Mongolia CENTRAL ASIA EAST Xinjiang Uighur[24] UIX 45 Kashen & Ili, Xinjiang, China Kazakhstan Uighur [2] UIG 55 Penjim, easternmost Kazakhstan, boundary of China Kazakh[2] KAZ 55 Kegen, easternmost section of Kazakhstan Lowland Kirghiz[2] LKIR 48 Talas, northernmost section of Kirghizstan Highland Kirghiz[2] HKIR 47 Sary-Tash, Kirghizstan, boundary with China and Tajikistan SOUTH[25] Shugnan SHU 44 High Pamirs, Tajikistan Hunza Burusho HUN 44 Northern Pakistan, Karakorum Mountains Kalash KAL 44 Northwest Frontier Province WEST[25] Uzbek UZB 42 Surkhandarya, Uzbekistan Altaic Turkmen TUK 41 Turkmenistan Altaic Kurdish KUT 32 Turkmenistan SOUTH ASIA Indian[26] INDIA 177 India INDUS VALLEY[25] Hazara HAZ 23 Northwest Frontier Province and Balochistan Pathan PAT 44 Northwest Frontier Province and Balochistan Baluch BAL 39 Southwestern Pakistan, Baluchistan Brahui BRA 38 Southwestern Pakistan, Baluchistan Parsi PAR 44 Southeastern Pakistan, Karachi Sindhi SIN 23 Southeastern Pakistan, Sindh Pakistania PAR 100 Karachi, Southeastern Pakistan IRANIAN PLATEAU[25] Gilaki GIR 37 Northern Iran, southwestern Caspian Sea area Mazandarian MAZ 21 Northern Iran, southeastern Caspian Sea area Kurdish KUI 20 Western Iran Indo-European Lur LUR 17 Southwestern Iran (Zagros Mountains) ANATOLIAN/ CAUCASUS Turkish[25] TUI 40 Mostly eastern and western Azerbaijan Altaic Turkish[27] TUT 45 Turkey EUROPE Swiss[28] SWI 74 Switzerland Italian[29] ITT 51 Italy Basque[29] BAS 45 Basque areas in Spain and France British[30] BRI 100 England Ancient populations EAST ASIA Jiangjialiang[31] Jiangjialiang 10 Hebei Province, China NORTH ASIA Xiongnu[32] Xiongnu 46 Northern Mongolia, Egyin Gol Valley CENTRAL ASIA Yuansha（this study） Yuansha 15 Center of Taklamakan Desert, Xinjiang, China Subeixi & Jiaohe[33,34] Turfan 10 Turfan Basin, Xinjiang, China Yanghai[33] Yanghai 10 Turfan Basin, Xinjiang, China Luobenoer[33] Luobenoer 10 Luobenoer region, Xinjiang, China Chawuhu[35] Chawuhu 8 South of Tian Moutain, Hejing county, Xinjiang, China


branches of Europe and India, is closer to the latter. Com- pared with the North and East Asian branches, the Yuansha branch appears more closely aligned to the South Asian branch, occupying the position between the branches of South Asia and Europe. This indicates that to the ancient Yuansha population the genetic contribu-tions from Southwestern Asia are relatively significant, as compared to East or North Asia.

2.3 Multidimensional scaling analysis

To further investigate the relationships between the Yuansha and other populations, multidimensional scal-ing analysis (Figure 3) was performed on dA values. The MDS analysis shows how the populations can be grouped together according to their geographic distribu-tions. Among ancient populations, the Yuansha has a closer affinity with the Chawuhu and Turfan populations,

while the genetic distance between the Yuansha and Yanghai is relatively far. Compared with these modern populations, the Yuansha has a closer relationship with Tajikistan (SHU) and Uzbekistan (UZB), as well as with the populations in the Indus Valley, including the Balu-chistan (BAL) and Karachi (PAT) populations.

3 Discussion

MtDNA data was obtained from the remains of fifteen ancient individuals at the Yuansha site. The ancient and modern mtDNA data of Xinjiang and Central Asian populations were used to explore the population history of origin, migration, and admixture in the Tarim Basin. The period of the Yuansha site is not later than the Western Han Dynasty. The analysis of physical anthro-pology reveals that the Yuansha population is classified

Figure 2 The NJ tree based on the dA values. (b) is the larger view of the box in (a). See Table 4 for codes.

Figure 3 Two-dimensional MDS plot. The codes of the populations are consistent with those in Table 4.


as the Indo-Afghan sub-type of the European type[36], which links the origin of the Yuansha population to the migration of the second period in the Tarim Basin. The results of the phylogenetic and multidimensional scaling analysis reveal that the Yuansha population appears more closely aligned to the modern Tajikistan popula-tion (SHU) in Southern Central Asia and the Indus Val-ley population, which is consistent with the results of the physical anthropological analysis. However, because of the lack of mtDNA data from the ancient Bactrian and Andronovo populations, the genetic relationships be-tween the Yuansha and those populations cannot be in-vestigated by molecular genetic analysis.

The research of Hemphill and Mallory[36] in 2004 suggests that the Gumugou population possessed a close affinity neither with the populations of the steppe nor with those of the oasis, but instead with the populations of the Indus Valley, a conclusion obtained by comparing the Tarim Basin samples from the Gumugou, Alagou, and Loulan Bronze Age sites with Bactrian, Afanasievo, and Andronovo Central Asian Bronze Age samples, as well as with ancient samples from the Indus Valley, Iran, and the Caucasus. The authors contended that, given the paucity of available evidence and appropriate compara-tive samples, the Gumugou population showed a close affinity with Indus Valley populations. However, all re-sults indicated that later Tarim Basin samples from Alagou (800BC―200AD) and Loulan (202BC―220AD) are more closely affiliated with earlier Bactrian popula-tions. None of the results revealed affinities between Tarim Basin samples and populations attributed to Bac-trian and Andronovo culture. The present study suggests that the Yuansha samples show a close affinity with those modern populations from South Central Asia and Indus Valley, which supports Hemphill’s results in a certain degree. Nevertheless, due to the absence of an-cient mtDNA data from South Central Asia and the In-dus Valley populations, the present results cannot sup-port a close affinity between the Yuansha population and Bactrian populations through direct evidence.

To further investigate the history of population origin, migration, and admixture in the Xinjiang region, ancient mtDNA data from Xinjiang, including Yuansha, Turfan Basin, Yanghai, Luobenoer, and Chawuhu, were compared with other modern and ancient DNA data from the circumjacent regions (Figures 2 and 3). The results show that the Xiongnu population (300BC―2AD) of Mongo-

lia exhibits a close affinity to North Asian populations, while the Jiangjialiang (about 3000BC or so) population of Hebei shares a close affinity with North of East Asian populations. These two populations are both identified as possessing a distant relationship with the Yuansha population. The Luobenoer population (about 1800BC or so) classified as of a proto-European type and linked to steppe culture shares a close affinity with populations of Europe and the Iranian Plateau, and it possesses a relatively remote affinity with the Yuansha population. Affinities are closest between the Yuansha population and the Chawuhu population, followed by the Turfan Basin populations. It is worth pointing out that, due to the similarities of the Yuansha and Chawuhu cultures, the former is assigned to the Zaghunluq type of the Chawuhu culture, and this is supported by the mtDNA results of this study. Compared with the Yuansha popu-lation, the Turfan Basin populations share a relatively close relationship with North and East Asian populations, which may be attributed to the contributions of Mongol-oid lineage to the gene pool of the Turfan Basin. This indicates to some extent that a westward admixture be-tween Mongolian and local lineages existed in the Tarim Basin at that time.

The trends of population migration can be inferred to some degree from analysis of the phylogenetics and multidimensional scaling. The results show that the gene pool of East Central Asia has been affected significantly by North and East Asian lineages. In the MDS plot, the Lowland Kirghiz population exhibits a closer affinity to the Mongolian population. The phylogenetic analysis shows that the Xinjiang Uighur population and the Lowland Kirghiz of East Central Asia cluster with the populations of North and East Asia. This influence tends to become relatively weak as one goes westward. Because of the insignificant influence of the population expansions of East and North Asia, the gene pool of South Central Asian populations remains relatively un-changed. Compared to those of the modern ethnic popu-lations in Xinjiang, the gene pools of the ancient Yuan-sha, Chawuhu, Turfan, and Luobenoer populations have been impacted only weakly by North and East Asian matrilines. In the phylogenetic and multidimensional scaling analysis, these ancient populations appear to be more closely aligned with European populations, which reflects an early population expansion from west to east.


In summary, the Yuansha population is identified as exhibiting its closest affinity to South Central Asian populations, followed by the Indus Valley populations, which is consistent with the results from the physical anthropological analysis. Among the ancient populations,

the Yuansha population shares its closest affinity with the Chawuhu population. The phylogenetic and multidimen-sional scaling analyses reveal that the ancient Yuansha population mainly reflects an early eastward population migration in Eurasia.

1 Han K X. Anthropological research of ancient populations in Xinjiang

and physical characteristics of Uigur. West Region Res, 1991, 2: 1―13

2 Comas D, Calafell F, Mateu E, et al. Trading genes along the silk road: mtDNA sequences and the origin of Central Asian populations. Am J Hum Genet, 1998, 63: 1824―1838

3 Comas D, Plaza S, Wells R S, et al. Admixture, migrations, and dis-persals in Central Asia: Evidence from maternal DNA lineages. Eur J Hum Genet, 2004, 1―10

4 Calafell F, Comas D, Pe´rez-Lezaun A, et al. Genetics and population history of Central Asia. Cambridge: McDonald Institute for Ar-chaeological Research, 2000

5 Francalacci P. DNA analysis of ancient desiccated corpses from Xin-jiang. J Indoeur Stud, 1995, 23: 385―389

6 Francalacci P. DNA analysis on ancient desiccated corpses from Xinjiang (China): Further results. Washington, DC: Institute for the Study of Man, 1998

7 Lalueza-Fox C, Sampietro M L, Gilbert M T P, et al. Unravelling migrations in the steppe: mitochondrial DNA sequences from ancient Central Asians. Proc R Soc Lond B, 2004, 271: 941―947

8 Mallory J P, Mair V H. The Tarim mummies. London: Thames and Hudson, 2000

9 Li S. The interaction between Northwest China and Central Asia during the second millennium B.C.: An archaeological perspective. In: Boyle K, Renfrew C, Levine M, eds. Ancient Interactions: East and West in Eurasia. Cambridge, McDonald Institute for Archaeological Research, 2002. 171―182

10 Barber E W. The mummies of Ürümchi. New York: Norton, 1999 11 Lin M C. Origin and migration of Tocharian. West Region Res, 2003,

3: 9―23 12 Han K X. The physical anthropology of the ancient populations of

the Tarim Basin and surrounding areas. In: Mair V H, eds. The Bronze Age and early Iron Age peoples of eastern Central Asia. Philadelphia, University of Pennsylvania Museum Publications, 1998. 558―570

13 Yao Y G, Kong Q P, Wang C Y, et al. Different matrilineal contribu-tions to genetic structure of ethnic groups in the silk road region in China. Mol Biol Evol, 2004, 21: 2265―2280

14 O’Rourke D H, Hayes M G, Carlyle S W. Ancient DNA studies in physical anthropology. Ann Rev Anthropol, 2000, 29: 217―242

15 Higgins D, Sharp G. CLUSTAL: A package for performing multiple sequence alignment on a microcomputer. Gene, 1988, 73: 237―244

16 Saitou N, Nei M. The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol Biol Evol, 1987, 4: 406―425

17 Andrews R M, Kubacka I, Chinnery P F, et al. Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA. Nat Genet, 1999, 23: 147

18 Finnilä S, Lehtonen M S, Majamaa K. Phylogenetic network for European mtDNA. Am J Hum Genet, 2001, 68: 1475―1484

19 Yao Y G, Kong Q P, Bandelt H J, et al. Phylogeographic differentia-tion of mitochondrial DNA in Han Chinese. Am J Hum Genet, 2002, 70: 635―651

20 Horai S, Murayama K, Hayasaka K, et al. MtDNA polymorphism in East Asian populations, with special reference to the peopling of Ja-pan. Am J Hum Genet, 1996, 59: 579―590

21 Yao Y G, Nie L, Harpending H, et al. Genetic relationship of Chinese ethnic populations revealed by mtDNA sequence diversity. Am J Phys Anthropol, 2002, 118: 63―76

22 Tajima A, Hayami M, Tokunaga K, et al. Genetic origins of the Ainu inferred from combined DNA analyses of maternal and paternal lineages. J Hum Genet, 2004, 49: 187―193

23 Kolman C J, Sambughin N, Bermingham E. Mitochondrial DNA analysis of Mongolian populations and implications for the origin of New World founders. Genetics, 1996, 142: 1321―1334

24 Yao Y G, Lv X M, Luo H R, et al. Gene admixture in the Silk Road region of China-evidence from mtDNA and melanocortin I receptor polymorphism. Genes Genet Syst, 2000, 75: 173―178

25 Quintana-Murci L, Chaix R, Wells R S, et al. Where West Meets East: The complex mtDNA landscape of the Southwest and Central Asian Corridor. Am J Hum Genet, 2004, 74: 827―845

26 Kivisild T, Rootsi S, Metspalu M, et al. The genetic heritage of the earliest settlers persists both in Indian tribal and caste populations. Am J Hum Genet, 2003, 72: 313―332

27 Comas D, Calafell F, Mateu E, et al. Geographic variation in human mitochondrial DNA control region sequence: The population history of Turkey and its relationship to the European populations. Mol Biol Evol, 1996, 13: 1067―1077

28 Pult I, Sajantila A, Simanainen J, et al. Mitochondrial DNA sequences from Switzerland reveal striking homogeneity of European popula-tions. Biol Chem Hoppe Sayler, 1994, 375: 837―840

29 Di Rienzo A, Wilson A C. Branching pattern in the evolutionary tree for human mitochondrial DNA. Proc Natl Acad Sci USA, 1991, 88: 1597―1601

30 Piercy R, Sullivan K M, Benson N, et al. The application of mito-chondrial DNA typing to the study of white Caucasian genetic iden-tification. Int J Legal Med, 1993, 106: 85―90

31 Wang C, Cui Y Q, Duan R H, et al. The ressearch of the Neolithic bones from the Jiangjialiang site, Yangyuan, Hebei Province. Ar-chaeology, 2001, 654: 74―81

32 Christine K T, Eric C, Bertrand L. Nuclear and Mitochondrial DNA Analysis of a 2000-year-old Necropolis in the Egyin Gol Valley of Mongolia. Am J Hum Genet, 2003, 73: 247―260

33 Cui Y Q, Duan R H, Zhou H, et al. Analysis of genetical structure of ancient Xinjiang population. Chem J Chin U, 2002, 23: 2270―2280

34 Cui Y Q, Duan R H, Zhou H, et al. Analysis of mitochondrial DNA from the ancient tombs of Turfan. Chem Res Chin U, 2002, 18: 419―423

35 Xie C Z, Liu S B, Cui Y Q, et al. Analysis of mitochondral DNA from the ancient tombs of Chawuhu. J Jilin Univ (science edition), 2005, 43: 538―540

36 Hemphill B E and Mallory J P. Horse-mounted invaders from the Russo-Kazakh steppe or agricultural colonists from western central Asia? A craniometric investigation of the Bronze Age settlement of Xinjiang. Am J Phys Anthropol, 2004, 124: 199―222

Mitochondrial DNA analysis of human remains from the Yuansha site in Xinjiang, China

Documents

Transcript of Mitochondrial DNA analysis of human remains from the Yuansha site in Xinjiang, China