Where West Meets East: The Complex mtDNA Landscape of the Southwest and Central Asian Corridor

Abstract

The southwestern and Central Asian corridor has played a pivotal role in the history of humankind, witnessing numerous waves of migration of different peoples at different times. To evaluate the effects of these population movements on the current genetic landscape of the Iranian plateau, the Indus Valley, and Central Asia, we have analyzed 910 mitochondrial DNAs (mtDNAs) from 23 populations of the region. This study has allowed a refinement of the phylogenetic relationships of some lineages and the identification of new haplogroups in the southwestern and Central Asian mtDNA tree. Both lineage geographical distribution and spatial analysis of molecular variance showed that populations located west of the Indus Valley mainly harbor mtDNAs of western Eurasian origin, whereas those inhabiting the Indo-Gangetic region and Central Asia present substantial proportions of lineages that can be allocated to three different genetic components of western Eurasian, eastern Eurasian, and south Asian origin. In addition to the overall composite picture of lineage clusters of different origin, we observed a number of deep-rooting lineages, whose relative clustering and coalescent ages suggest an autochthonous origin in the southwestern Asian corridor during the Pleistocene. The comparison with Y-chromosome data revealed a highly complex genetic and demographic history of the region, which includes sexually asymmetrical mating patterns, founder effects, and female-specific traces of the East African slave trade.

Introduction

The southwestern Asian corridor is a wide geographical area that extends from Anatolia and the trans-Caucasus area through the Iranian plateau to the Indo-Gangetic plains of Pakistan and northwestern India. This region is characterized by a patchwork of different physical-anthropology types with complex boundaries and gradients and by the coexistence of several language families (e.g., Indo-European, Turkic, and Sino-Tibetan) as well as relict linguistic outliers. The southwestern Asian corridor, located at the crossroads of major population expansions, was the first portion of Eurasia to be inhabited by the Homo sapiens sapiens population(s) that left Africa ∼60,000 years before the present (YBP) (Tishkoff et al. 1996; Watson et al. 1997; Quintana-Murci et al. 1999), and from this region modern humans migrated to the rest of the world. Although Paleolithic and Mesolithic people left their mark in the area, major prehistorical and historical events with possible genetic consequences occurred during the Neolithic period and later. Important agricultural developments occurred in the eastern horn of the Fertile Crescent ∼8,000 YBP, notably in Elam (southwestern Iran). The highly urban Elamite civilization had close contacts with Mesopotamians but exhibited an extensive differentiation from the rest of the Fertile Crescent populations, including a language that is thought to belong to the Dravidian family. It is hypothesized that the proto-Elamo-Dravidian language (McAlpin 1974, 1981), spoken by the Elamites in southwestern Iran, spread eastwards with the movement of farmers from this region to the Indus Valley and the Indian subcontinent (Cavalli-Sforza et al. 1994; Cavalli-Sforza 1996; Renfrew 1996). Starting ∼5,000 YBP, animal domestication, particularly the horse, gave the inhabitants of the Central Asian steppes the opportunity to expand geographically in different directions (Zvelebil 1980). These Central Asian nomads, probably from the Andronovo and Srubnaya cultures, migrated through Iran and Afghanistan, reaching Pakistan and India, and their arrival is contemporaneous with the decline of the strong agricultural South Asian civilizations, such as the Harappans. Most likely, their arrival on the Iranian plateau ∼4,000 YBP brought the Indo-Iranian branch of the Indo-European language family and, eventually, caused the replacement of Dravidian languages in Iran, Pakistan, and most of northern and central India (Renfrew 1987, 1996; Cavalli-Sforza 1996). Starting in the 3rd century b.c., the eastern part of the Eurasian steppes witnessed similar pastoral movements. By the time of the 3rd century a.d., Turkic-speaking peoples from the Altai region began to migrate westwards, replacing Indo-European languages in parts of Central Asia and, eventually, in what is now modern Turkey. Later, the Mongols also moved westward and, by the 13th century a.d., established their rule over a vast region, including parts of India, Pakistan, and Iran and reaching as far west as the Caucasus and Turkey (Cavalli-Sforza et al. 1994).

In the past decade, studies of mtDNA variation have provided a substantial contribution to the understanding of human origins and diffusion patterns. mtDNA surveys in worldwide populations have shown a continent-specific distribution of mtDNA lineages (Wallace et al. 1999; Ingman et al. 2000; Maca-Meyer et al. 2001; Herrnstadt et al. 2002; Mishmar et al. 2003). African populations are characterized by the oldest superhaplogroups, L1, L2, and L3 (Bandelt et al. 1995, 2001; Chen et al. 1995, 2000; Graven et al. 1995; Soodyall et al. 1996; Bandelt and Forster 1997; Watson et al. 1997; Alves-Silva et al. 2000; Torroni et al. 2001b; Salas et al. 2002), but it seems that only L3 radiated out of Africa, mainly in the form of haplogroups M and N, ∼60,000 YBP, giving rise to the extant Eurasian variation (Watson et al. 1997; Quintana-Murci et al. 1999; Wallace et al. 1999). Most western Eurasians are characterized by clades within haplogroup N (Torroni et al. 1996; Macaulay et al. 1999; Richards et al. 2000), whereas N and M contributed almost equally to the current eastern Eurasian mtDNA pool (Stoneking et al. 1990; Ballinger et al. 1992; Torroni et al. 1993; Horai et al. 1996; Kolman et al. 1996; Comas et al. 1998; Starikovskaya et al. 1998; Redd and Stoneking 1999; Schurr et al. 1999; Derbeneva et al. 2002; Kivisild et al. 2002; Yao et al. 2002).

Despite the major role played by the transect between the Near East and India in human origin and population dispersals, the extent and nature of mtDNA variation in the populations of the area are still not well resolved. In this context, mtDNA studies have focused on the western and eastern extremities of the southwestern Asian corridor, including the Near East/Caucasus region (Macaulay et al. 1999; Comas et al. 2000; Richards et al. 2000; Tambets et al. 2000; Nasidze and Stoneking 2001) and India (Mountain et al. 1995; Kivisild et al. 1999a, 1999b; Bamshad et al. 2001; Roychoudhury et al. 2001; Kivisild et al. 2003). In addition, Central Asian mtDNA variation is poorly characterized and is based only on HVS-I sequence data (Comas et al. 1998). Some populations of the region have been also analyzed for Y-chromosome variation, including Iranian (Quintana-Murci et al. 2001), Pakistani (Qamar et al. 2002), and, especially, Central Asian populations (Pérez-Lezaun et al. 1999; Karafet et al. 2001; Wells et al. 2001; Zerjal et al. 2002). To obtain a global mtDNA perspective of the entire region, we have now analyzed 910 mtDNAs from 23 different populations, located mainly in the southwestern Asian corridor but also, for comparison, in Central Asia. As a first step in the study, we performed high-resolution RFLP analysis and control-region sequencing of 208 mtDNAs, 108 from the western part of the corridor (Anatolia and the Caucasus), and 100 mtDNAs from its southeastern counterpart (southeast Pakistan). This allowed a clear-cut definition of the haplogroups (and their diagnostic markers) existing in the area. The phylogenetic information retrieved from this initial data set, together with previously published RFLP and HVS-I data, was then used to classify an extended collection of 702 newly obtained HVS-I sequences from the Iranian plateau, the Indus Valley, and Central Asia. The observed patterns of variation revealed different genetic contributions from western and eastern Eurasians and South Asians and evince complex demographic processes in some specific populations, including sexually asymmetrical mating patterns, founder effects, and differential migration patterns.

Material and Methods

Population Samples

The approximate location of the 23 populations from which the 910 mtDNAs were sampled is shown in figure 1. Each sample comprises unrelated healthy donors from whom appropriate informed consent was obtained. For the preliminary part of the study, 208 individuals from three different geographic regions were analyzed: 58 individuals from the Caucasus, 50 from Turkey and 100 from southeastern Pakistan. The three samples were heterogeneous; the sample from the Caucasus region was made up of three different ethnic groups: Georgians, Balkarians, and Chechens. The sample from Turkey was collected mainly in Konya (Anatolia). The Pakistani sample was collected in Karachi and comprised mainly Sindhis, who are a mix of tribes of different religions and ethnicities from the southeastern province of Sindh. The extended population sample included 702 individuals from 20 different populations living in the Iranian plateau, the Indus Valley, the Karakorum and Hindu Kush mountains, and Central Asia. Further details of the whole sample collection are reported in table 1 and in the work of Wells et al. (2001) and Qamar et al. (2002). The term “Makrani” refers to the so-called “Negroid Makrani” population living in the Makran coast of Baluchistan, distinct from the Makrani Baluch population, which is not considered in this study.

Figure 1.

Map of the southwestern and Central Asian corridor, showing the samples analyzed in the present study. Population codes are as reported in table 1. Boxed populations are those used for the initial step of the study (see the “Materials And Methods” section). Pie charts show the distribution of the main mtDNA lineage groups in the populations studied. Colored sections reflect the frequency of different haplogroup clusters, which group the western Eurasian (HV, pre-HV, N1, J-T, U-K, I, W, and X), the South Asian (M*, U2a-c, U9, R*, R1-R2, R5-R6, N1d, and HV2), the eastern Eurasian (M-CDGZ, A, B, F, and N9a) and the sub-Saharan African (L1, L2, and L3A) lineages.

Table 1.

Description of the Populations Included in the Study

Population	Code	n	Location	Language Family
Turkisha	AN	50	Anatolia, Turkey	Altaic
Caucasusa	CC	58	Georgia	Caucasian (north/south)
Persian	PE	42	Central and southern central Iran	Indo-European
Turkish	TI	40	Mostly eastern and western Azerbaijan	Altaic
Gilaki	GI	37	Northern Iran, southwestern Caspian Sea area	Indo-European
Mazandarian	MA	21	Northern Iran, southeastern Caspian Sea area	Indo-European
Kurdish	KI	20	Western Iran	Indo-European
Lur	LU	17	Southwestern Iran (Zagros Mountains)	Indo-European
Baluch	BA	39	Southwestern Pakistan, Baluchistan	Indo-European
Brahui	BR	38	Southwestern Pakistan, Baluchistan	Dravidian
Parsi	PA	44	Southeastern Pakistan, Karachi	Indo-European
Sindhi	SI	23	Southeastern Pakistan, Sindh	Indo-European
Pakistania	KA	100	Karachi, Southeastern Pakistan	Indo-European
Pathan	PT	44	North West Frontier Province and Balochistan	Indo-European
Makrani	MK	33	South Pakistan, Makran Coast	Indo-European
Hazara	HZ	23	North West Frontier Province and Balochistan	Indo-European
Hunza Burusho	HU	44	Northern Pakistan, Karakorum Mountains	Language isolate
Kalash	KL	44	North West Frontier Province	Indo-European
Gujarati	GJ	34	Northwestern India, Gujarat	Indo-European
Uzbek	UZ	42	Surkhandarya, Uzbekistan	Altaic
Turkmen	TK	41	Turkmenistan	Altaic
Kurdish	KT	32	Turkmenistan	Indo-European
Shugnan	SH	44	High Pamirs, Tajikistan	Indo-European

^aThese samples consist of mixed groups (see the “Materials and Methods” section for a detailed description) that were used in the initial step of the study.

mtDNA Analysis

High-resolution RFLP haplotypes were determined for the samples from the Caucasus region, Anatolia and Karachi. The entire mtDNA of each subject was PCR amplified using primer pairs and procedures previously described (Torroni et al. 1997). Each of the PCR segments was then digested with 14 restriction endonucleases (AluI, AvaII, BamHI, DdeI, HaeII, HaeIII, HhaI, HincII, HinfI, HpaI, MspI, MboI, RsaI, and TaqI). In addition, all mtDNAs were screened for the NlaIII sites at nucleotide positions (nps) 4216 and 4577. The presence/absence of the BstOI/BstNI site at np 13704, the AccI sites at nps 14465 and 15254, the BfaI site at np 4914, the XbaI site at np 7440, the MseI sites at nps 14766 and 16297, the MnlI site at np 10871, the MboII site at np 12703, and the HphI site at np 10237 were also analyzed in all the Pakistani-Karachi mtDNAs but only hierarchically in the mtDNAs from the Caucasus and Anatolia. Polymorphisms at nps 12308 and 11719 were also tested, the first by use of a mismatched primer that generates a HinfI site when the transition at 12308 is present (Torroni et al. 1996) and the second by use of a mismatched primer that generates a HaeIII site when the transition at 11719 is present (Saillard et al. 2000). The sequencing of the mtDNA control-region in the 208 individuals from the Caucasus region, Anatolia, and Karachi was performed as described elsewhere (Torroni et al. 2001a) and, in most cases, encompassed a large region (generally from np 16000 to nps 100–200). For the remaining 702 individuals, sequence data encompassed a shorter region (from np 16000 to np 16401), which includes the entire HVS-I, and variable positions were determined between nps 16024–16383, relative to the reference sequence (Anderson et al. 1981; Andrews et al. 1999). The published RFLP data (Macaulay et al. 1999; Quintana-Murci et al. 1999; Richards et al. 2000) and the new data obtained from the high resolution RFLP analyses of the 208 mtDNAs (see appendix A [online only]) were used to identify the RFLP and HVS-I sites (fig. 2), which are diagnostic of the main haplogroups and subhaplogroups within the mtDNA phylogeny. These markers were then selectively assayed, on the basis of the HVS-I information, in the remaining 702 mtDNAs by PCR amplification of the appropriate fragment and digestion with the informative restriction enzyme.

Figure 2.

Schematic phylogenetic tree of mtDNA haplogroups observed in the populations analyzed. The diagnostic mutations used to classify the whole data set are reported on the branches. Restriction enzyme sites are numbered from the first nucleotide of the recognition sequence. A plus sign (+) indicates the presence of a restriction site; a minus sign (−) indicates the absence of such a site. The restriction enzymes are designated by the following single-letter codes: a, AluI; b, AvaII; c, DdeI; e, HaeIII; f, HhaI; g, HinfI; h, HpaI; i, MspI; j, MboI; k, RsaI; l, TaqI; m, BamHI; n, HaeII; o, HincII; p, BstOI; q, NlaIII; r, BfaI; s, AccI; t, MboII; u, MseI; v, HphI; z, MnlI. Mutations in the HVS-I region are transitions unless the base change is specified explicitly. Boxes indicate novel information.

Data Analysis

Descriptive statistical indexes, the Tajima’s D (Tajima 1989) and Fu’s F_S (Fu 1997) neutrality tests, and the analysis of molecular variance (AMOVA) (Excoffier et al. 1992) were calculated using the Arlequin software, version 2.001 (Schneider et al. 2000). For the AMOVA analysis, we used the number of pairwise differences for the HVS-I sequence data and haplogroup frequencies for haplogroup data. We performed the AMOVA analyses either with all populations in a single group or divided into several groups, according to their geographic location or linguistic affiliation. For the geographic grouping, we divided populations into four regions: the Anatolian/Caucasus region (Anatolians and Caucasus populations), the Iranian plateau (Persians, Iranian Turks, Lurs, Iranian Kurds, Mazandarans, and Gilaks), the Indus Valley (Baluchi, Brahui, Parsi, Sindhi, Pakistani-Karachi, Pathans, Makrani, Hazara, and Gujarat) and Central Asia (Uzbeks, Turkmen, Kurds from Turkmenistan, Shugnan, Hunza Burusho, and Kalash). For the linguistic division, we grouped populations according to their linguistic affiliation: Indo-Europeans (Persians, Lurs, Iranian Kurds, Mazandarans, Gilaks, Baluchi, Parsi, Sindhi, Pakistani-Karachi, Pathans, Makrani, Hazara, Shugnan, Kalash, and Gujarat), Altaic (Anatolian, Iranian Turks, Turkmen, and Uzbek), Dravidian (Brahui), Caucasian (Caucasus), and language isolates (Burusho). The population genetic structure was also explored through the spatial analysis of molecular variance (SAMOVA) approach (Dupanloup et al. 2002), which defines groups of populations that are geographically homogeneous and maximally differentiated from each other. This method is based on a simulated annealing procedure that aims at maximizing the proportion of total genetic variance due to differences between groups of populations without any a priori definition of groups of populations that is based on geographic or linguistic features. The SAMOVA analyses were based on HVS-I sequence data and were done using the SAMOVA 1.0 software.

Median-joining networks (Bandelt et al. 1995, 1999) were constructed by hand and confirmed by the Network program (A. Röhl; Shareware Phylogenetic Network Software Web site). For network construction of some specific lineages, sequence data from other populations were taken from the literature. From the Anatolia/Caucasus region, we included Armenians (AM), Azerbaijanis (AZ), Turks (TR), and Kurds (KR) from Richards et al. (2000); Turks (TC) from Calafell et al. (1996); Turks (TT) from Tambets et al. (2000); and Kurds (KC) from Comas et al. (2000). From the Middle East/Arabian Peninsula, we included Iraqis (IQ), Syrians (SY), Yemenites (YM), Palestinians (PL), and Druze (DZ) from Richards et al. (2000); individuals from Dubai (DB) from A.T. (unpublished data); and Egyptians (EG) from Krings et al. (1999). From Pakistan/India, we included Pakistanis (PK) and Indians from Andhra Pradesh (AP), Gujarat (GK), Haryana (HY), Kashmir (KS), Maharashtra (MH), Punjab (PN), Rajasthan (RJ), Uttar Pradesh (UP), and Tamil Nadu (TN) from Kivisild et al. (1999a); and Indians (IN) from Mountain et al. (1995). From Central Asia, we included Kirghiz (KG), Uighur (UG), and Kazakh (KZ) samples from Comas et al. (1998). From western Eurasia, we included Basques (BS), Sicilians (SC), Bulgarians (BL), and Italians from Tuscany (TS) from Richards et al. (2000); Russians (RS) from Malyarchuk et al. (2002); Mansi (MN) from Derbeneva et al. (2002); and Sardinians (SD) from Di Rienzo and Wilson (1991). We also included Chinese (CH) from Yao et al. (2002). The time to the most recent common ancestor of some clades and their SEs were calculated by means of the estimator ρ, the averaged distance to a specified founder haplotype, and were determined as described by Forster et al. (1996) and Saillard et al. (2000). Time estimates were also calculated, using the Network program. Principal-components (PC) analyses were performed using SPSS version 10.0.7 software, with basal mtDNA haplogroup frequencies as input vectors. Admixture proportions (mY) and their SEs were calculated, using information from all haplogroups, by means of the program Admix 2.0 (Dupanloup and Bertorelle 2001), on the basis of 1,000 bootstraps. The parental populations used for the analysis were Iranian populations and Gujarati for the Parsi population, and Pakistani populations (excluding the Makrani) and a geographically dispersed set of sub-Saharan African samples (Krings et al. 1999; Brakez et al. 2001; Brehm et al. 2002, Salas et al. 2002) for the Makrani population.

Results

The Topology of the Southwest and Central Asian mtDNA Tree

The complete high-resolution RFLP haplotypes and HVS-I sequence data of the 208 individuals from the Caucasus region, Anatolia, and southeastern Pakistan and the detailed haplogroup classification and HVS-I sequence data of the extended database of 702 individuals are reported in the online-only material.

The phylogenetic relationships of the 51 different named haplogroups observed in the 910 samples, along with the diagnostic sites used for the mtDNA haplogroup classification, are shown in figure 2. The vast majority of the mtDNAs clustered into macrohaplogroups M, N, and R, but a limited number were found to belong to the sub-Saharan haplogroups L1a, L2a, L3b, and L3d. Five haplogroups, N1d, HV2, U9, R5, and R6, are defined here for the first time, whereas others (U2a, U2b, and U2c) represent newly identified subclades. Moreover, for some previously known haplogroups (R2 and U8b), we detected diagnostic coding-region markers that allow a better definition of the haplogroup topology within the tree.

Macrohaplogroup N in southwestern and Central Asia is partitioned into several branches: N1 (which also encompasses haplogroup I), N9a, A, W, X, and R. Within the N trunk, the new haplogroup N1d stems from the node of N1 and is defined by three characteristic RFLP sites (−951MboI, −5003DdeI, −8616MboI) and two HVS-I transitions (nps 16301 and 16356). The internal topology of superhaplogroup R has also been improved. The novel lineage R5 is defined by −8592MboI and transitions at nps 16266 and 16304, whereas the new other haplogroup, R6, is characterized by −12282AluI and transitions at nps 16129 and 16362. Moreover, the R2 mtDNAs, previously recognizable only by the HVS-I transition at np 16071, are now identifiable through the diagnostic coding-region motif +4216NlaIII, +4769AluI, −14304AluI. It is worth noting that +4216NlaIII is also one defining mutation of the lineage-cluster J-T (fig. 2). However, the comparison of entire mtDNA sequences belonging to both R2 and J-T (A.T., unpublished data) indicates that +4216NlaIII has indeed occurred independently on the two branches of the phylogeny. An improvement of the classification within HV was also obtained. A haplogroup, named HV2, was found to bear the HVS-I transition at np 16217 and most likely is also characterized by +9336RsaI, since 16 of the 20 mtDNAs with the HVS-I 16217 mutation harbor this coding region site. This lineage corresponds to an internal node of HV that Tambets et al. (2000) tentatively identified as P*. The improvement of the haplogroup U subclassification was even more extensive. This major western Eurasian haplogroup is also found in the Middle East and India and, at lower frequencies, in northern and eastern Africa, but the frequency distributions of its subclades appear to differ considerably among geographical regions (Kivisild et al. 1999a, 2003; Macaulay et al. 1999; Richards et al. 2000). Subclade U2 (characterized by the rather variable HVS-I transition at np 16051) was previously subdivided into two branches, the “European” U2e characterized by a further HVS-I transversion at np 16129, and the “Indian” U2i lacking such a transversion (Kivisild et al. 1999a). We show that U2e also harbors the distinguishing RFLP motif +13730HinfI, +15907RsaI, and U2i is indeed made up of three clusters, here termed “U2a,” “U2b,” and “U2c.” U2a is characterized by the rare and stable HVS-I transversion 16206C, U2b is defined by the diagnostic site −15047HaeIII, and U2c harbors the RFLP motif +5789TaqI, +8020MboI/+8022TaqI, −15060MboI (fig. 2). A subset of U that was already known but is now better defined is U8b. Finnilä et al. (2001) observed that, on the basis of the shared transition at np 9698, haplogroup U8 formed a sister clade with haplogroup K. Our data reveal that at least a subset of U8, here termed “U8b,” is also characterized by −9052HaeII, which is indeed also the diagnostic marker of haplogroup K. This observation strengthens the sister haplogroup status of U8b and K. Finally, our data reveal the presence of a new—and rare—previously undefined subgroup of U, termed “U9,” that is characterized by −6383HaeIII. This haplogroup does not correspond to the U9 of Herrnstadt et al. (2002), which, in reality, corresponds to a subset of the previously defined U3.

The comparison of the RFLP and HVS-I data obtained from our data set identified some pitfalls when classifying the internal lineages within some haplogroups (e.g., J and M) on the basis of the HVS-I sequence data alone. Thus, we classified all our J mtDNAs according only to their differential RFLP status (fig. 2), and, since an accurate RFLP classification of the South Asian branches of haplogroup M remains to be defined, we adopted a conservative classification and merged all South Asian M mtDNAs into M*.

Haplogroup Profile Distribution

The haplogroup repertoire present in the study populations is shaped mainly by the presence of lineages that can be attributed to eastern Eurasia, South Asia, and western Eurasia (fig. 1; table 2). Sub-Saharan African lineages, represented by haplogroups L1, L2, and L3A and their internal derivatives, are virtually absent from all populations analyzed except the Makrani from southern Pakistan, among whom they reach high frequencies (39%).

Table 2.

mtDNA Haplogroup and Subcluster Frequencies for the 23 Study Populations^[Note]

	Estimated Frequency in Population(%)
HaplogrouporSubcluster	AN	CC	PE	TI	GI	MA	KI	LU	BA	BR	PA	SI	KA	PT	MK	HZ	HU	KL	GJ	UZ	TK	KT	SH
Sample Size	50	58	42	40	37	21	20	17	39	38	44	23	100	44	33	23	44	44	34	42	41	32	44
L1a	0	0	0	0	0	0	0	0	0	0	0	0	0	0	3.0	0	0	0	0	0	0	0	0
L2a	0	0	0	0	0	4.8	0	0	0	0	0	0	1.0	0	9.1	0	0	0	0	0	0	0	0
L3b	0	0	0	0	0	0	0	0	0	0	0	0	0	0	9.1	0	0	0	0	0	0	0	0
L3d	0	0	0	0	0	0	0	0	2.6	2.6	0	0	0	0	18.2	0	0	0	0	0	0	0	0
M*	0	1.8	4.8	0	2.7	0	0	0	33.3	21.1	54.5	30.4	47.0	29.5	9.1	13.0	22.7	0	44.1	11.9	4.9	0	0
M-C	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2.3	0	0	2.4	7.3	9.4	18.2
M-D	2.0	0	0	0	2.7	0	0	0	0	5.3	0	0	0	0	0	0	0	0	0	9.5	22.0	0	0
M-G	2.0	0	0	0	0	0	0	0	0	0	0	0	1.0	0	0	4.3	0	0	0	2.4	0	0	0
M-Z	0	0	0	0	0	0	0	0	0	0	0	0	1.0	0	0	13.0	0	0	0	0	0	0	0
N*	0	0	2.4	0	0	0	0	0	0	0	0	0	0	4.5	0	0	2.3	0	2.9	0	0	0	0
N1	0	0	0	0	0	0	0	0	2.6	0	0	0	0	0	0	0	0	0	0	0	0	0	0
N1a	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2.4	0	0
N1b	2.0	6.9	2.4	0	2.7	9.5	0	0	0	0	0	0	0	2.3	0	0	0	0	0	0	0	0	0
N1c	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	3.1	0
N1d	0	0	0	0	0	0	0	0	2.6	2.6	0	0	3.0	0	3.0	0	0	0	0	0	0	0	0
N9a	0	0	0	2.5	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	7.1	0	0	0
I	2.0	1.8	2.4	5.0	0	4.8	5.0	0	0	0	0	8.7	0	2.3	0	0	4.5	0	0	0	0	3.1	2.3
A	4.0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2.3	0	0	7.1	2.4	3.1	2.3
W	0	5.2	2.4	2.5	0	9.5	10.0	0	0	0	0	17.4	1.0	4.5	0	0	2.3	0	8.8	2.4	0	0	4.5
X	6.0	8.6	0	0	0	0	0	0	0	0	0	0	1.0	0	0	0	2.3	0	0	0	2.4	6.3	2.3
R*	0	0	0	0	0	0	0	0	0	0	0	0	2.0	2.3	6.1	8.7	0	0	8.8	0	0	0	0
R1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	9.4	0
R2	0	1.8	0	0	0	9.5	0	0	7.7	7.9	0	0	1.0	0	3.0	0	0	0	0	0	0	0	0
R5	0	0	2.4	0	0	0	0	0	0	0	0	0	2.0	2.3	0	0	2.3	0	0	2.4	0	0	0
R6	0	0	0	0	0	0	0	0	0	0	0	0	3.0	0	0	0	0	0	0	0	0	0	0
B	0	0	0	0	0	0	0	5.9	0	0	0	0	0	0	0	8.7	2.3	0	0	0	2.4	6.3	0
F	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	8.7	0	0	0	2.4	2.4	0	0
(pre-HV)1	4.0	3.4	0	0	0	0	0	11.8	0	0	0	0	2.0	6.8	0	0	0	22.7	0	0	4.9	0	0
HV*	6.0	0	16.7	5.0	13.5	9.5	10.0	5.9	0	5.3	2.3	0	4.0	2.3	0	4.3	0	4.5	0	4.8	2.4	0	4.5
HV1	0	0	0	2.5	2.7	0	5.0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
HV2	0	0	2.4	0	8.1	0	5.0	5.9	10.3	0	9.1	4.3	0	0	6.1	0	0	0	0	2.4	2.4	0	0
H	26.0	22.4	14.3	30.0	13.5	14.3	10.0	17.6	20.5	26.3	2.3	8.7	12.0	4.5	3.0	13.0	6.8	4.5	5.9	21.4	22.0	12.5	29.5
V	0	0	0	2.5	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
U*	0	0	0	2.5	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2.3
U1	6.0	5.2	0	7.5	2.7	0	0	5.9	0	0	4.5	0	0	0	3.0	4.3	6.8	0	0	0	0	0	0
U2*	0	0	2.4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
U2a	0	0	0	0	0	0	0	0	2.6	5.3	0	4.3	5.0	6.8	3.0	0	2.3	0	8.8	0	0	0	0
U2b	0	0	0	0	0	0	0	0	0	0	0	4.3	4.0	6.8	3.0	0	9.1	0	5.9	0	0	0	0
U2c	0	0	0	0	0	0	0	0	0	0	0	8.7	2.0	2.3	3.0	0	0	0	2.9	0	0	0	0
U2e	2.0	3.4	2.4	2.5	0	0	0	0	0	0	0	0	0	0	0	0	0	15.9	0	0	2.4	6.3	6.8
U3	2.0	3.4	4.8	2.5	2.7	0	0	17.6	0	0	0	0	0	0	0	0	2.3	0	0	2.4	0	3.1	0
U4	2.0	3.4	2.4	0	0	0	0	5.9	2.6	0	13.6	0	0	0	0	8.7	4.5	34.1	0	4.8	2.4	0	4.5
U5	8.0	1.8	2.4	2.5	0	0	5.0	0	2.6	0	0	4.3	0	2.3	0	8.7	2.3	0	0	0	0	0	4.5
U7	2.0	0	2.4	2.5	10.8	0	20.0	5.9	2.6	10.5	2.3	8.7	5.0	0	3.0	4.3	6.8	2.3	8.8	4.8	0	0	0
U8b	2.0	0	0	0	2.7	0	10.0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2.3
U9	0	0	0	0	0	0	0	0	0	0	0	0	1.0	2.3	9.1	0	0	0	0	0	0	0	0
K	6.0	8.6	7.1	5.0	2.7	0	10.0	5.9	2.6	0	0	0	0	2.3	0	0	4.5	0	0	0	0	12.5	9.1
J1	6.0	8.6	16.7	7.5	16.2	19.0	10.0	5.9	5.1	7.9	0	0	1.0	6.8	6.1	0	6.8	2.3	2.9	7.1	9.8	3.1	4.5
J2	2.0	0	0	2.5	0	4.8	0	0	2.6	0	0	0	0	0	0	0	0	9.1	0	0	0	3.1	0
T*	8.0	10.3	7.1	12.5	13.5	9.5	0	5.9	0	0	4.5	0	1.0	4.5	0	0	2.3	4.5	0	4.8	4.9	18.8	2.3
T1	0	3.4	2.4	2.5	2.7	4.8	0	0	0	5.3	6.8	0	0	4.5	0	0	2.3	0	0	0	2.4	0	0

Note.— Population codes are as in table 1.

The eastern Eurasian component is represented by haplogroups A, B, F, and N9a, all of which belong to the major N trunk, and the East Asian branches of macrohaplogroup M, such as the C, D, G, and Z haplogroups. The latter lineages are particularly widespread among northern and East Asians and, to a lesser extent, Central Asians (Torroni et al. 1993, 1994a, 1994b; Kivisild et al. 2002; Yao et al. 2002; Kong et al. 2003). The eastern Eurasian lineage cluster shows, with some exceptions, a decreasing gradient of frequencies towards the west (fig. 1; table 2). The highest frequencies of these branches were found among the Central Asian populations, reaching their maximum in the Turkmen and Uzbeks (37% and 31%, respectively). Interestingly, Kurds from Turkmenistan showed the lowest frequencies of eastern Eurasian lineages (9%) in Central Asia, in sharp contrast to the local Turkmen population. These eastern Eurasian–specific lineages were absent—or at very low frequencies—in populations from the Anatolian/Caucasus region, the Iranian plateau, and the Indus Valley, with one exception: the Hazaras from northern Pakistan, among whom they reach 35%.

The South Asian influence is mainly represented by the nodal type of macrohaplogroup M (M*) and the three sister clades U2a, U2b, and U2c. The M* haplogroup is absent or infrequent in all the populations west of the Indus Valley and is present at low frequencies in our Central Asian populations (<12%). Conversely, it is present at high frequencies (30%–55%) in populations living in the southern coasts of Pakistan and northwestern India. The three sister clades U2a, U2b, and U2c show a similar geographic pattern to that of haplogroup M*, although their distribution is somewhat more restricted to the Indo-Pakistani region. Also, N1d and HV2 and some lineages within paragroup R* are at higher frequencies in populations located east of the Iranian plateau, and this will be discussed in more detail below.

The proportion of western Eurasian lineages (HV, pre-HV, N1, J-T, U-K, I, W, and X) showed the opposite pattern of that exhibited by eastern Eurasian lineages (fig. 1; table 2). They exhibit their highest frequencies in the Anatolian/Caucasus and Iranian regions and their prevalence decreases eastwards. Despite this decreasing frequency cline towards the East, they are still present at relatively high frequencies in the Indus Valley and Central Asia. Indeed, the western Eurasian presence in the Kalash population reaches a frequency of 100%, the most prevalent haplogroups being U4, (pre-HV)1, U2e, and J2.

Phylogeography of Specific Haplogroups

The phylogeography of several haplogroups suggests that they are either autochthonous to the southwestern Asian corridor or that at least they underwent a major expansion in this region. Among these lineages, haplogroup U7 presents the most widespread distribution. U7 is virtually absent in western and eastern European populations and is present at low frequencies (2%–4%) in the Near East, the Caucasus region, Central Asia, and the Indian subcontinent (Kivisild et al. 1999a, 2003; Macaulay et al. 1999; Richards et al. 2000; Tambets et al. 2000; Malyarchuk and Derenko 2001; Malyarchuk et al. 2002). Our data show that this haplogroup is present in most of the populations linking the Near East with Central and South Asia, reaching its highest frequencies in some Iranian and Indus Valley populations (table 2), in agreement with recent data reporting a frequency of 9% in a composite Iranian sample (Kivisild et al. 2003). Figure 3 shows the median-joining network for this haplogroup. The topology of the network shows that this haplogroup is divided into two major well-defined star-like subclades separated by a transition at np 16309. The time-depth calculated for paragroup U7* (without 16309) is 35,100 ± 8,500 years, whereas that for U7a (with 16309) is 22,500 ± 5,400 years. These coalescence times support the idea that the 16318T mutation is indeed the ancestral feature of U7. The overall coalescence time calculated for U7 is 38,200 ± 13,900 years.

Figure 3.

Network of the U7 lineage. Circle areas are proportional to haplotype frequency. Population codes are as reported in table 1 and in the “Materials And Methods” section. Mutated sites (−16,000) are indicated along the branches. The number sign (#) indicates the assumed root.

The phylogeography of haplogroups HV2 and R2 resembles that of U7 but has a more restricted geographic distribution. Both haplogroups are concentrated in southern Pakistan and India, with some overflow into adjacent areas, including the Near East/Caucasus region, the Iranian plateau, the Arabian Peninsula, and Central Asia, where most of the derived types are observed (fig. 4a and 4b). The coalescence times were estimated at 27,700 ± 9600 years for HV2 and 31,200 ± 8200 years for R2.

Figure 4.

Networks of (a) HV2 and (b) R2 lineages

The distribution of the three sister clades within haplogroup U2 (U2a, U2b, and U2c) is essentially restricted to the Indo-Pakistani regions (fig. 5a–c). They have not been observed in Europe and the Near East and, according to our data, they are absent in the Iranian plateau and Central Asian populations. They are, however, common in populations from Pakistan and India. The estimated coalescence times for these haplogroups are: 45,700 ± 14,400 years for U2a, 35,900 ± 9000 years for U2b, and 45,200 ± 10,400 years for U2c. The R5 lineage showed a similar distribution to the U2 subclades (fig. 5d), but its root types are more concentrated in the Indus Valley region, with the derivatives in central and southern India. The estimated time depth of this lineage is 51,800 ± 13,800 years.

Figure 5.

Networks of (a) U2a, (b) U2b, (c) U2c, and (d) R5 lineages

Finally, three small haplogroups (R6, N1d, and U9) have been observed so far only in south Pakistan. R6 was found in three individuals from the mixed sample from Karachi; N1d in one Baluchi, one Brahui, one Makrani, and three individuals from Karachi; and U9 in three Makrani, one Pathan, and one individual from Karachi.

Population Diversity and Demographic Regimes

HVS-I sequences have also been used to gain information on the internal population diversity (table 3). Most populations showed similar sequence diversity values, with the Kalash showing the lowest (0.830) and the Indian Gujarati the highest (0.998). The low diversity exhibited by the Kalash population is also evident in the low mean number of pairwise differences (3.857). This is the lowest value of all the populations studied, which otherwise ranged from 4.399 in the Baluchi and the Caucasus populations to 6.633 in the Makrani. As shown in table 3, most populations yielded significantly negative values for both Tajima’s D and Fu’s F_S neutrality tests. The only exceptions were the Mazandarians, the Kurds from Turkmenistan, and the Kalash. The former two groups exhibited significantly negative Fu’s F_S values and unimodal mismatch distributions (not shown) but the Tajima’s D statistic was not significantly different from 0. This contrasting pattern may be the result of mutation rate heterogeneity along the HVS-I region; this effect has been shown to confound the signature of population expansion in Tajima’s test, leading to higher D values (Aris-Brosou and Excoffier 1996). For the Kalash population, both neutrality tests gave nonsignificantly negative values (table 3), and the mismatch distribution was unequivocally multimodal (data not shown).

Table 3.

Diversity Indices and Neutrality Tests for the Study Populations

Population	Code	H (SE)a	K (K/n)b	Sc	Pi (SE)d	Tajima’s De	Fu’s FSe
Turkish	AN	.992 (.008)	46 (92)	62	5.758 (2.803)	−2.04	−25.34
Caucasus	CC	.992 (.007)	50 (86)	65	4.399 (2.987)	−1.93	−25.23
Persian	PE	.992 (.008)	38 (90)	54	5.787 (2.825)	−1.91	−25.32
Turkish	TI	.963 (.023)	32 (80)	49	4.717 (2.358)	−2.10	−25.61
Gilaki	GI	.990 (.009)	31 (84)	44	5.363 (2.646)	−1.76	−25.33
Mazandarian	MA	.995 (.017)	20 (95)	36	6.310 (3.117)	−1.45 (P = .058)	−14.74
Kurdish	KI	.995 (.018)	19 (95)	36	6.132 (3.044)	−1.57	−13.68
Lur	LU	.978 (.031)	15 (88)	34	5.515 (2.789)	−1.85	−8.24
Baluch	BA	.974 (.012)	26 (67)	39	4.399 (2.219)	−1.84	−18.02
Brahui	BR	.952 (.019)	22 (58)	39	4.834 (2.412)	−1.70	−10.14
Parsis	PA	.943 (.017)	20 (45)	36	4.660 (2.328)	−1.50	−6.50
Sindhi	SI	.992 (.015)	21 (91)	35	5.573 (2.778)	−1.58	−15.95
Pakistani-Karachi	KA	.992 (.003)	77 (77)	89	5.572 (2.699)	−2.21	−25.29
Pathan	PT	.993 (.007)	38 (86)	54	5.562 (2.724)	−1.95	−25.38
Makrani	MK	.975 (.015)	24 (73)	53	6.633 (3.213)	−1.81	−11.81
Hazara	HZ	.992 (.015)	21 (91)	39	6.205 (3.060)	−1.60	−14.74
Hunza Burusho	HU	.980 (.010)	32 (73)	59	6.462 (3.117)	−1.86	−20.92
Kalash	KL	.830 (.032)	11 (25)	17	3.857 (1.975)	−.04 (P = .559)	−.25 (P = .479)
Gujarati	GJ	.998 (.008)	33 (97)	53	6.298 (3.063)	−1.88	−25.22
Uzbek	UZ	.991 (.009)	37 (88)	54	5.403 (2.656)	−2.02	−25.43
Turkmen	TK	.989 (.009)	35 (85)	59	5.989 (2.915)	−2.02	−25.28
Kurdish	KT	.972 (.014)	21 (66)	41	6.494 (3.154)	−1.32 (P = .081)	−7.81
Shugnan	SH	.985 (.008)	33 (75)	49	5.958 (2.897)	−1.65	−24.23

^aSequence diversity (H) and standard error (SE).

^bNumber of different haplotypes and percentage of sample size in parentheses.

^cNumber of segregating sites.

^dAverage number of pairwise differences (Pi) with standard error (SE).

^eAll P values are < .05 (for Tajima's D) and <.02 (for Fu's F_S), except where noted.

Population Relationships

The basal mtDNA haplogroup frequencies of the 23 populations were used as input vectors to perform a PC analysis. Figure 6 shows the PC plot for the first two principal components, which account for 43% and 12% of the total variation respectively. Leaving aside the two outliers, the Kalash and the Makrani, geographic grouping of populations are apparent in the diagram. The first PC (PC1) mainly reveals a west-to-east cline by separating a group of closely related populations from the Iranian plateau from those inhabiting the Indus Valley and northwest India. The Central Asian Uzbeks, Turkmen, and Shugnan tend to be closer to populations from the Anatolian/Caucasus/Iranian regions, rather than to Indus Valley populations, as a consequence of the high prevalence of western Eurasian lineages observed in these populations. The Hazara from Pakistan shows an intermediate position between populations from the Indus Valley and those from Central Asia. PC2 essentially displays the outlier genetic position of the Makrani and the Kalash populations, who are separated from the rest of populations of the Iranian plateau and the Indus Valley.

Figure 6.

PC plot based on haplogroup frequencies for the 23 population samples (population codes are as in table 1).

Population Genetic Structure: AMOVA and SAMOVA Analyses

We investigated how the proportion of variance, based on haplogroup (main lineages) and haplotype (HVS-I sequences) frequencies, was distributed in a hierarchical mode by an AMOVA analysis (Excoffier et al. 1992). When the 23 populations were treated as a single group, populations turned out to show overall differentiation: the F_ST value for the haplogroup data was 0.067 (P<.001) and the φ_ST for the sequence data was 0.032 (P<.001). The fraction of genetic variance due to differences among linguistic groups (see the “Materials and Methods” section) was not statistically different from 0, independently of the genetic system used (i.e., haplogroup or sequence data), indicating that genetic variance within any population or among populations within groups was larger than that between groups and, therefore, that the division by linguistic affiliation is not reflected in mtDNA variation. Finally, when populations were regrouped into four geographic groups (see the “Materials and Methods” section), a small but significant differentiation among groups was detected (F_CT=0.043 and P<.001 for haplogroup data and φ_CT=0.016 and P<.001 for HVS-I data).

To investigate in greater detail the genetic structure of the populations and the amount of genetic variation due to differences among population groups, we applied the SAMOVA algorithm (Dupanloup et al. 2002), on the basis of HVS-I data, searching for two, three, and four groups. The inclusion of the Kalash population, which is among the most differentiated (table 3; fig. 6), gave inconsistent results (data not shown), so this population was excluded from further analyses. A search for two significantly differentiated population clusters revealed one group consisting of all populations from the Anatolian/Caucasus region and the Iranian plateau (including the Kurds from Turkmenistan), and a second group made up of populations from the Indus Valley and Central Asia (F_CT=0.021; P<.001). In the three-group search, the previous two remained unchanged, and a third group, represented by the Hazara, emerged from the analysis (F_CT=0.021; P<.001). Finally, the search for four groups revealed the Makrani of southern Pakistan as the fourth most differentiated group (F_CT=0.022; P<.001).

Discussion

This study provides the first comprehensive survey of mtDNA variation in a part of the world that was among the first regions to be inhabited after the “out of Africa” exit, and has subsequently experienced numerous waves of migration during the last 50,000 years. We now discuss the events, both ancient and modern, that are likely to have led to the current mtDNA distribution, and compare the mtDNA data with that from other loci, particularly the Y chromosome.

The mtDNA Landscape of the Southwestern Asian Corridor

A simple pattern underlies the mtDNA variation in this region: a west-to-east divide with a sharp boundary. Populations located west of the Indus basin, including those from Iran, Anatolia and the Caucasus, exhibit a common mtDNA lineage composition, consisting mainly of western Eurasian lineages, with a very limited contribution from South Asia and eastern Eurasia (fig. 1). Indeed, the different Iranian populations show a striking degree of homogeneity. This is revealed not only by the nonsignificant F_ST values and the PC plot (fig. 6) but also by the SAMOVA results, in which a significant genetic barrier separates populations west of Pakistan from those east and north of the Indus Valley (results not shown). These observations suggest either a common origin of modern Iranian populations and/or extensive levels of gene flow amongst them. There is a virtual absence of both common South Asian lineages (M*, U2a, U2b, and U2c) and the more autochthonous U9, R*, R2, R5, R6, N1d, and HV2 lineages in the Anatolian/Caucasus region and Iranian plateau, whereas these lineages make up >50% of the haplogroup profile in the adjacent Indus Valley. Most of these lineages appear to be restricted to the eastern part of the corridor (fig. 1). Whereas geographical clustering and the coalescent age of U7 (∼38,000 YBP; see table 2 and fig. 3) suggest that it is the most widespread local lineage of Pleistocene origin connecting the western and eastern extremes of the corridor, the Indus Valley and India show signals of an in situ differentiation of deep-rooting lineages (HV2, R2, R5, U2a, U2b, and U2c), the distribution of which appears to be limited to this region (figs. 4 and 5). All these lineages have high time depths (28,000–52,000 YBP), similar to haplogroup M* (32,000–53,000) in the region (Kivisild et al. 1999b; Quintana-Murci et al. 1999; Roychoudhury et al. 2001). Notably, haplogroup M* has also not penetrated west of the Indus Valley, although it is present at high frequencies in south Pakistani and Indian populations. Thus, the distribution and ages of these lineages suggest that they are the legacy of the first inhabitants of the southwestern Asian region who underwent important expansions during the Paleolithic period. It is interesting that the newly identified haplogroup U9, found in the Indus Valley, has also been observed in Ethiopia (A.T., unpublished data), supporting the link between East Africa and the southwestern and southern coasts of Asia (Kivisild et al. 1999a; Quintana-Murci et al. 1999). The absence of these lineages west of Pakistan may be due either to limited gene flow from the Indus basin or to important demographic expansions in the Fertile Crescent (including its eastern lobe, represented by present-day Iran), associated with a substantial increase in frequency and diversity of western Eurasian lineages. Geographical features such as the Dash-e Kavir and Dasht-e Lut deserts in Iran could have acted as significant barriers to gene flow, and this is consistent with Y-chromosomal data (e.g., the distribution of lineage R-M17) from these regions (Quintana-Murci et al. 2001; Wells et al. 2001; Qamar et al. 2002).

Gene flow from the Fertile Crescent to India has, however, been more common than that from east to west (fig. 1). Eastern populations within the corridor mostly exhibit a rich variety of west Eurasian lineages at high frequencies (26%–57%), with a gradient towards the Indian subcontinent, with lower frequencies in caste (<10%) and tribal groups (<1%) (Kivisild et al. 1999a; 2003; Bamshad et al. 2001). The substantial western Eurasian presence in the Indus Valley and northwestern India may have been the result of repeated gene flow received from further west at different periods, including the first Paleolithic arrivals to the corridor region from the Middle East and subsequent dispersals associated with Neolithic urban civilizations, such as Mesopotamians and Elamites, who may have carried farming towards the eastern part of the corridor. The exact mode and tempo in which the different western Eurasian lineages reached the Indo-Gangetic plains remains to be elucidated. However, it appears that J, T1, and U3, which have been proposed as the main marker haplogroups for the Neolithic diffusion of agriculture from the Middle East to the West (Richards et al. 2000, 2002), did not play an equivalent role in the diffusion of farming toward the East. The eastern Eurasian contribution to the west, in contrast, is negligible (fig. 1), in agreement with HVS-I sequence data in Turkish populations (Calafell et al. 1996; Comas et al. 1996, 1998). This pattern may seem surprising in view of the historically documented repeated waves of Altaic-speaking nomads (e.g., Turks, Huns, and Mongols) starting in the 3rd century a.d., who traveled from east to west, imposing Altaic languages in some western regions (e.g., Anatolia and Azerbaijan), probably through an elite-dominance process. In this context, it is interesting that five of the seven individuals belonging to eastern Eurasian lineages west of the Indus Valley are Turkic-speaking. The east-to-west differences in the genetic contribution of the eastern invaders along the corridor may be due to the existing population densities in these regions at the time of arrival. The genetic contribution of the newcomers was strong in the sparsely populated arid lands of eastern Central Asia. Conversely, the eastern influence in western territories was much lower, since the invading eastern nomads probably found higher population densities. Our results reinforce recent Y-chromosome data from Central Asia, which show that the paternal genetic contribution of the eastern invaders is barely detectable west of Uzbekistan (Zerjal et al. 2002).

The Effects of Admixture and Drift: Demographic Events in Central Asia

Central Asians exhibit high frequencies of East Asian lineages, which are otherwise virtually absent in populations from the Indo-Gangetic region and westwards, concomitantly with a high prevalence of lineages of western Eurasian origin (fig. 1). Two explanations have been put forward: Central Asians could represent an early incubator of Eurasian variation, or their current genetic diversity could result from later admixture between western and eastern Eurasian populations. Y-chromosome data have been interpreted as indicating that Central Asian populations are amongst the oldest on the continent and were the source of at least three major migration events (Wells et al. 2001) but were also a receiver of migrations (Zerjal et al. 2002). mtDNA studies (Comas et al. 1998) based on HVS-I variation in four populations of Central Asia found that they contained both European and East Asian motifs. This was interpreted as evidence for admixture between Europeans and East Asians, a conclusion that is substantiated by our more thorough analysis. Indeed, if Central Asia had been the source of modern Eurasian diversity, one would expect to observe (i) substantial overlap between present-day western and eastern Eurasian haplogroups and (ii) extensive divergence between the HVS-I types found in Central Asia and those observed in western and eastern Eurasia. This is not the case. Our data, which take into consideration coding-region information and provide a more clear-cut phylogeography, show a major demarcation in the Eurasian landscape between European and East Asian mtDNA lineages within both the R and N branches, and with M playing virtually no role in western Eurasia. Moreover, most Central Asian HVS-I types match sequences that are observed today in either western or eastern Eurasians, suggesting recent arrival in Central Asia.

The complexity of the peopling of the region is well illustrated by the Kalash population from the Hindu Kush valleys, where western Eurasian mtDNAs reach fixation with no detectable East or South Asian lineages (fig. 1 and table 2). Their outlying genetic position is seen in all analyses (table 3 and fig. 6). Moreover, although this population is composed of western Eurasian lineages, the most prevalent (i.e., U4, (pre-HV)1, U2e, and J2) are rare or absent in the surrounding populations and usually characterize populations from Eastern Europe, the Middle East, and the Caucasus (Macaulay et al. 1999; Richards et al. 2000; Tambets et al. 2000). Also, the internal HVS-I sequence diversity within each of these haplogroups was surprisingly low: 12 of 15 samples belonging to U4 were associated with the motif 16134–16356, all (pre-HV)1 samples harbored 16362, all U2e samples were characterized by the motif 16051-16129C-16154-16248-16362, and all J2 mtDNAs showed the motif 16069-16126-16193-16274-16278. These sequence motifs are almost entirely restricted to the Kalash community, except for those associated with U4. All these observations bear witness to the strong effects of genetic drift on the Kalash population. This distinctive demographic scenario is supported by the nonsignificantly negative values of Tajima’s D and Fu’s F_S neutrality tests (table 3), which reject the hypothesis of population growth, the unambiguous multimodal mismatch distribution (not shown), and the small census size of the population, 3,000–6,000. It has been suggested that this population descends from Greeks or from Slavic peoples, and they claim descent from a place called Tsyam, possibly in Syria (Robertson 1896; Decker 1992). The strong effects of drift and the small population size make genetic inference about the geographic origin of the Kalash difficult. However, a western Eurasian origin for this population is likely, in view of their maternal lineages, which can ultimately be traced back to the Middle East.

Correlation of Genes and Languages in the Southwestern Asian Corridor

The study of the mtDNA pool of present-day populations living in the southwest and Central Asian corridor shows that the linguistic differences in these regions (i.e., mainly Indo-European vs. Altaic) are not reflected in the patterns of mtDNA diversity. However, there are two linguistic outliers that merit further consideration: the Hunza Burusho and the Brahui. The Hunzas live mainly in the remote Hunza Valley of northern Pakistan and speak Burushaski, a language isolate of uncertain origin. Our analysis shows that the Hunza mtDNAs, like the Y haplotypes (Qamar et al. 2002), are shared with neighboring populations, particularly with southern Pakistanis (see PC plot in fig. 6). This genetic pattern could have been established before the linguistic differentiation took place, or there could have been substantial gene flow with neighboring populations. In any case, no distinctive genetic signature accompanies the linguistic and geographic isolation of the Hunza Burusho population, in agreement with recent data based on 182 autosomal microsatellites (Ayub et al. 2003).

The second linguistic outlier is the Brahui population, located in central Baluchistan, which represents a Dravidian-speaking enclave outside India. Historical records indicate that the Brahui are descendants of Turko-Iranian tribes from west Asia (Hughes-Buller 1991). Today, Dravidian languages are essentially restricted to south India and Sri Lanka, but the proto-Elamo-Dravidian hypothesis (McAlpin 1974, 1981) proposes that they originated in the Iranian province of Elam and were once spoken over a much larger area, including Iran, Pakistan, Afghanistan, and all India. The Brahui population is characterized by high prevalences (55%) of western Eurasian mtDNAs and the lowest frequency in the region (21%) of haplogroup M*, which otherwise is common (∼60%) among Dravidian-speaking Indian populations. As shown in the PC1 (fig. 6), the Brahui lie in an intermediate position between Iranian and Indus Valley populations, far from the Gujaratis and even farther from Dravidian-speaking Indian groups (results not shown). These observations exclude the possibility that the Dravidian presence in Baluchistan has resulted from recent incursions of Dravidian speakers from India and show that the maternal gene pool of the Brahui is similar to that of Indo-Iranian speakers from the southwestern Asian corridor. Although the present Brahui population could represent an ancient Indian Dravidian-speaking population relocated to Pakistan, where they admixed with local populations, no historical record supports this hypothesis. Thus, this suggests that they are the last northern survivors of a larger Dravidian-speaking region predating the arrival of Indo-Iranian speakers, thus reinforcing the proto-Elamo-Dravidian hypothesis (McAlpin 1974, 1981).

Traces of Recent and Sexually Asymmetrical Events

The phylogeographical cross-comparison of mtDNA and Y-chromosomal data is very useful for tracing differential male and female histories. Some populations studied here (Iranian, Pakistani, and Central Asian) have been analyzed previously for Y-chromosomal variation (Quintana-Murci et al. 2001; Qamar et al. 2002; Zerjal et al. 2002). In most cases, mtDNA variation is in good agreement with the Y-chromosomal data, suggesting that the patterns reflect general population processes. A good, although surprising, example of concordance between the two systems is the Hazara, who claim to be the direct male-line descendants of Genghis Khan’s army. The presence and time depth of the Y-chromosomal haplogroup C* (xC3c) in the Hazara, along with its absence from neighboring populations, has been interpreted as the genetic legacy of Genghis Khan and his male relatives (Qamar et al. 2002; Zerjal et al. 2003). Our results indicate that the Hazara are also characterized by very high frequencies of eastern Eurasian mtDNAs (35%, table 2, fig. 1), which are virtually absent from bordering populations, suggesting that the male descendants of Genghis Khan, or other Mongols, were accompanied by women of East Asian ancestry.

In contrast to the parallelism between mtDNA and Y-chromosomal data in most populations, the Parsis and the Makrani both show a sharp contrast between these loci. The Parsis live in southeastern Pakistan, and historical records indicate an Iranian origin (Nanavutty 1997). These followers of the prophet Zoroaster started their migration from Iran in the 7th century a.d., settling in the northwestern Indian province of Gujarat around 900 a.d. and eventually moving to Mumbai in India and Karachi in Pakistan. Y-chromosome data show that they resemble Iranian populations rather than their neighbors in Pakistan: an admixture estimate of 100% from Iran was obtained (Qamar et al. 2002), supporting the historical records. However, when the Parsi mtDNA pool was compared with those of the Iranians and Gujaratis (their putative parental populations), a strong contrast with the Y-chromosomal data emerged. About 60% of their maternal gene pool belongs to South Asian haplogroups, which make up only 7% of the combined Iranian sample (table 2). The very high frequency of haplogroup M among the Parsis (55%), similar to those of Indian populations and much higher than that of the combined Iranian sample (1.7%), highlights their close affinities with India (fig. 6). Our results lead to an admixture estimate of 100% from Gujarat and provide a strong contrast between the maternal and paternal components of this population. Although the small population size of the Parsis (a few thousand) may have distorted haplogroup frequencies in this population, diversity of both Y-chromosome and mtDNA lineages remains high, making a strong drift effect unlikely. Our results therefore support a male-mediated migration of the ancestors of the present-day Parsi population from Iran to India, where they admixed with local females, or directional mating in Gujarat between Iranian males and local women, leading ultimately to the loss of mtDNAs of Iranian origin.

Another example of an unequal sex-specific contribution is seen in the so-called “Negroid” Makrani of Baluchistan. This population lives in the Makran coastal region and shows distinct African physical traits (Sultana 1995). We observed a high presence (39%) of lineages L3d, L3b, L2a, and L1a, generally restricted to sub-Saharan African populations (Chen et al. 1995, 2000; Salas et al. 2002) and otherwise present in only 4 of the remaining 877 individuals examined. The presence of African mtDNAs among the Makrani seems to be of recent origin, since the Makrani haplotypes are identical to those observed in modern sub-Saharan African populations (Salas et al. 2002), particularly in Bantu-speaking populations from Mozambique. Indeed, all but one of the Makrani L1, L2, and L3A types matched Mozambique sequences, and these were the most frequent haplotypes in the Mozambique samples (L1a2, L2a1a, and L2a1b) (Pereira et al. 2001; Salas et al. 2002). Our results contrast with the Makrani Y-chromosome profile, which is similar to that of other Pakistani populations and is dominated by western Eurasian lineages (Qamar et al. 2002). The sub-Saharan African male-specific contribution, represented primarily by Hg E-M2, occurred at only 9% in the Makrani and is also present in neighboring populations, although at a lower prevalence (2%–4%). We estimated the maternal and paternal contributions of sub-Saharan Africans to the current Makrani gene pool, using information from all haplogroups, at 12% (±7%) for the Y chromosome and 40% (±9%) for the mtDNA. These findings must be interpreted in the light of known historical data. Forced migration from Africa began in the 7th century and increased considerably during the Omani Empire. The latter formed a strong slave-trade connection between the Makran port of Gwadar, the principal ports of Oman, and ports located in East Africa, including Mozambique (Clarence-Smith 1989; Sultana 1995). In the 16th and 17th centuries, the Portuguese also traded between Mozambique and southwestern Asia. The African component in the Makrani community may therefore represent the genetic legacy of this slave trade. Whereas the Atlantic slave trade dealt mainly with male labor, the East African slave trade seemingly favored females over males (Lovejoy 2000). Slave women were mainly domestics and/or concubines, and children fathered by the master were freed. In addition, strong cultural barriers hindered male slaves from fathering children, a situation exacerbated by the proportion of slaves imported as eunuchs (Lovejoy 2000). As a consequence of these practices, the contribution of paternal African genes to the population is expected to be low. Indeed, the contrast between male and female African contributions observed among the Makrani strongly supports historical records of a female sex bias during the East African slave trade. Other factors, such as asymmetrical mating patterns between African women and autochthonous males during the process of genetic admixture, and/or unequal reproductive success among Makrani males, might have accelerated the loss of African Y chromosomes from the population. In this context, a similar pattern has been reported recently in the Yemeni Hadramawt population (Richards et al. 2003), geographically adjacent to East Africa, where the African maternal contribution has also been interpreted as the result of the East African slave trade. Our data not only confirm a female-biased slave trade towards the East but also show that this pattern, which includes differential mating patterns between the sexes, extended to the eastern limits of the East African slave trade.

Conclusions

Our analysis of mtDNAs from the southwestern and Central Asian corridor shows that the highest variation is observed in populations located in the Indus Valley and Central Asia, highlighting this region as the place where western Eurasian lineages meet both the South Asian and eastern Eurasian genetic strata, respectively. The amalgamation of different genetic components in this area may have resulted from the successive and continuous waves of migration from diverse geographical sources at different time periods, from the early human settlements in the region after the “out of Africa” dispersal to migrations associated with the diffusion of new technologies, such as farming and/or pastoral nomadism, and accompanied by new languages, like the incursions of Indo-Iranian speakers from the northwest. In addition, the Indo-Gangetic region is characterized by the presence of autochthonous genetic footprints of Pleistocene origin and traces of recent historical events, such as the East African slave trade. This extraordinarily rich and heterogeneous genetic portrait testifies to the numerous and complex movements in the region and evinces more subtle demographic episodes in some populations, including founder effects and sexually asymmetrical events associated with differential migration patterns between males and females.

Supplementary Material

Table A1.

RFLP and Control Region Variation in Anatolian, Caucasus, and Pakistani-Karachi Samples

	Haplotypeb
Haplogroup and Samplea	RFLPc	Control Regiond	Read from Start np/End np
L2a:
KAR04	+3592h +4769e +10394c −10871z −12703t +13803e +16389g/−16390b	16189 16223 16234 16249 16278 16294 16295 16390 73 143 146 152 195	16000/00219
M*:
KAR46	−8005l −8783g +10394c +10397a −10871z −12703t +13366m/−13367b/+13367j −15925i +16389g/−16390b +16517e	16048 16129 16223 16390 16519 73	16033/00160
KAR47	−8005l −8783g +10394c +10397a −10871z −12703t +13366m/−13367b/+13367j −15925i +16389g/−16390b +16517e	16048 16129 16223 16390 16519 73	16001/00137
KAR61	−7859j +8249b/−8250e +10394c +10397a −10871z −12703t +15500e +16517e	16223 16324 16357 16519 16527 73	16004/00138
KAR56	−6850i −8882c +10394c +10397a −10871z −12703t +16517e	16129 16223 16265C 16519 73	16002/00136
KAR50	−5978a −7474a +9299a +10394c +10397a −10871z −12703t −16049k +16517e	16051 16183C 16189 16293 16316 16519 73 147 153	16001/00214
KAR26	−3337k −4990a +10394c +10397a −10598a −10871z −12703t +16517e	16092 16223 16289 16342 16519 73	16000/00132
KAR16	−4990a +8156k +10394c +10397a −10871z −12703t +16517e	16129 16223 16519 73	16034/00140
KAR45	−4990a +8156k +10394c +10397a −10871z −12703t +16517e	16129 16223 16519 73	16000/00130
KAR05	−4990a +8156k +10394c +10397a −10871z −12703t +16517e	16129 16223 16519 73	16033/00130
KAR43	−4577q +10394c +10397a −10871z −12703t −16310k +16517e	16126 16223 16311 16519 73	16026/00156
KAR14	−4577q +10394c +10397a −10871z −12703t +16517e	16126 16223 16519 73	16008/00142
KAR22	−4577q +10394c +10397a −10871z −12703t +16517e	16126 16223 16519 73	16000/00131
KAR69	−4577q +10394c +10397a −10871z −12703t +16517e	16126 16223 16519 73	16004/00138
KAR82	−4577q +10394c +10397a −10871z −12703t +15882b/−15883e −16310k +16517e	16126 16223 16311 16519 73	16000/00150
KAR63	−3534c/−3537a +10394c +10397a −10871z −12703t +13085k +16389g/−16390b +16517e	16111 16223 16231 16272 16362 16390 16519 73	16001/00232
KAR68	+9871a +10394c +10397a −10871z −11362a −12703t +16517e	16129 16519 73	16007/00141
KAR86	+8678a +8704k +10394c +10397a −10871z −12703t	16111 16353 56+T 59del 60del 66+T 73	16000/00225
KAR90	+8439a −8838e +9612k −9644a +10394c +10397a −10871z −12703t −13404l	16129 16141 16223 16271 73 143 146 152	16000/00160
KAR11	+8249b/−8250e +10394c +10397a −10871z +12234k (A12234G G12236A) −12703t +16517e	16166del 16223 16519 73	16000/00130
KAR89	+6618e −7859j +10394c +10397a −10871z −12703t −16310k +16517e	16086 16145 16223 16261 16311 16519 73	16000/00130
KAR64	+6618e −7859j +10394c +10397a −10871z −12703t −16310k +16517e	16145 16223 16261 16266 16291 16311 16519 73	16001/00134
KAR18	+6618e −7859j +10394c +10397a −10871z −12703t −13704p −16310k +16238l +16517e	16145 16223 16240 16261 16311 16519 73	16003/00137
KAR85	+6618e −7859j +10394c +10397a −10871z −12703t +15754c −16310k +16517e	16145 16176 16223 16261 16311 16519 73	16000/00144
KAR91	+4140l +10394c +10397a −10871z −12703t −16310k +16494i	16129 16223 16311 16357 16497 73 146 152	16000/00170
KAR59	+6333j −8616j −8882c +10394c +10397a −10871z −12703t +16517e	16129 16223 16265C 16519 73	16003/00128
KAR37	+3759o +10394c +10397a −10871z −12703t −15254s +16517e	16223 16302 16519 73	16000/00131
KAR39	+3759o +10394c +10397a −10871z −12703t −15254s +16517e	16223 16302 16519 73	16001/00130
KAR13	+3391e +10394c +10397a −10871z −12703t +16517e	16158 16223 16234 16319 16348 16362 16519 73	16003/00137
KAR100	+3391e +10394c +10397a −10871z −12703t +16389g/16390b +16517e	16126 16162 16172 16223 16256 16342 16390 16519 73	16000/00160
KAR41	+245l +9070l +10394c +10397a −10871z −12703t +14583c	16223 16260 16318T 16325 73	16001/00132
KAR92	+245l +10394c +10397a −10871z −12703t +16517e	16093 16134 16223 16318C 16519 73 93	16000/00131
KAR83	+245l +10394c +10397a −10871z −12703t +15487b +16517e	16223 16319 16519 73	16000/00133
KAR77	+10394c +10397a −10871z −12703t −16310k +16517e	16129 16223 16255 16311 16519 73	16000/00105
KAR95	+10394c +10397a −10871z −12703t −16310k +16389g/16390b +16517e	16093 16129 16223 16311 16390 16519 73	16000/00131
KAR10	+10394c +10397a −10871z −12703t +16517e	16111 16223 16519 73 114	16000/00130
KAR20	+10394c +10397a −10871z −12703t +16517e	16129 16172 16223 16519 73 146	16003/00220
KAR48	+10394c +10397a −10871z −12703t +16517e	16129 16172 16223 16519 73 146	16002/00210
KAR19	+10394c +10397a −10871z −12703t +16517e	16129 16172 16223 16519 73	16003/00135
KAR65	+10394c +10397a −10871z −12703t +16517e	16145 16223 16519 73	16004/00138
KAR32	+10394c +10397a −10871z −12703t +16517e	16223 16230 16234 16519 73	16003/00123
KAR27	+10394c +10397a −10871z −12703t +16517e	16223 16234 16519 73	16000/00130
KAR70	+10394c +10397a −10871z −12703t +16517e	16223 16234 16519 73	16002/00136
KAR58	+10394c +10397a −10871z −12703t +16517e	16223 16289 16519 73	16003/00137
KAR15	+10394c +10397a −10871z −12703t +16517e insertion of 11 Cs at np 5900	16223 16234 16519 73	16000/00132
KAR12	+10394c +10397a −10871z −12703t +12810k +16517e	16223 16270 16274 16292 16319 16352 16519 73	16000/00130
KAR97	+10394c +10397a −10871z −12703t +12810k +16517e	16223 16270 16319 16352 16519 73	16000/00132
KAR71	+10394c +10397a −10871z +11001n/+11002f +16275k +16389g/16390b	16223 16256 16275 16327A 16390 73	16003/00137
GEO15	+10394c +10397a −10871z −12703t −13103g/+13104j	16223 16362 16526 73 263	16000/00280
D:
ANA16	−5176a +10394c +10397a −10871z −12703t	16174 16223 16362 16468 73	15997/00200
G:
KAR36	+4830n/+4831f +10394c +10397a −10871z −12703t −15925i	16223 16274 16362 73 195	16000/00250
ANA53	+4830n/+4831f +10394c +10397a −10871z −12703t +15494c	16223 16227 16278 16362 73 152 207 263	16000/00290
Z:
KAR53	+2349j −2776b −4360g +10394c +10397a −10871z −12703t −16297u −16310k +16517e	16185 16223 16260 16298 16311 16519 73 151 152	16004/00218
A:
ANA07	+663e −12703t	16182C 16183C 16189 16223 16290 16319 16362 73 93 150 152 195	16000/00215
ANA50	+663e −5823a −12703t	16189 16223 16290 16319 16362 73 152	16000/00205
N1b:
BAL01	−1715c +8249b/−8250e +10237v −11362a −12703t +16176j +16389g/−16390b+16517e	16129 16145 16176G 16223 16309 16382 16390 16413	16031/16496
CHE14	−1715c +8249b/−8250e +10237v −11362a −12703t +16176j +16389g/−16390b +16517e	16145 16176G 16223 16291 16390 16519 73	16003/00151
ANA09	−1715c +8249b/−8250e +10237v −11362a −12703t +16176j +16389g/−16390b +16517e	16145 16176G 16223 16309 16390 16519 73 151 152	16000/00200
CHE17	−1715c +8249b/−8250e +10237v −11362a −12703t +16176j +16389g/−16390b+16517e	16145 16176G 16223 16390 16519 73 146 150 152	15997/00200
BAL12	−1715c +8249b/−8250e +10237v −11362a −12703t +16176j +16389g/−16390b+16517e	16145 16176G 16223 16390 16519 73 146 150 152 199	15997/00200
N1d:
KAR30	−951j −1715c −5003c −8616j +10237v −12703t	16223 16301 16356 73	16011/00141
KAR62	−951j −1715c −5003c −8616j +10237v −12703t	16223 16301 16356 73	16004/00136
KAR31	−951j −1715c −5003c −8616j +10237v −12703t −16310k	16223 16301 16311 16356 73	16000/00130
I:
GEO08	−1715c +8249b/−8250e −8572e +10032a +10237v +10394c −12703t −16310k +16389m/+16390j/−16390b +16517e	16129 16223 16259 16264 16311 16319 16362 16391 16519 73 199 204	15997/00225
ANA37	−1715c +8249b/−8250e +10032a +10237v +10394c −12703t +16389m/+16390j/−16390b +16517e	16086 16129 16223 16391 16519 73 152 199	16000/00200
W
KAR81	+8249b/−8250e −8994e −10830g −12703t −13704p +16517e	16183C 16189 16223 16292 16354 16519 73 143 189 194 195 196 204 207	16000/00220
GEO01	+4092e +8249b/−8250e −8994e −12703t +16517e	16173 16223 16292 16325 16352 16513+T 16519 73	16000/00120
GEO19	+4092e +8249b/−8250e −8994e −12703t +16517e	16192 16223 16292 16325 16362 16519 73 185 189	15997/00190
GEO20	+4092e +8249b/−8250e −8994e −12703t +16517e	n.d	-
X
KAR29	−1715c −12703t +14465s −16310k +16517e	16172 16189 16278 16311 16519 73 195	16001/00208
BAL16	−1715c −7853o +11329a −12703t +14465s −16310k +16517e	16183C 16189 16223 16265 16278 16311 16519 73 153 195	15997/00208
GEO05	−1715c −3944l −12703t +14465s +16517e	16183C 16189 16223 16278 16519 73 153 195	16000/00245
ANA38	−1715c −12703t +14465s+16517e	16108 16183C 16189 16209 16278 16325 16519 73 153 195	15996/00205
BAL20	−1715c −12703t +14465s +16517e	16093 16183C 16189 16223 16278 16519 73 153 195	16000/00208
BAL19	−1715c −12703t +14465s +16517e	16104 16189 16223 16278 16519 73 153 195	16000/00215
ANA27	−1715c −12703t +14465s +16517e	16189 16218 16223 1627816519	16025/16519
ANA32	−1715c −12703t +14465s +16517e	16189 16223 16278 16519 73 153 195	16000/00200
CHE18	−1715c −12703t +14465s +16303k +16517e	16129C 16189 16223 16278 16304 16519 73	16000/00124
R*:
KAR51	−5584a +10143a +10289e −12406o/−12406h −13404l +13826g +16145e	16086 16146 16260 16261 16319 16362 73 207	16000/00250
KAR80	+16494i +16517e	16136 16292 16497 16519 73	16000/00132
R2:
KAR72	+4216q +4769a −14304a −16310k +16517e	16071 16108 16111 16278 16311 16519 73	16004/00138
GEO11	+4216q +4769a −14304a	16071 16355 16357 73 81	16003/00120
R5:
KAR94	−8592j +13633e −16303k −16310k +16517e	16266 16304 16311 16356 16519 16524 73 152	16000/00250
KAR76	+4051k −8150i −8592j −16303k +16517e	16266 16304 16519 73 93	16000/00132
R6:
KAR54	−12282a +16389g/−16390b +16517e	16129 16213 16259A 16274 16319 16362 16390 16519 73	16001/00115
KAR52	+10394c −12282a +15112a +16517e	16129 16213 16362 16519 73	16004/00138
KAR73	−12282a +16517e	16114G 16129 16362 16519 73	16004/00138
(pre-HV)1:
ANA12	+11718e	16126 16362	16023/16383
KAR87	+11718e +16517e	16093 16126 16362 16519 57+C 64 195	16000/00225
KAR24	+10727c +11718e	16126 16362 60+T 64 152 153	16000/00250
ANA25	−9266e +11718e −15812k	16126 16168 16264 16295 16362 64	15997/00200
BAL11	+4769a −5176a +11718e	16126 16355 16362 58 64 146	15997/00200
CHE05	+11074c +11718e +16517e	16114 16126 16362	16032/16420
HV*
ANA02	+6582j +11718e −14766u	CRS	16000/00140
KAR21	+1973g +11718e −14766u	CRS	16000/00132
KAR23	+11718e −14766u	CRS	16003/00137
KAR55	+11718e −14766u	16319	16002/00136
ANA19	−5584a +11718e −14766u −16310k	16311	16023/16383
ANA39	+11718e −14766u −16208k	16189 16209 16320 16356 44+C 57+C 214	16000/00242
HV2:
KAR93	+11718e −13704p −14766u +16517e insertion of 11 Cs at np 5900	16217 16519 73 152	16000/00160
H:
KAR78	−951j −7025a +11718e −14766u.	16354	16000/00132
KAR79	−951j −7025a +11718e −14766u.	16354	16000/00160
BAL09	−951j +4769a −5176a −7025a +11718e −14766u +16517e	16292 16354 16519	15997/00200
KAR06	−7025a +11718e −14766u	16111 16256 16257	16008/00142
KAR02	−7025a +11718e −14766u −14858n/−14859f −16297u +16517e	16218 16297A 16519 73	16000/00130
KAR03	−7025a +11718e −14766u −14858n/−14859f −16297u +16517e	16218 16297A 16519 73	16000/00130
KAR60	−5003c −7025a +11718e −14766u	CRS	16004/00138
ANA51	−7025a +11718e −14766u −14869j/+14869g	16148 16256 16319	16023/16383
KAR49	−1004o −7025a −11365a +11718e +13366m/−13367b/+13367j −14766u −14869j/+14869g +16517e	16519	16002/00125
KAR08	−1004o −7025a +11718e −14766u −14869j/+14869g +16517e	16243 16519	16008/00142
KAR98	−1004o −7025a +11718e +12946c/+12949n/+12950f −14766u −14869j/+14869g +16517e	16243 16519	16000/00160
KAR96	−1004o −6260e −7025a +11718e −14766u −14869j/+14869g +16517e 9bp del	16519	16000/00131
ANA21	+4793e −7025a −8729j −10364e +11718e −14766u +16517e	16221 16519 152 263	16000/00290
KAR75	+4793e −7025a +11718e −14766u +16517e	16519	16007/00141
ANA23	+4332b −5837e −7025a +11718e −14766u −16303k	16304 73 151	16011/00220
KAR09	+4332b −7025a −8150i +8439a +10407k −10631c/+10634e +11718e −14766u −16303k	16304	16000/00130
ANA01	−7025a −9753g +11718e +14279e −14766u	16328A	15997/00200
ANA30	−7025a −9380f +11718e −14766u	16362 16482	16000/00200
CHE07	−7025a −9380f +11718e −14766u +16517e	16297 16362 16519	16002/00124
GEO12	−7025a −9052n/−9053f +11718e −14766u −16303k	16304 16335	16003/00075
ANA36	−7025a +9157e +11718e −14766u +16517e	16172 16266 16519 93	15997/00200
ANA31	−7025a +11718e −14766u	16232 16260 16278	15997/00200
BAL14	−7025a +11718e −14766u	16324 150	15997/00200
ANA52	−7025a +11718e −14766u	16366	16023/16383
CHE16	−7025a +11718e −14766u	55 57	15997/00204
GEO02	−7025a +11718e −14766u −16310k	16311 16527	15997/00200
BAL10	−7025a +11718e −14766u +16517e	16468 16519 183	15997/00200
CHE10	−7025a +11718e −14766u +16517e	16519	15997/00200
CHE03	−7025a +11718e −14766u +16517e	16519	15997/00200
ANA15	−7025a +11718e −14766u +16517e	16086 16519	16024/16540
ANA42	−7025a +11718e −14766u +16517e	CRS	16023/16383
ANA46	−7025a +11718e −14766u +16517e	CRS	16023/16383
ANA03	−7025a +11718e −14766u +15317c +16517e	16519 16546	16026/00019
ANA17	−7025a − 11326c +11718e −14766u −16303k +16517e	16271 16304 16519	15997/00200
BAL15	−5176a −7025a +11718e −14766u	16192	16000/00200
GEO03	+54k −7025a +11718e −14766u	57 57+G	15997/00200
BAL05	+5295a −7025a +11718e −14766u −16303k −16310k	16304 16311 16400 16527het	16000/00040
GEO17	+3744e +4732k/+4735k −7025a +11718e −14766u +16517e	16244 16292 16519 152	15997/00200
U1:
ANA18	−4990a +12308g −13103g/+13104j +14068l	16008 16182C 16183C 16189 16249 73	15997/00200
ANA34	−4990a +12308g −13103g/+13104j +14068l −16310k	16166del 16183C 16189 16249 16271 16311 73 146 152	16000/00200
BAL02	−4990a +12308g −13103g/+13104j +14068l	16086 16183C 16189 16245 16249 73	16000/00236
CHE08	−4990a +12308g −13103g/+13104j +14068l	16183C 16189 16249 73	16000/00218
ANA04	−4990a −7417l −7461l +12308g +8615g/−8616j −13103g/+13104j +14068l	16183C 16189 16249 16269+A 73	16018/00200
BAL13	−4990a +8911k/+8944k +12308g −13103g/+13104j +14068l	16129 16183C 16189 16224 16249 16288 16295 16527 73 150 195	16000/00208
U2a:
KAR34	−5552c −7055a +12308g −16049k −16303k −16310k +16517e	16051 16206C 16230 16304 16311 16519 73	16001/00132
KAR42	+5260b/−5261e +8569c/−8572e +12308g −16049k −16310k +16517e	16051 16154 16206C 16230 16311 16519 73	16026/00156
KAR33	+5260b/−5261e +8569c/−8572e +12308g −16049k −16310k +16517e	16051 16154 16206C 16230 16311 16519 73	16000/00250
KAR84	+5260b/−5261e +8569c/−8572e +10394c +12308g −16049k −16310k +16517e	16051 16154 16206C 16230 16311 16519 73	16000/00140
KAR40	+3391e +9135j +12308g +14465s −16049k −16310k	16051 16206C 16242 16291 16311 73	16001/00132
U2b:
KAR35	−8994e +12308g −15047e −16049k −16208k	16051 16086 16209 16239 16352 16353 16355 73	16001/00132
KAR17	+12308g −15047e −16049k	16051 16086 16129 16259A 16267 16291 16326 73	16001/00108
KAR01	+12308g −15047e −16049k −16208k	16051 16209 16239 16244 16352 16353 73 146 152 234	16000/00250
KAR44	+12308g −15047e −16049k −16208k	16051 16209 16239 16352 16353 73	16000/00130
U2c:
KAR25	−5789l +8020j/+8022l +8197a +12308g −15060j −16049k +16517e	16051 16179 16234 16240C 16519 73	16000/00132
KAR07	+4720b −5789l +8020j/+8022l +10806g +12308g −15060j −16049k	16051 16189 16234 73 146 152	16000/00250
U2e:
ANA24	−984c +12308g +13730g +15907k −16049k	16051 16129C 16182C 16183C 16189 16256 16266 73 152	16000/00208
BAL18	−984c +12308g +13730g +15907k −16049k +16517e	16030 16051 16129C 16189 16256 16519 73 147 152	16000/00280
GEO10	+12308g +13730g +15907k −16049k +16517e	16051 16129C 16362 16519 73	16000/00200
U3:
ANA26	−104i +12308g +16517e	16189 16343 16519	15997/00054
CHE09	+2293a +9189k/+9192k −9266e +12308g +14138j	16193 16249 16343 16413A	16023/16500
GEO14	+12308g +14138j	16343 73 150	15997/00200
U4:
ANA45	+4643k −4808c/−4810g +4984b +10907c +11329a +12308g +16517e	16179 16356	16023/16383
CHE11	+4643k +7702k −7853o +11329a +12308g +16517e	16356 16519 73 195	15997/00200
CHE06	+4643k +11329a +12308g +16517e	16356 16519 73 152 195	15997/00200
U5:
ANA05	+12308g −16310k	16174 16189 16192 16270 16311 16368 73 150	16021/00219
ANA49	+12308g+16398e	16145 16183C 16189 16192 16256 16270 16399 73 197T	16023/00200
GEO06	+12308g −16303k +16398e	16169 16192 16256 16270 16304 16399 73 150	15997/00200
ANA28	+12308g +16398e	16256 16270 16399 73	16000/00200
ANA35	+12308g +16143l	16144 16189 16270	16000/00070
U7:
KAR66	+12308g +16517e	16086 16309 16318T 16519 73 151 152	16003/00212
KAR67	+12308g +16517e	16140 16214 16309 16318T 16362 16519 73	16007/00141
KAR28	+12308g +16517e	16173 16309 16318T 16362 16519 73 152	16000/00250
ANA10	+12308g +16517e	16309 16318T 16343 16519 73 151 152	15997/00200
KAR74	+12308g +15754c +16517e	16207 16309 16318C 16519 73	16004/00138
KAR38	+12308g +12372a (G12372A A12373G) +16517e	16172 16309 16318C 16519 73	16001/00131
U8b:
ANA33	−9052n/−9053f +12308g −16208k	16092 16169 16189 16192 16210 16234 16259 73	15997/00191
U9
KAR57	−1004o −6383e −11350a +12308g	16242 73 195	16000/00250
K:
ANA14	−8391e −9052n/−9053f +10394c +12308g −13957e −16310k +16517e	16093 16224 16311 16519 73	16000/00080
GEO07	−9052n/−9053f +10394c +12308g −16310k +16517e	16145 16224 16311 16519 73 198	15997/00200
CHE02	−9052n/−9053f +10394c +12308g −16310k +16517e	16224 16261 16271 16311 16519 73	15997/00200
GEO09	−9052n/−9053f +10394c +12308g −16310k +16517e	16224 16298 16311 16519 73	15997/00200
ANA20	−9052n/−9053f +10394c +12308g −16310k +16517e	16224 16311	16023/16383
ANA11	−9052n/−9053f +10394c +12308g −16208k −16310k +16517e	16093 16145 16209 16224 16311	16023/16383
CHE13	−6260e −9052n/−9053f +10394c +12308g −16310k+16517e	16172 16224 16311 16352 16519	16003/16550
GEO04	+8249b/−8250e −9052n/−9053f +10394c +12308g −16310k+16517e	16224 16234 16260 16311 16318T 16519 73 152	15997/00200
J1:
BAL04	+4216q +10394c −13704p −16065g +16517e	16069 16126 16145 16261 16519 73	16022/00120
KAR99	+4216q −7598f +10394c −13704p −16065g	16069 16126 16145 16172 16261 73 146 152	16000/00160
ANA40	+4216q +8466a/+8470a +10394c −13704p −16065g	16069 16126 16145 16172 16183C 16189 16222 16235 16249 16261 73	16000/00200
BAL06	+4216q +8466a/+8470a +10394c −13704p+15500j −16065g	16069 16126 16145 16189 16222 16261	15997/16436
CHE12	+4216q +10394c −13704p −16065g	16069 73	16011/00075
ANA13	+4216q +10394c −13704p −16065g −16310k	16069 16126 16291 16311	16023/16383
ANA08	+4216q +10394c −13704p −16065g +16517e	16069 16126 16519 73 185 188	15997/00200
BAL07	+4216q +10394c −13704p −16065g +16517e	16069 16126 16193 16214 16519 73	16000/00200
BAL17	+4216q +10394c −13704p −16065g +16517e	16069 16126 16193 16519 64 73	16023/00130
J2:
ANA48	+4216q −7474a +10394c +11001n/+11002f −13704p −15254s −16065g	16069 16126 16169	16023/16383
T*:
KAR88	+4216q +4914r −13259o +13366m/−13367b/+13367j −13704p +15606a −15925i +16517e	16126 16294 16519 73	16000/00130
ANA44	+4216q +4914r −6260e +13366m/−13367b/+13367j −13704p +14068c +15606a +15925i −16310k	16126 16292 16294 16311	16023/16383
CHE04	+4216q +4914r −12406h −13259o +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16222 16294 16296 16519 73	15997/00200
ANA22	+4216q +4914r +9253e +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16163 16183C 16189 16243 16294 16519 73 150	16023/00220
GEO13	+4216q +4914r +8093a +8270k +13366m/ −13367b/+13367j +15606a −15925i+16517e	16126 16183C 16189 16291 16294 16295 16296 1651973	16000/00204
ANA43	+3391e +4216q +4914r +13366m/−13367b/+13367j +14277c +15606a −15925i −15812k +15346k +16517e	16126 16235 16294 16296	16023/16383
CHE15	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16163 16172 16189 16243 16294 16519 73	16008/00113
GEO18	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16294 16296 16519 73 200	15997/00203
GEO16	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16163 16189 16243 16294 16519 73	15997/00101
ANA06	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16294 16296 16519 73	15997/00200
CHE01	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16294 16296 16519 73	15997/00200
T1:
BAL03	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i −16310k +16517e	16126 16163 16186 16189 16294 16311 16519	16000/00019
BAL08	+4216q +4914r +13366m/−13367b/+13367j +15606a −15925i +16517e	16126 16163 16186 16189 16294 16355 16356 16519	16014/00030

^aANA = Anatolian; BAL = Balkarian; CHE = Chechens; GEO = Georgians; KAR = Pakistani from Karachi.

^bCoding-region sites and control-region motifs diagnostic of each haplogroup are shown in boldface.

^cSites are numbered from the first nucleotide of the recognition sequence. A plus sign (+) indicates the presence of a restriction site, a minus sign (−) the absence of such a site. The explicit indication of the presence/absence of a site implies the absence/presence in haplotypes not so designated. The restriction enzymes used in the analysis are designated by the following single-letter codes: a, AluI; b, AvaII; c, DdeI; e, HaeIII; f, HhaI; g, HinfI; h, HpaI; i, MspI; j, MboI; k, RsaI; l, TaqI; m, BamHI; n, HaeII; o, HincII; p, BstOI; q, NlaIII; r, BfaI; s, AccI; t, MboII; u, MseI; v, HphI; z, MnlI. A slash (/) separating states indicates the simultaneous presence or absence of restriction sites that can be correlated with a single-nucleotide substitution.

^dOnly those nucleotide positions that differ from the Cambridge Reference Sequence (CRS) (Anderson et al. 1981; Andrews et al.1999) are shown. Mutations are transitions, unless the base change is specified explicitly. A plus sign (+) indicates an insertion, and “del” indicates a deletion.

Table A2.

Haplogroup Affiliation and Sequence Variation of the 702 Individuals Analyzed in the Present Study

		Population a
Haplogroup or Subscluster	HVS-I Sequence Variationc	PE	TI	GI	MA	KI	LU	BA	BR	PA	SI	PT	MK	HZ	HU	KL	GJ	UZ	TK	KT	SH
Haplogroupb		42	40	37	21	20	17	39	38	44	23	44	33	23	44	44	34	42	41	32	44
L1a	148 172 187 188G 189 223 230 311 320												1
L2a	182C 183C 189 223 278 290 294 309												2
L2a	223 274 278 286 294 309				1
L2a	223 278 286 294 309												1
L3b	093 124 223 278 362												2
L3b	124 223 278 311 362												1
L3d	124 212 223 319								1
L3d	124 223 318T 319												4
L3d	124 223 319							1					2
M*	038 178 223 288																1
M*	051 183C 189 223 319																	1
M*	066 126 154 223								1
M*	086 209 223 278									1
M*	093 129 223 311 357	1
M*	093 223 234											1			1
M*	103C 223																	1
M*	108 129 223 291 298														1
M*	111 168 189 192 223 264 275 300 352									1
M*	111 223 239																		1
M*	126 147 223														4
M*	126 154 223										2
M*	126 162 172 189 223 256																		1
M*	126 169 223											1
M*	126 221 223 311								1
M*	126 223							1		2		3
M*	126 223 266 311													1
M*	126 223 289											1
M*	126 223 311									2		2					1
M*	126 289														2
M*	127 223 266 311													1
M*	129 172 223											1
M*	129 189 223 249 311																1
M*	129 223							4	5			1	2
M*	129 223 265C							1		5
M*	129 223 304																1
M*	129 223 311							1
M*	129 223 311 362																	1
M*	145 176 223 261 266 291 311			1													1
M*	145 176 223 261 290 311											1
M*	145 176 223 261 311							1						1
M*	145 223 243 261 311							1
M*	166del 223														1		1
M*	168 239									1
M*	179 223 311																1
M*	179del 223									1
M*	179del 223 288A 302							1
M*	184 223 298 319																	1
M*	184 223 311												1
M*	185 223 289 362											1
M*	188 223 231 362										1
M*	223																1
M*	223 234									2
M*	223 234 274																1
M*	223 234 295G 311 320										1
M*	223 270 319 352	1							1		2						1
M*	223 275 327A 362									1							1
M*	223 289							1							1		1
M*	223 295																	1
M*	223 304									5
M*	223 304 359									2
M*	223 311							2		1		1
M*	223 311 362										1
M*	223 318T																1
M*	223 324 357																1
M*	223 360 363																1
C	093 223 261 288 298																			3
C	093 223 298 327																	1
C	129 223 298 327																				1
C	223 239 298 327 357																				3
C	223 239 327																				1
C	223 239 327 357																				3
C	223 297 298 301 327														1
C	223 298 311 327																		1
C	223 298 327																		1
C	223 298 327 344A 357																		1
D	042 093 214 223 256 362																	1
D	092 170 223 265C 311 362																	1
D	129 174 223 239 362																		1
D	129 223 249 311 362			1
D	153 221 223 290 301 362								1
D	172 182C 183C 189 223 362								1d
D	223 224 245 292 362																		1
D	223 242 362																	1
D	223 245 362																		1
D	223 245 362 368																	1	4
D	223 362																		2
G	223 227 265C 278 362																	1
G	223 227 274 278 362													1
Z	111 136 223 260 298													2
Z	185 209 260 298													1
N*	086 172 187 189 217 223											1
N*	092											1
N*	093 292																1
N*	111 144 223 261 311	1
N*	148 223														1
N1	093 223 292 309 311							1
N1a	147A 172 223 248 320 355																		1
N1b	075 145 148 176G 186 223 311			1
N1b	126 145 176A				1
N1b	145 176G 209 223											1
N1b	145 176G 223	1
N1b	145 176G 223 311				1
N1c	184 201 223 265 286																			1
N1d	223 301 356							1	1				1
N9a	093 111 129 223 257A 261		1
N9a	111 129 223 257A 261																	2
N9a	223 257A 261																	1
I	129 168 172 173 223					1
I	129 178 223 296 311										1
I	129 183C 189 223 311											1
I	129 186 223 271																				1
I	129 223		2												2
I	129 223 311	1			1						1									1
A	125 223 274 290 311 319 362														1
A	223 242 278 290 319																	1
A	223 242 290 319																		1	1
A	223 290 319 362																	2			1
W	093 192 223 292 325 326										1
W	129 223 292										1						1
W	145 183C 189 223 292 320										1
W	145 189 223 29																1
W	172 223 292											1
W	183C 189 223 292 320																1
W	189 223 292														1
W	189 223 292 325 355										1
W	192 223 292 311 325											1
W	192 223 292 325				1
W	209 223 255 292 381G																				2
W	212 223 265C 292 303				1
W	223 286 292					1
W	223 292	1	1			1												1
X	086 189 223 278 309																			1
X	162 179 183C 189 223 278																				1
X	183C 189 223 278														1					1
X	189 223 278																		1
R*	093 126 325 355 362												1
R*	093 179 227 245 266 278 362																1
R*	172 220C 265 298 362												1
R*	179 227 245 266 278 362																1
R*	304													2
R*	362											1
R*	CRS																1
R1	278 311																			3
R2	071							3	3				1
R2	071 234				1
R2	071 355 357				1
R5	093 266 304																	1
R5	266 304 311 356											1			1
R5	266 304 325 356	1
B	086 136 182C 183C 189 217																			2
B	093 140 182C 183C 189 217 274 335													1
B	136 154 183C 193+C 217 218						1
B	136 183C 189 217													1
B	140 182C 183C 189 243																		1
B	182C 183C 189 217														1
F	134 182C 183C 189 232A 249 304																	1
F	172 183C 189 232A 249 304 311																		1
F	183C 189 232A 249 304 311													1
F	183C 189 304													1
(pre-HV)1	093 126 355 362						1
(pre-HV)1	126 270 362																		2
(pre-HV)1	126 287 362											1
(pre-HV)1	126 301 362											2
(pre-HV)1	126 355 362						1
(pre-HV)1	362															10
HV*	037	1
HV*	086 183C 189				1
HV*	092																	1
HV*	155 189 311 316	1
HV*	158T 311													1
HV*	174					1	1
HV*	184 357			1
HV*	189				1																2
HV*	189 223																		1
HV*	193 311	1
HV*	223 324			1
HV*	230 325											1
HV*	260	1
HV*	295																	1
HV*	311	1	1	1
HV*	317			1
HV*	319								2
HV*	327A 362			1
HV*	356									1
HV*	362															2
HV*	CRS	2	1			1
HV1	067 179 355		1
HV1	067 355			1		1
HV2	093 217										1e
HV2	129 217 271 311						1
HV2	130 217 243																	1
HV2	168 189 217 287					1
HV2	189 217 343			1
HV2	214 217 335			2				3					1
HV2	217							1e											1e
HV2	217 243 309	1											1
HV2	217 309									4
H	086					1
H	086 311								1
H	093 129																			1
H	093 183C 189																		1
H	093 311							1
H	104 300 325 362													1
H	111 167 288 304 362																		1
H	111 239 362														1
H	124 184																				2
H	134																		1
H	145 227		1
H	168 239									1
H	168 288				1
H	172																	1
H	172 209																	1
H	172 274 319 362																			1
H	178														1
H	184																		1
H	188			1
H	189 356 362																				1
H	193 354																				1
H	201 256 319 362		1
H	209							2	1
H	218 328																		1
H	223 260																		1
H	243							3	6
H	256 352	1	1
H	256G																1
H	261																1
H	264																	1
H	291 324		1
H	294 304														1						1
H	300 325 362	1
H	304																		1		1
H	304 311													1
H	311											2						1			1
H	318T	1
H	324																		1
H	354		1	1												1					1
H	362	1									1							1	1
H	366																				2
H	CRS	2	7	3	2	1	3	2	2		1		1	1		1		4		2	3
V	153 298		1
U*	093 189		1
U*	362																				1
U1	129 169 183C 189 249 274												1
U1	145 182C 183C 189 249									1
U1	182C 183C 189 206 249						1
U1	182C 183C 189 249		1							1					3
U1	182C 183C 189 249 311			1
U1	183C 185 189 249													1
U1	183C 189 249		1
U1	189 249		1
U2*	051 093 184 189 234 294 342	1
U2a	051 093 206C														1
U2a	051 129 206C 362											1
U2a	051 145 206C 319								1				1
U2a	051 172 206C											1					1
U2a	051 181 206C											1
U2a	051 206C							1	1		1
U2a	051 206C 215 230 304 311																1
U2a	051 206C 230 304 311																1
U2b	051											1
U2b	051 168 210										1
U2b	051 189 209 239 352																1
U2b	051 209 239 244 352 353												1
U2b	051 209 239 311 352 353														4
U2b	051 209 239 352 353																1
U2b	051 239											1
U2b	051 239 352 353											1
U2c	051 126 189 234 247 264 304											1
U2c	051 189 234										1
U2c	051 234 247												1
U2c	051 234 304 317 343 360 362																1
U2c	051 247 254 355										1
U2e	051 092 129C 153 182C 183C 189 261 362																			1
U2e	051 092 129C 182C 183C 189 362																		1
U2e	051 092 129C 183C 189 248 362																			1
U2e	051 129C 145 183C 235 362																				1
U2e	051 129C 154 248 362															7
U2e	051 129C 183C 189 223 256 362																				1
U2e	051 129C 183C 189 311		1
U2e	051 129C 183C 189 311 362	1
U2e	051 129C 189 362																				1
U3	086 343			1
U3	093 343	1
U3	104 266A 311 343	1
U3	111 343																			1
U3	129 343														1
U3	168 343 355 362						1
U3	172 245 343						1
U3	203 343		1
U3	343						1											1
U4	092 311 356																	1
U4	134 172 356							1
U4	134 310 311 355 356						1
U4	134 356															12
U4	134 356 362															1
U4	136 356														2
U4	154 356 362																				1
U4	179 356	1
U4	287 356																		1
U4	298 356 362													1
U4	356									6				1		2		1
U4	356 362																				1
U5	093 189 270					1
U5	114A 192 256 270 294																				1
U5	114A 256 270 294	1
U5	129 192 256 270 292																				1
U5	192 256 270													1
U5	192 256 270 304														1
U5	192 256 270 362							1
U5	256 270		1								1			1
U5	256 270 309											1
U7	069 227 278 318C 359					1
U7	092 278 309 318T														1
U7	092 278 318T														1
U7	093 129 318T 362																	1
U7	111 129 318T	1
U7	126 148 309 318T		1
U7	129 146 318T													1				1
U7	129 309 318T									1
U7	129 318T						1
U7	176 309 318T							1
U7	189 207 309 318C								1
U7	189 207 318C								1
U7	192 309 318T					1
U7	207 309 318T																1
U7	233 309 318C																1
U7	243 309 318T					1
U7	309 318C			1
U7	309 318T			2		1			1		1				1		1
U7	318C								1				1
U7	318T			1
U7	318T 352										1					1
U8b	066 129 169 183C 189 234 311					1
U8b	111 172 183C 189 311					1
U8b	129 183C 189 266			1
U8b	172 189 234 311																				1
U9	051 355												3
U9	263 355											1
K	093 176 213 224 311						1
K	093 209 224 311																			3
K	093 224 295 311		1
K	093 224 311					1						1
K	093 224 311 319																				2
K	111 224 311	1
K	129 224 311					1
K	172 224 301 311																				1
K	216T 224 311 320														1
K	224 269 311		1
K	224 311	2		1																1
K	224 311 319							1
K	224 311 320														1						1
J1	051 069 126 193															1
J1	069 086 126 145 261											1
J1	069 086 145 261	1
J1	069 092 126 145 172 222 261 271														1
J1	069 111A 126 145 222 261 311	1
J1	069 126																		1
J1	069 126 145 163 222 261 296 311																				1
J1	069 126 145 172 189 222 261		1
J1	069 126 145 172 222 261	1															1	1
J1	069 126 145 172 222 261 278																		1
J1	069 126 145 172 222 261 362																				1
J1	069 126 145 172 261	1
J1	069 126 145 185 261														2
J1	069 126 145 222 235 261							2	3
J1	069 126 145 222 235 261 311					1
J1	069 126 145 222 256 261 278 365				1
J1	069 126 145 222 261		1	2	1
J1	069 126 145 222 261 278 287					1
J1	069 126 145 223 261																	1
J1	069 126 145 256 261	1
J1	069 126 145 261	1	1	1
J1	069 126 145 261 263+A																			1
J1	069 126 145 261 274				1
J1	069 126 145 261 278	1
J1	069 126 145 261 290						1
J1	069 126 185 298			1
J1	069 126 193			1								1						1	2
J1	069 126 193 266												1
J1	069 126 193 274 343			1
J1	069 126 261				1
J1	069 145 261												1
J1	126 221											1
J2	069 126 145 183C 189				1
J2	069 126 148 193 304																			1
J2	069 126 193 274 278															4
J2	069 126 193 278 295		1
J2	069 126 239 366							1
T*	086 126 294 296			1
T*	111A 126 292 294 296											1
T*	126 153 294 296											1
T*	126 163 209 294																	1
T*	126 163 294 309																	1
T*	126 172 183C 189 294				1
T*	126 189 294		1
T*	126 189 294 296 304				1
T*	126 274 292 294	1
T*	126 287 294 304 360	1
T*	126 292 294						1
T*	126 292 294 296		1	1
T*	126 294		1																	2	1
T*	126 294 296		1	2												2			1	2
T*	126 294 296 304																		1
T*	126 294 296 362			1
T*	126 294 302 342		1
T*	126 294 304														1
T*	126 296 362	1								2										2
T1	093 126 163 186 189 294									2
T1	126 163 172 186 189 294 298																		1
T1	126 163 186 189 193 294	1
T1	126 163 186 189 239 243 294 362									1
T1	126 163 186 189 294		1	1	1										1
T1	126 163 186 189 294 325								2			1
T1	126 163 186 189 294 360											1

^aPopulation codes as in table 1.

^bHaplogroup affiliation was determined according to the HVS-I motifs and RFLP diagnostic markers shown in figure 2.

^cMutation sites are −16,000.

^dThis sample, assigned to haplogroup D and harboring the RFLP diagnostic mutation of this lineage (−5176 AluI), does not present the mutation +10397 AluI characterizing the macro-haplogroup M.

^eThese samples assigned as haplogroup HV2 and harboring the HVS-I mutation 16217, which is diagnostic of this lineage, do not present the coding region site +9336 RsaI.