Y-chromosome Genotyping and Genetic Structure of Zhuang Populations

In this study, the Y-haplogroups were typed for eight branches of the Zhuang population with 13 biallelic markers and 7 Y-chromosome short tandem repeats STR, and for every haplogroup, t

Y-chromosome Genotyping and Genetic Structure of Zhuang Populations

CHEN Jing1,2, LI Hui2,3,①

, QIN Zhen-Dong2, LIU Wen-Hong2, LIN Wei-Xiong4, YIN Rui-Xing5, JIN Li2, PAN Shang-Ling1,①

1 Department of Pathophysiology, Guangxi Medical University, Nanning 530021, China;

2.MOE Key Laboratory of Contemporary Anthropology, School of Life Sciences, Fudan University, Shanghai 200433, China;

3 Department of Genetics, School of Medicine, Yale University, New Haven CT 06520-8005, USA;

4 Medical Research Center of Guangxi Medical University, Nanning 530021, China;

5 Institute of Cardiovascular Diseases, the First Affiliated Hospital, Guangxi Medical University, Nanning 530021, China

Abstract: Zhuang, the largest ethnic minority population in China, is one of the descendant groups of the ancient Bai-Yue

Linguistically, Zhuang languages are grouped into northern and southern dialects To characterize its genetic structure, 13 East Asian-specific Y-chromosome biallelic markers and 7 Y-chromosome short tandem repeat (STR) markers were used to infer the haplogroups of Zhuang populations Our results showed that O*, O2a, and O1 are the predominant haplogroups in Zhuang Frequency distribution and principal component analysis showed that Zhuang was closely related to groups of Bai-Yue origin and therefore was likely to be the descendant of Bai-Yue The results of principal component analysis and hierarchical clustering analysis contradicted the linguistically derived north-south division Interestingly, a west-east clinal trend of haplotype frequency changes was observed, which was supported by AMOVA analysis that showed that between-population variance of east-west division was larger than that of north-south division O* network suggested that the Hongshuihe branch was the center of Zhuang Our study suggests that there are three major components in Zhuang The O* and O2a constituted the original component; later, O1 was brought into Zhuang, especially eastern Zhuang; and finally, northern Han population brought O3 into the Zhuang populations

Key words: Y chromosome; Zhuang; internal genetic structure

Received: 2006-02-23; Accepted: 2006-04-07

This work was supported by the Priority Project of the National Natural Science Foundation of China (No 39993420), the Science Foundation of Guangxi Province (No.GSN0339041, GSY0542044) and the Genographic Project of National Geography

① Corresponding authors LI Hui, E-mail:LiHui.Fudan@gmail.com,Tel:+86-21-6564 2419;

PAN Shang-Ling, E-mail: s.pan@gxmu.net.cn ,Tel: +86-771-535 8292

The nonrecombining portion of the human

Y-chromosome (NRY), which is paternally inherited

and does not undergo recombination during cell

divi-sion, is prone to form population-specific

polymor-phisms In addition, single nucleotide polymorphisms

(SNP) on the Y-chromosome, which has a lower

probability of recurrent mutations and higher

reliabil-ity, are more group- and area-specific and can record

more accurately human historical migrations and

evolutionary events, a quality because of which they

are rapidly being accepted as one of the most

effec-tive markers for studying human evolution and origin[1,2] Using the techniques of denaturing high-performance liquid chromatography (DHPLC) and single-stranded conformation polymorphism (SSCP),

Underhill et al.[1,2] have investigated several Y-chro- mosome biallelic markers from populations world-wide in the past and constructed 131 Y-haplogroups and mapped human evolution genealogy In Asia, on the basis of 19 East Asian-specific polymorphic

markers on the Y-chromosome, Su et al.[3,4] estab-lished 17 Y-haplogroups, 7 of which were specific to

the East Asian population By clearly tracing the

pa-ternal migration route in East Asia and the Pacific

Region, Su et al found that the South-Asian group

had more Y-haplogroups compared with the North-

Asian group, indicating that the East Asians

origi-nated from the south On the basis of the data of Su et

southern China using these Y-SNP markers and

ex-plained their origin, migration, mixture, and evolution,

thereby adding important genetic information and

evidence to the origin of these groups

With a population of more than 16 million,

Zhuang is China’s largest minority population, with

94% of its population living in the Guangxi

autono-mous region The Zhuang language belongs to the

Kam-Tai linguistic family, the Tai-Kadai sublinguistic

family, and the Tai-Sek branch[7], which can be

sub-classified into Southern and Northern dialects

bounded by Yongjiang River (for details on the

dis-tribution of the Zhuang branches, see Fig 1) It

should be noted that the Bouyei ethnic group in the

Guizhou Province actually belongs to the same

popu-lation as Zhuang, as shown by their language and culture, and the so-called Shui Hu in the Yunnan Province, which is completely different from Shui people in Guizhou, is in fact Bouyei Therefore, all of these ethnic groups are considered Zhuang academi-cally Unfortunately, the Zhuang population does not have its own written script and has to use Han char-acters to record events historically, and these records might be incomplete On the basis of the few avail-able historical records, Zhuang can be traced back to the ‘Luo-Yue’ and ‘Xi’Ou’ groups, 2000 years ago[7] However, the origin of Zhuang might be more com-plicated than expected because Zhuang might have experienced complex evolvement and migration and

to a great extent, may have a close relationship with the origin of Thai and Lao in southeast Asia At the same time, Zhuang might be mixed with other sur-rounding ethnic groups, especially with Han There-fore, many questions emerge: what is the exact origin

of Zhuang? Is there any genetic evidence to support their historical migration events? Is it reasonable to classify Zhuang into South and North groups just

Fig 1 Distribution of Zhuang branches

North Zhuang dialect group: Guibei, Liujiang, Hongshuihe, Yongbei, Youjiang, Bouyei, Qiubei, Nhang, Tai-Mène South Zhuang

dialect group: Yongnan, Tày, Man-Caolan, Nung, YanGuang, WenMa, E (Wuse), Tsün-Lao

based on dialects without any genetic indication? Can

the genetic nature of Zhuang provide any information

for the study of the origin and migration of the

sur-rounding populations? Furthermore, when the early

east-Asian population moved northward via the

Guangxi region, were there any original genetic

ma-terials retained in the native Zhuangs?

Very few published reports are available about

Zhuang’s paternal genetic structure, and most of these

have mainly focused on some of its special branches

In this study, the Y-haplogroups were typed for eight

branches of the Zhuang population with 13 biallelic

markers and 7 Y-chromosome short tandem repeats

(STR), and for every haplogroup, the frequency was

calculated, the principal component analysis was

car-ried out, and the heredity framework was drawn The

authors of this study hope to determine Zhuang’s

in-herited structure at a genetic level and to provide

ge-netic data for further studies on Zhuang’s linguistics,

origin, transformation, and admixing with other

eth-nic groups

1 Materials and Methods

1 1 Sample collection and DNA extraction

A total of 129 blood samples were collected

from eight representative Zhuang-living areas in the

Guangxi Province, each representing eight Zhuang

branches DNA was extracted from white blood cells

using the traditional phenol-chloroform method[8]

Table 1 shows the details on the different branches

and the sample sizes All individuals who were

Zhuang for more than three generations and were

unrelated healthy males were asked to sign informed

consent at the time of recruitment

1 2 Y-chromosome biallelic typing

Two strategies were introduced to type Y-SNP

For SNPs with length variation, i.e., deletion or

inser-tion, fluorescence PCR (primer information shown in

Table 2) was used, the obtained product was

electro-phoresed on a 3100 genetic analyzer (ABI company,

USA) to determine the individual’s genotype For

Table 1 Sample distribution of eight branches of Zhuang from Guangxi province

Branches Abbreviation Size

Youjiang YJ 5

Hongshuihe HSH 39

Yongnan YN 19 Zuojiang/Tày ZJ 15 Dejing/Nung DJ 3

SNPs without length changes, i.e., substitution or transversion, PCR-RFLP (Restriction Fragment Length Polymorphism) was used[4,6,8- 10] The Y- hap-logroup of every subject was determined by the inte-grated analysis of the Y-SNP typing results of the two typing methods

For PCR, a primers mix was used, which con-tained M175, 0.04 μL; M121, 0.03 μL; M134, 0.03 μL; M117, 0.06 μL; M111, 0.08 μL; and M15, 0.02

μL Each PCR reaction (volume 5 μL) included KodDash polymerase (TOYOBA) 0.5 U, 0.26 μL Primer Mix, 10 ng of genomic DNA, and buffer The PCR conditions were as follows: 30 cycles of 98℃

for 10 s, 55℃ for 2 s, 74℃ for 2 s, and a final exten-sion at 4℃

To test the degree of variation among different branches, Y-STR was typed using the same method that was used for typing Y-SNP Information on Y-STR primers is also shown in Table 2 The volume

of each PCR system was 5 μL, containing KodDash polymerase, 0.05 μL; 10× buffer, 0.5 μL; dNTP (2.5 mmol/L each), 0.4 μL; genomic DNA, 10 ng; mixture

of primers, 0.4 μL Usually, PCR reaction was carried out in two panels Panel 1 included primer pairs DYS389, 0.05 μL × 2 (forward and reverse); DYS390, 0.07 μL × 2; DYS391, 0.08 μL × 2; with a total vol-ume of 0.4 μL Panel 2 included DYS388, 0.05 μL × 2; DYS392, 0.07 μL × 2; DYS393, 0.02 μL × 2;

DYS19, 0.06 μL × 2

Table 2 Y-Chromosome SNP, STR fluorescence primers

Product size (bp)

Wide type Mutant type M121 F:ACAAAGACCTGGACAGATTAC

R:CCCTTAAAAACAGCATGATA

FAM 123 118 M117 F:GTACGAAGAAAATCAAGGCTATTA

R:TTGGGTAGAAAAACTGCAAGTAG

FAM 317 313 M175 F:TTGAGCAAGAAAAATAGTACCCA

R:TTCAGTTAGCCTTGATTGACTGT

FAM 226 221 M134 F:AGAATCATCAAACCCAGAAGG

R:TCTTTGGCTTCTCTTTGAACAG

NED 232 231 M15 F:ACAAATCCTGAACAATCGC

R:GTCTGGGAAGAGTAGAGAAAAG

FAM 151 142 M111 F:TAACATAAACAGTATGCCAAA

R:TGCCCTAAAGTTAATACCAG

HEX 197 195 DYS388 F:GTGAGTTAGCCGTTTAGCGA

R:CAGATCGCAACCACTGCG

FAM DYS389 F:CCAACTCTCATCTGTATTATCTATG

R:TCTTATCTCCACCCACCAGA

FAM DYS390 F:TATATTTTACACATTTTTGGGCC

R:TGACAGTAAAATGAACACATTGC

NED DYS391 F:CTATTCATTCAATCATACACCCA

R:GATTCTTTGTGGTGGGTCTG

HEX DYS392 F:TCATTAATCTAGCTTTTAAAAACAA

R:AGACCCAGTTGATGCAATGT

NED DYS393 F:GTGGTCTTCTACTTGTGTCAATAC

R:AACTCAAGTCCAAAAAATGAGG

HEX DYS394 F:CTACTGAGTTTCTGTTATAGT

R:ATGGCATGTAGTGAGGACA

HEX F: forward; R:reverse.

1 3 Statistical analysis

The Y-haplogroup of every individual was

de-fined based on the experimental results of the authors

of this study and the NRY haplogroup tree of

East-Asia shown in reference [11]

In practice, to ensure that an effective genetic

sample size has been obtained, for every branch,

other published data such as those on Bouyeis,

Zhuangs, Suis in Yunnan[12], Bouyeis in Guizhou [13],

and unpublished data (Fudan University, data not

shown) such as those on Tianlin Zhuangs, Shangsi

Zhuangs, Wuse Zhuangs, and Man-Caolan Zhuangs

in Guangxi were added to the relevant Zhuang

branches

Hierarchical clustering analysis was carried out

to show the genetic distance (affinity) among

Guangxi Zhuang branches, using SPSS13.0 by

calcu-lating the frequency of different haplogroups in every

branch The phylogeny relationship among Zhuang

branches was carried out by combining analysis of hierarchical clustering, principal component analysis, and association analysis to determine the association between phylogeny and Y-chromosome haplogroups; this was then observed using the gradient distribution chart which was drawn using Surfer7.0 software for every principal component according to its geo-graphic distribution, in which principal component value served as the height values

Variances among and within populations result-ing from Y-haplotype frequencies of different

branches were calculated using AMOVA (Analysis of

MOlecular VAriance framework) in Arlequin1.1 software to further elucidate the phylogeny of differ-ent Zhuang branches Finally, Zhuang’s genetic framework was drawn using the network 11.0 soft-ware with the same Y-STR haplogroups of different branches to show the detailed difference and associa-tion among Zhuang branches

2 Results

2 1 Distribution of NRY haplogroups in Zhuang

branches

The haplogroup frequencies of different Zhuang

branches stemming from the typing results of Y-SNP

were calculated As shown in Table 3, Zhuang’s

Y-haplogroups mainly cluster around O*, O1, O2a,

and O3, the four most common haplogroups in East

Asia Haplogroup O* is the most frequent, followed

by O2a and O1, showing that Zhuang is a typical

southern group in East Asia and possesses more

an-cient Y-haplogroups Interestingly, O3, O3e, and

O3e1, the characteristic haplogroups for East Asian

northern group were also frequently found in

Zhuangs, showing a common gene communication

between the two populations

2 2 Principal component analysis

Principal component analysis was carried out

using SPSS 13.0 software, and the principal

compo-nent dot plot of Y-SNP frequencies was drawn (Fig 2)

according to Y-SNP typing results of different

branches, as obtained by the authors of this study, and

additional data of Yunnan Bouyeis, Zhuangs, Suis,

Guangxi Tianlin, Shuangsi, Wuse, Man-Caolan

Zhuangs, and Guizhou Bouyeis In this analysis,

Tianlin was merged with the Youjiang branch,

Shangsi was merged with the Zuojiang branch, and

Bouyei with the Guibian branch

The result of principal component analysis showed that the cumulative contribution of principal component 1 (pc1) and component 2 (pc2) accounted for 82.5% of the total difference It is obvious from Fig 2 that 12 Zhuang branches and other correlated groups mainly gathered into two bigger groups Hongshuihe, Guibei, Yongbei, Yongnan, and Man-Caolan were located in the upper part of the principal components plot and constituted the first group, whereas Guibian, Zuojiang, Youjiang branch, Yunnan Zhuang, and Yunnan Bouyei clustered in the lower part of the chart, and constituted the second group Yunnan Shui Hu seemed to be isolated from all branches, but was somewhat closer to the second group on pc2 As shown in Fig 2, it is the pc2 that separated these branches into two groups Consider-ing the geographic location of each branch, it was observed that the difference between the two groups was, genetically, an east–west rather than a south-north profile, contradicting the traditional south-south-north grouping of Zhuangs by linguistic factors

Using the values of pc1 and pc2 as height values, the principal component gradient distribution diagram was drawn in the same manner as drawing relief map with contours according to the geographic location of each Zhuang branch (Fig 3) In the relief map of pc1,

it was observed that the peak value was near the Guangxi-Vietnam border, which gradually changed

Table 3 Y-SNP haplogroup frequencies of branches of Zhuang

Branches Size C D D1a F* K* O O1 O2a* O2a1 O3 O3a O3e* O3e1

YB 23 8.70 4.35 4.35 21.704 17.39 8.70 21.74 8.70 4.35 4.35

HSH 39 2.56 5.13 5.13 23.08 10.26 5.13 7.69 20.51 5.13 15.38

GN 21 4.76 4.76 4.76 38.10 4.76 4.76 4.76 14.29 14.29

YN 19 5.26 5.26 15.79 10.53 10.53 10.53 31.58 5.26 5.26

Fig 2 The principal components plot of Y-SNP frequencies of Zhuang populations

Fig 3 Geographic map of Y-SNP principal components of Zhuang populations

X coordinate: longitude; Y coordinate: latitude The geographic map of Y-SNP principal components of Zhuang branches was drawn using principal component value as contour, wherein, the lower the value, the darker of the color For pc1 (left), the peak value was near the Guangxi-Vietnam border, which gradually changed north-eastward and north-westward The peak value of pc2 that classifies Zhuang branches into two main groups appeared in the east Hongshuihe basin (right), exhibiting an east-to-west gra-dient It is quite clear that contour lines run along rivers, indicating that native Zhuangs migrated upstream along rivers in Guangxi

in early times

north-eastward and north-westward This might be a

clue to the spreading of the East Asian population in

ancient Guangxi when it first entered East Asia

Be-cause pc2 classifies Zhuang branches into two main

groups, its significance might be more definite The

peak value of pc2 appeared in the east Hongshuihe

basin, exhibiting an east-to-west gradient In addition,

a higher value of pc2 was seen in northwest Guangxi

bordering the Yunnan Province This might be

influ-enced by data of the Yunnan Bouyeis Furthermore,

from Fig 3 it is quite clear that contour lines run

along rivers, indicating that native Zhuangs migrated

upstream along rivers in Guangxi in early times

2 3 Correlation analysis

To understand the meaning of each principal

component, correlation analysis was carried out to

seek the origin of each principal component Each

principal component was sorted out after calculating

the correlation coefficient for every haplogroup such

that the positively and negatively correlated

hap-logroups of the principal components could be

ob-served Theoretically, the more positive correlated

haplogroup a principal component has, the more

dis-tinct is its genetic structure, and the more practical

meaning it has For more detail, see Figs 4 and 5

Apparently, the structure of pc2 is clearer than

that of pc1, after comparing the correlation

coeffi-cient values of the two principal components There

were many contradictions in the variables among

positively correlated haplogroups in pc1, whereas in

pc2, most positively associated variables fell within

the positive correlation area Moreover, pc2 was

markedly positively correlated with longitude,

show-ing that pc2 was more significant than pc1

Further analysis of pc1 showed that the number

of positive correlated haplogroups, despite their weak

correlations, was bigger than that of the negative

groups This is attributed mainly to the difference

between O2a and O3e, where O2a was a southern

aboriginal haplogroup in the East Asian population

and O3e was probably a northern haplogroup In this

study, it was observed that pc1 was positively

associ-ated with O2a (r = 0.69, p = 0.02) and that the values

of pcl were all positive Therefore, the meaning of pc1 underscores the positive correlation of Zhuang branches and Yunnan Shui Hu with O2a, showing that these ethnic groups are all typical southern groups of East-Asian population, a conclusion that is consistent with the historical records of Zhuang’s Baiyue origin

For pc2, the number of positively correlated haplogroups was almost the same as that of the nega-tive ones, with an opposite trend being found between O* and O2a: haplogroup O* was positively but O2a was negatively associated with pc2 Both pc2 and haplogroup O* were clearly related to longitude, im-plying that haplogroup O* was the main component

of pc2, i.e., it is O* that separates the Zhuang branches into two main groups With the westward movement of haplogroup O*, O2a retreated in the same direction This process agrees with the lower frequency of O2a in east Zhuang branches

No significant correlation was found between other haplogroups and the two principal components

2 4 Hierarchical clustering analysis

To further elucidate the relationship among Guangxi Zhuang branches, hierarchical clustering analysis was carried out with average linkage (be-tween groups) in SPSS 13.0 software and the results are shown in Fig 6, in which Mien and Yi are the foreign groups It is clear that the center of all Zhuang branches emerges in the Hongshuihe area, evolving gradually toward Yongbei, Yongnan, and Guibian along the Hongshuihe river, then toward Guibei and Man-Caolan, and finally toward Zuojing, Youjiang, and Yunnan, providing strong evidence once again that the difference among Zhuang branches is funda-mentally east-west and not south- north, as tradition-ally believed Intriguingly, Wuse Zhuangs living in Yongle, Rongshui, a northwest county in Guangxi, are genetically close to the remote Zuojiang Zhuangs who live in southwestern Guangxi, implying a special moving event in ancient times

Fig 4 Correlations among principal components, Y-SNP frequencies, longitude, and latitude of Zhuang branches

Principal components were ranked after their correlation coefficient for relevant haplogroup were calculated to estimate the posi-tively and negaposi-tively correlated haplogroups of the principal components Each correlation coefficient was marked with relevant color according to its value There were many contradictions in the variables among positively correlated haplogroups in pc1 (left), whereas in pc2, most positively associated variables fell within the positive correlation area (right).

Fig 5 Correlations and statistical significance among principal components, Y-SNP frequencies, longitude, and latitude of Zhuang branches

r: correlation coefficient; P: probability value; long.: longitude; lat.: latitude Chromatism staff gauge shows different p values pc1

was positively associated with O2a, a typical south aboriginal haplogroup in East Asian population, underscoring the positive cor-relation of Zhuang branches and Yunnan Shui Hu with O2a For pc2, the number of positively correlated haplogroups was almost the same as that of the negative ones, with an opposite trend being observed between O* and O2a, in which haplogroup O* was positively but O2a was negatively associated with pc2

Fig 6 Dendrogram of Y-SNP of Zhuang branches

Yi and Mien, two non-Baiyue populations living in southwest

China, were introduced to illustrate the cluster of Zhuang and

its relationship with surrounding non-Baiyue groups This

dendrogram showed that the ancestor of Zhuang gathered in

Hongshuihe area first, then spread toward Yongjiang and

You-jiang basin, and the surrounding area Genetically, Zhuang was

much far away from Yi and Mien

2 5 AMOVA analysis

Linguists divide the Zhuangs into the southern

group and northern group, according to their dialects

If this classification has a genetic basis, the variance

among populations should be greater, whereas the

variance within populations should be smaller in the

molecular variance framework analysis As suggested

by the pc2 results, the AMOVA results were

com-pared between the southern and northern groups and

between the eastern and western groups (Table 4)

However, no differences were observed, as once

an-ticipated Instead, the variance among populations of

the eastern-western groups was far greater than that

of the northern-southern groups, whereas the variance

within populations was quite the opposite, further suggesting that the classification of east and west Zhuang is more reasonable than that of the south- north grouping if these groups need to be distin-guished However, because the east-west difference is graded due to Zhuang population migration, it is very difficult to differentiate Zhuangs into two distinct groups genetically

haplogroup of Guangxi Zhuang branches

The combination of several short tandem repeats (STR) in NRY comprises another type of Y- chromo-some haplogroup, which reflects minute genetic variances in different populations with the same Y-SNP This differs from Y-SNP haplogroup, which can only be roughly used in the classification of sys-tematic phylogeny After sorting out different Zhuang branches on the basis of Y-SNP information, the ge-netic network was drawn using Network 11.0 with Y-STR data from individuals of all branches to show the relationship and differences between branches with the same Y-SNP haplogroups

As shown in Fig 7, the information of STR network of haplogroup O* is more abundant due to its higher frequency in all Zhuang branches Nine individuals from Hongshuihe branch with haplogroup O* had 8 Y-STRs, the highest Y-STRs frequency of all, three of which were shared by Guibian, Guibei, and Yongbei individuals In addition, Hongshuihe Zhuangs were closely related with individuals from other haplogroups, especially Yongbei and Guibei individuals, establishing its central position in all

Table 4 Results of AMOVA of northern-southern group and eastern-western group of Zhuang branches

Eastern–western Northern-southern Eastern: HSH, YN, YB, CL, GN Northern: YB, YJ, HB, HSH, GB

Groups

Western: ZJ, YJ, GB, E, DJ Southern: YN, ZJ, DJ, CL, E

Fig 7 STR network of haplogroups O*, O1, O2a, and O3* of Zhuang branches

In the network of haplogroups O* and O1, Hongshuihe branch had more STR haplogroup polymorphisms and was closely related with others, especially Yongbei and Guibei individuals, establishing its central position in all Zhuang branches In the O2a network, Zuojiang branch had more Y-STRs In the O3* network, no shared STR haplogroup was seen.

Zhuang branches Only one Guibei individual shared

the same Y-STR haplogroups with that of Hongshuihe

although Guibei branch also had 8 STR haplogroups

that were distributed at the edge of the O* haplogroup

network More linkage was also found among

Zuoji-ang, YoujiZuoji-ang, and Dejing branches, which, as a

closed group, was more isolated from others Taken

together, all these findings verify and agree with

re-sults from the principal component analysis and the

above-mentioned clustering analysis

Likewise, in the O1 network, Hongshuihe branch had more STR haplogroup polymorphisms than others and occupied the center of the framework However, it failed to maintain its leading position in O2a and gave way to Zuojiang branch as a result of fewer individuals carrying the O2a haplogroup There was no shared STR haplogroup or any remarkable difference among branches in the O3* network One possible explanation is that O3* is not the dominant and characteristic haplogroup in Zhuangs