E. Gomez-Casado J. Martnez-Laso J. Moscoso J. Zamora M. Martin-Villa M. Perez-Blas M. Lopez-Santalla P. Lucas Gramajo C. Silvera E. Lowy A.

Arnaiz-Villena
Key words: Amerindians; Esquimo; HLA; Mayans; Na Dene Acknowledgments: This work was supported by grants from the Spanish Ministry of Science PM-199923 and BMC-20011299.

Origin of Mayans according to HLA genes and the uniqueness of Amerindians

Abstract: The HLA allele frequency distribution of the Mayans from Guatemala was studied and compared with those of other First American Natives and worldwide populations (a total of 12,364 chromosomes and 6182 individuals from 60 different populations). The main conclusions were (1): the closest Amerindian group to Mayans is the Arhuacs, who were the first recorded Caribbean Islands inhabitants (2). Mayans are not so close to Mesoamerican Zapotec, Mixe and Mixtec Amerindians, who genetically cluster together. Mixe had been related to Mayans only on linguistic bases (3). DRB1*0407 and DRB1*0802 alleles are found in 50% of Mayans; these alleles are also found in other Amerindians, but the Mayans high frequencies may be showing a founder effect for this Mesoamerican-Caribbean population (4). Extended Mayan specific HLA haplotypes are described for the first time (5). Language and genes do not completely correlate in microgeographical studies (6). Significant genetic input from outside is not noticed in Meso and South American Amerindians according to the genetic analyses; while all world populations (including Africans, Europeans, Asians, Australians, Polynesians, North American Na-Dene Indians and Eskimos) are genetically related. Meso and South American Amerindians tend to remain isolated in the neighbour joining analyses.

Authors affiliations: E. Gomez-Casado1, J. Martnez-Laso1, J. Moscoso1, J. Zamora1, M. Martin-Villa1, M. Perez-Blas1, M. Lopez-Santalla1, P. Lucas Gramajo2, C. Silvera1, E. Lowy1, A. Arnaiz-Villena1
1

Department of Immunology and Molecular Biology, H 12 de Octubre, Universidad Complutense, 28041 Madrid, Spain Department of Laboratory, Hospital Regional de Xela, Quetzaltenango, Guatemala Correspondence to: Antonio Arnaiz-Villena Departamento de Microbiologa I (Inmunonolog a) Facultad de Medicina Universidad Complutense 28040-Madrid Spain e-mail: aav@efd.net website: http://chopo.pntic. med.es/biomol

It has been established that more than one wave of people gave rise to the American Indians (1, 2). The timing of the peopling of the Americas is under debate, but the most ancient human settlements are found in South and not North America (Monteverde, Chile; Pedra Furada, Brazil). Greenberg put forward the theory that American Indians came from Asia through a Bering land bridge 30,000 6000 years ago in three waves (1, 3): Amerindians (most North and South American Indians) (2), Na-Dene speakers, including Athabascans, Navajo and Apache Indians, and (3) Eskimos. This has come under revision because (a) Asian haplotypes are rarely seen if
Received 2 October 2002, revised 24 December 2002, accepted for publication 28 December 2002 Copyright Blackwell Munksgaard 2003 Tissue Antigens. ISSN 0001-2815 Tissue Antigens 2003 61: 425436 Printed in Denmark. All rights reserved

The contribution by E. Gomez-Casado and J. Martnez-Laso was equal and the order of authorship is arbitrary.

425

Gomez-Casado et al : HLA genes in Mayans

DQB1 and DQA1 are taken into account, (b) the common Asian B blood group is not found among Amerindians (2, 3), (c) Meso and South America Amerindians tend to remain isolated in all genetics analyses carried out (2), and (d) it seems probable that more than one peopling wave existed (1, 3). The Amerindian Mayans lived in the area represented in Fig. 1 when the Spanish invaders arrived in their lands in 1523. Pedro de Alvarado (Hernan Cortes lieutenant) was put in charge of the Mayan land conquest, which took 20 years (4). Swadesh (5) has classified the following languages within the Macro-Mayan group: Leuca (Honduras), Xinca (El Salvador), Totonac (Veracruz, Mexico), Mixe (Oaxaca, Mexico), Huastec (North Gulf coast, Mexico) and approximately 28 Mayan languages spoken in the area depicted in Fig. 1 in addition to some already extinguished ones. The Mayan culture is considered a continuation of the Olmec and other more ancient MesoAmerican cultures (4), but they developed a sophisticated writing

system which has not been deciphered. Only, the Spanish priests took care of teaching (5) how to write in Spanish the Mayan language. Two types of books have brought us most of the Mayan cosmogony and others histories: (a) a collection of books found in the Yucatan Peninsula (Chilan-Balan) and (b) one manuscript found in Chichicastenango (Guatemala): the sacred Popol-Vuh. Numbers and other doubtful translations have been put forward (6, 7). The Mayan civilization flourished between 1500 years BC and 1400 AD (8) when Toltecs firstly and Aztecs later helped to destroy the Mayan civilization. The Spanish conquerors found a very impoverished Mayan civilization compared with the Aztecan one. Nowadays, Mayan languages are spoken in Yucatan, Mexico, Guatemala, Honduras and El Salvador (Fig. 1). In the present work we aim (1): to determine the HLA class I (A and B) and class II (DRB1 and DQB1) Mayan alleles by using DNA typing, including automated sequencing, and (2) to compare the Mayan HLA profile with the other First American Natives and other worldwide populations in order to investigate the origin of the Mayans and the much-debated peopling of the Americas.

Materials and methods

Population samples
One hundred and thirty-two healthy unrelated Maya individuals from Quetzaltenango village (Xela in Mayan language, Guatemala, Fig. 1) were used for HLA genotyping and phylogenetic calculations. Each individual was born in the Guatemala Mayan area and had a Mayan appearance. Their four grandparents had been born in the same area and spoke the Mayan language. The origin of all other populations used for comparisons is detailed in Table 1. In total 12,364 chromosomes were studied, including populations from different origins (Caucasoids, Orientals, Negroids, Polynesians, Micronesians, Na-Dene, Eskimos and Amerindians). In particular, the Amerindian group includes tribes from the linguistic families of Macro-mixteco (Mixtecan and Zapotecan), Macro-Maya (Mixe), Macro-Yuma (Seris), Chibcha (Arsario, Kogi, Arhuaco and Cayapa), and Ge Pano Caribe (Xavantes, Mataco and Toba) (9, 10).

HLA genotyping, DNA sequencing and statistics

Generic HLA class I (A and B) and high resolution HLA class II (DRB1 and DQB1) was performed by using a reverse dot-blot technique with the Automated Innolipa system (Innogenetics N.V., Zwijndrecht, Belgium). HLA-A, -B, -DRB1, and -DQB1 allele DNA
Fig. 1. Map showing the geographic location of the Mayan population studied and Mayan languages. White names indicate countries. Black boxes with white letters indicate capital cities.

sequencing was only performed when indirect DNA typing yielded ambiguous results (11). Statistical analysis was performed with Arlequin v1.1 software kindly provided by Excoffier and Slatkin (12).

426

Tissue Antigens 2003: 61: 425436

Gomez-Casado et al : HLA genes in Mayans

Populations used for the present work Id. Number 1 2 3 4 5 6 7 Id. Number 32 33 34 35 36 37 38

Population Mayans Mazatecans Seri Mixe Mixteco Zapoteco Mexican Mestizo

n 132 89 100 55 103 75 99

Reference Present Study 2 25 25 25 25 Vargas-Alarcon et al. (unpuplished results)

Population Japanese Khalk-Mongolians Tuvinians Khoton-Mongolians Germans Sardinians Italians

n 493 202 197 85 295 91 284

Reference 16 59 60 59 16 16 16

8 9 10 11 12 13 14 15 16 17 18

Wayu Arhuaco Kogi Arsario Cayapa Xavantes Guarani Toba-Pilaga Mataco-Wichi Eastern-Toba Jaidukama

112 123 67 20 100 74 32 19 49 135 39

56 56 56 56 57 44 25 44 44 44 Martinez-Laso et al. (unpublished results)

39 40 41 42 43 44 45 46 47 48 49

French Spaniards Spanish-Basques Algerians Berbers (Souss) Moroccans Western Samoa Madang Rabaul New Caledonia Fidji

179 176 80 102 98 98 102 65 60 65 57

16 61 61 62 63 64 65 66 66 66 66

19 20 21 22 23 24 25 26 27 28 29 30 31

Eskimos Athabascans Tlingit Nivkhs Udegeys Koryaks Chukchi Kets Evenks Singapore-Chinese Buyi Manchu Koreans

35 124 53 32 23 92 59 22 35 71 70 50 100

58 51 16 58 58 58 58 58 58 16 16 16 16

50 51 52 53 54 55 56 57 58 59

Papua New Guinean Central Desert Yuendumu Kimberley Cape York South American Blacks North American Blacks Hottentots San (Bushmen) Ainu

57 152 119 82 80 59 447 65 103 50

66 67 67 68 68 16 16 16 16 69

n number of individuals analyzed for each population.

Table 1

In summary, this program calculated HLA-A, -B, -DRB1, and -DQB1 allele frequencies, HardyWeinberg equilibrium and the linkage disequilibrium between two alleles at two different loci. Linkage disequilibrium [D , also named LD (13)] and its level of significance (p) for 2 2 comparisons were determined using the formulae of Mattiuz and coworkers (14) and the 11th International Histocompatibility Workshop methodology (13).
0

In addition, the most frequent complete haplotypes were deduced following a methodology used in the 11th International Histocompatibility Workshop (1): the 2, 3, and 4 HLA loci haplotype frequencies (13, 15) (2); the previously described haplotypes in other populations (13); and (3) haplotypes if they appeared in two or more individuals and the alternative haplotype was well defined (13). To compare phenotype and haplotype HLA frequencies with other populations,
Tissue Antigens 2003: 61: 425436

427

Gomez-Casado et al : HLA genes in Mayans

the reference tables of the 11th and 12th International HLA Workshops were used (16 and 17; also see Table 1). Phylogenetic trees (dendrograms) were constructed with the allelic frequencies by using the Neighbour-joining (NJ) method (18) with the standard genetic distances (SGD; 19), by using the software DISPAN, which contained the programs GNKDST and TREEVIEW (20, 21). Correspondence analysis in three dimensions and its bidimensional representation was carried out by using the VISTA v. 5.02 computer program (22, http:/forrest.psych.unc.edu) with the DA genetic distance (20, 21, 22,). Correspondence analysis consists of a geometric technique that may be used for displaying a global view of the relationships among populations according to HLA (or other) allele frequencies. This methodology is based on the allelic frequency variance among populations (similarly to the classical components methodology) and on the display of the statistical visualization of the differences.

With regard to HLA class II alleles, 28 different DRB1 alleles were found in the Mayans. Two of these alleles are very frequent in Amerindians, and are remarkably frequent in Mayans, accounting for more than 50% of the total: DRB1*0407(38.6%) and DRB1*0802(18.9%). The other class II alleles had a frequency below 5% (see Table 2). In general, the Amerindian DRB1 alleles found in Mayans are also present in Meso and South-American Indians with a different distribution: DRB1*0408 is also present in Zapoteco; DRB1*0410 was also found in Mixe, Zapoteco and Mixteco and was initially described in one Australian aborigine; and DRB1*0411 is also present in many South-Americans and Meso-Americans but Mixe and Seri (2, 25, 27). On the other hand, the presence of DRB1*03011, *0402 and DRB1*07 alleles indicate that our Mayan sample has had an admixture with Europeans (probably Spaniards). DQB1 allele frequencies reflect the DRB1 locus allele distribution as a result of the strong linkage disequilibrium between these two loci. The expected heterozigosity is higher than observed for HLA-A (P 0.033), -B (P 0.000009), and -DRB1 (P 0.000009) loci. This

Results
Characteristic HLA allele frequencies of the Mayan population compared with other populations
Table 2 shows the HLA class I (A and B) and class II (DRB1 and DQB1) allele frequencies found in the Mayan population. Some of the class I alleles have been fully characterized and they are included as such in order to increase the information about HLA genetic profile of the Mayan population. Fourteen different HLA-A alleles were found in the Mayan ethnic group. HLA-A*02, A*24, A*31 and A*68 had frequencies higher than 10%, and these are the most frequent in other Amerindians (2). Four out of the 18 HLA-B alleles found in Mayans were the most frequent ones found in other Amerindian groups; B*15, B*40 and especially B*35 and B*39, which have a frequency higher than 63%. These alleles comprise several subtypes at the DNA level; the complete characterization of these alleles in Amerindians would probably distinguish different Amerindian groups. Some of the class I alleles found in Mayans are: A*2403, which has also been found in Caucasoids and Orientals (23, 24, 25); A*2404 and A*2410, both of which are also present in Orientals (23); A*2414, which has an unknown origin (23); A*6805. which is also present in other Amerindians (23); B*1507, initially described in American Indians (23); B*1522, which is widely distributed in Amerindians (26, 27); B*4008, which is present also in Caucasoids (23); B*4011, also found in Aztecs (28); B*5108 and B*5703, which are found in Caucasoids and Blacks/Caucasoids, respectively (23); B*3512 and B*3514, also found in Meso-Americans (26, 27, 28); and B*3520, which is only found in Terena Indians from SouthWest Brazil (24).

fact may be because of the high frequency of a few HLA-A, -B and DRB1 alleles and haplotypes, a certain degree of recent admixture, or, less likely, selection. In fact a 19.5% of the DRB1 alleles seems to be of non-Amerindian origin in the present Mayan population. Two types of analyses were carried out in order to compare Mayan HLA frequencies with those of other populations (1): with pooled generic DRB1 and DQB1 data, and (2) with high-resolution DRB1 data. It was not possibly to carry out a study comparing HLA class I (A and B) allele frequencies or HLA class I and II conjointly because of the lack of class I studies in many Amerindians. The single DRB1 study was carried out in order to compare the American Indian HLA population frequencies with those of Polynesians, Melanesians and Micronesians who lacked DQB1 analyses (see Table 1). HLA-DRB1, -DQB1 NJ tree (data not shown) grouped together the Amerindians and separated them from Na-Dene and Eskimo Native American groups, and also from the Orientals and Caucasoids; this is also seen in the correspondence analysis (Fig. 2). When the Polynesians, Melanesians, Micronesians and Negroids are included (see Table 1) the topology of the tree (DRB1 alone) does not change (Fig. 3). Genetic distances between Mayans and other populations show that Mayans are both close to Meso-Americans and South-Americans (see Table 3 and Fig. 4). Indeed, Arhuaco (Colombia, Chibchan family) show the closest genetic distance to Mayans, followed by Mixteco (Mexico, Oto-Manguean family), Mexican-Mestizo, Kogi (Colombia, Chibchan family), Seri (Mexico, Hokano-Huitlaka family), Mazatecans and Zapoteco (Mexico, Oto-Manguean family), Wayu (Colombia, Macro-Arhuac family), Eastern-Toba (Brazil), Arsario (Colombia, Malayo family), Mixe (Mexico, Mixe-Zoque family), Cayapa (Ecuador, Barbacoan family), etc. North American Na-Dene (Athabascans) are
Tissue Antigens 2003: 61: 425436

428

Gomez-Casado et al : HLA genes in Mayans

HLA-A, -B, -DRB1 and -DQB1 allele frequencies in the studied Mayan population Alleles
HLA-A*

Allele frequencies %

Alleles
1801 35

Allele frequencies %
0.3 40.1 0.3 0.3 1.9 0.3 1.1 1.5 0.7 16.3 1.1 0.3 0.3 4.2 1.1 0.3 0.7 3.0 0.3 0.3 0.3 0.7 1.1 3.4 0.3 0.3 0.3 1.5 0.3 0.3 0.3 0.7 0.7

Alleles
0407 0408 0410 0411 07 0801 0802 08041 09012 1001 1101 1104 1201 1301 1302 1303 1304 1401 1402 1501 1602

Allele frequencies %
38.6 0.7 0.7 2.6 4.5 0.4 18.9 0.4 0.7 0.7 1.5 0.7 0.4 1.1 1.5 0.4 0.4 1.1 4.1 2.3 2.6

01 0101 02 03 0301 0302 11 2301 24 2403 2404 2410 2414 2501 26 29 3001 3002 31 31012 33 68 6801 6802 6803 6805 8001 Nd

1.5 0.7 36.3 1.1 0.4 0.4 1.5 0.4 21.2 0.7 0.4 0.4 0.4 0.4 0.7 2.3 0.7 0.7 0.7 10.2 1.5 4.5 1.9 1.1 7.9 0.4 0.4 0.7

35/15 3508 3512 3514 3517 3520 3801 39 3905 3906 3908 40 4002 4008 4011 44 44031 4501 49/50 4901 50 51 5101 5102 5108 5301 57 5701 5703

HLA-DQB1*

02 0301 0302 03032 0304 0305/0402 0402 0501 05031 0602

6.4 13.3 48.1 11.4 0.8 1.5 19.7 3.0 1.5 1.9 1.9 0.7 0.3

HLA-B*

5801 nd

07 0702 08 0801 13 14 1402 15 1507 1517 1522

1.5 0.7 0.3 1.5 0.7 0.7 1.1 1.1 1.1 0.3 4.2 0101 0102 03011 04 0402 0403 0404 0405 0.7 1.1 2.3 0.7 0.4 2.7 3.0 0.4 HLA-DRB1*

0603 0604 nd

Alleles DQB1*0201 and 0202 were all assigned as DQB1*02. Allele frequencies may not sum 100% because one decimal is considered. Not determined (ND) because of an unavailable amount of DNA.

Table 2
Tissue Antigens 2003: 61: 425436

429

Gomez-Casado et al : HLA genes in Mayans

Fig. 2. Correspondence analysis showing a global view of the relationship among Amerindian (green), Siberian, Na-Dene and Eskimo (red), Asian (orange) and European/North African populations (yellow).

as genetically distant from Mayans (and other Amerindians) as Tuvinians or Caucasoids; the same is observed for Eskimos. It is also remarkable that the genetic distance Mayans-Guarani (Brazil) is the highest found between Mayans and other Amerindians, Athabascans and Eskimos. This may be because of the admixture of the Guarani sample with Caucasoids or, otherwise, to this Mayan sample admixture with Europeans. Genetic distances, NJ and correspondence analyses indicate that Mayans do not follow a clear pattern of close language and geographic relationships; genes and languages do not correlate at the microgeographical level (29)

frequency in Meso and South-American Indians: Chiriguanos (Argentina), Wayu (DRB1*0802-DQB1*0402 in low frequency, Colombia), Yukpa (Venezuela), Seri (Mexico), Zapoteco (Mexico), Mixteco (Mexico) and Mixe (Mexico) (25). A*02-B*15-DRB1*0404-DQB1*0302 (1.5%) has also been found to be one of the most frequent haplotypes in Mayans (Table 4). The partial DRB1*0404-DQB1*0302 haplotype was found in MesoAmerican Zapotec (15.7%), Mixtec (1.0%) and Mixe (5.8%) Indians, although the extended haplotype with A2-B39 is only present in Zapotec and Mixe (27). Also, Mazatecans show a high frequency (9.8%) of the partial haplotype DRB1*0404-DQB1*0302, which is associated to A24-B35 and gives rise the extended haplotype of

HLA-A, -B, -DRB1 and -DQB1 linkage disequilibria analysis compared with other populations
HLA-A-B, -B-DRB1, -DRB1-DQA1, and -DQA1-DQB1 two loci and -DRB1-DQA1-DQB1 with linkage disequilibrium were calculated, as described in the Materials and methods section (not shown). The nine most frequent different extended haplotypes are depicted in Table 4. Mayan-extended HLA haplotypes have been obtained for the first time, allowing their comparison with previously reported ones in other populations (Table 4 and its footnote). The DRB1*0407DQB1*0302 haplotype gives rise to seven of the most frequent haplotypes in combination with A*02-B*35, A*68-B*39, A*24-B*35, A*02-B*39, A*31-B*35, A*24-B*15 and A*68-B*35 (Table 4). These haplotypes account for 33.4% of the total haplotypes. Also, the DRB1*0802-DQB1*0402 haplotype is combined with A*02-B*35 and A*24-B*35, adding up to 12.6%. Both DRB1*0407-DQB1*0302 and DRB1*0802-DQB1*0402 haplotypes have a count of 46% of the Mayan haplotypes. In fact, these haplotypes are present in high

lower frequency [2.5% (2)]. A*02-B*44-DRB1*07-DQB1*02(1.5%), A*01-B*08-DRB1*03011DQB1*02 (0.7%) and other low-frequency haplotypes (Table 4 footnote) show that our Mayan sample have had a certain degree of admixture with Spaniards or other Europeans (19.5% according to DRB1 allele analysis). The observed high-variability class I and class II extended haplotypes are compatible with a 12% recombination rate between these two MHC regions.

Historical background and discussion Peopling of the Americas occurred in more than one immigration wave
The first Amerindian Natives are believed to have come from Asia through the Bering land bridge between 30,000 and 12,000 years

430

Tissue Antigens 2003: 61: 425436

Gomez-Casado et al : HLA genes in Mayans

Genetic distances between Mayans and other populations (x 10-2) obtained by using HLA-DRB1 high-resolution allele frequencies(see Table 1 for population identification) Population Arhuaco Mixteco Mexican Mestizo Kogi Seri Mazatecan Zapoteco Wayu EasternToba Arsario Mixe Cayapa TobaPilaga Xavantes Jaidukama Mataco Wichi Tuvinians Athabascan KhalkMongolian French German Spaniards Italians Koreans N. Amer. Blacks Japanese Moroccans Berbers (Souss) DRB1 (DA) 14.04 14.57 16.21 16.49 17.24 18.11 18.76 18.76 19.58 21.14 22.58 26.23 28.45 35.61 38.74 41.38 42.28 43.91 44.70 46.88 47.70 47.72 49.50 51.25 51.29 52.34 52.63 53.33 55.83 Population Algerians Tlingit Ainu S. Amer. Blacks Eskimos Hottentots Manchu Udegeys Chukchi Guarani Spanish-Basques Khoton Mongolian Singapore-Chinese Sardinians Koryaks Western Samoa Buyi San (Bushmen) Nivks Fidji New Caledonia Mandang Evenks Kets Papua New Guinean Cape York Central Desert Yuendumu Kimberley DRB1 (DA) 56.01 56.80 57.58 57.86 58.64 58.83 58.87 58.87 60.21 60.95 62.48 63.43 64.24 65.07 65.36 65.42 65.62 68.14 70.37 72.13 74.21 74.96 76.05 77.26 81.70 85.34 87.74 89.39 95.11

Fig. 3. Neighbour-joining dendrogram showing relatedness between Mayans and Amerindian, Na-Dene and Eskimo, Asian, Polynesians and Caucasoid/North African populations. Genetic distances between populations (DA) were calculated by using high-resolution HLA-DRB1 genotyping. Data from other populations were taken from references detailed in Table 1.

Algerians
DA genetic distances.

Table 3

populations (33, 34) supported the three-wave model, as did Wallaces before the present (BP). These conclusions have been based on cultural, morphological and genetic similarities between American and Asian populations (3). Both Siberia (3) and Mongolia (30, 31) have been put forward as the most likely places of origin in Asia. Greenberg first postulated the theory of triple migrations for explaining the peopling of the Americas (32): Amerindians (most North and South American Indians; 12,000 BP), Na-Dene (Athabascans, Navajo, Apache; 8000 BP) and Eskimo-Aleuts (6000 BP; see below, Fig. 3). Studies carried out before the widespread use of mitochondrial, Y chromosome and other nuclear DNA markers for the study of mtDNA study (35). However, other mtDNA studies have not (36, 37); more recently, only one wave coming from Mongolia/North China has been proposed as giving rise to the First Native American ancestors (30, 38). The study of Y-chromosome DNA markers seemed to suggest the existence of a single major paternal haplotype in both North and South American Native populations (39, 40). However, more recent studies of the Y chromosome have shown that more than one paternal founder haplotype arrived in America during different migrations (1), probably from Siberia (41). All these discrepancies may be the result of methodological differences and also of
Tissue Antigens 2003: 61: 425436

431

Gomez-Casado et al : HLA genes in Mayans

Fig. 4. World map showing the HLA genetic relationships between the populations studied in the present work (see Table 1). In green, Meso and South American Indians are shown as a separate group. All other populations show a geographically driven relatedness with a gradient from red (North American Na-Dene and Eskimos) to orange (Australians and Polynesians). Yellow (Caucasoid Europeans and North Africans) and white (Black Africans). Blue colour represents areas from where no data is shown. The map illustrates the hypothesis that more than one immigrant wave took place for the peopling of the Americas and that Amerindians are fully differentiated from the other world populations.

the different genealogical history and phylogenetic usefulness of different DNA markers. For instance, functional molecules (mit cyt b) are used against an admixture of intronic and exonic markers, as in the Alu or STR studies. DNA nuclear investigations have also been carried out to ascertain the origin of First Americans (Alu insertions) (42). The results are not concordant with the multiplewave migration hypothesis; instead three identifiable clusters of people are found, reflecting the geographical distribution. A surprisingly short genetic distance between Chinese and Native Americans was found and explained by a recent gene flow from Asia (42). A trans-Pacific route of American peopling from Asia or Polynesia has been suggested because HTLV-1 virus strains shared identical sequences in Japan and in the northern coast of South America (43) and some HLA alleles may have been introduced by the same TransPacific route (44). Finally, both genetic (45) and archaeological (46)

evidence suggests that a two-way trans-Atlantic traffic occurred before Columbus discovered America; archaeologists in New Mexico have recently found tools used 20,000 years ago in Spain (46). On the other hand, the Mayan origins have been subject to all kind of science fiction-speculations, relating this people with Atlantis (Mayans). However, it seems established that the writing and culture of Mayans comes from the old Olmecs with their own peculiarities (7). Arhuacs are the first reported people that populated the Caribbean Islands and the Caribbean shores of the continent of the Americas. By using HLA genetic calculations, and as seen in Figs 2 and 3 and Table 3, it is clear that Arhuacs (latter substituted by Caribs in the Caribbean Islands) (4) are the closest genetic group to Mayans. The peopling of the Americas occurred in more than one immigration wave, probably coming from different places. The relative

432

Tissue Antigens 2003: 61: 425436

Gomez-Casado et al : HLA genes in Mayans

Most frequent HLA-A, -B, -DRB1, -DQA1 and -DQB1-extended haplotypes in the Mayan population Haplotypes A*02-B*35-DRB1*0407-DQB1*0302
a

full-blown Olmec, Toltec, Mayan and Peruvian cultures appeared without external contact. Moreover, the most ancient archaeological American sites are distant from the postulated entrance door: the Bering Strait (Monteverde, Chile; Pedra Furada, Brazil 50).

HF (%) 10.6 8.4 6.4 5.0 4.2 4.2 2.6 2.3 2.3 1.5 1.5

Possible origin Amerindian Amerindian Amerindian Mayan Amerindian Amerindian Mayan Mayan Amerindian Mayan Spanish/Caucasoid

A*02-B*35-DRB1*0802-DQB1*0402b A*68-B*39-DRB1*0407-DQB1*0302c A*24-B*35-DRB1*0407-DQB1*0302d A*24-B*35-DRB1*0802-DQB1*0402e A*02-B*39-DRB1*0407-DQB1*0302f A*31-B*35-DRB1*0407-DQB1*0302

g

American uniqueness
Our data demonstrate how Amerindians show a relative homogeneity as opposed to other First Native American groups: Table 3 shows a discontinuity between (a) Meso and South American Indians and (b) North American Na-Dene and Eskimo groups. Data are lacking from many North American Amerindian groups that are needed to confirm or reshape our analysis. Moreover, Fig. 2 shows that Amerindians cluster separately from Na-Dene and Eskimo North American Indians. A simple interpretation is that Amerindians are less related to the Na-Dene-speaking (Athabascan) and Eskimo groups than among themselves; it suggests that the Amerindian is a more homogeneous group, with a different origin from the Na-Dene and Eskimo groups. This is also supported by other genetic (classical markers) and cultural data (33). However, the genetic relatedness among the alleles of the Amerindian groups (Green in Fig. 2) does not correlate with the language groups (i.e., Macro-Mixteco-speaking ethnic group in Oaxaca State: Mazatecans, Mixteco, Zapoteco; Macro-Mayaspeaking group in Guatemala and Mexico: see Mixe ethnic group (9); Chibcha-speaking group: Arhuaco, Kogi, Arsario, Cayapa). A different NJ analysis was carried out by amalgamating many worldwide and American populations (Fig. 3). It was only possible to use DRB1 genes in this case because the following Pacific Islanders and Australian Aborigine populations only had DRB1 typing: Kimberley, Central Desert, Yuendumu, Cape York, Papua New Guinean, New Caledonia, Fidji, Rabaul, Mandang and western Samoa; but this dendrogram and their genetic distances (Table 3) show again that Amerindians (Meso and South American) are not related to Na-Dene (Athabascan) and Eskimos (Guarani group: the large distance may have resulted from the Caucasoid admixture in this sample). Also, Amerindians do not show relationships with Polynesians, Australians [thus, almost discarding a massive Pacific colonisation, as suggested by Cerna and coworkers (44)], Caucasoids or African Blacks. However, two of the most common DR-DQ Amerindian haplotypes are found in the Na-Dene (DRB1*0802-DQA1*0401DQB1*0402, DRB1*1402- DQA1*0501-DQB1*0301) and one of them in the Japanese (DRB1*1406- DQA1*0501-DQB1*0301). Taking together both the relative unrelatedness of Meso and South America Amerindians and the presence of the latter three HLA haplotypes in the Na-Dene and Japanese at a low frequency (51), it would seem that a migrational wave would have started in South America towards North America and Asia (Boas theory, 3). This is further supported
Tissue Antigens 2003: 61: 425436

A*24-B*15-DRB1*0407-DQB1*0302h A*68-B*35-DRB1*0407-DQB1*0302
i

A*02-B*15-DRB1*0404-DQB1*0302j A*02-B*44-DRB1*07-DQB1*02
k

HF haplotype frequency. Mayan haplotypes may be shared with other Amerindians, but this is not yet described. a Found in Mazatecans (2.5%) from Mexico (2). b Also found in Nahuas (Aztecs) from Mexico (6.1%, unpublished results). c Found in Mazatecans (3.3%) with A24 instead. d Not found in other Amerindian groups. e Found in Peruvian Indians (2.4%, unpublished results). f Also found in Mazatecans (10.8%). g,h Not found in any other Amerindians. i Found in Nahuas (2%). j Not found in any other Amerindians. k Found in Caucasoids. Other haplotypes found are: A*02-B*35-DRB1*1402-DQB1*0301(1.1%), A*02-B*35DRB1*1406-DQB1*0301(1.1%), A*29-B*44-DRB1*07-DQB1*02(1.1%), A*68-B*40DRB1*0411-DQB1*0302(1.1%), A*01-B*08-DRB1*03011-DQB1*02(0.7%), A*01-B*57DRB1*1301-DQB1*0603(0.7%), A*02-B*15-DRB1*0802-DQB1*0402(0.7%), A*02-B*35DRB1*0403-DQB1*0302(0.7%), A*02-B*40-DRB1*0407-DQB1*0302(0.7%), A*02-B*51DRB1*0802-DQB1*0402(0.7%), A*24-B*07-DRB1*07-DQB1*02(0.7%), A*24-B*35DRB1*0403-DQB1*0302(0.7%), A*24-B*35-DRB1*1406-DQB1*0301(0.7%), A*24-B*35DRB1*16021-DQB1*0301(0.7%), A*24-B*39-DRB1*0802-DQB1*0402(0.7%), A*30-B*13DRB1*0411-DQB1*0305(0.7%), A*31-B*35-DRB1*0802-DQB1*0402(0.7%), A*31-B*39DRB1*0407-DQB1*0302(0.7%), A*31-B*40-DRB1*0404-DQB1*0302(0.7%), A*68-B*07DRB1*1501-DQB1*0602(0.7%), and A*68-B*35-DRB1*1101-DQB1*0301(0.7%).

Table 4

strength of marker discrimination for explaining the different relatedness found in First American Natives is difficult to ascertain. However, classical mtDNA and Y markers have given controversial results (see above). Alu repeats studies have even found a close relatedness between the Mesoamericans and the Chinese (see above). HTLV-1 virus subtype frequencies in populations suggest a close relatedness between the Amerindians and the Japanese (see above). All of these data should not be disregarded, because they all should help to account for the true peopling history and First Natives relatedness. In fact, the true scenario may altogether be different and more complicated than foreseen, as there are archaeological indications that Caucasoids (47), African Blacks (48), and Mongoloids from northern China (45, 49 may have all been among the First Native Americans; Negroids were particularly seen by Spanish conquerors in Meso-America, Caribbean and the north-western area of South America (48). In fact, many scholars are increasingly doubtful that

433

Gomez-Casado et al : HLA genes in Mayans

by the placement of the most ancient archaeological American sites (Chile and Brazil; 50, 52). On the other hand, Meso and South American Indians could have come from Asia and their HLA antigenic profile could have been changed as a result of the severe bottleneck (53) that they underwent after the European invasions in 1492: 80,000,000 people died because of microbia (measles, influenza, smallpox) and war borne by Europeans (54). It has been proposed that hybrid HLA genes resistant to the European-borne diseases resulted from EuropeanAmerindian contact and subsequent intragenic gene conversion (3, 52, 55). Indeed, these new molecules have been found more frequently in Amerindian groups living below the USMexican border on the Rio Grande (unpublished results); other authors have postulated that the Amerindian uniqueness may have resulted from the interplay of random genetic drift, including successive founder and bottleneck effects, and balancing selection (together with a relative genetic isolation). However, the fact that Amerindians were susceptible to Caucasoid diseases suggests that their original set of HLA molecules was very different from the Eurasian sets. Thus, the problem of the Amerindian origin is still open; they are probably different from all world populations (Figs 24). However, Na-Dene North American Indians and Eskimos show an altogether different HLA profile: they are related to some Asian groups (Table 3; Figs 2 and 3). If Meso and South American Indians come from Asia, they must have originated from a very different Asian people to those existing nowadays. The physical anthropology of South Amerindians is very different from Asians and also more varied [dolicocephalia, skin colour, etc. (33)]. Our analysis in Fig. 3 indicates that, while representative populations from most world ethnic groups (Fig. 4) are related, Amerindians cluster into a separate group; correspondence analyses (Fig. 2) also support these findings. Finally, the peopling of the America sequence may have been more complicated than previously thought: it seems that Mongoloids from China/Mongolia (but not from Siberia, see above) are found to have been in both America or the Middle Atlantic (Azores) before Columbus (see Introduction and 45).

The Spanish priest Diego de Landa Mayan alphabet has been the basis of many linguistic investigations (4, 6, 7, 8); in this respect the Mayan glyphs resemble the Egyptian glyphs, which represent a mono or bi-silabic alphabet (29). However, the Landas based Mayan language does not translate all Mayan glyphs found in all places (7), suggesting that different languages and writings existed; what we now call Mayan language was the imposition of Spanish rulers. Several languages (more or less related) were spoken in the nowadays Mayan area that has been defined on cultural parameters, which sometimes do not differ so much from the features of other ancient people from the Gulf Coast [i.e. the Olmecs (5, 7)]. Proto-Mayan artifacts and culture as ancient as 2600 years BC (7) have been found in the Mesoamerican area (Fig. 1). However, other scholars give data for the Mayan culture life span as being between 1500 BC and 1500 AD, having examples of the different cultural periods in North Yucatan, Middle Yucatan, Mexico or Belize and the southern area [Mexico, Guatemala, Honduras and El Salvador (5)]. In summary, the many Mayan languages or dialects have appeared with time and isolation (Mayan), the studied Mayans in the present paper came from Xela, Guatemala (the southern area), and their closest genetic relatives are tribes living nowadays in the Caribbean area of Colombia, i.e. Arhuac, Kogi, and Arsario. Mixe Indians from Mexican Mesoamerica are not genetically close to Mayans in spite of the Mixe who have been linguistically included within the Mayan family in the Mexican Penutian group (10). The Mixe ethnic group is genetically closer to the OaxacanMexican groups (Mazatecan, Zapoteco and Mixtecan, Figs 1 and 2 and Table 3). Thus, it may be postulated that Mayans are an autochthonous population related to the first-recorded Caribbean inhabitants, the Arhuac, and whose culture flourished between 1500 BC and 1500 AD by not fully explained reasons. The genetic homogeneity has not yet been established but there is linguistic heterogeneity (Fig. 1). Arhuacs were later substituted in the Caribbean Islands by Taino Indians, who were found when Columbus arrived in 1492 AD (5). The different evolutionary rate of genes and languages does not give a gene-language correlation when microgeographical studies are performed [this also occurs in the Caucasus (10)]; only when large, geographically extended groups of population are used and corrections (sometimes artificial) are applied is some correlation found (10, 33).

Mayans: culture, language and genes

Mayans are classified together with Mixe Indians from the northern Mexican Gulf Coast in linguistic bases (10). However, classification of the Mayan languages is controversial, are not deciphered and all dialects may represent altogether different languages [Fig. 1 (7, 8)].

References
1. Karafet TM, Zegura SL, Posukh O et al. Ancestral Asian source (s) of New World Y-chromosome founder haplotypes. Am J Hum Genet 1999: 64: 81731. 2. Arnaiz-Villena A, Vargas-Alarcon G, Granados J et al. HLA genes in Mexican Mazatecans, the peopling of the Americas and the uniqueness of Amerindians. Tissue Antigens 2000: 56: 40516. 3. Crawford MH. The Origins of Native Americans: Evidence from Anthropological Genetics. Cambridge: Cambridge University Press, 1998.

434

Tissue Antigens 2003: 61: 425436

Gomez-Casado et al : HLA genes in Mayans

4. Mundo. Maya. [editorial]. Madrid: Guas Acento, Acento, 1995. 5. Ruz A. El Pueblo Maya. Mexico: Salvat, 1993. 6. Coe MD. Breaking the Maya Code. New York, USA: Thames and Hudson, 1999. 7. Hasselkus HM. Estructura de la Glfica Maya. Mexico, DF: Hasselkus, 1998. 8. Valverde-Valdes MC. In: Consejo para la Cultura y las Artes, ed. Los Mayas. Mexico DF: Consejo para la Cultura y las Artes, 2000. 9. Swadesh M. Indian Linguistic Groups of Mexico. Mexico DF: Escuela Nacional de Antropologia e Historia, 1959. 10. Ruhlen MA. Guide to the Worlds Languages: I Classification. London: Edward Arnold, 1987. 11. Arnaiz-Villena A, Timon M, Corell et al. Primary immunodeficiency caused by mutations in the gene encoding the CD3-G subunit of the T-Lymphocyte receptor. N Engl J Med 1992: 327: 52933. 12. Excoffier L, Slatkin M. Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population. Mol Biol Evol 1995: 12: 9217. 13. Imanishi T, Akaza T, Kimura A, Tokunaga K, Gojorobi T. Estimation of allele and haplotype frequencies for HLA and complement loci. In: K. Tsuji, M. Aizawa, T. Sasazuki, eds. HLA 1991, Vol I. Oxford: Oxford University Press, 1992: 769. 14. Mattiuz PL, Ihde D, Piazza A, Ceppelini R, Wodmer WF. New approaches to the population genetics and segregation analysis of the HLA system. In: Histocompatibility Testing, 1970. Copenhagen: Munksgaard, 1970: 193206. 15. Arnaiz-Villena A, Martinez-Laso J, GomezCasado E et al. Relatedness among Basques, Portuguese, Spaniards and Algerian studied by HLA alelic frequencies and haplotypes. Immunogenetics 1997: 47: 3743. 16. Imanishi T, Akaza T, Kimura A, Tokunaga K, Gojobori T. Allele and haplotype frequencies for HLA and complement loci in various ethnic groups. In: K. Tsuji, M. Aizawa, T. Sasazuki, eds. HLA 1991, Vol 1. Oxford: Oxford University Press, 1992: 1065220. 17. Clayton J, Lonjou C. Allele and Haplotype frequencies for HLA loci in various ethnic groups. In: D. Charron, ed. Genetic Diversity of HLA, Funcional and Medical Implications. Paris: EDK, 1996: 650820. 18. Saitou N, Nei M. The neighbour-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 1997: 4: 40625. 19. Nei M. Genetic distances between populations. Am Nat 1972: 106: 283. 20. Nei M. Analysis of gene diversity in subdivided populations. Proc Natl Acad Sci USA 1973: 70: 33213.

21. Nei M, Tajima F, Tateno Y. Accuracy of estimated phylogenetic trees from molecular data II. Gene frequency data. J Mol Evol 1983: 19: 15370. 22. Young FW, Bann CM. A Visual Statistics system. In: RA Stine, J Fox, eds. Statistical Computing Environments for Social Researches. New York: Sage Publications, 1996: 20736. 23. Robinson J, Waller MJ, Parham P, Bodmer JG, Marsh SGE. IMGT/HLA Database a sequence database for the human major histocompatibility complex. Nucl Acids Res 2001: 29: 2103. 24. Marcos CY, Fernandez-Vina MA, Lazaro AM, Moraes ME, Moraes JR, Stastny P. Novel HLA-A and HLA-B alleles in South American Indians. Tissue Antigens 1999: 53: 47685. 25. Petzl-Erler ML, Gorodezky C, Layrisse Z. et al. Anthropology report for the Latin-American Region: Amerindian and admixture populations. In: D. Charron, ed. Genetic Diversity of HLA, Funcional and Medical Implications. Paris: EDK, 1997: 33745. 26. Gomez-Casado E, Vargas-Alarcon G, Martinez-Laso J et al. Generation of the HLAB35-B5-B16, and B15 groups of alleles studied by intron 1 and 2 sequence analysis. Immunogenetics 1997: 46: 46976. 27. Hollenbach JA, Thomson G, Cao K et al. HLA Diversity, differentiation, and haplotype evolution in Mesoamerican natives. Hum Immunol 2001: 62: 37890. 28. Vargas-Alarcon G, Gomez-Casado E, Martinez-Laso J et al. Description of a new HLA-B40 allele (B*4011) found in a Mexican individual of Nahua (Aztec) descent. Immunogenetics 1997: 46: 35960. 29. Arnaiz-Villena A, Alonso-Garca J. Egipcios, Bereberes, Guanches y Vascos. Madrid: Editorial Complutense, 2001. 30. Kolman CJ, Sambuughin N, Bermingham E. Mitochondrial DNA analysis of Mongolian populations and implications for the origin of New World founders. Genetics 1996: 142: 132134. 31. Merriwether DA, Hall WW, Vahlne A, Ferrell RE. mtRNA variation indicates Mongolia may have been the source for the founding population for the New World. AM J Hum Genet 1996: 59: 20412. 32. Greenberg JH, Turner CG, Zegura SL. The settlement of the Americas: a comparison of the linguistic, dental and genetic evidence. Curr Anthropol 1986: 27: 47798. 33. Cavalli-Sforza LL, Menozzi P, Piazza A. The History and Geography of Human Genes. Princeton: Princeton University Press, 1994.

34. Parham P, Ohta T. Population biology of antigen presentation by MHC class I molecules. Science 1996: 272: 6774. 35. Wallace DC, Torroni A. American Indian prehistory as written in the mitochondria DNA. a review. Hum Biol 1992: 64: 40316. 36. Horai S, Kondo R, Nakagawa-Hattori Y, Hayashi S, Sonoda S, Tajima K. Peopling of the Americas, founded by four major lineages of mitochondrial DNA. Mol Biol Evol 1993: 10: 2347. 37. Torroni A, Sukernik RI, Schurr TG et al. mtDNA variation of aboriginal Siberians reveals distinct genetic affinities with Native Americans. Am J Hum Genet 1993: 53: 591608. 38. Merriwether DA, Rothhammer F, Ferrel RE. Distribution of the four founding lineage haplotypes in Native Americans suggests a single wave of migration for the New World. Am J Phys Anthropol 1995: 98: 41130. 39. Santos FR, Rodriguez-Delfin L, Pena SDJ, Moore J, Weiss KM. North and South Amerindians may have the same major founder Y chromosome haplotype. Am J Hum Genet 1996: 58: 136970. 40. Karafet TM, Zegura SL, Vuturo-Brady J et al. Y chromosome markers and trans-Bering Strait dispersals. Am J Phys Anthropol 1997: 102: 30114. 41. Santos FR, Arpita P, Tyler-Smith C et al. The Central Siberian origin for Native American Y chromosomes. Am J Hum Genet 1999: 64: 61928. 42. Novick GE, Novick CC, Yunis J et al. Polimorphic Alu insertions and the Asian origin of Native American populations. Hum Biol 1998: 70: 2339. 43. Leon SFE, Ariza-Deleon A, Leon SME, Ariza CA. Peopling of Americas. Science 1996: 273: 7235. 44. Cerna M, Falco M, Friedman H et al. Differences in HLA class II alleles of isolated South American Indian populations from Brazil and Argentina. Hum Immunol 1993: 37: 21320. 45. Bruges-Armas J, Martinez-Laso J, Martins B et al. HLA in the Azores Archipielago: possible presence of Mongoloid genes. Tissue Antigens 1999: 54: 34953. 46. Holden C. Were Spaniards among the first Americans? Science 1999: 286: 14678. 47. Morell V. Kennewick Mans trials continue. Science 1998: 280: 1902. 48. Alcina-Franch J. Los Origenes de America. Madrid: Alhambra, 1985.

Tissue Antigens 2003: 61: 425436

435

Gomez-Casado et al : HLA genes in Mayans

49. Neves WA, Pucciareli HM. Morphological affinities of the First Americans: an exploratory analysis based on early South American remains. J Hum Evol 1991: 21: 26173. 50. Gibbons A. 150th bash draws a crowd. Science 1998: 279: 13067. 51. Monsalve MV, Edin G, Devine DV. Analysis of HLA class I and class II in Na-Dene and Amerindian populations from British Columbia, Canada. Hum Immunol 1998: 59: 4855. 52. Olivo-Diaz A, Gomez-Casado E, Gorodezky C et al. A new HLA B-15 allele (B*1541) found in a Mexican of Nahua (Aztec) descent. Immunogenetics 1998: 48: 14851. 53. Ayala F. The myth of Eve: molecular biology and human origins. Science 1995: 270: 19306. 54. Dobbins F. Disease transfer contact. Annu Rev Anthropol 1993: 22: 27391. 55. Gomez-Casado E, Martinez-Laso J, GonzalezHevilla M et al. A novel HLA-A*6816 allele possibly generated by a point mutation in a Chilean from Punta Arenas (Magellan Strait). Immunogenetics 2000: 51: 25760. 56. Yunis JJ, Ossa H, Salazar M et al. Major histocompatibility complex class II alleles and haplotypes and blood groups of four Amerindians tribes of northern Colombia. Hum Immunol 1994: 41: 24858.

57. Titus-Trachtenberg EA, Richards O, De Stefano GF, Erlich HA. Analysis of HLA class II haplotypes in the Cayapa Indians of Ecuador: a novel DRB1 alleles Reveals Evidence for convergent Evolution and Balancing Selection at position 86. Am J Hum Genet 1994: 55: 1607. 58. Grahovac B, Sukernik RI, OhUigin C et al. Polymorphism of the HLA class II loci in Siberian populations. Hum Genet 1998: 102: 2743. 59. Munkhbat B, Sato T, Hagihara M et al. Molecular analysis of HLA polymorphism in Khoton-Mongolians. Tissue Antigens 1997: 50: 12434. 60. Martinez-Laso J, Sartakova M, Allende L et al. HLA molecular markers in Tuvinians: a population with both Oriental and Caucasoid characteristics. Ann Hum Genet 2001: 65: 24561. 61. Martinez-Laso J, De Juan D, Martinez-Quiles N, Gomez-Casado E, Cuadrado E, Arnaiz-Villena A. The contribution of the HLA-A-B-C and -DR -DQ DNA typing to the study of the origins of Spaniards and Basques. Tissue Antigens 1995: 45: 23745. 62. Arnaiz-Villena A, Benmamar D, Alvarez M et al. HLA allele and haplotype frequencies in Algerians: relatedness in Spaniards and Basques. Hum Immunol 1995: 43: 25968.

63. Izaabel H, Garchon HJ, Caillat-Zucman S et al. HLA class II DNA polymorphism in a Moroccan population from the Souss, Agadir area. Tissue Antigens 1998: 51: 10610. 64. Gomez-Casado E, del Moral P, Martinez-Laso J et al. HLA genes in Arabic-speaking Moroccans: close relatedness to Berbers and Iberians. Tissue Antigens 2000: 55: 23949. 65. Gao X, Veale A, Serjeantson SW. HLA class II diversity in Australian aborigines: unusual HLA-DRB1 alleles. Immunogenetics 1992: 36: 333. 66. Gao X, Bhatia K, Trent RJ, Serjeantson SW. HLA-DR, DQ nucleotide sequence polymorphisms in five Melanesian populations. Tissue Antigens 1992: 40: 317. 67. Lester S, Cassidy S, Humphreys I et al. Evolution in HLA-DRB1 and major histocompatibility complex class II haplotypes of Australian aborigines. Definition of a new DRB1 allele and distribution of DRB1 gene frequencies. Hum Immunol 1995: 42: 15460. 68. Gao X, Zimmet P, Serjeantson SW. HLA-DR, DQ sequence polymorphisms in Polynesians, Micronesians and Javanese. Hum Immunol 1992: 34: 15361. 69. Bannai M, Tokunaga K, Imanishi T et al. HLA class II alleles in Ainu living in Hidaka district, Hokkaido, Northern Japan. Am J Phys Anthropol 1996: 101: 19.

436

Tissue Antigens 2003: 61: 425436