The HLA-DQA1 and HLA-DQB1 alleles and haplotypes of two Brazilian Indian tribes. Evidences of conservative evolution of HLA-DQ.

Reference: Sotomaior VS, Faucz FR, Schafhauser C, Janzen-Dück M, Boldt AB, Petzl-Erler ML. HLA- DQA1 and HLA-DQB1 alleles and haplotypes in two Brazilian Indian tribes: evidence of conservative evolution of HLA-DQ. Hum Biol. 1998 Aug;70(4):789-797. PubMed PMID: 9686487. The HLA-DQA1 and HLA-DQB1 alleles and haplotypes of two Brazilian Indian tribes. Evidences of conservative evolution of HLA-DQ. V. S. Sotomaior, F. R. Faucz, C. Schafhauser, M. Janzen-Dück, A.B.W. Boldt, M. L. Petzl-Erler Laboratório de Genética Molecular Humana, Departamento de Genética, Setor de Ciências Biológicas, Universidade Federal do Paraná, Brazil Key words: Amerindians, MHC, HLA-DQ, DQA1, DQB1, polymorphism. Short title: HLA-DQA1 and HLA-DQB1 in Brazilian Indians Corresponding author: Prof. Dr. M. Luiza Petzl-Erler Department of Genetics Federal University of Paraná Caixa Postal 19071 81531-990 Curitiba, Brazil e-mail: perler@ufpr.br 1

Abstract Nucleotide sequence polymorphism of the HLA-DQA1 and HLA-DQB1 class II genes was analyzed in the Kaingang and Guarani Amerindians from Southern Brazil, using PCR-SSO typing methods. Four different DQA1-DQB1 haplotypes were found: DQA1*0401- DQB1*0402 (associated to DRB1*0802), DQA1*0501-DQB1*0301 (associated to DRB1*1602 and DRB1*1413), DQA1*03-DQB1*0302 (associated to DRB1*0404 and DRB1*0411) and DQA1*03-DQB1*03032 (associated to DRB1*0901). These HLA-DQA1 and HLA-DQB1 alleles and haplotypes are common in many other populations of all major ethnic groups. Alleles/haplotypes introduced in the populations by post-columbian admixture were seen at low frequency, both in the Kaingang (3.2%), and in the Guarani (3.8%). No novel HLA-DQA1 and HLA-DQB1 alleles have thus far been identified in Amerindians. This differs from previous results for HLA-DRB1, another class II locus presenting novel alleles (i.e, alleles not found in other ethnic groups, and probably generated after migration of Paleo- Indians to the American continent) in the Guarani and in other South American Indian populations. The distribution of the DQ alleles and haplotypes in Amerindians indicate a weaker diversifying selective pressure on the HLA-DQ genes in comparison to HLA-DRB1 and HLA-B. The more conservative evolution of HLA-DQA1 and HLA-DQB1 in comparison to HLA-DRB1 is a strong evidence of (still not well defined) functional differences of these class II genes. 2

The Guarani and the Kaingang are two anthropologically well defined and genetically distinct endogamic Amerindian tribes. The Guarani belong to the Tupi linguistic group, while the Kaingang are Gê-speaking people. In spite of living side by side for many centuries these populations still differ in many aspects of their culture. The study of their HLA genes has been contributing new insights regarding the mechanisms of generation and maintenance of the HLA polymorphism (Belich et al. 1992; Petzl-Erler et al. 1993; Petzl-Erler and McDevitt 1994; Pando et al. 1994; Parham et al. 1997). The two populations exhibit great difference regarding the sets of alleles and allelic frequencies at almost all thus far examined MHC loci (Guerra et al. 1992; Petzl-Erler et al. 1993; Messias et al. 1993; Petzl-Erler and McDevitt 1994; Weg-Remers et al. 1997; Petzl-Erler et al. 1997). Thus, the clear cultural and linguistic distinctiveness of these populations is paralleled by a great genetic distance between them (Petzl-Erler et al. 1993; Salzano et al. 1997). The class II region of the human major histocompatibility complex (MHC) consists of a series of closely linked genes mapped to 6p21.3. While some MHC class II loci are essentially monomorphic or bear few common alleles, others are highly polymorphic. The HLA-DQA1 and HLA-DQB1 genes - which encode, respectively, the and the chain of the HLA-DQ molecule - are among the polyallelic and polymorphic ones. Eighteen DQA1 and 31 DQB1 alleles had been identified until 1996. Variability of the HLA-DQ molecule is increased further, since heterodimers may result from the chains encoded by DQA1 and DQB1 alleles in both cis and trans positions (Kwok et al. 1993). The genetic diversity of the MHC genes is of interest for its functional and evolutionary significance. The analysis of these genes has been proving useful in unraveling the historical relationships between distinct ethnic groups. Furthermore and equally important, the population studies are giving important insights regarding the microevolution of the MHC itself. The relative contribution of point mutation and interallelic gene conversion as mechanisms of generation of new HLA alleles, and of different forms of selection and genetic drift in shaping the diversity of these genes at the population level, are subject of controversies. Searching for evidences concerning the evolutionary mechanisms acting on the HLA-DQ diversity, we studied the second exon nucleotide sequence polymorphism of the HLA-DQA1 and HLA-DQB1 genes in the two major Amerindian tribes of Southern Brazil, the Kaingang and the Guarani. All individuals examined belong to the population samples investigated by Petzl-Erler et al. (1993) and were typed serologically for HLA-A, B, C, DR and DQ. A subgroup of them was typed for DRB by direct DNA analysis (Petzl-Erler and McDevitt 1994, and unpublished results). The population sample analyzed in this study was of 200 individuals (92 Guarani and 108 Kaingang). Standard methods were used for DNA purification and PCR-SSO typing. The sequences of the primers and probes correspond to those recommended by the 11th and 12th International Histocompatibility Workshops (Kimura and Sasazuki 1992; Fernandez-Viña and Bignon 1994). Twenty-one and 36 different SSO were used for HLA-DQA1 and HLA-DQB1 typing, respectively. Positive and negative controls for the several probes were International Histocompatibility Workshop reference cell lines and individuals from the local cell panel. 3

Allelic frequencies were obtained by direct counting. Genotypic frequencies, equally obtained by direct counting, were tested for Hardy and Weinberg equilibrium fit. The statistical significance of all deviations was verified by chi-square tests with Yates correction, if pertinent. The DRB1-DQA1-DQB1 haplotypes were identified by analysis of homozygous individuals, and by segregation analysis in families. Information on the DRB1 alleles of these populations was available from our previous studies (Petzl-Erler and McDevitt 1994, and unpublished results). The SSO hybridization patterns resulted in detection of the following DQA1 and DQB1 alleles and haplotypes, all of which shared by both tribes: DQA1*0501-DQB1*0301, DQA1*03-DQB1*0302, DQA1*03-DQB1*03032 and DQA1*0401-DQB1*0402, besides a few alleles/haplotypes introduced in the populations by post-columbian admixture (Tables 1 and 2). The approach used did not allow discrimination between DQA1*03011 and DQA1*0302 since the difference between these alleles is located in exon 3, not examined in our study. However, in the Amerindian, Asian, Oceanian, European and African populations analyzed for the subtypes of DQA1*03, DQA1*03011 was the DQA1*03 allele associated with all DR4 alleles, excepting DRB1*0405 (Fernandez-Viña et al. 1994). Further, DQB1*0302 has been reported to be in strong positive linkage disequilibrium with DQA1*03011. Therefore, DQA1*03011-DQB1*0302 is the most likely combination carried by the DR4-DQ8(3) haplotypes. For similar reasons, probably the DQA1 allele of the DRB1*0901-DQB1*03032 haplotype is DQA1*0302. The number of different DQ alleles and haplotypes in these two tribes is low, in comparison to that seen in exogamic populations (e.g., Petzl-Erler et al. 1997). This is in accordance with results for other HLA and non-hla loci (see above), and with the hypothesis of a founder effect during or before the colonization of America. The significant difference of allelic frequencies between the two tribes (p < 10-6 for both loci) is related to the great genetic distance between them (Petzl-Erler et al. 1993; Salzano et al. 1997), and probably results mainly from random genetic drift. Non-Amerindian alleles were easily identified due to their absence from other non-admixed Amerindian populations, and to the extended MHC haplotypes bearing these alleles (data not shown). Their cumulated frequency gives an estimate of the degree of admixture with non- Amerindians (Caucasoids and Negroids), which is 3.2% for Kaingang and 3.8% for Guarani, being similar to that calculated previously (Petzl-Erler et al. 1993). The genotypic frequencies did not depart from Hardy and Weinberg expectations. A non-significant excess of heterozygotes was seen for DQA1 and DQB1 in both populations. The same four DQA1-DQB1 haplotypic combinations found to be common in the Guarani and Kaingang are common in other Amerindians (Layrisse et al. 1995; Yunis et al. 1994; Petzl-Erler et al. 1997), as well as in non-amerindians (e.g. Petzl-Erler et al. 1997). Linkage disequilibrium between DQB1 and DQA1 is absolute in the populations studied. Sporadically, unusual combinations of DQA1 and DQB1 alleles in otherwise Amerindian haplotypes (i.e., in haplotypes not introduced in the populations by post-columbian genetic admixture) have been reported (Trachtenberg et al. 1995, 1996). They were reported to occur at low frequencies, besides the usual DQA1-DQB1 haplotypes. The possibility of mistyping or exchange of biological samples must be considered when occasional unexpected haplotypic combinations are observed. Careful validation of any unusual haplotype in well-characterized 4

endogamic populations, such as the many Amerindian tribes, is essential before any inference about the evolution of MHC diversity is made. A few unusual DQA1-DQB1 haplotypes seen during the typings for this study could be ruled out after comparison of molecular and previous serological typings, by retyping of samples presenting the seemingly unusual combinations and, whenever possible, by segregation analysis in informative families. In all cases, the most usual haplotypic combination was confirmed. Associations between DRB1 and DQ alleles were also absolute (Table 3). Every DRB1 allele was seen in cis with just one DQA1-DQB1 haplotype, though two of the DQA1- DQB1 allelic combinations were found associated to more than one DRB1 allele. However, differently than the DQA1-DQB1 haplotypic combinations, some of the DRB1-(DQA1- DQB1) haplotypes seen in these Amerindians have a restricted geographical distribution. Absolute linkage disequilibrium between DR2 and DQ7(3) (haplotype DRB1*1602- DQA1*0501-DQB1*0301) was first described in Warao Amerindians (Layrisse et al. 1988). DRB1*1602 is the only DR2 allele as yet found in Amerindians, and the vast majority of Amerindian groups have the same haplotype also found in the Kaingang and Guarani (e.g. Layrisse et al. 1995; Petzl-Erler et al. 1997). Interestingly, haplotype DRB1*1602- DQA1*0501-DQB1*0301 has not been found in populations of Asian, European or African ancestry. Allele DRB1*1413, which was described among the Guarani (Petzl-Erler and McDevitt 1994; Pando et al. 1994) and not yet found in other populations, is inserted in a DQA1*0501-DQB1*0301 haplotype - the same found in combination with other DR14(6) alleles of Amerindians, and with the DRB1*1602 allele as commented above. Haplotype DRB1*0404-DQA1*03011-DQB1*0302 has a wider geographic distribution, being found also in Eastern Asia and Oceania (Gao et al. 1991, 1992a, 1992b). DRB1*0411- DQA1*03011-DQB1*0302 was seen only in Australian Aborigines (Gao et al. 1991) and in South-American Indians (Petzl-Erler et al. 1997). Absolute linkage disequilibrium, like that observed between the DRB1, DQA1 and DQB1 alleles in these Amerindian populations, may result from action of stochastic evolutionary factors (founder effect and genetic drift). However, the concomitant influence of natural selection against alternative haplotypic combinations sporadically arising by crossingover cannot be ruled out. Diversity of DQ alleles and haplotypes is greater in Asians, Europeans and Africans than in native Americans. Nevertheless, the DQA1-DQB1 haplotypes found in absolute linkage disequilibrium in Amerindians, present strong positive disequilibrium in all other populations in which the same alleles occur. The conservation of DQA1-DQB1 haplotypes in populations of all continents may indicate that DQA1 and DQB1 genes are coevolving under natural selection. It is well-known that these loci are in tight genetic linkage, and that many generations are needed to restore equilibrium, because the recombination fraction is low. Nonetheless, it has been shown that while HLA-DQ heterodimers may result from alleles located both in cis and in trans, not all HLA-DQ and polypeptides interchange freely. Patterns of preferential heterodimer formation were demonstrated by Kwok et al. (1993). So, there are structural constraints on dimers formation, which supports the hypothesis of coevolution of the DQA1 and DQB1 genes, maintaining strong positive linkage disequilibrium between combinations of co-adapted 5

alleles. Also DRB1-DQ haplotypic combinations present strong linkage disequilibrium in all populations, which however is, on average, less pronounced than for DQA1-DQB1. Further, and more important in this context, DRB1-DQ combinations in cis are more variable in our species than DQA1-DQB1 combinations, so they probably are less subjected to evolutionary restrictions by natural selection. Their strong conservation among South- and North- American Indians suggests the major role of founder effects and random genetic drift in shaping the diversity of the class II haplotypes in native Americans. The HLA-DQ genes showed a lower degree of allelic polymorphism than HLA-B, HLA-DPB1, and HLA-DRB1 in the same Amerindian populations (Petzl-Erler et al. 1993, 1997; Petzl-Erler and McDevitt 1994; Parham et al. 1997). This does not differ from what is seen in other populations of all continents. In addition, no "unusual" alleles, i.e, alleles not found in Asian, European or African populations were seen at the HLA-DQ loci in the Kaingang and Guarani. This contrasts with results for another class II gene, HLA-DRB1, for which unusual and unique alleles have been reported in the same and other American Indians (Cerna et al. 1993; Petzl-Erler and McDevitt 1994; Pando et al. 1994). The different patterns of polymorphism between the highly homologous HLA class II loci reflect the distinct evolutionary pathways of these genes, which in turn are restrained by the functional and structural differences of the corresponding gene products. Although the functional differences between MHC molecules of the same class are less well defined than the structural ones, dissimilar roles have been suggested for HLA-DQ and HLA-DR molecules. It has been proposed that HLA-DQ is more involved in T suppressor cell induction (Hirayama et al. 1987; Sasazuki 1990), whereas HLA-DR upregulates the immune response by activating T helper lymphocytes. HLA-DQ molecules should be important not for antigen presentation to mature CD4+ T cells but for its contribution to positive and negative thymic selection. Apart from corroborating functional differences between the HLA-DQ and HLA-DR molecules, the conservation of the same HLA-DQ alleles in distantly related populations such as Europeans and American Indians, and others, indicates that no strong environmental diversifying selection pressure is operating on the HLA-DQ genes. In fact, the intensity of selection operating at DQA1 and DQB1 was estimated to be lower than that acting on the DRB1 (and HLA-B) locus (Satta et al. 1994), which is in accordance with our conclusions based on the HLA diversity in Amerindians relative to other human populations. In summary, our results indicate a more conservative evolution of DQA1 and DQB1 alleles and haplotypes in comparison to DRB1 alleles and DRB1-DQ haplotypes. The relatively reduced allelic and haplotypic class II diversity seen among Amerindians, in comparison to outbred Asian, European, African, and neo-american populations, is due mainly to founder effects and genetic drift. Nevertheless, concomitant influence of conservative selection on DQ alleles and DQA1-DQB1 combinations in cis, and stronger diversifying selection over DRB1 alleles is likely. 6

Literature Cited Belich, M.P., J.A. Madrigal, W.H. Hildebrand, J. Zemmour, R.C. Williams, R. Luz, M.L. Petzl-Erler, and P. Parham. 1992. Unusual HLA-B alleles in two tribes of Brazilian Indians. Nature 357:326-329. Cerna, M., M. Falco, H. Friedman, E. Raimondo, A. Maccagno, M.A. Fernandez- Viña, and P. Stastny. 1993. Differences in HLA class II alleles of isolated South American Indian populations from Brazil and Argentina. Hum. Immunol. 37:213-220. Fernandez-Viña, M., and J.D. Bignon 1994. 12th IHWC HLA class II reference protocol. Addendum I/95. In: Technical Handbook. Twelfth International Histocompatibility Workshop (Charron, D., Fauchet, R., eds.). HLA et Médicine, Paris. Fernandez-Viña, M., M. Falco, X. Gao, M. Cerna, Y. Sun, E. Raimondi, and P. Stastny. 1994. DQA1*03 subtypes have different associations with DRB1 and DQB1 alleles. Hum. Immunol. 39:290-298. Gao, X., K. Bhatia, R.J. Trent, and S.W. Serjeantson. 1992a. HLA-DR,DQ nucleotide sequence polymorphisms in five Melanesian populations. Tissue Antigens 40:31-37. Gao, X., and S.W. Serjeantson. 1991. Diversity in HLA-DR4-related DR, DQ haplotypes in Australia, Oceania, and China. Hum. Immunol. 32:269-276. Gao, X., P. Zimmet, and S.W. Serjeantson. 1992b. HLA-DR,DQ sequence polymorphisms in Polynesians, Micronesians, and Javanese. Hum. Immunol. 34:153-161. Guerra, L.K., I.R. Rehfeldt, M.T. Lima, M.L. Petzl-Erler, and I.J.T. de Messias. 1992. BF and C3 genetic polymorphisms in Kaingang Indians from Southern Brazil. Human Heredity 42:153-156. Hirayama, K., S. Matsushita, I. Kikuchi, M. Iuchi, N. Ohta, and T. Sasazuki. 1987. HLA-DQ is epistatic to HLA-DR in controlling the immune response to schistosomal antigen in humans. Nature 327:426-430. Kimura, A., and T. Sasazuki. 1992. Eleventh International Histocompatibility Workshop reference protocol for the HLA DNA-typing technique. In: HLA 1991. Proceedings of the Eleventh International Histocompatibility Workshop and Conference (Tsuji, K., Aizawa, M, Sasazuki T., eds.). Oxford University Press, New York, pp. 397-419. Kwok, W.W., S. Kovats, P. Thurtle, and G.T. Nepom. 1993. HLA-DQ allelic polymorphisms constrain patterns of class II heterodimer formation. J. Immunol. 150(6):2263-2272. 7

Layrisse, Z., Y. Guedez, E. Domingues, F. Herrera, M. Soto, O. Balbas, M. Matos, J.C. Alfonzo, J. Granados, and J. Scorza. 1995. Extended HLA haplotypes among the Bari Amerindians of the Perija range. Relationship to other tribes based on four-loci haplotype frequencies. Hum. Immunol. 44:228-235. Layrisse, Z., H.D. Heinen, and O. Balbas. 1988. Unique HLA-DR/DQ associations revealed by family studies in Warao Amerindians. Haplotype and homozygosity frequencies. Hum. Immunol. 23:45-57. Messias, I.T., M.L. Petzl-Erler, I.R. Rehfeldt, M.T. Lima, and L.K. Guerra. 1993. Variability of complement factor B and complement component C3 among Guarani Indians from Southern Brazil. Brazil J. Genetics. 16(3):803-809. Parham, P., K.L. Arnett, E.J. Adams, A.-M. Little, K. Tees, L.D. Barber, S.G.E. Marsh, T. Ohta, T. Markow, M.L. Petzl-Erler. Episodic evolution and turnover of HLA-B in the indigenous human populations of the Americas. Tissue Antigens 50:219-232. Pando, M., G. Theiler, R. Melano, M.L. Petzl-Erler, and M.L. Satz. 1994. A new HLA-DR6 allele (DRB1*1413). found in a tribe of Brazilian Indians. Immunogenetics 39:377. Petzl-Erler, M.L., R. Luz, and V.S. Sotomaior. 1993. The HLA polymorphism of two distinctive South-American Indian tribes: The Kaingang and the Guarani. Tissue Antigens 41:227-237. Petzl-Erler, M.L., C. Gorodezky, Z. Layrisse, W. Klitz, L. Fainboim, C. Vullo, J. Bodmer, E. Egea, C. Navarrete, E. Infante, C. Alaez, A. Olivo, H. Debaz, N. Bautista, G. de la Rosa, M.N. Vazquez, J.L. Navarro, M.J. Pujol, C. Duran, C. Schafhauser, F.R. Faucz, M. Janzen, P. Maciag, A.B.W. Boldt, P.S.A. Souza, C.M. Probst, G.F. Silva, N. Makhatadze, E. Dominguez, S. Montagnani, M. Matos, A. Martinez, F. Herrera, J. Hollenbach, G. Thomson, M. Pando, L. Satz, J. Larriba, G. Fernandez, S.A. Pesoa, A. Borosky, G. Garavito, L. Angel, J. Brown, E. Llop. 1997. Anthropology Component Report for Region Latin-America: Amerindian and Admixed Populations. In: Charron, D. (Ed.) Genetic Diversity of HLA. Functional and Medical Implications. EDK, Sèvres. Petzl-Erler, M.L., and H.O. McDevitt. 1994. Molecular analysis of the HLA-DRB genes in two tribes of Brazilian Indians. Hum. Immunol. 41:180-184. Salzano, F.M., S.M. Callegari-Jacques, T.A. Weimer, M.H.L.P. Franco, M.H. Hutz, and M.L. Petzl-Erler. 1997. Electrophoretic protein polymorphisms in Kaingang and Guarani Indians of Southern Brazil. Am. J. Hum. Biol. 9:505-512. Sasazuki, T. 1990. HLA-linked immune suppression genes. Jinrui Idengaku Zasshi 35:1-13. 8

Satta, Y., C. O huigin, N. Takahata, and J. Klein. 1994. Intensity of natural selection at the major histocompatibility complex loci. Proc. Natl. Acad. Sci. USA 91:7184-7188. Trachtenberg, E.A., H.A. Erlich, O. Rickards, G.F. DeStefano, and W. Klitz. 1995. HLA class II linkage disequilibrium and haplotype evolution in the Cayapa Indians of Ecuador. Am. J. Hum. Genet. 57:415-424. Trachtenberg, E.A., G. Keyeux, J.E. Bernal, M.C. Rhodas, and H.A. Erlich. 1996. Results of expedicion humana. I. Analysis of HLA class II (DRB1-DQA1-DQB1-DPB1). alleles and DR-DQ haplotypes in nine Amerindian populations from Colombia. Tissue Antigens 48:174-181. Weg-Remers, S., M. Brenden, E. Schwarz, K. Witzel, P.M. Schneider, L.K. Guerra, I.R. Rehfeldt, M.T. Lima, D. Hartmann, M.L. Petzl-Erler, I.J.T. Messias, G. Mauff. Major histocompatibility complex (MHC) class III genetics in two Amerindian tribes from southern Brazil: The Kaingang and the Guarani. Hum. Genet. (in press) Yunis, J.J., H. Ossa, M. Salazar, M.B. Delgado, R. Deulofeut, A. de la Hoz, D.H. Bing, O. Ramos, E.J. Yunis, and E.J. Yunis. 1994. Major histocompatibility complex class II alleles and haplotypes and blood groups of four Amerindian tribes of Northern Colombia. Hum. Immunol. 41:248-258. Acknowledgements This research was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and Universidade Federal do Paraná. We thank the populations of Ivaí and Rio das Cobras for their cooperation. 9

Table 1. HLA-DQA1 and HLA-DQB1 allele frequencies. Allele Guarani (n = 184) n i f Locus HLA-DQA1 Kaingang (n = 216) n i f *0101 or *0104 1 0.005 7 0.032 *0102 2 0.011 0 0 *0201 1 0.005 0 0 *03 @ 56 0.304 54 0.250 *03011 34 0.185 41 0.190 *0302 9 0.049 5 0.023 *0401 24 0.130 111 0.514 *0501 100 0.544 44 0.204 Locus HLA-DQB1 *0201 4 0.022 0 0 *03 # 153 0.832 98 0.454 *0301 97 0.527 44 0.204 *0302 34 0.185 41 0.219 *03032 9 0.049 5 0.023 *0402 24 0.130 111 0.514 *0501 1 0.005 7 0.032 *0602 2 0.011 0 0 @ DQA1*03 broad comprises alleles *03011 and *0302, as well as alleles not subtyped (8 in the Kaingang and 13 Guarani, which could be either *03011 or *0302). # DQB1*03 broad comprises alleles *0301, *0302 and *03032, as well as alleles not subtyped (8 in the Kaingang and 13 Guarani, which could be either *0302 or *03032). n: Total number of examined alleles; n i : number of alleles, f: relative frequency. In bold: alleles previously found in genetically isolated South-American and Mexican Amerindian populations. 10

Table 2. DQA1-DQB1 haplotypes and their frequencies (%). nr. Haplotype Serological Specificity Guarani Kaingang (n = 92) (n = 108) 1 DQA1*03011 - DQB1*0302 DQ8(3) 18.5 19.0 2 DQA1*0302 - DQB1*03032 DQ9(3) 4.9 2.3 3 DQA1*0401 - DQB1*0402 DQ4 13.0 51.4 4 DQA1*0501 - DQB1*0301 DQ7(3) 52.7 20.4 * DQA1*03 - DQB1*0302 or *03032 DQ8(3) or DQ9(3) 7.1 3.7 other** DQ1 and DQ2 3.8 3.2 * This could be either haplotype 1 or haplotype 2. ** Haplotypes introduced in the populations by admixture with non-amerindians. 11

Table 3. DRB-DQA1-DQB1 haplotypes. Population Haplotype Serological Specificities K and G DRB1*1602 - DRB5*02 - DQA1*0501 - DQB1*0301 DR16(2), DR51, DQ7(3) K DRB1*0404 - DRB4*0101 - DQA1*03(a) - DQB1*0302 DR4, DR53, DQ8(3) G DRB1*0411 - DRB4*0101 - DQA1*03(a) - DQB1*0302 DR4, DR53, DQ8(3) G DRB1*1413 - DRB3*0101 - DQA1*0501 - DQB1*0301 DR14(6), DR52, DQ7(3) K and G DRB1*0802 - DQA1*0401 - DQB1*0402 DR8, DR52, DQ4 K and G DRB1*0901 - DRB4 nt - DQA1*03(b)- DQB1*03032 DR9, DR53, DQ9(3) Data on DRB genes from Petzl-Erler and McDevitt (1994). (a) Most likely DQA1*03011 and (b), most likely DQA1*0302. See text for more information. nt: not tested; K: Kaingang; G: Guarani. 12