J. Renquin A. Sanchez-Mazas L. Halle S. Rivalland G. Jaeger K. Mbayo F. Bianchi C. Kaplan Key words: Aka Pygmies; Africa; anthropology; Bantu Congolese; genetic diversity; HLA class II; linguistics; MHC selection; PCR-dot blot reverse; PCR-SSP Acknowledgments: This work was supported by a Swiss FNRS grant (.3100-49771.96) to A. Sanchez-Mazas and by the French DGRST (No. 77.7206; 78.7.2456; 79.7.0456). HLA class II polymorphism in Aka Pygmies and Bantu Congolese and a reassessment of HLA-DRB1 African diversity Abstract: HLA-DRB1, -DQB1 and -DPB1 polymorphisms were investigated in two African populations, the Basse Lobaye Aka Pygmies of the Central African Republic, and a Bantu-speaking group from the Democratic Republic of Congo Kinshasa. Allelic and haplotypic frequency distributions reveal marked differences between the two populations in spite of their geographical proximity: the Aka exhibit high frequencies for several alleles, especially at the DPB1 locus (0.695 for DPB1*0402), probably due to rapid genetic drift, while the Bantu distributions are more even. Genetic distances computed from DRB1 allelic frequencies among 21 populations from North and sub-Saharan Africa were applied to a multidimensional scaling analysis. African populations genetic structure is significantly shaped by linguistic differentiation, as confirmed by an analysis of molecular variance. However, selective neutrality tests indicate that many African populations exhibit an excess of heterozygotes for DRB1, which is likely to explain the genetic similarity observed between some North African and Bantu populations. Overall, this study shows that natural selection must be taken into account when interpreting the patterns of HLA diversity, but that this effect is probably minor in relation to the stochastic events of human population differentiations. Authors’ affiliations: J. Renquin1, A. Sanchez-Mazas1, L. Halle2, S. Rivalland2, G. Jaeger3, K. Mbayo4, F. Bianchi2, C. Kaplan2 1 Laboratory of Genetics and Biometry (LGB), Department of Anthropology and Ecology, University of Geneva, Geneva, Switzerland, 2 Institut National de la Transfusion Sanguine (INTS), Laboratoire d’Immunologie Plaquettaire, Paris, France, 3 Centres Européens Associées de Biologie Humaine (CEABH), Paris, France, 4 Cliniques Universitaires de Kinshasa, Kinshasa, Democratic Republic of Congo Due to its high level of polymorphism revealed by high-resolution DNA molecular typings, the human major histocompatibility complex (MHC) plays a key role in anthropology for investigating the genetic relationships between populations. Previous studies related to HLA diversity in sub-Saharan Africa indicate a high level of heterogeneity among populations within this continent, and a correlation with geographic and/or linguistic differentiations (1–5). However, some populations, like the Pygmies, were not represented. The Received 13 July accepted for publication 17 September 2001 Copyright c Munksgaard 2001 Tissue Antigens . ISSN 0001-2815 Tissue Antigens 2001: 58: 211–222 Printed in Denmark . All rights reserved aim of this study is to investigate the HLA class II polymorphism in two African populations: Aka Pygmies (6) from the Central African Correspondence to: Department of Anthropology and Ecology 12 rue Gustave-Revilliod 1227 Geneva Switzerland Tel: π41 22 702 69 84 Fax: π41 22 300 03 51 e-mail: alicia.sanchezmazas/anthro.unige.ch or Cécile Kaplan INTS 6 rue Alexandre Cabanel 75739 Paris France Tel: π331 4449 3067 Fax: π331 4567 1930 e-mail: cecile.kaplan/teaser.fr Republic (CAR), and a group of Bantu individuals originating from the Democratic Republic of Congo Kinshasa (DRC). 211 Renquin et al : HLA class II in Aka Pygmies and Congolese Fig. 1. Present geographic location of the Aka Pygmies and the Bantu Congolese under study. (Kinshasa) to Congo (Brazzaville) and the second, approximately 150 Aka Pygmies years ago, left Congo (Brazzaville) to reach the village of Bagandu From 1975 to 1980, the EurAfrican Centre for Human Biology (CE- situated in the southwest of the CAR. ABH) in cooperation with the Central African government, under- An initial study, published by the INTS group in 1981 (9) on took a comprehensive study of the Aka Pygmies (Basse Lobaye HLA class I serological typing revealed the presence of HLA-B37 region) and their environment (7). This population is of particular (8%) with A30-B37 linkage disequilibrium (D⫽0.035, PⰆ0.001). interest because it constitutes an isolate of nomadic hunter-gath- This allele is very rare or absent in other sub-Saharan populations erers in a primary environment. These Pygmies move around con- except the Fulani (17%) (4). Sixty samples typed again in 1996 stantly in the southern part of Central Africa and the northern part revealed the presence of B41 (2.44%) and B42 (3.44%). of the People’s Republic of Congo Brazzaville (8). Their territory (3æ5 N and 17æ30–18æ E) is defined by the natural borders of the Oubang- Bantu Congolese ui river and its two tributaries the Lobaye and the Ibanga. Every year the Aka leave their environment for a few weeks to work for The Democratic Republic of Congo Kinshasa (DRC) covers an the Bantu population in exchange for fulfilment of their needs. Ac- area of some 2,400,000 square kilometres stretched out between a cording to the Pygmies, they have been two waves of migration: latitude of 5æ2ø north and a latitude of 13æ15ø south. Because of the original older migrant group travelled eastward from Congo their geographical situation in the centre of Africa, the autoch- 212 Tissue Antigens 2001: 58: 211–222 Renquin et al : HLA class II in Aka Pygmies and Congolese thonous population and more specifically the Pygmies were faced (14) was used for samples with at least 106 of PBMC, and the Pro- with Bantu migration waves towards the end of the first milleni- teinase K lysis technique (15) was used for samples with less than um (10). These Bantus migrated along 3 paths: the eastern route, 106 of PBMC. leading to settling in savannahs of the Katanga and Kivu re- HLA class II alleles for loci DRB1, DQB1 and DPB1 were deter- gions, the western route, resulting in settlement in the Lower mined by polymerase chain reaction (PCR)-reverse dot blot hybridi- Congo and Bandundu regions, and finally the central route, end- zation INNO-LIPA (Innogenetics N.V., Zwijnaarde, Belgium) (16). ing in the Equator and Bandundu regions. Later came Sudanese DQA1 and DRB1*15 high-resolution were analysed by PCR-SSP (17) migration to the north-eastern part of the country (Ubangi-Kibali- (Dynal France SA., Compiègne France). The HLA nomenclature was Ituri) leading to interbreeding with the Bantus. Finally, coming used according to S.G.E. Marsh for the WHO Nomenclature Com- from the Ethiopian plateaux, Nilotics travelled up the Nile valley mittee for Factors of the HLA system (January 2001 update; http:// to settle near the great lakes in the Kivu region. The Congolese www.anthonynolan.com/HIG/nomenc.htm/). included in this study (Fig. 1) belong to the Bantu-speaking group (11, 12) which is 35% composed of Kongo, 29% of Luba (75% from the provinces of Eastern-Kasaı̈ and 25% from Katan- Statistical analyses ga), 16% of Mongo, 8% of Yaka, 6% of Yanzi and 4% of Northwest Bantu (Sakata Teke). This work follows an initial study of Allele and haplotype frequencies were estimated by an EM algo- the markers HLA class I and class II typed by serological tech- rithm (18) using the computer package Arlequin (19). The hypoth- niques in 1994 (13). esis of Hardy-Weinberg equilibrium was tested by two independent The geographic location of the two sampled populations is shown in Fig. 1. approaches (exact test and Chi-square) as explained by SanchezMazas et al. (20). Two-locus haplotypes were tested for linkage disequilibrium. For each haplotype AB, we computed the normalised linkage disequilibrium D’AB, and we tested the null hypothesis of Material and methods no linkage disequilibrium by c2 ⫽ 2nD2AB (1 degree of pA(1ªpA)pB(1ªpB) freedom), where 2n is the number of gametes (21). Global linkage Population samples disequilibrium between each pair of loci was tested both by a Chisquare and a likelihood ratio test (both tests implemented in Arle- Pygmies quin). Allelic and haplotypic frequencies observed in the sampled Blood samples were taken from 543 Pygmies and were transported populations were compared to those estimated in the North and to Paris within 36 h. The lymphocytes were separated at National sub-Saharan African population samples available at the 12th Inter- Institute of Blood Transfusion (INTS) using ficoll hypaque density national Histocompatibility Workshop and later publications (4, 5, gradient centrifugation and then cryopreserved in liquid nitrogen. 22–29) (Sanchez-Mazas, personal results). For that purpose, pair- A total of 93 samples were taken from unrelated Aka Pygmies, wise linearized Fst’s or co-ancestry coefficients (30) were computed including both sexes, and covering the 20 to 70 age group. The a) on the basis of a limited number of DRB1 alleles (13 broad speci- samples were chosen without common ancestry going back at least ficities plus a null allele) to allow the comparison of many popula- three generations. tions (n⫽21 African samples), and b) on the basis of most DRB1 alleles (65 DRB1 subtypes) for a reduced set of 13 populations. The choice of populations was done on the basis of high sample size Congolese and, for the second analysis, high resolution of DRB1 typings. Gen- Samples from 85 unrelated healthy volunteer blood donors living in etic distances (co-ancestry coefficients) were applied to a multidi- Kinshasa or its immediate surroundings were collected on EDTA. mensional scaling analysis by using the NTSYS computer package These were shipped to Paris for DNA extraction. (31). Fst values were tested for significance by a permutation procedure, and were used to investigate population genetic structure DNA extraction and HLA molecular typing by an analysis of variance (ANOVA) (32). Selective neutrality of DRB1 distributions was tested in all compared populations by Genomic DNA was extracted from frozen peripheral blood cells Ewens-Watterson’s (33, 34) and Slatkin’s (35, 36) statistical tests. All (PBMC) by two different techniques: the Miller modified method these tests were carried out by using Arlequin. Tissue Antigens 2001: 58: 211–222 213 Renquin et al : HLA class II in Aka Pygmies and Congolese Allelic frequencies and Hardy-Weinberg equilibrium significance in Aka Pygmies and Bantu Congolese Aka Pygmies (n⫽93) Bantu Congolese (n⫽85) DQB1*0501 0.218 0.171 DRB1*0101 0.005 0.018 DQB1*0502 0.000 0.012 DRB1*0102 0.000 0.041 DQB1*05031 0.000 0.012 DRB1*0301 0.118 0.065 DQB1*0601 0.013 0.000 DRB1*0302 0.000 0.029 DQB1*0602 0.182 0.299 DRB1*0401 0.000 0.012 DQB1*0603 0.038 0.006 DRB1*0403 0.005 0.000 DQB1*0604 0.064 0.035 DRB1*0405 0.011 0.012 DQB1*0605 0.000 0.012 DRB1*0406 0.000 0.006 DQB1*0606 0.005 0.000 DRB1*0701 0.241 0.065 DQB1*0608 0.011 0.000 DRB1*08 0.000 0.053 DQB1*0609 0.000 0.032 DRB1*0901 0.011 0.053 null 0.009 0.023 DRB1*1001 0.011 0.018 DRB1*11 0.065 (0.1821) A) DRB1 DRB1*1101 0.094 DRB1*1102 0.059 DRB1*1104 0.029 DRB1*1201 0.129 0.029 DRB1*1202 0.016 0.000 DRB1*1301 0.161 0.065 DRB1*1302 0.048 0.111 DRB1*1303 0.000 0.006 DRB1*1308 0.000 0.006 DRB1*1401 0.038 0.006 DRB1*15 (0.1301) 0.205 DRB1*1502 0.007 DRB1*1503 0.123 DRB1*16 DRB1*1601 null 0.018 0.005 0.006 0.000 HWE2: P-value by exact test 0.15 (n.s.) 0.42 (n.s.) HWE: P-value by c2 test 5.10ª10(3) 0.94 (n.s.) DQB1*02 0.353 0.171 DQB1*0301 0.043 0.168 DQB1*0302 0.032 0.006 DQB1*03032 0.021 0.018 DQB1*0402 0.011 0.035 B) DQB1 1 2 3 4 Aka Pygmies (n⫽93) HWE: P-value by exact test 0.16 (n.s.) 0.13 (n.s.) HWE: P-value by c2 test 0.14 (n.s.) 0.78 (n.s.) C) DPB14 (n⫽81) (n⫽83) DPB1*0101 0.049 0.229 DPB1*0201 0.030 0.175 DPB1*0301 0.025 0.060 DPB1*0401 0.031 0.054 DPB1*0402 0.695 0.199 DPB1*1101 0.000 0.030 DPB1*1301 0.006 0.048 DPB1*1401 0.000 0.006 DPB1*1701 0.000 0.042 DPB1*1801 0.049 0.073 DPB1*1901 0.006 0.012 DPB1*2901 0.000 0.006 DPB1*3201 0.000 0.006 DPB1*3401 0.006 0.000 DPB1*3901 0.068 0.012 DPB1*4001 0.031 0.012 DPB1*4901 0.000 0.024 DPB1*5101 0.000 0.006 DPB1*5501 0.000 0.006 null 0.004 0.000 0.73 (n.s.) 0.39 (n.s.) 0.99 (n.s.) 0.99 (n.s.) HWE: P-value by exact test 214 Tissue Antigens 2001: 58: 211–222 2 HWE: P-value by c test Computed as the sum of the subtypes frequencies HWE: Hardy-Weinberg equilibrium hypothesis P∞0.001 but two rare phenotypes (DRB1 1001/0101 and DRB1 1502) account for 68% of the total c2 Not the same sample sizes for DPB1 as for DRB1 and DQB1 Table 1 Bantu Congolese (n⫽85) Renquin et al : HLA class II in Aka Pygmies and Congolese tively). Finally, a drastic difference is observed for allele DPB1*0402, Results and discussion as the frequency found in the Aka reaches 69.5% (against 19.9% in the Congolese). More even frequencies characterise the DPB1 dis- Allelic frequencies and Hardy-Weinberg equilibrium tribution estimated in the Congolese. Overall, the Aka exhibit a Allele frequencies are presented in Table 1. The hypothesis of lower level of HLA genetic diversity than the Bantu, and especially Hardy-Weinberg equilibrium can be accepted at all loci (DRB1, so at the DPB1 locus (observed heterozygosities in Aka versus Con- DQB1, DPB1) in both Aka and Bantu Congolese. This result is quite golese: 86.1% vs. 90.9% at the DRB1 locus, 78.6% vs. 81.9% at the surprising for the Congolese sample, which is constituted of individ- DQB1 locus, and 50.4% vs. 85.9% at the DPB1 locus). uals coming from many different tribes. It suggests that the test for For some alleles, the frequencies observed in the Aka are in fact Hardy-Weinberg equilibrium is conservative, but also that Bantu close to those found in East Africans: DRB1*0701 and DQB1*02 populations are genetically homogeneous. are frequent in Oromo and Amhara (where the frequency of In spite of being geographically close, the two populations ex- DRB1*0701 is about 20%, and that of DQB1*02 more than 30%). hibit marked differences in their allelic profile. The most contrasted For other alleles, Aka frequencies are closer to those found in Bantu frequencies are those found for DRB1*0701, 0301, 1201 and 1301, populations: DRB1*1301 (16% in the Aka) is frequent in the Bubi much higher in the Aka (24.1%, 11.8%, 12.9% and 16.1%, respec- (23%), whereas another DRB13 subtype, DRB1*1302, is frequent in tively), as well as DRB1*15, 11 and 1302, predominant in the Congo- East Africans, and DQB1*0501 (22% in the Aka) is generally fre- lese (20.6%, 18.2%, and 11.2%, respectively). At the DQB1 locus, quent in the Bantu (15–20%). From these results, it is not clear we mostly note higher DQB1*02 and 0501 and lower (although high) whether the Aka are genetically closer to Afroasiatic or Bantu popu- DQB1*0602 frequencies in the Aka (35.3%, 21.8%, 18.2%, respec- lations. DRB1-DQB1 most frequent haplotypes (frequency ⬎3% in at least one population) and linkage disequilibrium in Aka Pygmies and Bantu Congolese Aka Pygmies (n⫽93) Dø Bantu Congolese (n⫽85) 2 c DRB1-DQB1 Frequency 0102–0501 – – 0301–02 0.112 0.929 0701–02 0.219 0.863 08–0301 – – 0901–02 0.011 1.000 11–0301 0.043 1.000 1101–0301 – – 1102–0301 – – 1104–0301 – 11–0602 0.011 1101–0602 – – 1201–0501 0.129 1.000 1301–0602 0.059 0.226 1301–0603 0.032 0.830 1301–0604 0.032 0.404 4.27 1302–0501 0.011 0.005 ⬍0.001 1302–0604 0.032 0.644 25.3 1401–0501 0.032 0.817 15–0602 – – 1503–0602 0.112 0.893 – ª0.084 Table 2 Frequency – 0.035 0.827 9.59 ** 11.2 ** 0.065 1.000 22.02 ** 19.8 ** 0.057 0.849 15.87 ** – – 0.047 0.866 13.92 ** 1.19 n.s. 0.053 1.000 18.01 ** ** – – – – 0.024 0.099 – – 0.059 1 – 54.3 Dø c2 Significance Significance – – 0.32 n.s. 20.59 ** – – 0.012 0.279 0.8 * 0.008 n.s. – – – – – – 0.071 0.643 5.45 * ** 0.021 0.638 4.07 * n.s. 0.018 ª0.087 0.01 n.s. ** 0.006 1.000 6.76 ** * – – – n.s. 0.062 0.460 8.05 ** ** 0.018 0.437 4.05 * 6.56 * – – – – – – 0.200 0.959 ** – – 33.5 2.82 10.5 33.6 – 26.5 – ** – **P⬍0.01; *P⬍0.05; frequencies in bold: ⬎10%. Tissue Antigens 2001: 58: 211–222 215 Renquin et al : HLA class II in Aka Pygmies and Congolese DRB1-DQB1-DPB1 most frequent haplotypes (frequency ⬎3%) and linkage disequilibrium significance in Aka Pygmies and Bantu Congolese Genetic distances among African populations Linkage disequibrium significance The multidimensional scaling analyses (MDS) based on DRB1 gen- Frequency DRB1-DQB1 DQB1-DPB1 DRB1-DPB1 etic distances and carried out on the broad (21 populations) and 0701–02–0402 0.216 ** n.s. n.s. shown in Figs. 2 and 3, respectively. An MDS analysis carried out 1201–0501–0402 0.111 ** n.s. n.s. on DRB1-DQB1 haplotypes on 18 populations gives very close re- 03011–02–0402 0.075 ** n.s. n.s. sults to those presented in Fig. 2 (results not shown). 1503–0602–0402 0.072 ** n.s. n.s. Except for the Aka, the populations were clustered according to 11–0301–0402 0.037 ** n.s. n.s. the linguistic phylum to which their language belongs. The Merina 1301–0602–3901 0.031 n.s. ** n.s. from Madagascar are peculiar as they speak an Austronesian lan- 1503–0602–3901 0.031 ** ** * guage of Asian origin. They are clearly discriminated genetically DRB1-DQB1-DPB1 haplotype reduced (13 populations) data sets (see Material and methods) are A) Aka Pygmies (n⫽81) from all other African populations due to the main contribution of B) Bantu Congolese (n⫽83) 15–0602–0201 0.100 ** * ** South-East Asian ancestors to their present genetic pool (39). In 15–0602–0402 0.056 ** n.s. n.s. spite of their relative overlap, speakers of the other linguistic groups 1101–0602–0101 0.030 * n.s. n.s. represented (Afroasiatic, Niger-Congo, Khoisan) can be discrimi- **P⬍0.01; *P⬍0.05; n.s. not significant nated genetically from each other, suggesting a significant corre- Table 3 lation between genetic and linguistic differentiations, in keeping with previous results (1, 4, 25). This is confirmed statistically here by the ANOVA analysis, which shows that the variance due to differentiations among linguistic groups, Fct, is highly significant Haplotype frequencies and linkage disequilibrium (Fct⫽0.016, PⰆ0.0001) (the Aka were included in the Niger-Congo cluster for this analysis). A reasonable interpretation of the corre- In Tables 2 and 3 are presented the estimated frequencies and link- lation found between genetic and linguistic diversity in Africa can age disequilibrium significance for DRB1-DQB1 and DRB1-DQB1- be found in previous publications (2, 3, 5). DPB1 haplotypes, respectively. The frequency differences observed in Table 2 between Aka and Bantu Congolese are in keeping with However, we also note variable degrees of genetic heterogeneity within each linguistic group: what we expected from the observed allelic frequencies (Table 1) and the tight linkage disequilibrium generally found between loci DRB1 and DQB1 (37, 38). – among the Afroasiatic, the Ethiopian Oromo and Amhara (subSaharan populations) are grouped together and genetically differ- According to Table 3A, the 4 most frequent DRB1-DQB1 haplo- entiated from the North African populations. Geographic differen- types found in the Aka are observed in combination with tiation, with limited gene flow, on both sides of the Sahara Desert DPB1*0402. This is clearly due to the extremely high frequency may explain this result. observed for this allele (almost 70%), rather than to linkage disequi- – among the Khoisan, the San, with a very high frequency of librium. Indeed, linkage disequilibrium is found to be non signifi- DRB1*0401 (42%), are highly differentiated from the Khoi and cant between DPB1 and the other loci for most haplotypes (Table from all other populations, probably as a result of a rapid genetic 3). Moreover, global tests of linkage disequilibrium indicate a sig- drift. On the other hand, the Khoi exhibit some genetic affinities nificant association between DRB1 and DQB1 in both populations with Afroasiatics, and more particularly with the East Africans (PⰆ0.001 by both methods in both populations), while only one of Oromo and Amhara (high frequency of DRB1*1302 and DRB*15), the 8 tests of linkage disequilibrium involving DPB1 (2 methods, as shown in Figs. 2 and 3. This corroborates the hypothesis that applied to the 2 loci pairs DRB1-DPB1 and DQB1-DPB1, in the 2 the ancient Khoisan homeland was much more extended than populations) is found significant (DQB1-DPB1 by the exact test in today and included East Africa (2). Contrary to the San, the Khoi the Aka). The almost independence of DPB1 relatively to the other adopted pastoralism and probably kept more contacts with their HLA loci has been suggested before (38). neighbors. As noted above for DPB1 taken alone, the DRB1-DQB1-DPB1 – among the Niger-Congo, the West African Mandenka and the distribution observed in the Congolese is more even than in the Aka Bubi (who live on an island at an intermediate location between (Table 3B). western and southern Africa), are both genetically close to each 216 Tissue Antigens 2001: 58: 211–222 Renquin et al : HLA class II in Aka Pygmies and Congolese Fig. 2. Multidimensional scaling analysis of 21 African populations based on the frequencies of 14 DRB1 allele frequencies (DRB1*01, 03, 04, 07, 08, 09, 10, 11, 12, 13, 14, 15, 16, null). Population samples are: 93 Aka Pygmies from the Central African Republic (present study), 85 Bantu Congolese for Congo Kinshasa (present study), 198 Mandenka from Eastern Senegal (5), 101 Bubi from the island of Bioko, Equatorial Guinea (22)*, 167 Banzabi from Gabon (24)*, 72 Matabele and 82 Shona from Zimbabwe, 199 Zulu from South Africa, 75 Sotho from Lesotho, 163 Merina from Madagascar (4)*, 107 M’Zab from Algeria, 101 Beduin from Siwa (Egypt), 119 Egyptians (Mansoura), 107 Egyptians (Assouan), 99 Tunisian (25)*, 101 Algerian (25, 26), 98 Moroccan (25*, 27), 98 Amhara and 83 Oromo from Ethiopia (28)*, 77 San and 91 Khoi from Namibia (29). Population names are bolded and/or underlined according to the results of Ewens-Watterson’s and Slatkin’s tests for selective neutrality. Bold: P⬍0.05 by both tests; underlined: P⬍0.05 by one test and P⬎0.05 by the other test; neither bolded nor underlined: P⬎0.05 by both tests. Non-significant Fst’s are found for the following population pairs: a) P⬎0.05: Amhara-Oromo, Algerian-Moroccan, Algerian-Tunisian, Moroccan-Tunisian, Sotho-Matabele, Matabele-M’Zab, Egyptian (South)-Beduin; b) 0.01⬍P⬍0.05: Sotho-Zulu, Sotho-Shona, Sotho-Algerian, Sotho, Egyptian (Delta), Sotho-Egyptian (South), Sotho-Beduin, Congolese-Matabele, Congolese-Beduin, CongoleseBanzabi, Zulu-Matabele, Zulu-Tunisian, Matabele-Beduin, Algerian-Egyptian (South), Algerian-Beduin. *Raw data from XIIth International Histocompatibility Workshop and Conference, personal communication to A. Sanchez-Mazas. other and differentiated from the other populations (Fig. 2). This ment with geography. Coming back to Fig. 2, we note that most is due to a high frequency of DRB*13 (37%). However, the Bantu populations (Matabele, Shona, Sotho, Zulu, Banzabi and DRB*13 subtype involved is in fact not the same in the two popu- Congolese) are genetically close to each other, in keeping with lations: DRB1*1304 is the most frequent in the Mandenka (25% their common origin. On the other hand, they are also close to (5)), and DRB1*1301 the most frequent in the Bubi (23% (22)). Afroasiatics, and, more particularly, North Africans. As indicated This is revealed by Fig. 3 (where fewer populations but more in the Legends for Figs. 2 and 3, significance tests on Fst’s be- HLA alleles are considered), showing that the Bubi are genetically tween populations indeed reveal many non-significant values intermediate between the Mandenka and the Congolese, in agree- among Bantu and North Africans. This unexpected result is hardTissue Antigens 2001: 58: 211–222 217 Renquin et al : HLA class II in Aka Pygmies and Congolese Fig. 3. Multidimensional scaling analysis of 13 African populations based on the frequencies of 65 DRB1 allele frequencies (DRB1*0101, 0102, 0103, 0104, 0301, 0302, 0303, 0304, 0305, 0401, 0402, 0403, 0404, 0405, 0406, 0407, 0408, 0410, 0412, 0413, 0414, 0417, 0422, 07, 08, 0901, 1001, 1101, 1102, 1103, 1104, 1105, 1107, 1109, 1113, 1116, 1117, 1201, 1202, 1301, 1302, 1303, 1304, 1305, 1307, 1308, 1309, 1310, 1311, 1312, 1314, 1315, 1319, 1401, 1402, 1403, 1404, 1407, 1408, 1409, 1414, 1418, 15, 16, null). Population samples are listed in the Legend for Fig. 2. Population names are underlined according to the results of Ewens-Watterson’s and Slatkin’s tests for selective neutrality. Underlined: P⬍0.05 by one test and P⬎0.05 by the other test; not underlined: P⬎0.05 by both tests. ly explained by historical relationships, and may involve selective populations. Pygmies are presently scattered across many regions effects, as discussed below. Finally, the Aka Pygmies, although of Central Africa, with some groups (either Pygmies or ‘‘Pygmo- included in the Bantu cluster from a linguistic point-of-view, are ids’’) living in eastern regions located in Sudan, Uganda and Ethi- genetically isolated. This is due, for example, to their high fre- opia (6). They probably inhabited a larger part of Central Africa quency of DRB1*1201, very rare in other populations. Their gen- before, and underwent several migrations between eastern and etic divergence is easily explained by their peculiar mode of life western regions. (as hunters-gatherers) and their low population size (small tribes), which probably lead to a rapid genetic drift. However, they also Neutrality tests share genetic affinities with both Afroasiatics and Bantu, as mentioned before on the basis of allelic frequencies. Besides a possible Both tests of selective neutrality (Ewens-Watterson’s and Slatkin’s) convergence effect due to natural selection, as explained below, it used in the present study give non-significant results at the 3 HLA is thus likely that the Aka Pygmies, who now speak a Bantu class II loci for both Aka and Congolese (all P⬎0.05). This is not language, were subject to some admixture with neighboring remarkable for DPB1, which generally exhibits skewed and neutral- 218 Tissue Antigens 2001: 58: 211–222 Renquin et al : HLA class II in Aka Pygmies and Congolese like frequency distributions (37, 38, 40). At the other HLA class II populations projected at the periphery of the multidimensional scal- loci, the present results confirm that rejection of the neutral hypoth- ing analysis exhibit a neutral-like DRB1 distribution, while the esis is not systematic. populations projected at the centre, where Afroasiatic and Bantu We also applied neutrality tests to the DRB1 distributions of all linguistic groups overlap, exhibit an excess of heterozygotes. From African populations included in our genetic distance analyses. The these results, we may suggest that balancing selection, already results are given in Table 4 and reported in Figs. 2 and 3, where shown to be acting on MHC (5, 20, 37, 38, 41), maintained a high the populations for which we observe a significant departure from level of polymorphism in many populations, often leading to ap- neutrality due to a lower homozygosity are bolded and/or under- parent genetic affinities between populations which are distantly lined. related from a historical and/or a geographical point-of-view. This According to Fig. 2, where we consider many populations but a would explain in part the genetic proximity of Afroasiatics and limited number of alleles, most Bantu and Afroasiatic populations Bantu. On the other hand, the results are quite different when we exhibit a significant excess of heterozygotes. The exceptions are the consider a lower number of populations and a maximal number of Banzabi (Bantu from Gabon (24)), the Mandenka (5), both Khoisan alleles (Table 4 and Fig. 3): only non-significant or contradictory populations, and the Aka Pygmies. For some populations (Bubi, results are observed. We note from Table 4 that the fact of including Matabele, Beduin, Oromo, Tunisian, and Merina), Ewens-Watter- a large number of alleles reduces the expected homozygosity, while son’s and Slatkin’s tests give contradictory results (one rejection and the observed homozygosity does not change substantially. The rejec- one acceptance of the null hypothesis). Overall, we note that the tion of the neutral hypothesis, in the first analysis, may be biased Results of Ewens-Watterson’s and Slatkin’s selective neutrality tests applied to the DRB1 frequency distributions of 21 African population samples DRB1 broad specificities (max. number⫽14) Populations N n Fobs DRB1 subtypes (max. number⫽65) Fexp PEW PSL N n Fobs Fexp PEW PSL 0.543 Aka Pygmies 93 12 0.162 0.255 0.082 0.255 93 17 0.139 0.183 0.236 Bantu Congolese 85 13 0.136 0.238 0.047 0.028 85 22 0.091 0.134 0.094 0.050 Mandenka 196 13 0.207 0.280 0.263 0.118 196 22 0.115 0.166 0.141 0.109 Bubi 101 13 0.194 0.251 0.312 0.042 101 20 0.097 0.157 0.051 0.006 San 77 11 0.277 0.267 0.648 0.600 77 14 0.221 0.210 0.666 0.417 Khoi 91 13 0.176 0.241 0.238 0.165 91 19 0.139 0.163 0.395 0.155 Amhara 98 10 0.145 0.319 0.005 0.005 98 20 0.117 0.154 0.210 0.336 Oromo Tunisians 83 11 0.156 0.279 0.034 0.092 83 19 0.116 0.157 0.207 0.133 101 13 0.135 0.243 0.026 0.056 101 26 0.092 0.114 0.262 0.040 Algerians 100 13 0.121 0.248 0.009 0.008 100 28 0.077 0.102 0.150 0.054 Egyptians (Delta) 119 14 0.126 0.239 0.035 0.017 119 52 0.055 0.050 0.776 0.732 Egyptians (South) 107 14 0.124 0.230 0.021 0.014 107 37 0.060 0.075 0.231 0.223 Beduin 102 13 0.135 0.245 0.023 0.054 102 27 0.083 0.112 0.163 0.140 Sotho 75 13 0.112 0.227 0.003 0.017 Shona 82 14 0.127 0.221 0.036 0.036 199 13 0.135 0.279 0.013 0.006 Zulu Matabele Banzabi Moroccans 72 13 0.143 0.223 0.071 0.012 167 13 0.191 0.270 0.223 0.247 95 13 0.133 0.246 0.019 0.026 M’Zab 107 13 0.123 0.248 0.005 0.005 Merina 163 14 0.179 0.256 0.214 0.015 N: Sample size (number of individuals); n: number of alleles; Fobs: observed homozygosity; Fexp: expected homozygosity; PEW: P-value obtained by Ewens-Watterson’s test (1000 permutations); PSL: P-value obtained by Slatkin’s test (1000 permutations) P-values in bold are significant at the 5% level. Table 4 Tissue Antigens 2001: 58: 211–222 219 Renquin et al : HLA class II in Aka Pygmies and Congolese by considering only broad specificities. Alternatively, its acceptance, in the second analysis, may be due in part to a reduced power of Concluding remarks the test to reject the null hypothesis when the number of alleles is high, the sample size is low, and the selective effect is weak (41). Thanks to the present HLA class II analyses of two new African We should thus be care when interpreting the results of selective population samples, the Aka Pygmies and the Bantu Congolese, and neutrality tests applied to HLA. the effort of previous HLA workshops to gather available data for According to these results, balancing selection apparently influ- many populations (23, 29, 45), a clearer picture is emerging on HLA ences the pattern of HLA diversity in African populations, but this diversity at the continental level of Africa. Generally speaking, the effect is not very extensive. First, we have shown that DRB1 popula- patterns of HLA diversity reflect population history as for most tion genetic structure is significant and highly correlated to linguis- classical polymorphisms, which exhibit high correlations with tic differentiations, which indicates that population past relation- linguistic and/or geographic differentiations (46, 47). Unlike other ships largely influenced the pattern of HLA genetic diversity. Sec- systems, however, a higher level of diversity than expected under a ond, a departure from neutrality is not always demonstrated, or neutral mode of evolution is observed in many populations, al- demonstrable, statistically. For many samples, statistical signifi- though to variable extents, and not at all HLA loci (37, 40, 41, 48). cance depends on the number of alleles considered (differences be- This diversifying selection, to which some kind of pathogen re- tween Figs. 2 and 3). For others, we never reject the neutral hypoth- sistance may be due (49), is expected to shift the migration-drift esis either considering a large number of alleles or not. In the latter balance towards the maintenance of low frequencies for a high num- case, a possible explanation is that the effect of genetic drift ber of alleles, and to act as an opposite force to population differen- counterbalanced that of balancing selection, especially in Khoisan tiation. We have indeed shown here that selection may be a possible and Pygmies who have been isolated for a long time from other explanation for the occurrence of some low genetic distances be- populations. However, the Senegalese Mandenka, although being an tween geographically or linguistically distant populations. The endogamous community (42), probably underwent a recent popula- overall pattern of HLA diversity is however not drastically affected tion expansion (43), and their genetic pool may be representative of by selective effects, and the HLA polymorphism may still be con- broader West African genetic diversity (44). sidered as a useful tool in anthropology. References 1. Dard P, Sanchez-Mazas A, Tiercy J-M et al. Joint report: HLA-A, -B, and DR differentiations among North and West African populations. In: Tsuji K, Aizawa M, Sasazuki T, eds. HLA 1991. Oxford: Oxford University Press, 1992: 632–636. 2. Excoffier L, Pellegrini B, Sanchez-Mazas A, Simon C, Langaney A. Genetics and history of sub-Saharan Africa. Yearbook Phys Anthrop 1987: 30: 151–194. 3. Excoffier L, Harding RM, Sokal RR, Pellegrini B, Sanchez-Mazas A. Spatial differentiation of RH and GM haplotype frequencies in sub-Saharan Africa and its relation to linguistic affinities. Hum Biol 1991: 63: 273–307. 220 4. Hammond MG, Du Toit ED, Sanchez-Mazas A et al. HLA in sub-Saharan Africa: 12th International Histocompatibility Workshop: SSAF report. In: Charron D, ed. Genetic diversity of HLA: functional and medical implication. Paris: EDK, 1997: 345– 353. 5. Tiercy J-M, Sanchez-Mazas A, Excofffier L et al. HLA-DR polymorphism in a Senegalese Mandenka population: DNA oligotyping and population genetics of DRB1 specificities. Am J Hum Genet 1992: 51: 592–602. 6. Cavalli-Sforza LL. African Pygmies. Orlando: Academic Press, 1986. 7. Ngaro SB, Jaeger G, Pinerd G. Study of a tropical forest ecosystem on the basis of human data. Nat Resource 1977: 13: 8–14. Tissue Antigens 2001: 58: 211–222 8. Bahuchet S. Notes pour l’histoire de la région de Bagandu. In: Bahuchet S, ed. Pygmées de Centrafrique. Paris: SELAF, 1976: 51–72. 9. Muller JY, Hallé L, Jaeger G. HLA-A and B antigens in AKA pygmies. Tissue Antigens 1981: 17: 372–375. 10. Ndaywel è Nziem I. L’espace et les hommes. In: Ndaywel è Nziem I, ed. Histoire générale du Congo. De l’héritage ancien à la République Démocratique. Paris: De Boeck et Larcier, 1998: 41–56. 11. Ndaywel è Nziem I. Le Congo ancien méridional. In: Ndaywel è Nziem I, ed. Histoire générale du Congo. De l’héritage ancien à la République Démocratique. Paris: De Boeck et Larcier, 1998: 79–225. 12. Ruhlen M. A Guide to the World’s Languages. V.1, Classification. London: Edward Arnold, 1987. Renquin et al : HLA class II in Aka Pygmies and Congolese 13. Hallé L, Mbayo K, Lurhuma Z et al. HLAA, B, C, DR and DQ polymorphism in zairians. Tissue Antigens 1994: 44: 196– 199. 14. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988: 16: 1215. 15. Tagliaferro L, Corbelli M, Maietta G, Pellegrino V, Pignatelli P. Use of a rapid and simple method to extract proviral DNA in the identification of HIV-1 by PCR. New Microbiol 1995: 18: 303–306. 16. Buyse I, Decorte R, Baens M et al. Rapid DNA typing of class II HLA antigens using the polymerase chain reaction and reverse dot blot hybridization. Tissue Antigens 1993: 41: 1–14. 17. Olerup O, Aldener A, Fogdell A. HLADQB1 and DQA1 typing by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours. Tissue Antigens 1993: 41: 119–134. 18. Excoffier L, Slatkin M. Maximum-likehood estimation of molecular haplotype frequencies in a diploid population. Molecular Biol Evol 1995: 12: 921–927. 19. Schneider S, Roessli D, Excoffier L. Arlequin: A software for population genetics data analysis. ver. 2.0. Geneva: Genetics and Biometry Laboratory, Department of Anthropology, University of Geneva, 2000. 20. Sanchez-Mazas A, Steiner QG, Grundschober C, Tiercy J-M. The molecular determination of HLA-Cw alleles in the Mandenka (West Africa) reveals a close genetic relationship between Africans and Europeans. Tissue Antigens 2000: 56: 303– 312. 21. Weir BS. Genetic data analysis II. Sunderland, MA: Sinauer Associates, 1996. 22. de Pablo R, Garcia-Pacheco JM, Vilches C. HLA class I and class II allele distribution in the Bubi population from the island of Bioko (Equatorial Guinea). Tissue Antigens 1997: 50: 593–601. 23. Bodmer J, Cambon-Thomsen A, Hors J, Piazza A, Sanchez-Mazas A. Anthropology report: introduction. In: Charron D, ed. Genetic diversity of HLA: functional and medical implication. Paris: EDK, 1997: 269– 284. 24. Migot-Nabias F, Fajardy I, Danze PM et al. HLA class II polymorphism in a Gabonese Banzabi population. Tissue Antigens 1999: 53: 580–5. 25. Hors J, El Chenawi F, Djoulah S et al. HLA in North African populations: 12th International Histocompatibility Workshop: NAFR report. In: Charron D, ed. Genetic diversity of HLA: functional and medical implication. Paris: EDK, 1997: 328– 334. 26. Djoulah S, Sanchez-Mazas A, Khalil L et al. HLA-DRB1, DQA1 and DQB1 DNA polymorphisms in healthy Algerian and genetic relationships with other populations. Tissue Antigens 1994: 43: 102–109. 27. Izaabel H, Garchon H-J, Caillat-Zuccman S et al. HLA class II DNA polymorphism in a Moroccan population from the Souss, Agadir area. Tissue Antigens 1998: 51: 106– 110. 28. Fort M, de Stefano GF, Cambon-Thomsen A et al. HLA class II allele and haplotype frequencies in Ethiopian Amhara and Oromo populations. Tissue Antigens 1998: 51: 327–36. 29. Imanishi T, Akaza T, Kimura A, Tokunaga K, Gojobori T. Allele and haplotype frequencies for HLA and complement loci in various ethnic groups. In: Tsuji K, Aizawa M, Sasazuki T, eds. HLA 1991. Oxford: Oxford University Press, 1992: 1065–1220. 30. Reynolds J, Weir BS, Cockerham CC. Estimation for the coancestry coefficient: basis for a short-term genetic distance. Genetics 1983: 105: 767–779. 31. Rohlf FJ. NTSYS-PC, Numerical Taxonomy and Multivariate Analysis System. v. 2.02g. Setauket, NY: Exeter Software, Applied Biostatistics, Inc., 1998. 32. Excoffier L, Smouse P, Quattro J-M. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 1992: 131: 479–491. 33. Watterson GA. The homozygosity test of neutrality. Genetics 1978: 88: 405–417. 34. Watterson GA. The homozygosity test after a change in population size. Genetics 1986: 112: 899–907. 35. Slatkin M. A correction to the exact test based on the Ewens sampling distribution. Genet Res 1996: 68: 259–260. 36. Slatkin M. An exact test for neutrality based on the Ewens sampling distribution. Genet Res 1994: 64: 71–74. 37. Bugawan TL, Klitz W, Blair A, Erlich HA. High-resolution HLA class I typing in the CEPH families: analysis of linkage disequilibrium among HLA loci. Tissue Antigens 2000: 56: 392–404. 38. Sanchez-Mazas A, Djoulah S, Busson M et al. A linkage disequilibrium map of the MHC region based on the analysis of 14 loci haplotypes in 50 French families. Eur J Hum Genet 2000: 8: 33–41. 39. Soodyal H, Jenkins T, Hewitt R, Krause A, Stoneking M. The peopling of Madagascar. In: Boyce AJ, Mascie-Taylor CGN, eds. Molecular biology and human diversity. Cambridge: Cambridge University Press, 1996. 40. Grundschober C, Sanchez-Mazas A, Excoffier L, Langaney A, Jeannet M, Tiercy J-M. HLA-DPB1 DNA polymorphism in the Swiss population: linkage disequilibrium with other HLA loci and population genetic affinities. Eur J Immunogenet 1994: 21: 143–157. 41. Sanchez-Mazas A. African diversity from the HLA point of view: influence of genetic drift, geography, linguistics, and natural selection. Hum Immunol 2001: 62: 937– 948. 42. Blanc M, Sanchez-Mazas A, Hubert van Blyenburgh N, Sevin A, Pison G, Langaney A. Inter-ethnic genetic differentiation: Gm polymorphism in eastern Senegal. Am J Hum Genet 1990: 46: 383–392. 43. Excoffier L, Schneider S. Why huntergatherer populations do not show signs of Pleistocene demographic expansions. Proc Nat Acad Sci U S A 1999: 96: 10597– 10602. 44. Excoffier L, Poloni ES, Santachiara Benerecetti S, Semino O, Langaney A. The molecular diversity of the Niokholo Mandenkalu from Eastern Senegal: an insight into West Africa genetic history. In: Boyce AJ, Mascie-Taylor CGN, eds. Molecular biology and human diversity. Cambridge: Cambridge University Press, 1996: 141–155. 45. Clayton J, Lonjou C, Whittle D. Allele and haplotype frequencies for HLA loci in various ethnic groups. In: Charron D, ed. Genetic diversity of HLA: functional and medical implication. Paris: EDK, 1997: 665– 820. 46. Cavalli-Sforza LL, Menozzi P, Piazza A. The history and geography of human genes. Princeton, NJ: Princeton University Press, 1994. Tissue Antigens 2001: 58: 211–222 221 Renquin et al : HLA class II in Aka Pygmies and Congolese 47. Chen J, Sokal RR, Ruhlen M. Worldwide analysis of genetic and linguistic relationships of human populations. Hum Biol 1995: 67: 595–612. 222 48. Sanchez-Mazas A, Tiercy J-M, Kervaire B, Langaney A. HLA-DPB1 in anthropology: A unique polymorphism [Abstract]. Hum Immunol 2000: 61 (Suppl 1): S127. Tissue Antigens 2001: 58: 211–222 49. Hill AVS, Sanchez-Mazas A, Barbujani G et al. Human genetic variation and its impact on public health and medicine. In: Stearns SC, ed. Evolution in health and disease. Oxford: Oxford University Press, 1999: 62– 74.
© Copyright 2026 Paperzz