Population distribution of the human vitamin D binding protein Anthropological considerations.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 68:107-122 (1985) Population Distribution of the Human Vitamin D Binding Protein: Anthropological Considerations J. CONSTANS, S. HAZOUT, R.M.GARRUTO, D.C. GAJDUSEK, E.K. SPEES Centre d’Hkmotypobgie C.N.R.S., CHU Purpan, 31300 Toulouse (J.C.), AND Unite de Recherche5 Bwmathkmatiques et Biostatistiques, U.263 INSERM, Uniuersitk Paris 7, Paris (S. H.), France; National Institutes of Health, Bethesda (R.M. G., D. C.G.), Diuiswn of Transplantation Surgery, Baltimore City, Baltimore (E.K.S.) DBP subtypes, Rare alleles, Worldwide distribution, KEY WORDS Migrations, Heterogeneity, Clusters ABSTRACT The polymorphism of the serum vitamin D binding rotein (DBP) in humans is based on the existence of three common alleles, Gclf, Gc”, and Gc2, and 84 rare alleles. The geographical distribution of GdF, Gc”, and Gc2 alleles shows north to south clines, together with a balanced equilibrium between the GclF or GclS allele frequency and the Gc2 frequency. The distribution of the FSTvalues shows high variability within a geographical area. For European and North Asiatic groups, the FSTvalues are the lowest observed, and the reason may be a long process of homogenization. Aboriginal populations from Australia and New Guinea and groups from both North Africa and South America show the greatest heterogeneity of their allele frequencies. Systematic factors such as genetic drift and selection may account for this distribution. In contrast with the three main DBP alleles, the distribution of the rare alleles corresponds to patterns of human migrations that occurred during prehistoric and historic periods. Thus, the rare mutants are of particular relevance to anthropological and genetical investigations. The polymorphism of the vitamin D binding protein (DBP) also called group-specific component, had been restricted to three phenotypes, Gcl-1, Gc2-1 and Gc2-2, and to six rare mutants (Constans and Viau, 1977). The electrophoretic methods used were immunoelectrophoresis, agarose electrophoresis, and immunofixation (Hirschfeld et al., 1960; Johnson et al., 1975). In recent years, the electrophoretic procedures employed in the study of protein polymorphism in human groups and animal species have significantly improved with the development of the techniques of isoelectric focusing (IEF on polyacrylamide and agarose gels) (Righetti, 1979; Radola, 1980). Since its first description, IEF itself has been improved, and further advances are expected (Gorg et al., 1983). Despite IEF’s being a n indispensable tool, such traditional procedures as horizontal or vertical electrophoresis on agarose and polyacrylamide gels are ‘c 1985 ALAN R LISS, INC still useful for the standardization of new data, and their use complements IEF (Constans and Cleve, 1979; Cox, 1981). The proteins are known t o differ in charge, conformation, or both, and the charges distributed on the protein surface are those that contribute the most to electrophoretic mobilities. Detergents, such as urea and SDS, are now more frequently introduced when studying tissue and serum protein polymorphisms in order to reveal charge differences concealed by the three-dimensional configuration of the proteins (Aquadro and Avise, 1981). These different procedures were used in various combinations when investigating the polymorphism of the serum DBP in human populations. In addition to three main alleles, GdF, Gc“, and Gc2, 84 rare mutants Received August 9, 1983; revised April 3, 1985; accepted April 29, 1985. 108 J. CONSTANS ET AL were detected and defined (Constans et al., 1983). This study presents the results of the analysis of the polymorphism of the serum DBP determined on more than 9,000 serum specimens from French, African, Indian, Asiatic, and Amerindian groups. Unpublished data on western and southern Pacific regions are also included; altogether 54 populations have been investigated. am* m m m t-t-hrn 0t-e-l mNNm 000 0000 m ro m ro N*Ln m 0 0 t- *0 9 0 MATERIALS AND METHODS Electrophoresis in the form of isoelectric focusing electrophoresis was performed on a polyacrylamide gel using a 4-6.5 pH range ampholyte solution, as previously described (Constans and Viau, 1977). After migration, the DBP phenotypes were determined using the technique of print-immunofixation. The Gc frequences given in this study for French groups, including a n additional sample of 190 Spanish Basques (Table 1) and some African populations, were published elsewhere (Constans et al., 1980al. The materials are composed of four series. The first series consisted of a total of 647 sera (Table 2) obtained from the following groups: three samples from Africa: (1) Erythreans (Ethiopia),(2) a Bantu-speaking group from Cameroon (Bamilikk tribe), and (3) a limited sample of Bushmen from South-Africa; samples from two tribes living on the southeast and the west coasts of Malagasy; and a sample of blacks from Baltimore, Maryland. C N z W u3* 00 8 88 0 0 t- * N x x mww * o w d c met-? rnrnrod 000 gqgc oooc m*m wt-ma gwg The second series included a total of 1,530 sera (Table 3) from the following groups: one tribe, the Jirel-Sherpa of Nepal; one sample obtained among the Dogpa tribe of Tibet; Tamil individuals living in Pondicherry, India; a sample collected from Bangkok blood donors in Thailand; South Vietnamese refugees sampled a t random on the occasion of a medical examination for a n epidemiological survey; a group of refugees belonging to a H’Mong tribe from Laos; and the people of the Karangasem community from Bali. The third series, a total of 1,870 sera, had been collected among the following groups: h N 25 24 23 21 22 20 19 18 17 15 16 14 13 12 11 10 Algeria Harratin and M' Rabtin (Saoura) Tuareg Isseqamaren (Hoggar) Mali Tuareg Kel-Kummer (Menaka) Ethiopia Erythrean Djibouti Afar Issa North Yemen: Bedouins Iraq: Kurds Senegal Peulh Fula Cameron Bantu Bamileke Central Africa Pygmies Bi-Aka Sara South Africa Bushmen Malagasy West coast East coast United States (Black people, Baltimore) 0.629 31 0.727 0.641 0.667 0.835 291 120 121 126 0.608 0.825 914 123 0.780 0.270 0.224 135 58 357 0.463 0.430 0.500 0.315 0.425 0.083 0.150 0.167 0.306 0.091 0.185 0.077 0.115 0.589 0.596 0.358 0.430 0.357 0.677 0.541 GC'F __ 0.446 0.482 95 92 126 260 160 N 161 Population 0.029 0.030 0.032 0.005 0.034 0.004 0.027 0.004 0.008 0.005 0.004 0.004 0.010 0.004 0.010 0.005 0.008 0.016 0.008 0.008 0.008 0.004 0.004 0.090 0.005 TABLE 2. DBP allele frequencies among North African and West African groups, Malagasy, and Middle East populations 0.015 0.008 0.003 0.149 0.171 0.130 0.065 0.069 0.083 0.086 0.053 0.137 0.172 0.179 0.125 0.131 0 0.015 0.054 110 J. CONSTANS ET AL. TABLE 3. DBP allele frequencies in Asiatic groups Population 26 27 28 29 30 31 32 Nepal (Jirel Sherpa) Tibet (Dogpa) India (Pondicherry Tamil) Thailand (Bangkok Thais) South Vietnam Laos (H-mong) Bali (Karangasem) N a l F a l S 195 231 112 199 186 293 294 0.251 0.364 0.143 0.405 0.492 0.469 0.677 0.518 0.376 0.625 0.354 0.271 0.003 0.377 0.252 &'A2 &'A3 &'A9 a l C 7 a 2 0.005 0.226 0.009 0.002 0.249 0.232 0.005 0.236 0.234 0.154 0.071 *two tribes (Pima and Lumbees) living; in Arizona and in North Carolina respectively; one tribe from Guatemala, the Ixils; a Quechua-speaking group living in Tarabuco (Bolivia) and a cross-bred Indian (Spanish-Quechua) group from Auchapata (Bolivia); and different groups from Brazil: to the ones already published (Constans and Salzano, 198Ocbthe Gorotire, Kraho, and Gaingang-to samples were added-a Caraibspeaking tribe, the Macushi; and the ICana River Indians, now speaking a Tupi language, Nyengatu, introduced by missionaries (Salzano, 1980). FSTIover all loci is by convention simply the arithmetic mean: The fourth series was made up of a total of 855 samples from both Southeast Asia and South Pacific Islands (Table 5). This series came from the following groups: Principal component analysis The method described by Morrison (1976) was used to represent the distribution of the different populations investigated according to their GclF, Gc", and Gc2 allele frequencies. The two F1 and F2 axes summarized 98% of the total variability of the allele frequencies and represent, respectively, the opposite distribution of the samples according to GclF and Gels and to (GclF, GclS)and G c ~ . The relations used to locate the points on Figure 1are as follows: individuals from Australia, Papua New Guinea (Fore-speaking group), and from Irian Jaya (a Moni-speaking group); individuals from the New Hebrides, Banks, and Torres Islands and from Anuta in the Solomon Islands (the samples from the Solomon, Banks, and Torres Islands, collected in the course of a multidisciplinary survey, were previously described by Blake et al., 1983); people living in Ifalik and Fais Islands and also Chamorros from Guam, Micronesia; two additional samples came from people living in the Marquesas and the Cook Islands, Polynesia. FST analysis The FSTparameter (Wright, 1969) is used to measure the level of differentiation between the geographical groups studied. In the case of the three alleles, we choose to use the formulation proposed by Wright (1978). where g equals the number of loci. Since all variable forms of the DBP have been shown to be the products of codominant alleles a t the same locus, gene counting was used to assess the frequency of the alleles assuming a distribution of these alleles in the samples according to Hardy Weinberg equilibrium. F1 = 0.776 x GclF - 0.60 X Gels - 0.192 x Cic2 - 0.0095 Fs = 0.240 X GclF + 0.563 x Gels - 0.791 x &2 - 0.01581 Correlation coefficients were also calculated between the frequencies of the three alleles. To compare the variation of the Gels and Gc2 allele frequencies in the populations investigated, the frequency of the GclF being considered as constant, the partial correlation coefficient was calculated according to the following relation: 42 43 44 37 38 39 40 41 36 35 33 34 53 54 50 51 52 49 48 47 45 46 0.231 0.178 0.277 0.197 0.330 0.200 0.321 0.448 0.135 0.365 0.474 0.198 &lF 0.637 0.714 0.540 0.426 0.290 0.480 0.464 0.453 0.610 0.510 0.414 0.475 &lS ~ 1 A 2 ~ 1 A 3 &'A4 0.082 ~ 1 A 8 TABLE 4. DBP allele frequencies in Amerindian groups Oceanic Groups Australia: Aborigines Melanesia Papua New Guinea: Fore (Kuru Area) Irian Jaya: Moni New Hebrides Banks Islands Torres Islands Solomon Islands: Anuta Island Micronesia: Western Caroline Islands Ifaluk Island Fais Island Mariana Islands: Guam (Chamoros) Polynesia Marquesas Islands Cook Islands (Mauke moun) Population 253 171 161 155 136 106 165 211 104 108 58 242 N 0.009 0.003 a l A 9 0.443 0.398 0.592 0.538 0.375 39 131 88 166 49 0.288 0.385 0.244 0.379 80 66 141 66 0.311 GC'F 29 N 0.298 0.316 0.214 0.187 0.364 0.386 0.303 0.250 0.364 0.587 GC'S 0.042 0.163 0.030 0.034 &A1 0.034 GlC2 TABLE 5. DBP allele frequencies among Melanesian, Micronesian, and Polynesian samples North America (U.S.) Pima Lumbee Central America (Guatemala): Ixil Guyana (F): South America Palikour Brazil Gorotire Kraho Caingang Iqana River Macushi Bolivia Aymara (Altiplano) Quechua (Tarabuco) Metis (Auchapata) Population 0.011 ~ 2 A 1 0 0.026 ~ l C 7 0.259 0.286 0.194 0.275 0.250 0.284 0.311 0.343 0.227 0.034 GC2 0.002 &ClO 0.123 0.105 0.183 0.351 0.380 0.320 0.215 0.099 0.173 0.125 0.112 0.325 2 112 J.CONSTANS ET AL EAST and NORTH AFRICA SUNDA ISLAND -0.4 -0.3 EUROPE / -0.1 -0: 0.1 0.5 0.2 I 0.b w AXE 1 + y ABORIGINES --I f-0.3 Fig. 1. Eigenveckor diagram for the populations investigated in this study using the two principal axes F1 and F2(refer to Materials and Methods). GEOGRAPHICAL DISTRIBUTION OF THE Gc", Gc2 ALLELES a'*, Student's test was used with a df equal to n - 2. RESULTS AND DISCUSSION The gene frequencies obtained in this study and those already published are assembled in five tables (Tables 1-5). The frequencies for alleles GclF, GclS, and Gc2 are summarized in Figure 1.The different samples are represented in a diagram that reveals three axes and several clusters (Fig. 1).The distribution of the rare mutants is represented in two maps (Figs. 2,3). The results show two main features of the DBP polymorphism: (1)the worldwide presence of the three &IF, and Gc2 alleles, except in one isolated group; (2) the existence of more than 84 rare mutants. Their distribution coincides with historic and prehistoric isolation and human migration patterns. European and african groups (Fig. 1) The French groups are characterized by a Gc2 frequency of 0.20 to 0.30, a GclF frequency of 0.10 to 0.20, and a Gels frequency of 0.55 to 0.60 (Table 1).Similar values are found among Danish (Thymann, 1979), German (Weidinger, 1981), Belgian and English groups (Papiha, 1982a), as well as among Europeans living abroad (Dykes et al., 1981; Nicholls, 1982). The Basques are of a particular interest with the highest Gc2 and the lowest GclF gene frequencies. In general, European values tend to cluster tightly (Fig. 1) with high Gc2 and low GclF frequencies. Populations from the Middle East, such as the Bedouins from North Yemen and the Kurds from Iraq, are characterized by lower Gc2 and markedly higher GclF allele frequencies than are found in Europeans. They constitute a different cluster in which the Druses from Israel (Cleve et al., 1978) and Fig. 2. Distribution of rare Gc mutants among Asiatic and Amerindian groups. Fig. 3. Distribution of the rare mutants among African groups and Black people living in the United States. The Aboriginal populations living in Australia New Guinea, and New Hebrides (Vanuatu) are located on this map as a result of the presence of the GciA1mutant. DBP POLYMORPHISM IN HUMANS the Berbers from Tunisia (Lefranc et al., 1981) could be included. In contrast with other North Africans, the Berbers are genetically related to Middle Eastern Groups. This observation is in agreement with their assumed anthropological origins. The main difference betwen North Africans and Middle Eastern groups is a lower Gc2 and a higher GclS allele frequencies (Table 2). Populations from North Africa are located in the diagram (Fig. 1)in an intermediate position between the two clusters of Europe and Africa. A similar genetic pattern is observed for the Tuareg from Algeria or Mali, the Erythreans, and the two tribes from Djibouti. This occurs in spite of their cultural differences and their geographical locations. In this cluster the influence of the Gc2 allele is so weak that it may disappear under an important founder effect, as in the Kel Kummer Tuareg. African populations are characterized by the highest GclF frequencies observed and a Gc2frequency not very different from the one present among Saharan groups. Africans make up one cluster, but they can be divided into two subgroups-one represented by the populations living in West or Central Africa, and the tribes from the west coast of Malagasy; the second composed of Pygmies and Bushmen, the tribes living on the east coast of Malagasy and also African people living in America (Baltimore).The subgroups differ from one another by their GclF allele frequencies (Table 2). The genetic pattern of the African groups from the United States (Kueppers and Harpel, 1979; Dykes et al., 1983b) show gene frequencies similar to the ones obtained in our study. These additional black American groups are distributed in the two clusters. Different anthropological investigations have suggested the presence of approximately 20% European genes among the American Black populations (see summary in Bodmer and Cavalli-Sforza, 1976) but this is not evident from our data on DBP polymorphism. Asian and Amerindian groups A very scattered pattern corresponds to the distribution of these groups despite common geographical locations or anthropological origins. The most original position is occupied by the Karangasem community from Bali: with their high GclF and low Gc2. allele frequencies, they look like an African group in the diagram (Fig. 1).Our observation is 115 confirmed by the data published by Tan et al. (1981) on Bidayuk and Ibans from Malaya, or on Indonesians. South Vietnamese, Thais, and Tibetans present very similar allele frequencies to those observed among Chinese from Hong Kong and Taiwan or among Japanese groups (Matsumoto, 1980; Kwok, 1981; Shibata 1983).This group is characterized by a high Gc2 frequency (Table 3). The Jirel tribe from Nepal (Fig. 11, is located in an intermediate position between the above small Asiatic group and the Tamils Pondicherry). Newars from Nepal are very similar to the Jirel Sherpa (Yuasa et al., 1983). Papiha et al. (1985) present a very large set of data on different Indian groups. They confirm the Gc frequencies obtained among the Tamils and the location of the Indians in a cluster much like the European one. The other samples belonging to South Asian regions are widely distributed in the diagram. Nevertheless, it is possible to delineate three clusters. One is made of the two Polynesian samples (Cook and Marquesas Islands) and of the Chamorros from Guam overlaps the cluster of the North Asiatic groups. The location of the Anuata sample (Fig. 1) points to a genetic similarity with the Polynesians. This observation would confirm the hypothesis proposed by Blake et al. (1983) concerning the origins of human settlements in this island. According to Bayard (1976), linguistic affinities are observed between Anuata and Tonga, which belongs to the Polynesian cultural influence. The possible presence of genetic Melanesian origins on Anuata should also be considered (Blake et al., 1983). In general the DBP polymorphisms of the Polynesians and North Asians do not differ. The two Micronesian samples show a higher GclF allele frequency, which makes them more similar to the Sunda Islands groups. The last groups studied belonging to these South Asian regions comprise the small Aboriginal populations living in Melanesia, New Hebrides, and in Australia. Kambok and Kirk (personal communication)obtained similar DBP allele frequencies for groups belonging to this geographical area. It is interesting to notice that the Melanesians in the Bank, Torres, and Solomon Islands are located in a cluster clearly distinct from the others made of Micronesian and Polynesian 116 J. CONSTANS ET AL TABLE 6. The FST values calculated from the GcfF,Gc", and Gc2 allele frequencies for eight main geographical areas FST Geographical zones I I1 I11 Europe North Asia South Africa India South Asia North Africa Australia and New Guinea South America 1F 1s 2 Nonweighted mean value ,023 ,007 ,045 ,036 ,048 ,040 ,013 ,044 ,012 ,010 ,001 ,024 ,017 ,029 ,046 ,066 ,061 0.015 0.005 0.036 0.024 0.031 0.045 0.045 0.054 samples (Fig.1). From our data it is not possible to delineate a clear difference between groups speaking Austronesian languages and the ones belonging to the Papuan linguistic zone (Tyron, 1979). It seems that the great heterogeneity of the allele frequencies among the South Asian populations is, perhaps, due to different anthropological origins but also to significant genetic differentiation among those following ancient migrations and experiencing extreme isolation over long periods of time. The Amerindian populations can be gathered into two clusters in the diagram (Fig. 1). One is composed of Bolivian Indians. Palikours from Guyana are not very far from this cluster. The second cluster is composed of Gorotire, Kraho, and Caingang from Brazil and also Lumbees from the United States. Amerindian groups of this last cluster are characterized by high Gc2 frequencies, while the Andeans present the highest GclS frequencies. Iqana Indians are located at the mean distance between the Brazilian and the Bolivian groups. It seems that a third group of Amerindians could be considered, made of Pima from the United States (Dykes et al., 1983), Macushi from Brazil, and Ixils from Guatemala. Other Indian tribes such as Dogribs and Athapaskan-speaking groups (Szathmary et al., 1983) may be related to this third Amerindian cluster. These data show very different genetic patterns for Asians and Amerindians, a finding consistent with other genetic systems. (For a review, refer to Bodmer and Cavalli-Sforza, 1976; Mourant, 197621). One possible explanation for the differences observed among Amerindian groups is a founder effect among the first migrants to the New World as a ,008 ,038 ,017 .015 .049 ,056 .058 result of isolation within a vast geographical region during a relatively short period of time (MacNeish 1977; Dumond, 1980). Besides, the Amerindians may have different Asiatic origins. If we consider the Gc2 frequencies (0.10 to 0.20 and about 0.30 to 0.40) at least two separate populations may have participated in the settlements of America. Among them the Eskimos from Alaska, regarded as latecomers, are, according to the DBP data (Matsumoto et al., 19801, very similar to Andeans. Genetic Heterogeneity of Gcl< Gc", G 2 Alleles Distributions The Gc polymorphism among human populations shows that the range of variation of the gene frequencies differs with the populations. This variability can be estimated by the FSTvalues determined for each allele in eight main geographical areas (Table 6). The lowest nonweighted mean values for FSTare observed among European (0.015) and North Asian groups (0.005). The genetic differentiation between the three main alleles of the DBP appears to be very small among these populations. This may be due to frequent interpopulation migrations within the concerned geographical zones over a large period of their past history giving rise to a process of homogenization. A second group consists of populations living in India, South Africa, and in the Soutern part of Asia. In these regions, the FST values range from 0.024 to 0.036. The third group (Table 6) is characerized by the highest values (about 0.050) of the nonweighted FST parameter. This result may be explained by genetic drift in groups that share a similar pattern of social and cultural isolation. A second reason may be a great variation of DBP POLYMORPHISM IN HUMANS TABLE 7. The correlation coe rcients calculated with the frequencies of the three Gc", and Gr? alleles in the different human groups ,ff. Correlation coefficient between GclF and Gc" GclF and Gc2 Gc" and Gc2 The data obtained in this study - + 0.847* 0.577* 0.083 N S Our data and including those summarized by Papiha et al. (19851 0.905* 0.649* + 0.291* - *p < 0.01. population sizes in their past history (bottleneck effect). Populations from Australia, New Guinea, and groups living in South Asia show very different FST values. These differences can only be the result of factors such as selection or genetic drift. The present anthropological and historical knowledge of those populations supports the hypothesis of genetic drift. But the influence of factors related to selection is the most likely explanation, especially if we consider the existence of strong geographical associations owing to the presence of regular clinal variations of the DBP frequencies. From Europe to Africa and from North to South Asia the G c ' ~ allele frequency decreases. These observations are corroborated by the values of the correlation coefficients calculated between the three DBP alleles (Table 7). GclF and GclS frequencies are negatively correlated as well as GclF and Gcz frequencies, but on the contrary, GclS and Gc2 frequencies are positively correlated when additional data (Papiha et al., 1985) are included in the comparison. The partial correlation coefficient calculated between GclS and Gc2 when GclF is maintained constant is -0.915, which is highly significant (p < 0.01) for a negative variation between the frequencies of these two alleles. From these comparisons we can consider the distribution of the DBP polymorphism as corresponding to the presence of a balanced equilibrium: a n increased frequency of the GclF allele is followed by a decrease in the frequency of the GclS and Gc2 alleles. These data confirm that the gradient of the DBP allele frequencies corresponds to the existence of clinal variations in relation to geographical parameters. To explain the distribution of the Gc alleles, one has to consider the influence of selective fac- 117 tors associated with the biological activity of the proteins produced by the three main alleles. It then remains to show in which way the three DBP proteins may be involved in different biological activities. It has been shown by Constans (1978b) that the serum DBP level is higher among Gc' carriers and that the GclS allele is associated with higher levels than the GclF allele. In addition, there is, probably, a significant difference of affinity for the vitamin D metabolites between the GclF, Gc", and Gc2 proteins (Constans et al., 1980b). Recently it was demonstrated by Daiger et al., 1984 that the serum DBP levels were under the control of the three main alleles, a polygenic component, and environmental effects. Different comparisons have been made between the distribution of the Gc allele frequencies and the influence of geographical factors with respect to skin pigmentation (Mourant et al., 1976b)and latitudes (Walter, 1969). In this study a n examination of the two geographical North-South clines corresponding to the GclF and the Gc2 gene frequencies demonstrates a clear convergence between skin pigmentation (Loomis, 1967)in the populations concerned and the DBP alleles, with, however, two exceptions: the Asiatic Indians and the Amerindian groups. Moreover, a recent investigation of the first hereditary DBP deficiency failed to reveal any biological or pathological abnormality in the family studied (Vavrusa, 1983). The above observation, therefore, shows that the most significant biological activity of the protein is yet to be elucidated in order to understand in which way selection may act differently in humans according to the presence of the GclF, Gc", and Gc2 proteins. An unexplored pathway for this biological activity may be found in the DBP binding to the lymphocyte and in the involvement of the DBP in the immune defence, since recent findings show that the most active vitamin D metabolite (1.25 - (OHIz-DS)plays a part in the regulating mechanisms of the cell division and differentiation (Shavit et al., 1983; Frampton et al., 1983). WORLDWIDE DISTRIBUTION OF RARE MUTANTS Eighty-four rare mutants of the DBP have now been detected. In contrast to the three frequent alleles, the distribution of the rare mutants is not strictly limited to geographical zones but is more a reflection of the an- 118 J. CONSTANS ET AL. thropological relationship between groups, their exchanges, and their migrations. Anthropological relationship between Asian groups and Amerindian tribes As represented in Fig. 2 the following rare mutants G~1A2,1A3 &1A4,1 A8,1A9 , and Gc1c2,1c3, are obse&ed in the Northern Asian groups of Japan, Korea, and Taiwan (Ishimoto et al., 1979; Omoto and Miyake, 1979; Matsumoto et al., 1980) and also in some Southeast Asian regions of Borneo, Java, Indonesia, and Thailand (Matsumoto et al., 19801, where Chinese communities are numerous. The same mutants, except for the Gclc2 and GclC7alleles, were detected among Eskimos from Alaska and Greenland. The presence of the GclA4mutant was confirmed in a recent study of Eskimos from Alaska (Dykes et al., 1983).Among the South American Indians only the Gc1A3, GclA9, and GclC7 alleles appear to be present, but no rare Gc mutant was found in the two samples from North American Indians. This observation is now confirmed by recent data on South and North Amerindians (Dykes et al., 1983). Three significant features emerge from these results: The presence of the same mutant, GclC7, among a Dogpa group living in the Tibetan mountains and a Gorotire Indian tribe from Brazil (Constans and Salzano, 1980~) is probably the first example of a possible anthropological relationship between a Southern Mongoloi'd group and a n Amerindian tribe. The G d A 3 , 1 A 8 and GcIA9mutants are found not only among Aymara and Quechua Indians from Peru and Bolivia but also among Palikours, (an Arawak group) from French Guyana. They are the first genetic mutants found in groups that are culturally different. It is possible that a t the time of the first settlements in South America, these rare mutants and others were present in a population located on the Eastern slopes of the Andean zone, near the Amazon River. From this area individuals formed into smaller groups that moved northward to the Orinoco swamps. These groups could have included individuals with rare mutants and interbred with ancient and already settled Caribbean tribes. The GclA2*lA4mutants, detected only among Eskimos from North America, may confirm that these groups belonged either to the last major migrations coming from Asia before the Bering bridge was submerged, or to Asian populations (North Mongoloids) of a genetic origin other than that of the American Indians (Dumond, 1980). The GclAl0 mutant observed for the first time among the Chippewa Indians (Cleve et al., 1963)has not been observed in any other Amerindian group. This rare mutant might be called a "private" variant according to the hypothesis developed by Nee1 (1973). It can be deduced from the distribution of these rare mutants that there was probably considerable differentiation among the Asian populations when the Americas were occupied. This would also explain why the rare protein mutants such as those discussed here are more commonly distributed among North Asian groups and South American Indians than they are among the South Asian groups (Thailand, Indonesia, and Polynesia). One can also expect to find new DBP mutants with a very limited geographical distribution in Melanesian and Micronesian groups which are anthropologically different from the North Asian populations. A very rare example of mutant distribution is illustrated by GclA12and GclC1', because they are found in Indonesia (Bali) and India (Papiha, personal communication). Their presence in these two groups is in agreement with the tales and mythologies of the Balinese group, which indicate a n Indian origin (Breguet et al., 1982b). That these mutants are also present in Europe constitutes a n interesting genetic evidence of the Indo-European migrations. Diffusion of the African rare mutants The second map (Fig. 3) is essentially devoted to the populations that originated from Africa or are still living there. The GclC3mutant is detected among North African groups and in Europe. The same mutant is also present among the Erythreans, but so far it has not been observed among black African populations. The G c is ~an- ~ other mutant with a n approximately similar geographical distribution and is found in Mali, Senegal, Gambia, and Ethiopia. It is difficult to ascertain precisely where these two mutants originated from, but they seem to be absent from Central and South Africa. They were probably scattered throughout the Mediterranean region by population exchanges and in North Africa, from east to west, by trade and slave caravans. Anthropological and cultural relationships between Ethiopians, Somalians, Malagasy ~ DBP POLYMORPHISM IN HUMANS 119 cording to Schanfeld et al. (1980), the Gml;17,-5,10,11,13,14 when associated with the Gm23antigen is characteristic of the African populations. The same haplotype without the Gm23 antigen was found by the same author among Indian and Aboriginal populations. With the transferrin system, the TfD1mutant is known to be present in all African groups but also in Southeast Asia (Bali, Aboriginal groups from Australia a n Melanesia).The partial structure of this protein was determined (Wang et al., 19671, and no peptide difference was found between the proteins obtained from a n African and from a n Australian sample. In red cell enzymes, rare mutants such as PGM: and PGMZ have been found in India, Australia, and recently in a Balinese group (Tenganam) and they are also among Pygmies (Vergnes et al., 1979). Electrophoretic comparisons have so far failed to show any difference between the proteins of mutants from the different ethnic groups (Blake and Omoto, 1975; Kirk et al., 1977; SantachiaraBenerecetti et al., 1980; Breguet 1982131. Owing to the simultaneous presence of these rare mutants in so many different groups, a multifocus hypothesis seems very unlikely. According to prehistoric and historic data, the hypothesis of a unique focus can be supported if we consider a migration A common ancestor for African and in the remote past of a common ancestor for Aboriginal groups the African and Aboriginal groups. This Since 1967, several authors have described event had to occur some 1or 1.5 million years a very high frequency of the GclA1 allele ago during the existence of the first homiamong African and Australian Aboriginal nids. Later, a local and independent evolupopulations (Cleve et al., 1967; Gajdusek and tion in Africa and Asia would give rise to Alpers, 1972). As samples from such popula- distinct groups such as the Pygmies, Bushtions became available, the mobility of the men, and Hottentots in Africa, and the AboGclA1proteins was compared using three dif- rigines and oceanic populations in Asia. ferent electrophoretic systems, namely, polyacrylamide gel, isoelectric focusing, and CONCLUSIONS isoelectric focusing in presence of 3 M urea The DBP polymorphism in humans pre(Constans et al., 1983).All the GclA1proteins obtained from the different geographical sents two patterns, one related to the distriplaces show a n identical electrophoretic mo- bution of the rare mutants and the second to bility. It can be assumed that the structure the occurence of three main alleles GclF, of the GclA1protein is the same in the Afri- Gc", and Gc2. The worldwide distribution of the rare alcan, Malagasy, Australian, New Guinean, leles is in agreement with the prehistoric and and Balinese samples. In addition, other examples of similar rare historic data on the migrations and intermutants in the different geographical groups breeding of human populations. A much better understanding of that distribution can be can be revised. In the Gm system, the Gm1,17,23,5,10,11,13,14 expected when the populations of India, the haplotype was detected in the community of Middle East, and China have been thorTenganam (Bali) (Breguet et al., 1982b). Ac- oughly sampled. (Lowlands) and Djibouti tribes are supported by the presence of the GclA6mutant in those groups. G c lC lO is another rare mutant restricted to populations living on the west coast of Africa. It was detected in Bantu and related groups from Zaire, the Ivory Coast, and Cameroon. Its presence among the Lumbee sample would correspond to a Negro'id admixture among them as indicated by human leukocyte antigen (HLA) typing (Grier et al., 1979). GclA1 was first described by MacDermid and Cleve (1972) under a "GcAb" (Aborigine) denomination. Since then has been found with the highest frequency in African groups living in tropical areas of Central Africa. Its frequency falls rapidly as one moves north, but it is still frequent in the south, among the Bushmen. The present authors have also observed the same GclA1and GcZA3 mutants among tribes living on the east and west coasts of Malagasy. In three U S . groups with African ancestry living in Baltimore, Philadelphia, and Georgia (Kueppers and Harpel, 1979), the GclA1, GcZA3,GcZA5, were observed. These data on the Gc mutants are in agreement with the historical and anthropological knowledge about the various origins of the Africans brought to America. 120 J. CONS? 'ANS ET AL donesia. 11. Haemoglobin types and red cell isozymes. Anthropological microdifferentiations withHum. Hered. 32:308-317. in the large populations or between villages H, Kirk, RL, Parker, WC, Bearn, AG, Schacht, in the same linguistic or cultural area could Cleve, LE, Kleinman, H, and Horsfall, WR (1963) Two genetic be investigated by comparing the distribuvariants of the group-specific component of human tions of these rare mutants. The polymorphserum: Gc Chippewa and Gc aborigine. Am. J. Hum. Genet. 25:368-379. ism of the DBP may also contribute to accurate genetic control of the pedigrees col- Cleve, H, Kirk, RL, Gajdusek, DC, and Guiart, J (1967) On the distribution of the Gc variant Gc aborigine in lected in the course of anthropological surMelanesian populations; Determination of Gc-types in veys. sera from Tongariki island, New Hebrides. Acta Genet. Base1 17:511-517. If the rare mutants show a limited geographical distribution or belong to related Cleve, H, Patutschnick, W, Nevo, S, and Wendt,GG (1978) Genetic studies of the Gc subtypes. Hum. Genet. groups, the three main alleles are observed 44: 117-122, in every human population studied. The Constans, J, and Viau, M (1977) Group-specific compoemerging pattern obtained from the distrinent: Evidence for two subtypes of the Gc' gene. Science 198:1070-107 1. bution of their gene frequencies shows the stability of these frequencies over large geo- Constans, J, and Viau, M (1978a) Study by isoelectric focusing of the serum protein which transports vitagraphical areas. A cline associated with inmin D (DBP) and of the polymorphism of Gc. Imporcreasing GclF and decreasing Gc2 gene tance in Anthropology. Cr. Acad. Sci. Paris 287,21:809frequencies is, without doubt, present be812. tween the northern and southern regions. Constans, J , Viau, M, and Ruffie, J (1978b) Study of the Gc protein in several French population samples: geThe present-day knowledge of the biological netic polymorphism by isoelectric focusing and quanactivity of DBP is still too limited for the titative results. Cr. Acad. Sci. (Paris) 287,22:1003-1006. hypothesis of a selective pressure based on Constans, J, and Cleve, H (1979) Group specific compothis activity to be definitely established. nent. Report on the first international workshop. Hum. ACKNOWLE,DGMENTS We are greatly indebted to Professor A.E. Mourant for his interest in this work and to Dr. Kawai (Osaka Medical School) for his active collaboration. Pr. L.Y.C. Lai is particularly acknowledged for his comments and fruitful discussions during the preparation of this paper. LITERATURE CITED Aquadro, CF, and Avise, J C (1981) Genetic divergence between rodent soecies asaessed bv usine two dimensional electrophoksis. Proc. N a i . AcG. Sci. USA 78:3784-3788. Bellwood, PS (1980) The peopling of Pacific. Sci. Am. J. 5t138-147. Blake, NM, and Omoto, K (1975) Phosphoglucomutase types in the Asian Pacific area: A critical review including new phenotypes. Ann. Hum. Genet. Lond. 38251-273. Blake, NM, Hawkins, BR, Kirk, RL, Bhatia, K, Brown, P, Garruto, RM, and Gajdusek, DC (1983) A population genetic study of the Banks and Torres islands (Vanuatu) and of the Santa Cruz islands and Polynesian outliers (Solomon islands). Am. J. Phys. Anthropol. 62343-362. Bodmer, WF, and Cavalli-Si'orza, LL (1976) In Genetics, Evolution and Man. W.H. Freeman and Company (Ed.) pp. 381-408. Breguet, G , Ney, R, Grimm, W, Hope, SL, Kirk, RL, Blake, NM, Narendra, IB, and Toha, A (1982a) Genetic survey of an isolated community in Bali, Indonesia. I. Blood groups, serum proteins and hepatitis B serology. Hum. Hered. 3252-61. Breguet, G , Ney, R, Kirk, RL, and Blake, NM (1982b) Genetic survey of an isolated community in Bali, In- Genet. 48:143-149. Constans, J, Lefevre-Witier, Ph, Richard, P, and Jaeger, G (1980a) Gc (vitamin D-Binding protein) subtype polymorphism and variants distribution among Saharan, Middle-East and African populations. Am. d. Phys. Anthropol. 52:435-441. Constans, J, Viau, M, and Bouissou, C (1980b) Affinity differences for the 25-OH-D3 associated with the genetic heterogeneity of the vitamin D binding protein. FEBS Lett. 11:107-111. Constans, J, and Salzano, F (1980~)Gc and transferrin isoelectrofocusing subtypes among Brazilian indians. J. Hum. Evol. 9:489-494. Constans, J, Viau, M, Jaeger, G , and Palisson, M J (1981) Gc, Tf, Hp subtype and a1 antitrypsin polymorphisms in the Pygmy Bi-Aka sample. Phenotype association between TfDl and GcAb(Gcl*l) variants. Hum. Hered. 31;129-137. Constans, J, and al. (1983) The polymorphism of the vitamin D-binding protein (Gc). Isoelectric focusing in 3 M urea a s additional method for identification of genetic variants. Hum. Genet.&: 176-180. Cox, DW (1981) New variant of 011 antitrypsin: A comparison of Pi typing techniques. Am. J. Hum. Genet. 33:354-365. Daiger, SP, Schanfield, MS, Cavalli-Sforza, LL (1975) Human group specific component (Gc) protein hinds vitamin D and 25 hydroxy vitamin D. Proc. Natl. Acad. Sci. USA 122076-2080. Daiger, SP, Miller, M, and Chakraborty, R (1984) Heritability of quantitative variation a t the Group-Specific Component (Gc)locus. Am. J. Hum. Gent. 363634574, Davie, MWJ, Lawson, DEM, Emberson, C, Barnes, JLC, Roberts, GE, and Barnes, ND (1982) Vitamin D from skin: Contribution to vitamin D status compared with oral vitamin D in normal and anticonvulsant treated subjects. Clin. Sci. 63t461-472. Dumond, DE (1980) The archeology of Alaska and the peopling of America. Science 209:984-991. DBP POLYMORPHISM IN HUMANS 121 the alleles of the Gc system of plasma proteins. Hum. Dykes, D, Polesky, H, and Cox, E (1981) Isoelectric focusing of Gc (vitamin D-binding globulin) in parentage Genet. 33:307-314. testing. Hum. Genet. 58:174-175. Neel, JV (1973) “Private” genetic variants and the freDykes, D, Copouls, B, and Polesky, H. (1983a) Descripquency of mutation among South American Indians. tion of six new Gc variants. Hum. Genet. 63:35-37. Proc. Natl. Acad. Sci. USA 70:3311-3315. Dykes, DD, Crawford, H, and Polesky, HF (1983b) Popu- Nicholls, C, and Mulley, J C (1982)Distribution of the Gc lation distribution in North and Central America of (group specific component) subtypes in cordbloods and PGMl and Gc subtypes as determined by isoelectric blood donors. Aust. J. Exp. Biol. Med. Sci. 60:427-431. focusing (IEF). Am. J. Phys. Anthropol. 62:137-145. Omoto, K, and Miyake, K (1979) The distribution of Frampton, RJ, Omond, SA, and Eisman, J A (1983)Inhisubtypes of serum Gc globulin in Japanese and the neighbouring populations. Jpn. J. Hum. Genet. 24:224bition of human cancer cell growth by 1,25-dihydroxy vitamin D3 metabolites. Cancer Res. 43:4443-4447. 225. Gajdusek, DC, and Alpers, MP (1972) Genetic studies in Omoto, K, and Misawa, S (1984) The distribution of Gc relation to kuru. I - Culture, historical and demosubtypes in the Philippines. Jpn. J. Hum. Genet. 29:223-224. graphic background. Am. J. Hum. Genet. 24,suppL:Sl38. Papiha, SS, Seyedna, Y, and Sunderland, E (1982a)Phosphoglucomutase (PGM) and Group-specific component Giirg, A, Postel, W, Weser, J, Weidinger, S, Patutschnick, (Gc)isoelectric focusing among Zoroastrians of Iran. W, and Cleve, H (1983) Isoelectic focusing in immobiAnn. Hum. Biol. 9:571-574. lized pH gradients for the determination of the genetic Pi (q-antitrypsin)variants. Electrophoresis4: 153-157. Papiha, SS, Roberts, DF, White, I, Chahal, SMS, and Asefi, JA (1982b) Population genetics of the GroupGrier, JO, Ruderman, FJ, and Johnson, AH (1979) HLA specific component (Gc) and phosphoglucomutase profile in the Lumbee Indians of North Carolina in (PGM) studied by isoelectric focusing. Am. J. Phys. transplantation proceedings 11,4:1767-1769. Anthropol. 59:l-7. Hirschfeld, J, Jonsson, B, and Rasmuson, M (1960)Inheritance of a new group specific system demonstrated in Papiha, SS, White, I, and Roberts, DF (1983) Some genetic implications of isoelectric focusing of human red normal human sera by means of an immunoelectrocell phosphoglucomutase (PGM1) and serum protein. phoretic technique. Nature 185:931-932. Group specific component (Gc):Genetic diversity in the Ishimoto, G, Kuwata, M, and Nakajima, H (1979) Grouppopulations of Himachal Pradesh, India. Hum. Genet. specific component (Gc)Polymorphism in Japanese: An 63:67-72. analysis by isoelectric focusing on polyacrylamide gels. Papiha, SS, Constans, J, White, I, and McGregor, IAM Jpn. J. Hum. Genet. 24.75-83. (1985) Group specific component (Gc)subtypes in GamJonhson, AM, Cleve, H, and Alper C (1975) Variants of bian and Transkeian populations: A description of a the group-specific component system as demonstrated new variant. Am. Hum. Biol. 1217-26. by immunofixation electrophoresis. Report of a new variant, Gc Boston (GcB). Am. J. Hum. Genet. 27:728Radola, BJ (1980) (Ed.) Electrophoresis ’79. W. de Gruy736. ter Publisher. Kirk, RL et al. (1977) Genes and people in the Caspian Righetti, PG (1979)(Ed.) Progress in isoelectric focusing littoral: A population genetic study in Northern Iran. and isotachophoresis. North Holland American ElseAm. J. Phys. Anthropol. 46:377-390. vier. Kueppers, F, and Harpel, B (1979)Group specific compo- Salzano, FM (1980) New studies on the color vision of nent (Gc)“subtype” of Gc’ by isoelectric focusing in Brazilian Indians. Rev. Brasil Genet. 11&3:317-327. US Blacks and Whites. Hum. Hered. 29:242-249. Santachiara-Benerecetti,AS et al. (1980) Population geKwok, KYY, and Lewis, WHP (1981) Group specific comnetics of red cell enzymes in Pygmies: A conclusive ponent (Gc) subtypes in the Chinese population of Hong account. Am. J. .Hum. Genet. 32:934-954. Kong. Hum. Genet. 59:72-74. Schanfield, MS, Alexeyeva, TE, and Crawford, MH (1980) Studies on the immunoglobulin allotypes of Asiatic Lefranc, MP, Chibani, J, Helal, AN, Boukef, R, Seger, J, populations. Hum. Hered. 343-349. and Lefranc, G (1981) Human transferrin (TO and Group specific component (Gc)subtypes in Tunisia. Shavit, ZB, Teitelbaum, SL, Reitsmer, P, Hall, A, Pegg, Hum. Genet. 59:60-63. LE, Trial, J, and Kahn, AJ (1983) Induction of monoLoomis, WF (1967) Skin pigment regulation of vitamin cytic differentiation and bone resorption by 1,25-dehyD biosynthesis in man. Sci. N.Y. 157:501-506. droxyvitamin Ds. Proc. Natl. Acad. Sci. US4 80:59075911. Mac Dermid, EM, and Cleve, H (1972) A comparison of the fast migrating Gc-variant of Australian aborig- Shibata, K (1983) Haptoglobin, group-specific compoines, New Guinean indigenes, South African Bantu nent, transferrin and antitrypsin subtype8 and new and Black Americans. Hum. Hered. 22:249-253. variants in Japanese. Jpn. J. Hum. Genet. 28, 1:17-28. Mac Neish, RS (1977) Les premiers Americains. La Re- Szathmary, EDE, Ferrell, RE, and Gershowitz H (1983) cherche 78:444-452. Genetic differentiation in Dogrib Indians: Serum proMatsumoto, H, Matsui, K, Ishida, N, Ohkura, K, and tein and erythrocyte enzyme variation. Am. J. Phys. Teng, YS (1980)The distribution of Gc subtypes among Anthropol. 62:249-254. the Mongoloid populations. Am. J. Phys. Anthropol. Tan, SG, Gan, YY, Asuan, K, Abdullah, F (1981) Gc 53,4:505-508. subtyping in Malaysians and in Indonesians from Morrison, DF (1976) Multivariate Statistical Methods. North Sumatra. Hum. Genet. 59/75-76, 2nd Edition, Mc Graw Hill, Chapter 8. Thorne, AG, and Wolpoff, M (1981) Regional continuity in Australian Pleistocene Hominid evolution. Am. J. Mourant, AE, Kopec, AC, Domaniewska-Sobczak, R Phys. Anthropol. 55r337-349. (1976a) Distribution of the Human Blood Groups and Other Polymorphisms. Oxford Press. Thomas, H (1978) L’Asie temoin de l’origine des Hominid&. La Recherche 9:916-919. Mourant, AE, Tills, D, and Domaniewska-Sobczak, R (1976b) Sunshine and the geographical distribution of Thymann, M (1979) Gc subtype determination by isoelec- 122 J. CONSTANS ET AL. tric focusing in agarose gels. In: Proceedings of the 8th Int. Congr. SOC. Forens Heniogenetics. Long, pp. 437448. Tryon, DT (1979)Remarks on the language situation in the Solomon islands. In SA Wurm (ed): New Guinea and Neighboring Areas: A Socio Linguistic Laboratory. Contributions to the Sociology of Language, Vol. 24, The Hague, Mouton, pp. 32-51. Vavrusa, B, Cleve, H, and Constans, J (1983) A deficiency mutant of the Gc system. Hum. Genet. 65:102107. Vergnes, H, Sevin, A, Sevin, J, and Jaeger, G (1979) Population genetic studies of the Bi-Aka Pygmies (Central Africa). A survey of red cell and serum enzymes. Hum. Genet. 48:343--355. Walter, H, and Steegmuller, H (1969) Studies on the geographical and racial dist,ribution of the Hp and Gc polymorphisms. Hum. Hered. 19:209-221. Wang-an Chuan, Sutton, HE, and Scott, ID (1967)Transferin D1 identity in Australian aborigenes and Ameri- can Negroes. Science 156t936-940. Weidinger, S, Cleve, H, Schwarzfischer, F, and Patutschnick, W (1981) The Gc system in paternity examinations. Application of Gc subtyping by isoelectric focusing in Biomathematical evidence of paternity. K. Hummel and J. Gerchow (Eds). Springer Verlag. pp. 113-121. White, JP, and O'Connell, JF (1979) Australian prehistory: New aspects of antiquity. Science 203:21-28. White, JP, and Allen J (1980) Melanesian prehistory: Some recent advances. Science 207:728-734. Wright, S (1969) Evolution and the genetics of population. Vol. 2, Chicago, University of Chicago. Wright, S (1978) Evolution and the genetics of populations. Vol. 4, Chicago, University of Chicago. Yuasa, I, Saneshige, J, Okamoto, N, Ikawa, S, Hikita, T, Ikebuchi, J, Inoue, T, and Okada, K (1983)Distribution of Hp, Tf, Gc and Pi polymorphisms in a Nepalese population. Hum. Hered. 33:302-306.