Marius J. Coetzee*, Sharon C. Bartleet, Michele Ramsay, and Trefor Jenkins. MRC Human ..... Luzzato L, Allen NC (1968) Relatioship between the genes for.
Hum Genet (1992) 89 : 111-113
9 Springer-Verlag1992
Glucose-6-phosphate dehydrogenase (G6PD) electrophoretic variants and the PvuII polymorphism in Southern African populations Marius J. Coetzee*, Sharon C. Bartleet, Michele Ramsay, and Trefor Jenkins MRC Human Ecogenetics Research Unit, Department of Human Genetics, School of Pathology, South African Institute for Medical Research and University of the Witwatersrand, Johannesburg, South Africa Received September 15, 1991 / Revised January 13, 1992
Summary. Southern African Bantu-speaking negroid and San populations were examined with regard to the glucose-6-phosphate dehydrogenase (G6PD) P v u I I restriction fragment length polymorphism (RFLP) showing alleles of 4 k b and 1.6 kb, called Type 1 and Type 2, respectively. The standardized disequilibrium coefficient for the electrophoretic G 6 P D types and P v u l I alleles in the Southern African population was 0.28. The molecular lesion causing the G d A mutation is the same in the San and Southern African negroid populations. G d A c h r o m o s o m e s are found in association with both the Type 1 and Type 2 alleles, whereas none of the 62 Gd B c h r o m o s o m e s from the Southern African populations had the Type 2 allele. Five of the 44 G d B chromosomes studied in the A m e r i c a n Black population had the Type 2 allele, indicating that the Gd B allele in the two populations may have different origins. The presence of all 3 A - deficiency mutations in the G 6 P D A gene, in a region where the ancestral population was thought to have predominantly G 6 P D B, m a y be explained by their origin in Africa after the divergence of the races.
Introduction Over 300 glucose-6-phosphate dehydrogenase (G6PD) variants have been described (Beutler 1989). In populations of African origin, the alleles for the G 6 P D B, A and A - enzyme forms reach polymorphic frequencies, G d B being the commonest. The G 6 P D A variant occurs in 20%, and the G 6 P D A - in 11% of A m e r i c a n Black males (Beutler 1989). They both have reduced activity when c o m p a r e d with G 6 P D B, namely 85% and 1 0 % 15%, respectively. Female G d A and G d B heterozygotes develop lower parasitaemia levels when exposed to Plasm o d i u m f a l c i p a r u m than do other individuals and enjoy a selective advantage in malaria areas (Luzzato 1985). * Present address: Department of Haematology, University of the Orange Free State, Bloemfontein, South Africa Offprint requests to: T. Jenkins, Department of Human Genetics, SAIMR, P. O. Box 1038, Johannesburg 2000, South Africa
This enzyme and its alleles offer some insight into the history of man in Africa, where few written records exist. The Gd A allele differs from the G d B allele by a single base transition that occurs at nucleotide 376 (Takizawa et al. 1987) and that creates an additional F o k I site; Gd Ais caused by an additional mutation that, in the majority of cases, occurs at nucleotide 202. Yoshida et al. (1988) have reported a P v u I I restriction fragment length polymorphism (RFLP) that is polymorphic in Negroids, but not in Caucasoids. The 4 kb P v u I I fragment was designated Type 1 and the 1.6 kb P v u I I fragment Type 2. The Type 2 allele has been shown to be in linkage disequilibrium with the electrophoretic G 6 P D A variant (Beutler and Kuhl 1990; Fey et al. 1990). The aim of the present study was to determine the relationship between the P v u I I R F L P and the electrophoretic G 6 P D variants in Southern African populations.
Subjects and methods After informed consent was obtained, a total of 276 random blood samples were collected from 16 San males (from Tsumkwe, Namibia) and 260 Southern African Black males from Johannesburg. Of the latter, 77 were from Bantu-speaking Blacks presenting for paternity tests at our laboratory, 50 were volunteer staff members, and 133 were donors from the Highveld Blood Transfusion Service. DNA was extracted from 102 randomly selected samples but, from the remaining 174 samples, DNA was extracted from only the 25 G6PD A individuals identified by starch gel electrophoresis. In addition, DNA was extracted from 39 Black female individuals (who were not typed for G6PD by electrophoresis) in order to increase the sample size for the calculation of the frequency of the PvuII alleles. Only the results from the random male subjects were used for calculating the frequencies of Gd B, G d A a n d G d A- and for examining their linkage disequilibrium with the two PvulI alleles. Females were excluded because G d A / G d A homozygotes cannot be distinguished electrophoretically from G d A / G d A- heterozygotes. Conventional spectrophotometric and starch gel electrophoretic methods were used to investigate the activity and mobility of the G6PD enzyme (Betke et al. 1967). DNA samples were digested with either PvuII or FokI, and the fragments were subjected to electrophoresis in agarose. They were then transferred onto nylon membranes according to the method of Southern (1975). The genomic G6PD probe (Takizawa et al. 1987) was kindly provided by Dr.Akira Yoshida, City of Hope Research Institute. A 1.6kb
112 BglII/SphI fragment was used for detecting the RFLP; the 0.45 kb Fok! fragment was used for detecting the FokI alleles. Both fragments were radio-labelled with the random primed DNA synthesis method (Feinberg and Vogelstein 1983). The association between the GdB, Gda and Gda alleles and the PvuII alleles was assessed by calculating the standardized linkage disequilibrium coefficient (Hill and Robertson 1986).
Table 1. Relationship between G6PD phenotypes and PvuII al-
Results and discussion
a Only males used in order to avoid incorrect assignment of female heterozygotes b Yoshida et al. 1988
The G6PD electrophoretic variants showed the same allele distribution as had been previously reported in Southern African Blacks (Nurse et al. 1985). The frequencies of the Gd A and Gd A alleles in our sample of Southern African Bantu-speaking Blacks (0.15 and 0.02, respectively) are lower than in American Blacks (0.20 and 0.11, respectively) (Beutler 1989), although the difference is not significant (P < 0.10). We pooled the 156 random male and female Negroid chromosomes (independently of enzyme variants) and obtained PvuII allele frequencies of 0.96 for Type 1 and 0.04 for Type 2. In a sample of 23 San X chromosomes, the PvuII allele frequencies were the same. Table 1 shows the correlation between the PvuII alleles and the G6PD phenotypes. Out of 75 Gd B chromosomes from Negroids, we found none with a Type 2 allele; among the 20 Gd A chromosomes, two had Type 2 alleles. Some 13 San males were G6PD B and all posessed the PvuII Type 1 allele. The Type 2 allele shows significant linkage disequilibrium with Gd A in American Blacks (Yoshida et al. 1988; Beutler and Kuhl, 1990; Fey et al. 1990). A total of 43 chromosomes (39 Gd A, 3 Gd A- and 1 Gd B) was examined for the additional F o k I site. It was present in all except the Gd B chromosome, in accordance with previous findings (Yoshida and Takizawa 1988). Malaria has been a major health problem in Africa for millenia, and mutations that ameliorate the effects of the disease often reach polymorphic frequencies. Examples of such polymorphisms include the thalassaemias, the Duffy blood group allele Fy, sickle and other structural haemoglobin variants and G6PD deficiency (Jenkins and Dunn 1981). Jenkins (1982) has postulated, on the basis of the frequencies of the alleles in the different geographical regions of sub-Saharan Africa, that the mutations for Od A, Gd A- and sickle haemoglobin occurred in this temporal order. The Bantu-speaking Negroes of Southern Africa probably originated from a nuclear population in the vicinity of the source of the Congo River sometime after Gd A had arisen. This, together with the fact that there is less endemic malaria south of the Limpopo River, may explain why the frequencies of Gd A and Gd A- are lower in this area than elsewhere in Africa, and why the sickle haemoglobin mutation is absent. The frequency of Gd a is, on average, 0.08 in Southern African Bantu-speakers today, although the present sample has a lower frequency. Fey et al. (1990) have found the Type 2 PvuII allele to be present at a frequency of 0.32-0.40 in Zambians, Kenyans, Nigerians and West Indians. This is considerably higher than the frequency of 0.04 in Southern African Bantu-speakers and San, and is a reflection of the
leles in randomly selected males Population
G6PD B
G6PD A
Total
Type I Type2 Type I Type2 Bantu-speaking Negroids~' 75 North American Blacksb 39
0 5
18 8
2 6
95 58
fact that PvuII Type 2 is only found in G6PD A individuals in Africa. The low frequency of Gd a and the Type 2 allele in the San is to be expected, as they were among the earliest inhabitants of Southern Africa and, because of a hunting and gathering lifestyle, were probably less exposed to P. falciparum malaria. Yoshida et al. (1988) reported a Type 2 allele frequency of 0.19 in a North American Black population, but their sample was not random, having been enriched by the addition of G6PD A individuals. The North American Black population originated from slaves taken from West and Central Africa where P.falciparum malaria is hyperendemic, this might explain the much higher frequency of both Gd a and Gd A- (0.22 and 0.20, respectively) in this population (Luzzato and Allen 1968). None of the 75 Gd B chromosomes in the Southern African Black population carried the Type 2 allele, in contrast to 5 out of the 44 (11.3%) in the North American Black sample. The difference is borne out by the odds ratio for the association of the alleles with the phenotypes; this is 20.41 in the Southern African and 5.49 in the North American samples. It is therefore possible that a second Gd B allele may be present in the North American sample. The source of this allele may not be African at all and may be the result of admixture. Previous work (Fey et al. 1990) has not provided an answer regarding whether Gd B occurs in association with both PvuII alleles in West Africa: the G6PD enzyme type was not established and the D N A was not examined for the FokI restriction enzyme site. The PvuII data show a significant linkage disequilibrium between Gd A and the Type 2 allele in Southern African Blacks (A = 0.28), which is not significantly different from the value observed in the North American Black sample (0.35), indicating that the Gd A mutation in the two populations is probably the same and that it occurred before their divergence. The fact that the three deficiency mutants, collectively called Gd A- and identified in African populations, all arose in a gene that was already Gd A suggested to Beutler (1989) that Gd A may have been the most common allele in Africa at the time that the deficiency mutations occurred. This seems an unlikely explanation in view of the fact that sequence studies on chimpanzee G6PD suggest that the ancestral human enzyme had the G6PD B sequence (Beutler 1991). Gd A would therefore have arisen in Africa, after the dispersion of modern man to the other continents.
113 Acknowledgements. We wish to thank the volunteers who donated blood samples and the staff of the Highveld Blood Transfusion Service for collecting many of these samples. M.J.C. was supported by a Research Fellowship of the South African Medical Research Council.
References Betke K, Brewer GJ, Kirkman HN, Luzzato L, Motulsky AG, Ramot B, Siniscalco M (1967) Standardisation of procedures for the study of glucose-6-phosphate dehydrogenase. WHO Tech Rep Ser 366: 30-43 Beutler E (1989) Glucose-6-phosphate dehydrogenase: new perspectives. Blood 73 : 1397-1401 Beutler E (1991) Glucose-6-phosphate dehydrogenase deficiency. N Engl J Med 324 : 169-174 Beutler E, Kuhl W (1990) Linkage between a PvuII restriction fragment length polymorphism and G6PD A-202A/376G:evidence for a single origin of the common G6PD A - mutation. Hum Genet 85 : 9-11 Beutler E, Kuhl W, Vives-Corrons J-L, Prchal JT (1989) Molecular heterogeneity of glucose-6-phosphate dehydrogenase A - . Blood 74 : 2550-2555 Feinberg AP, Vogelstein B (1983) A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132: 6-13 Fey MF, Wainscoat JS, Mukwala EC, Falusi AG, Vulliamy TJ, Luzzato L (1990) A PvuII restriction fragment length poly-
morphism of the glucose-6-phosphate dehydrogenase gene is an African specific marker. Hum Genet 84: 471-472 Hill WG, Robertson A (1968) Linkage disequilibrium in finite populations. Theor Appl Genet 38:226-231 Jenkins T (1982) Human evolution in Southern Africa. In: BonneTamir B (ed) Human genetics, part A: the unfolding genome. Liss, NewYork, pp 227-253 Jenkins T, Dunn DS (1981) Haematological genetics in the tropics. Clin Haematol 10:1029-1050 Luzzato L (1985) Malaria and the red cell. Recent Adv Haematol 4 : 109-126 Luzzato L, Allen NC (1968) Relatioship between the genes for glucose-6-phosphate dehydrogenase and for haemoglobin in a Nigerian population. Nature 219:1041-1042 Nurse GT, Weiner JS, Jenkins T (1985) The peoples of Southern Africa and their Affinities. Clarendon Press, Oxford Southern EM (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Bio198 : 503517 Takizawa T, Yoneyama Y, Miwa S, Yoshida A (1987) A single nucleotide base transition is the basis of the common human glucose-6-phosphate dehydrogenase variant A(+). Genomics 1:228-231 Yoshida A, Takizawa T, Prchal JT (1988) RFLP of the X chromosome-linked glucose-6-phosphate dehydrogenase locus in Blacks. Am J Hum Genet 42 : 872-876 Yoshida A, Takizawa T (1988) The same extra FokI Cleavage site exists in glucose-6-phosphate variants A(+) and A ( - ) . Am J Hum Genet 43 : 131-133
