b-globin Gene Cluster Haplotypes in Sickle Cell Patients From Southwest Iran

Transcription

1 American Journal of Hematology 74: (2003) b-globin Gene Cluster Haplotypes in Sickle Cell Patients From Southwest Iran Z. Rahimi, 1 M. Karimi, 2 M. Haghshenass, 2 and A. Merat 1 * 1 Department of Biochemistry, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran 2 Hematology Research Center, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran Sickle cell anemia in Iran is accompanied by a high level of HbF and mild clinical presentation. Here we report haplotypes of the b gene cluster found in 81 randomly selected sickle cell patients, including 47 sickle cell anemia (SS), 17 sickle cell trait (AS), and 17 sickle/thalassemia (S/thal) from southwest Iran. We found all five common typical haplotypes as well as five atypical haplotypes in our patients. Except for four patients with homozygous Benin haplotype, none of the other African typical haplotypes were found in a homozygous state. Arab-Indian was found to be the most prevalent haplotype in the study population. This haplotype accounted for 51.1% as the homozygous form in SS patients, where 69.1% of chromosomes in these patients had the Arab-Indian haplotype. Bantu A2 was the second most prevalent haplotype among all patients. The mean %HbF in SS patients was and in the homozygous Arab-Indian haplotype it was still higher (30.40%), while in AS patients the %HbF was only The high %Gg chain (71.81) in the Arab-Indian homozygous haplotype was concomitant with the presence of an Xmn I site in both chromosomes. The presence of the Arab-Indian haplotype as the predominant haplotype might be suggestive of a gene flow to/from Saudi Arabia or India. More haplotype investigations of a normal population can clarify the high incidence of Bantu A2 haplotype in our population. Am. J. Hematol. 74: , ª 2003 Wiley-Liss, Inc. Key words: sickle cell; haplotype; Iran INTRODUCTION The sickle cell gene is linked to several DNA structures which are characterized by specific patterns of restriction endonuclease sites in the b-globin gene cluster haplotype [1]. Five different haplotypes have been demonstrated to be the most common backgrounds associated with the b S mutation. These haplotypes are designated the Benin, Senegal, Bantu, Cameroon, and Arab-Indian types, according to the geographical area in which they are most commonly found [2,3]. Sickle cell disease with the Arab-Indian haplotype found in India and the Persian Gulf region is associated with milder clinical course and higher HbF levels ( %) as compared to those in Africa (8 12.3%) [4 7]. While the Arab-Indian haplotype contains the highest level of HbF among all known haplotypes, the Senegal haplotype remains associated with a higher level of HbF compared to other African haplotypes [6]. The Arab-Indian and Senegal haplotypes are both ª 2003 Wiley-Liss, Inc. associated with a C!T mutation at position 158 in the area 5 0 to the Gg gene [8,9]. Sickle cell anemia in Iran was reported for the first time by Haghshenass et al. [10] to be present with a mild phenotype and associated with a much higher HbF level (18%) as compared to sickle patients reported in other parts of the world. Since the hematological characteristics and clinical severity of sickle cell anemia are linked to the haplotype background Contract grant sponsor: Shiraz University of Medical Sciences. Contract grant number: *Correspondence to: Ahmad Merat, Department of Biochemistry, Medical School, Shiraz University of Medical Sciences, P.O. Box , Shiraz, Iran. amerat@sums.ac.ir Received for publication 30 December 2002; Accepted 15 July 2003 Published online in Wiley InterScience ( DOI: /ajh.10422

2 b Gene Haplotypes of Sickle Cell in Iran 157 [3 7], we tried to study the molecular basis which might explain high HbF and the benign clinical presentation of sickle cell disease in Iran. Therefore, as the b S mutation has occurred in various DNA backgrounds, we thought of the possibility of the presence of haplotype(s) linked to the b S mutation in our geographical region, other than those reported in the literature. We also looked at the %Gg chain in the HbF associated with each haplotype. To accomplish this goal, we determined the b S gene haplotypes and the Gg content of HbF in 81 sickle patients referred to our university hospitals. PATIENTS AND METHODS Patients A total of 81 nonconsanguineous sickle cell patients, consisting of 39 males and 42 females, ages 5 58 years, were recruited from the hospitals of Shiraz University of Medical Sciences. Patients were all from southwest Iran (Fars and neighboring provinces) and included 47 sickle cell anemia (SS), 17 sickle cell trait (AS), and 17 sickle/thalassemia (S/thal) cases. When sickle patients had low MCV and HbA2 levels >3.5% they were considered S/thal cases. Hematological Analysis Complete blood counts were carried out using an automated cell counter. HbF and HbA2 were determined from g/total non-a chains and DEAE-52 column elution [11] methods, respectively. Using a sickle preparation test and hemoglobin electrophoresis on cellulose acetate at alkaline and acid ph, phenotypes were determined [12]. The Gg and Ag chains of globin were separated by high performance liquid chromatography (HPLC) using a gradient of acetonitrile and 0.1% trifluoroacetic acid (TFA) in water and a C18 preparative column [13]. Haplotyping From EDTA-treated venous blood, DNA was prepared by the phenol-chloroform extraction method [14] with minor modifications and kept at 20 Cuntilanalyzed. Six regions around and within the b-gene cluster were amplified by polymerase chain reaction (PCR), using primers from TIB MOLBIOL listed in Table I. Primer sequences were those given by Old [15] with his advised minor modifications. For DNA from each patient the PCR products were treated with the appropriate enzymes and the resulting fragments were separated on a 3% multipurpose agarose gel containing ethidium bromide. The bands were visualized on a UV light box and photographed by a Uvitec Gel Documentation System. The polymorphic restriction sites studied were, 5 0 to E gene by Hind II, 5 0 to Gg gene by Xmn I, within IVS 2 of the Gg and Ag genes by Hind III, 3 0 to cb by Hind II, and IVS 2 of the b gene by Ava II as shown in Figure 1. To determine the heterozygous haplotypes, it was assumed that the heterozygous patient possessed one common haplotype and a rare haplotype, as assumed by Steinberg et al. [16], rather than two different rare haplotypes. TABLE I. Oligonucleotide Primers Employed for Haplotyping of b-globin Gene Cluster Restriction enzymes and sequence of primers Direction Position HindII 5 0 to " 5 0 TCTCTGTTTGATGACAAATTC! AGTCATTGGTCAAGGCTGACC XmnI 5 0 to Gg 5 0 AACTGTTGCTTTATAGGATTTT! AGGAGCTTATTGATAACCTCAGAC HindIII within IVS II Gg 5 0 AGTGCTGCAAGAAGAACAACTACC! CTCTGCATCATGGGCAGTGAGCTC HindIII within IVS II Ag 5 0 ATGCTGCTAATGCTTCATTAC! TCATGTGTGATCTCTCAGCAG HindII 3 0 to cb 5 0 GTACTCATACTTTAAGTCCTAACT! TAAGCAAGATTATTTCTGGTCTCT AvaII within IVS II b 5 0 GTGGTCTACCCTTGGACCCAGAGG! TTCGTCTGTTTCCCATTCTAAACT

3 158 Rahimi et al. Fig. 1. Locations of six regions for b-globin gene cluster treated with four restriction enzymes. RESULTS Results of the hematological studies including total Hb, HbF, MCV, and MCH for patients are shown in Table II. The %HbF, %Gg, and G/A ratios for SS, AS, and S/thal patients are depicted in Table III. The distribution of the five typical haplotypes (Benin, Senegal, Bantu, Cameroon, and Arab-Indian) and five atypical haplotypes present in the patient population is given in Table IV. Among all patients, the homozygous Arab- Indian was found to be the most predominant haplotype (29.6%), followed by Arab-Indian in combination with an atypical haplotype (19.8%) similar to that described as Bantu A2 by Srinivas et al. [17] and later as H haplotype by Konstantopoulos et al. [18]. There were four homozygous Benin and three homozygous Bantu A2 haplotypes among the patients. The haplotype backgrounds in the remaining patients were combinations of various typical Arab-Indian being the most frequent and atypical haplotypes found in this study. All 24 homozygous Arab-Indian haplotype (Table IV) were among the SS patient group. Out of the four cases with homozygous Benin background (Table IV), three belonged to SS patients and one was an S/thal patient. Among the 17 AS patients no homozygous Arab-Indian was found but five possessed the Arab-Indian combined with the atypical Bantu A2 haplotype. From the three cases of homozygous Bantu A2, two were among AS patients. The 17 S/thal patients showed a rather similar pattern of haplotype distribution to that of AS cases in having predominantly chromosomes carrying the Arab-Indian and Bantu A2 haplotypes. Patterns of some agarose gel electrophoreses for haplotyping are illustrated in Figure 2. DISCUSSION The b S mutation has occurred in many DNA structures creating different backgrounds for the b-gene haplotypes. Most of these haplotypes are rare [1]. Four out of five common haplotypes, known as Benin, Senegal, Bantu, and Cameroon, were found mostly as homozygous forms in Africa [2], where heterozygous forms of these haplotypes were rare. The fifth haplotype (Arab- Indian) was found in tribal populations of the subcontinent of India and an eastern province of Saudi Arabia [3]. TABLE II. Hb, HbF, MCV, and MCH Values for SS, AS, and S/thal Patients Patient Hb (mg/dl) %HbF MCV (fl) MCH (pg) SS ± ± ± ± 3.63 (47) (41) (39) (41) (41) AS ± ± ± ± 3.51 (17) (15) (8) (15) (15) S/thal 9.83 ± ± ± ± 6.29 (17) (16) (15) (16) (16) Values are mean ± SD. Numbers in parentheses represent the number of patients. Data for a few patients in each group were unavailable. SS ¼ Sickle cell anemia. AS ¼ Sickle cell trait. S/thal ¼ Sickle/thalassemia. TABLE III. %HbF, %Gg and the Ratios of G/A in SS, AS, and S/thal Patients Patients %HbF %Gg G/A Ratio SS ± ± ± 0.89 (39) (39) (39) SS ± ± ± 0.33 Arab-Indian homozygotes (22) (22) (22) SS ± ± ± 1.13 Heterozygotes Arab-Indian/Bantu A2 (6) (6) (6) AS 1.20 ± 0.72* ± ± (8) (10) (10) S/thal ± ± ± 1.23 (15) (16) (16) Values are mean ± SD. Numbers in parentheses show the number of patients. *Significantly lower than the values for SS and S/thal patients.

4 TABLE IV. Distribution of Haplotypes of b-globin Gene Cluster in Patients b Gene Haplotypes of Sickle Cell in Iran 159 Haplotype Total SS AS S/thal Arab-Indian/Arab-Indian (+++ ++/+++ ++) 24(29.6%) 24 Benin/Benin ( ++/ ++) 4(4.9%) 3 1 BantuA2/BantuA2 (+ +/+ +) 3(3.7%) 1 2 Arab-Indian/Benin (+++ ++/ ++) 4(4.9%) 3 1 Arab-Indian/BantuA2 (+++ ++/+ +) 16(19.8%) Benin/BantuA2 ( ++/+ +) 6(7.4%) Bantu/BantuA2 ( + +/+ +) 2(2.5%) 1 1 Arab-Indian/Senegal (+++ ++/ ++ ++) 6(7.4%) Arab-Indian/Cameroon (+++ ++/ ++++) 2(2.5%) 1 1 Arab-Indian/Bantu (+++ ++/ + +) 1(1.2%) 1 Benin/Cameroon ( ++/ ++++) 1(1.2%) 1 Cameroon/Bantu ( ++++/ + +) 1(1.2%) 1 Arab Indian/Bantu A2a (+++ ++/+ ) 4(4.9%) 3 1 Arab Indian/Bantu A4 (+++ ++/ ++ + ) 4(4.9%) Arab Indian/Bantu A1 (+++ ++/ +) 1(1.2%) 1 Arab Indian/Bantu A6 (+++ ++/ ++ +) 1(1.2%) 1 Bantu/Bantu A2a ( + +/+ ) 1(1.2%) 1 For each haplotype number of patients are given as percent of total haplotypes in parentheses. SS ¼ Sickle cell anemia. AS ¼ Sickle cell trait. S/thal ¼ Sickle/thalassemia. High levels of HbF associated with a mild clinical course of the disease have been known in Iran for many years [10]. But except for a brief report on three SS patients from the central part of Iran [19], no comprehensive information about the haplotype background of the b S mutation in this region is available. In the present effort to investigate the haplotype backgrounds of the b S mutation, in randomly selected sickle cell patients from southwest Iran, Arab-Indian was found to be the most frequent haplotype 24 homozygous and 39 heterozygous patients (Table IV). The homozygous form of Arab-Indian occurred only among SS patients, Fig. 2. An agarose gel electrophoresis pattern of some of RFLP products of b gene cluster. Lane 1 shows the DNA molecular weight marker with a 50 basepair ladder. Lane 2 is the DNA which is / for Hind III within IVS II Ag. Lane 3 is the DNA which is +/+ for Hind II 3 0 yb. Lane 4 is the DNA +/ for Hind III within IVS II Gg. Lane 5 is the DNA +/ for Xmnl 5 0 -Gg. accounting for 51.1% of the total haplotypes in these patients, followed by the heterozygous form of Arab- Indian/atypical Bantu A2 as the second prevalent haplotype (seven cases). Out of 94 chromosomes from SS patients, nearly 69% (65 chromosomes) carried the Arab-Indian haplotype. The very high frequency of Arab-Indian in our study population may be suggestive of a gene migration to/from Saudi Arabia and India with genetic admixture. Also, the presence of the homozygous Benin haplotype might be regarded as a reflection of gene flow from Africa through the possible routes suggested by Ragussa et al. [20] to southwest Iran. There were five atypical haplotypes in our study population. Four of these haplotypes corresponded to Bantu A1, Bantu A2, Bantu A4, and Bantu A6 based on the resemblance of polymorphic sites on the 5 0 region of the b gene, reported by Srinivas et al. [17]. However, the fifth atypical haplotype was similar to Bantu A2, but the Ava II site for this particular atypical haplotype was absent, and so it was designated Bantu A2a. It is notable that all five typical haplotypes were encountered in our study population, with Arab-Indian being the most and Cameroon being the least frequent haplotypes. The rather frequent atypical Bantu A2 haplotype found as both homozygous and heterozygous in our study group may indicate the possibility of its being a common haplotype in our population. Investigation on haplotype patterns of the normal population in this region will certainly provide further evidence to the present findings. There was a very wide range of the %HbF in each patient group, presumably due to factors other than haplotype, influencing HbF levels [16]. Nevertheless, the mean %HbF for SS patients was 27.83, which is higher than the 13.34% reported for Saudi Arabia [4]

5 160 Rahimi et al. and 16.4% reported for India [5]. Also, the AS patients had a mean %HbF of 1.20, which is higher than the 0.7 given for India [5] and slightly more than 1.09 for an eastern province of Saudi Arabia [9]. In SS patients, the homozygous form of the Arab- Indian haplotype was associated with high %HbF (30.40), as shown in Table III. Also, the %HbF tended to be high in the heterozygous cases with an Arab- Indian haplotype. This finding is consistent with the reported influence of the Arab-Indian haplotype on HbF level [9]. Nagel et al. [21] showed that the presence of only one chromosome with the Senegal haplotype, in comparison to combinations of other African haplotypes, is associated with a higher HbF level and amelioration of anemia. However, in our study population the HbF was significantly lower in the AS patient group as compared to the SS or S/thal groups by one-way ANOVA test (P < 0.001), while SS and S/thal did not have different HbF levels. Also, the %Gg and the G/A ratios were not statistically different in all three patient groups. No comparison was possible in the %Gg for various haplotypes in the SS, AS, and S/thal groups due to the small number of patients with each haplotype. However, the 22 homozygous Arab-Indian haplotype in SS patients had exactly the some mean value (71.78) for %Gg as the six heterozygous Arab-Indian/Bantu A2 patients (71.78) in this group. The high %HbF in our various haplotypes was associated with the presence of a Xmn I site that is the sequence 157 to 166 of the Gg gene if a T is present at position 158, as reported by Gilman and Huisman [22]. The presence of this site has been shown on chromosomes bearing the Senegal or Arab-Indian haplotypes [8,9]. We found the XmnI site to be present in patients with Arab-Indian, Senegal, and Bantu A4 atypical haplotypes. All other typical and atypical haplotypes lacked this site. However, haplotype is not the only factor in determining the HbF level other factors, including gender and genes not linked to the b-like gene cluster, are also involved [16]. REFERENCES 1. Antonarakis SE, Boehm CD, Serjeant GR, Theisen CE, Dover GJ, Kazazian HH. Origin of the b S -globin gene in blacks: the contribution of recurrent mutation or gene conversion or both. Proc Natl Acad Sci USA 1984;81: Pagnier J, Mears JG, Dunda-Belkhodja O, et al. Evidence for the multicentric origin of the sickle cell hemoglobin gene in Africa. Proc Natl Acad Sci USA 1984;81: Serjeant GR. The geography of sickle cell disease: opportunities for understanding its diversity. Ann Saudi Med 1994;14: Padmos MA, Roberts GT, Sackey K, et al. Two different forms of homozygous sickle cell disease occur in Saudi Arabia. Br J Haematol 1991;79: Kulozik AE, Kar BC, Satapathy RK, Serjeant BE, Serjeant GR, Weatherall DJ. Fetal hemoglobin levels and b S globin haplotypes in an Indian population with sickle cell disease. Blood 1987;69: Nagel RL, Fabry ME, Pagnier J, et al. Hematologically and genetically distinct forms of sickle cell anemia in Africa. N Engl J Med 1985;312: Nagel RL, Srinivas R, Dunda-Belkhodja O, et al. The hematologic characteristics of sickle cell anemia bearing the Bantu haplotype: the relationship between Gg and HbF level. Blood 1987;69: Labie D, Dunda-Belkhodja O, Rouabhi F, Pagnier J, Ragusa A, Nagel RL. The -158 site 5 0 to the Gg gene and Gg expression. Blood 1985;66: Miller BA, Salameh M, Ahmed M, et al. High fetal hemoglobin production in sickle cell anemia in the eastern province of Saudi Arabia is genetically determined. Blood 1986;67: Haghshenass M, Ismail-Beigi F, Clegg JB, Weatherall DJ. Mild sickle cell anemia in Iran associated with high levels of fetal hemoglobin. J Med Gen 1977;14: Fairbanks VF, Klee GG. Biochemical aspects of hematology. In: Burtis CA, Ashwood ER, editors. Tietz text book of clinical chemistry. Philadelphia: WB Saunders; p Schneider RG. Differentiation of electrophoretically similar hemoglobins- such as S, D, G and P; or A2, C, E, and O by electrophoresis of the globin chains. Clin Chem 1974;20: Huisman THJ. High-performance liquid chromatography as a method to identify haemoglobin abnormalities. Acta Haematol 1987;78: Old JM, Higgs DR. Gene analysis. In: Weatherall DJ, editor. Methods in hemotology, vol. 6. The thalassemias. London: Churchill Livingstone; p Old JM. Hemoglobinopathies. In: Elles R, editor. Methods in molecular medicine: molecular diagnosis of genetic disease. Totowa, NJ: Humana Press; p Steinberg MH, Hsu H, Nagel RL, et al. Gender and haplotype effects upon hematological manifestations of adult sickle cell anemia. Am J Hematol 1995;48: Srinivas R, Dunda O, Krishnamoorthy R, et al. Atypical haplotypes linked to the b S gene in Africa are likely to be the product of recombination. Am J Hematol 1988;29: Konstantopoulos K, Vulliamy T, Swirsky D, Reeves JD, Kaeda J, Luzzatto L. DNA haplotypes in Africa and West Indians with sickle cell anaemia or SC disease. Gene Geography 1996;10: Rahgozar S, Poorfathollah AA, Moafi AR, Old JM. b S gene in central Iran is in linkage disequilibrium with the Indian-Arab haplotype. Am J Hematol 2000;65: Ragusa A, Lombardo M, Sortino G, Lombardo T, Nagel RL, Labie D. b S gene in Sicily is in linkage disequilibrium with the Benin haplotype: implications for gene flow. Am J Hematol 1988;27: Nagel RL, Erlingsson S, Fabry ME, et al. The Senegal DNA haplotype is associated with amelioration of anemia in African- American sickle cell patients. Blood 1991;77: Gilman JG, Huisman THG. DNA sequence variation associated with elevated fetal Gg globin production. Blood 1985;66: