- Email: [email protected]
Haplotype analysis of beta thalassemia patients in Western Iran
Blood Cells, Molecules, and Diseases 42 (2009) 140–143
Contents lists available at ScienceDirect
Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y b c m d
Haplotype analysis of beta thalassemia patients in Western Iran Zohreh Rahimi a,b,c,⁎, Adriana Muniz d, Reza Akramipour e, Fareidon Tofieghzadeh e, Hadi Mozafari a, Asad Vaisi-Raygani b, Abbas Parsian f a
Medical Biology Research Center, Kermanshah University of Medical Sciences, Medical School, Daneshgah Avenue, Kermanshah, Iran Department of Biochemistry, Medical School, Kermanshah University of Medical Sciences, Kermanshah, Iran Department of Biochemistry, Medical School, Kurdistan University of Medical Sciences, Kurdistan, Iran d Department of Medicine, Division of Hematology; Albert Einstein College of Medicine, Bronx, NY, USA e Department of Pediatrics, Medical School, Kermanshah University of Medical Sciences, Kermanshah, Iran f Division of Neuroscience & Behavior, National Institutes of Health, Rockville, Maryland, USA b c
a r t i c l e
i n f o
Article history: Submitted 27 November 2008 Available online 13 January 2009 (Communicated by S. Orkin, M.D., 02 December 2008) Keywords: Haplotype β-thalassemia Mutation XmnI site Western Iran
a b s t r a c t β-thalassemia (β-thal) is the most common single gene disorder in Iran. To determine the chromosomal background of beta thalassemia mutations in Western Iran we studied β-globin gene cluster haplotypes in 314 β-thal and 70 βA chromosomes with a Kurd ethnic background from the province of Kermanshah, Iran using PCR-RFLP. β-thal mutations were analyzed using PCR-ARMS, RFLP and direct genomic sequencing. Haplotypes were constructed by analyzing the pattern of seven restriction sites through the β-globin gene cluster. Haplotype I was the most prevalent haplotype (35.7%) among β-thal chromosomes followed by haplotype III (28.6%). βA chromosomes similar to β-thal chromosomes were linked to diverse haplotypes but predominantly with haplotype I (42.9%). The predominant IVSII-1 (G → A) mutation in this population (33%) was strongly linked to haplotype III (66.1%) but was also found on chromosomes with haplotypes I, II, V, X and atypical. The second prevalent mutation was CD8/9 +G (13.5%) and showed a strong association with haplotype I (96.4%) and a weak association with haplotype V (3.6%). Haplotype background for Kurdish mutations among our studied population was similar to those among Kurdish Jews and people of Kurdistan of Iran. Identification of the most common mutations on different haplotype backgrounds can be explained by a variety of gene conversion and recombination events. © 2008 Elsevier Inc. All rights reserved.
Introduction Beta thalassemia is one of the most common genetic diseases worldwide resulting from aberrant beta-globin chain production. It is highly prevalent among people of Mediterranean, African, or Asian descent [1,2]. The presence of clinical heterogeneity among β-thalassemia patients has been paralleled by genetic heterogeneity [3]. β-thalassemia is the most common single gene disorder in Iran. More than 25,000 affected individuals have been reported. High prevalence of β-thalassemia (approximately 10%) occurs in Northern and Southern provinces, near to the Caspian Sea and Persian Gulf, respectively. The prevalence of β-thalassemia alleles in other parts of Iran has been estimated to be 4–8% [4]. In Kermanshah province the relative frequency of β-thalassemia is approximately 3% and according to the report of the Health Center of Kermanshah there are 280 β-thalassemia major registered patients.
⁎ Corresponding author. Medical Biology Research Center, Kermanshah University of Medical Sciences, Medical School, Daneshgah Avenue, Kermanshah, Iran. Fax: 0098 831 4276471. E-mail addresses: [email protected], [email protected] (Z. Rahimi). 1079-9796/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bcmd.2008.12.001
Orkin et al. [5] found strong linkage disequilibrium between polymorphic sites across the β-globin gene cluster (haplotypes) and β-thalassemia mutations. A limited number of haplotypes are found in each population, and each haplotype is usually associated with one specific type of thalassemia. When mutation analysis fails, haplotype analysis is useful in identifying a β-thal chromosome, with an accuracy of about 90% if the family is informative [6]. β-globin gene cluster haplotypes are useful in diagnosis of particular molecular defects in β-thalassemia, prenatal diagnosis of β-thalassemia, and elucidating population affinities [7]. Population of Iran consists of multiethnic groups, so each region of the country has their own distinct set of mutations. To our knowledge there is no report related to the haplotype/mutation association in Western Iran. This is the first paper to report the analysis of association of β-globin gene cluster haplotypes with β-thalassemia mutations among population of Western Iran with the Kurdish ethnic background to identify the origin and historical background of these mutations in this area. Materials and methods The patient group was comprised of 157 unrelated β-thalassemia major patients (87 males and 70 females) who referred to the Shahid
Z. Rahimi et al. / Blood Cells, Molecules, and Diseases 42 (2009) 140–143
Fahmideh Hospital of Kermanshah University of Medical Sciences for blood transfusion. The control group consisted of 35 healthy individuals with normal blood indices from the same area. Informed written consent was obtained from each individual before participation. The study was approved by Ethics Committee of Kermanshah University of Medical Sciences and was in accordance with the principles of the Declaration of Helsinki II. The levels of Hb A2 and HbF were determined by DEAE-52 microcolumn chromatography and alkaline denaturation [8,9], respectively. DNA was extracted from leukocytes of whole blood by phenolchloroform method [10]. For genotyping of variations, genomic DNA from patients was analyzed by PCR-ARMS and RFLP as previously described [11]. Unknown mutations were identified by direct genomic sequencing using an ABI 3730 DNA analyzer (Applied Biosystems, Foster City, Ca). Haplotypes were determined by analyzing the pattern of seven restriction sites through the β-globin gene cluster using primers reported by Old [12] and Lee et al. [13]. The polymorphic sites were, 5′ to ε gene by Hind II, within IVS II of the Gγ and Aγ genes by Hind III, within and 3′ to ψβ by Hind II, IVS 2 of the β gene by Ava II, and 3′ to the β gene by BamH I. To determine the heterozygous haplotypes in β-thal and βA chromosomes, it was assumed that the heterozygous individual possessed one common haplotype and a rare haplotype rather than two rare different haplotypes. The haplotype classification of Orkin [5] has been adopted.
We performed haplotype analysis using a total of 314 chromosomes and identified 17 different haplotypes. In order to define on which allele the polymorphic sites of the β-globin gene cluster is located the probands and both parents need to be analysed. Since the parents were not always available we studied 83 patients (166 chromosomes) homozygous for the same mutation. As presented in Table 1, haplotype I was the most prevalent haplotype (35.7%) among β-thal chromosomes followed by haplotypes III (28.6%), VII (10.2%), V (7.3%), IX (3.5%), A3 (3.2%), VI (2.9%), II (1.6%), IV (1%) and haplotypes VIII, X, A, and D each with 0.6%. Atypical haplotypes accounted for 3.6% of the total haplotypes.
Table 1 β-globin gene cluster haplotypes linked to β-thalassemia and βA chromosomes in Western Iran Haplotype I II III IV V VI VII VIII IX X A3 A D
II atypical Total
Table 2 Haplotypes associated with the same mutation Mutation (N)⁎
Haplotype (N)
IVSII.1 G:A (56)
III (37) V (8) Atypical (6) I (3) X (1) II atypical (1) I (27) V (1) I (12) III (2) V (1) A (1) VII (6) III (2) IX (2) I (4) Atypical (2) IX (1) A (1) I (8) VII (5) II atypical (2) D (1) I (5) VII (1) I (5) V (1) I (4) II (1) VIII (1) A3 (2) VI (2) A3 (2) III (2) IX (2) III (4)
CD 8/9 + G (28) CD 36/37-T (16)
CD 15 TGG: TAG (10)
IVSI.110 G:A (8)
IVSI.1 G:A (8) CD 39 C:T (8)
CD 44-C (6) IVSI.110 C:T (6) IVSI.5 G:C (4) CD 8-AA (4)
Results
+−−−−++ −++−+++ −+−+++− −+−++−+ +−−−−+− −++−−−+ +−−−−−+ −+−+−+− −+−++++ −+−−−−+ −++−+−+ −−−−−++ −−−−−−+ −++++++ −+−+++ −+++++− −++ + + + + − − − − − + ++ −++−−+− −++−−++ −++−++−
βA (%)
βThal (%)
30 5 4 – 10 – 3 – 1 – 1 – – 5 3 3
112 (35.7) 5 (1.6) 90 (28.6) 3 (1) 23 (7.3) 9 (2.9) 32 (10.2) 2 (0.6) 11 (3.5) 2 (0.6) 10 (3.2) 2 (0.6) 2 (0.6) 4 (1.3) – – 1 (0.3) – 3 (1) 3 (1) – 314
(42.9) (7.1) (5.7) (14.3) (4.3) (1.4) (1.4)
(7.1) (4.3) (4.3)
2 (2.9) 2 (2.9) 1 (1.4) 70 (100)
The order of the sites from 5 to 3 is as follows: Hind II 5′ to ε gene, Hind III within IVS 2 of the Gγ and Aγ genes, Hind II within and 3′ to ψβ gene, Ava II within IVS 2 of the β-globin gene, and BamH I 3′ to the β-globin gene.
141
IVSII.745 C:G (4) IVSI.108-25 (4) CD 22/23/24 del (4) N:number of chromosomes.
Distribution and frequency of β-globin gene cluster haplotypes for βA chromosomes are also depicted in Table 1. βA chromosomes similar to β-thal chromosomes were linked to diverse haplotypes but predominantly with haplotype I (42.9%). The magnitude of these frequencies, for other haplotypes including atypical, is also presented in Table 1. From the twenty different mutations that were analyzed, the mutation IVSII.1 (G → A) accounted for approximately 33% of studied thalassemic alleles followed by CD8/9 +G (13.51%). The results of scrutiny of individuals with homozygous mutations (166 chromosomes) and their linkage to corresponding haplotypes are demonstrated in Table 2. Discussion The linkage of β-thalassemia mutation with specific haplotypes has been reported previously [5]. In chromosomes of β-thalassemia patients from Mediterranean area, haplotype I of the β-thal gene was found to be the most prevalent haplotype (47%) followed by haplotypes II (17%), V (12%), III (8%), VI (6%), VII (6%), IX (3%), IV (1%) and VIII (1%) (Orkin et al., 1982). In the present study haplotype I was the most prevalent haplotype among β-thalassemia patients (35.7%), and normal individuals (42.9%). Previously we reported that haplotype I was the most prevalent among β-thalassemia minor individuals and normal controls from Southern Iran (46.2% and 43.3%, respectively) [14]. However, in the present study the second prevalent haplotype was haplotypes III (28.6%) and V (14.3%) for patients and controls, respectively. Interestingly, in Southern Iran the second prevalent haplotype among patients was haplotype V (15.4%) and in normal
142
Z. Rahimi et al. / Blood Cells, Molecules, and Diseases 42 (2009) 140–143
individuals was haplotype III (15.4%) [14]. Since haplotype I is the most prevalent haplotype in β-thal and βA chromosomes, it implies that β-thalassemia mutations might have arisen from the chromosomal background common in the population, rather than due to selection pressure or gene flow (migration). Recently, in screening 370 chromosomes from β-thalassemia patients of Kermanshah Province we identified twenty different mutations. The most common mutation was IVSII-1 (G → A) (32.97%) followed by, in order of frequency, CD8/9 +G (13.51%), IVSI-110 (C → T) (8.38%), CD 36/37-T (7.84%), CD8-AA (5.94%), CD15 (G → A) (4.86%), IVSI.1 (G → A) (4.59%), IVSI.6 (T → C) (3.78%), CD39 (C → T) (2.97%), IVSII.745 (C → G) (2.43%), CD44-C (2.43%), IVSI.5 (G → C) (2.16%), IVSI.108-25 bp (2.16%), CD82/83-G (1.08%), and CD 22/23/24 (−AAGTTGG) (1.08%). Rare mutations (b1%) were IVSII.2,3 (+ 11 − 2) (0.54%), IVSI.128 (T → G) (0.54%), CD6-A (0.27%), CD37 (G → A) (0.27%), and CD 9/10 + T (0.27%) (data not presented). The frequency of IVSII.1 (G → A) is high in Middle East (7–47%). This mutation is linked to various β haplotypes (I, III, IX or atypical) and results in a disease of variable severity [15]. The IVS II.1 (G → A) mutation has been reported as the most prevalent mutation among Iranians [16]. In the patient sample analyzed on the present study the prevalence of this mutation was 33%. This mutation was strongly linked to haplotype III (66.1%). However, it was also found on chromosomes with the haplotypes I, II, V, X and atypical. In β-thalassemia patients from the Mediterranean countries this mutation is associated with haplotypes III and V [17]. In contrast to our study Yavarian et al. [18] in small sample from South Western Provinces of Hormozgan (presumably Persian and Arab in origin) and Fars (Farsi speakers and presumably Persian in origin.) found that most of the patients with IVS II.1 (G → A) mutation (80%) had haplotype I and 20% of patients with this mutation had haplotype III. These differences could be attributed to the different ethnic backgrounds of patients in two studies. It has been suggested that a variety of gene conversion and recombination mechanisms could be responsible for heterogeneity of mutation/haplotype associations [15]. Haplotypes associated with IVSII.1 (G → A) mutation are also represented among the normal individuals indicating that the mutation arose on existing chromosomal background. The association between haplotype and mutation is not invariant, because the same haplotype, when present in two different ethnic groups, has been associated with different mutations. Also, mutation spread from one haplotype to another within a racial group has been found [3]. Very mild to moderate phenotypes are virtually found only among individuals who are homozygotes for the mutation in cis XmnI+ of haplotype III. In our study patients with haplotype background III and IX were XmnI positive. The presence of an XmnI cleavage site (+) causes a significant delayed decline of Hb F production in infant age, a high propensity of postnatal γ-globin expression and ameliorating the clinical phenotypes [19]. In Mediterranean area the polymorphism XmnI indicates linkage disequilibrium with haplotypes III, IV, and IX [20]. The CD 8/9 (+G) in Lebanon was linked to haplotype I [21] and in India to 5′ subhaplotype + −−− [22]. This mutation in our population indicates a strong association with haplotype I (96.4%) and a weak association with haplotype V (3.6%). Haplotype V has the same subhaplotype (+ − − − − +) as haplotype I. In India IVSI.5 (G → C) was linked to 5′ subhaplotype + − (found in haplotypes I, V and VII) [22] and to haplotypes VII, and X [17]. In present study this mutation was associated to haplotype I. In Lebanese population this mutation was associated with haplotype VII [20]. The frameshift 36/37 (−T) mutation that has been found in South Western province of Iran (Khuzistan) and also among Kurdish Jewish patients was found to be the fourth prevalent mutation (7.84%). In our studied patients with Kurdish ethnic origin this mutation was mostly associated with haplotype I (75%) and with lesser frequency was linked to haplotypes III, V, and A (Table 2). Among Kurdish Jews this mutation was associated with haplotype I [23].
In Mediterranean CD 39 (C → T) mutation was linked to haplotypes I, II, VII and haplotype IX [17]. In our patients, the CD 39 (C → T) was found with haplotype VII (62.5%) and also was associated with haplotypes II atypical and D. Similarly among Kurdish Jews this mutation was linked to haplotype VII. In Kurdistan of Iran 58.3% of the nonsense CD 39 mutation had haplotype background VII [23]. Conversely, the mutation was linked to haplotypes I, II in Lebanon, with haplotypes I, II, IX in Algeria and with haplotypes II, IV in Turkey [2]. It seems that the ancestral background for the frameshift mutations 36/37 (−T) and CD 39 (C → T) are haplotype I and VII, respectively. In Lebanon and among Kurdish Jews haplotype I was associated with IVS I.110 (G → A) mutation [2,23]. In the present study IVS I.110 (G → A) mutation was linked to haplotype I in 50% of patients and the remaining patients had haplotype backgrounds II, atypical, IX, and A (Table 2). In the Fars province of Iran this mutation was linked to haplotype I [14]. The IVS I.110 (G → A) mutation is tightly linked to haplotype I in Algeria (92%) and Turkey (93%). However a highest diversity of haplotypes linked to IVSI.110 (G → A) was observed in Turkey [2]. In Mediterranean area this mutation was associated with haplotypes I, II, and IX [20]. The CD 15 (G → A) mutation was associated with haplotype VII in 60% of chromosomes and the remaining 40% was linked to haplotypes III, IX. The mutation IVSII.745 (C → G) in Mediterranean area was linked to haplotype VII [17]. However, in the present study this mutation equally linked to haplotypes VI and A3 (Table 2). In Lebanese population the mutation IVSI.108-25 was found to be linked to haplotype IX [21] whereas, in our population this mutation is linked to haplotypes III and IX with equal frequency. Haplotype IX includes the same 5′ subhaplotype (− + − + + +) as haplotype III. The Mediterranean β-thalassemia patients with IVSI.1 (G → A) mutation are linked to haplotype V [17] like in the majority of the Lebanese β-thalassemia chromosomes, although, in one case linkage of chromosome to haplotype IX was detected. All β-thalassemia chromosomes of our study were in linkage disequilibrium with haplotype I. The FSC 8 (−AA) mutation in our study existed in three different haplotypes, II, VIII, and A3 while in the Mediterranean area this mutation is associated with haplotypes IV, VI and VII [20,21]. It has been demonstrated that a single mutation can be appear in association with different haplotypes. It has also been suggested that the association between specific haplotypes and a specific thalassemic mutation could be very high (85%). In fact the same haplotype could be associated with different genetic defects. In general, involvement of the same mutation to more than one haplotype is most consistent with the crossing-over of 5′ into the β-globin cluster, presumably within the region of increased recombination [17]. The linkage of CD44-C to haplotype I have been previously reported in the Kurdish population [23]. In Western Iran, 83.3% of chromosomes with this mutation that are present in individuals with Kurdish ethnic background were linked also to this haplotype. The minor association of CD44-C with haplotype VII observed in our study could be due to a meiotic recombination starting from haplotype I. CD44-C in two chromosomes from Lebanon was found on haplotypes I and on 5′ haplotype X, respectively [21]. The CD −22/23/24 deletion was in linkage disequilibrium with haplotype III. We found that several mutations can occur within a single haplotype. However, the association of one specific mutation to one haplotype is the most predominant. Conclusion Finding the most common mutations on different haplotype backgrounds can be explained by a variety of gene conversion and recombination events. However, almost all mutations have arisen on
Z. Rahimi et al. / Blood Cells, Molecules, and Diseases 42 (2009) 140–143
the same chromosomal background of normal population. Also, we demonstrated that haplotype background for Kurdish mutations among our studied population with ethnic background of Kurd is similar to those among Kurdish Jews and people of Kurdistan of Iran. References [1] M. Knott, K.M. Ramadan, G. Savage, et al., Novel and Mediterranean beta thalassemia mutations in the indigenous Northern Ireland population, Blood Cells Mol. Dis. 36 (2006) 265–268. [2] L. Zahed, J. Demont, R. Bouhass, et al., Origin and history of the IVSI.110 and codon 39 beta thalassemia mutations in the Lebanese population, Hum. Biol. 74 (2002) 837–847. [3] C. Wong, S.E. Antonarkis, S.C. Goff, et al., On the origin and spread of βthalassemia: recurrent observation of four mutations in different ethnic groups, Proc. Natl. Acad. Sci. U. S. A. 83 (1986) 6529–6532. [4] Z. Rahimi, A. Vaisi Raygani, A. Merat, et al., Thalassemic mutations in Southern Iran, Ir. J. Med. Sci. 31 (2006) 70–73. [5] S.H. Orkin, H.H. Kazazian Jr, S.E. Antonarakis, et al., Linkage of β thalassemia mutations and β globin gene polymorphism with DNA polymorphism in human β globin gene cluster, Nature 296 (1982) 627–632. [6] B.G. Forget, H.A. Pearson, Hemoglobin synthesis and the thalassemias, in: R.I. Handin, S.E. Lux, T.P. Stossel (Eds.), Blood: Principles and Practice of Hematology, J.B. Lippincott Company, Philadelphia, 1995, pp. 1525–1590. [7] R.L. Nagel, H.M. Ranney, Genetic epidemiology of structural mutations of the βglobin gene, Semin. Hematol. 27 (1990) 342–359. [8] V.F. Fairbanks, G.G. Klee, Biochemical aspects of haematology, in: C.A. Burtis, E.R. Ashwood (Eds.), Tietz Text Book of Clinical Chemistry, W.B Saunders, Philadephia, 1994, pp. 2041–2042. [9] K. Betke, H.R. Marti, I. Schlicht, Estimation of small percentages of foetal haemoglobin, Nature 184 (1959) 1877–1878.
143
[10] J.M. Old, D.R. Higgs, Gene analysis, in: D.J. Weatherall (Ed.), Methods in Hemotology. Vol:6. The Thalassemias, Livingstone, 1983, pp. 74–101. [11] J.M. Old, N.Y. Varawalla, D.J. Weatherall, Rapid detection and prenatal diagnosis of beta-thalassaemia: studies in Indian and Cypriot populations in the UK, Lancet 386 (1990) 834–837. [12] J.M. Old, Hemoglobinopathies, in: R. Elles (Ed.), Methods in Molecular Medicine: Molecular Diagnosis of Genetic Disease, Humana press Inc, Totowa, NJ, 1996, pp. 169–183. [13] Y.J. Lee, S.S. Park, J.Y. Kim, H.L. Cho, RFLP haplotypes of β-globin gene complex of β-thalassemic chromosomes in Koreans, J. Korean Med. Sci. 17 (2002) 475–478. [14] Z. Rahimi, A. Merat, M. Akhzari, et al., β-Globin gene cluster haplotypes in Iranian patients with β-Thalassemia, Int. J. Hematol. Oncol. 2 (2005) 30–34. [15] G.O. Tadmouri, N. Garguier, J. Demont, et al., History and origin of β-thalassemia in Turkey: sequence haplotype diversity of β-globin genes, Hum. Biol. 73 (2001) 661–674. [16] H. Najmabadi, R. Karimi-Nejad, S. Sahebjam, et al., The beta-thalassemia mutation spectrum in the Iranian population, Hemoglobin 25 (2001) 285–296. [17] S.E. Antonarakis, H.H. Kazazian Jr, S.H. Orkin, DNA polymorphism and molecular pathology of the human globin gene cluster, Hum. Gene 69 (1985) 1–14. [18] M. Yavarian, M. Karimi, E. bakker, et al., Response to hydroxyurea treatment in Iranian transfusion-dependent β-thalassemia patients, Haematologica 89 (2004) 1172–1178. [19] V. Viprakasit, W. Chinchang, L. Suwanthol, V.S. Tanphaichitr, Common origin of a rare beta-globin initiation codon mutation (ATG → AGG) in Asian, Clin. Lab. Haematol. 27 (2005) 409–415. [20] W. Lemsaddek, I. Picanco, F. Seuanes, et al., The β-thalassemia mutation/haplotype distribution in the Moroccan population, Hemoglobin 28 (2004) 25–37. [21] L. Zahed, M. Qatanani, M. Nabulsi, A. Taher, β-thalassemia mutations and haplotype analysis in Lebanon, Hemoglobin 24 (2000) 269–276. [22] A. Bandyopadhyay, S. Bandyopadhyay, M.D. Chowdhury, U.B. Dasgupta, Major beta-globin gene mutations in Eastern India and their associated haplotypes, Hum. Hered. 49 (1999) 232–235. [23] D. Rund, T. Cohen, D. Filon, et al., Evolution of a genetic disease in an ethnic isolate: β-thalassemia in the Jews of Kurdistan, Proc. Natl. Acad. Sci. 88 (1991) 310–314.
Contents lists available at ScienceDirect
Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y b c m d
Haplotype analysis of beta thalassemia patients in Western Iran Zohreh Rahimi a,b,c,⁎, Adriana Muniz d, Reza Akramipour e, Fareidon Tofieghzadeh e, Hadi Mozafari a, Asad Vaisi-Raygani b, Abbas Parsian f a
Medical Biology Research Center, Kermanshah University of Medical Sciences, Medical School, Daneshgah Avenue, Kermanshah, Iran Department of Biochemistry, Medical School, Kermanshah University of Medical Sciences, Kermanshah, Iran Department of Biochemistry, Medical School, Kurdistan University of Medical Sciences, Kurdistan, Iran d Department of Medicine, Division of Hematology; Albert Einstein College of Medicine, Bronx, NY, USA e Department of Pediatrics, Medical School, Kermanshah University of Medical Sciences, Kermanshah, Iran f Division of Neuroscience & Behavior, National Institutes of Health, Rockville, Maryland, USA b c
a r t i c l e
i n f o
Article history: Submitted 27 November 2008 Available online 13 January 2009 (Communicated by S. Orkin, M.D., 02 December 2008) Keywords: Haplotype β-thalassemia Mutation XmnI site Western Iran
a b s t r a c t β-thalassemia (β-thal) is the most common single gene disorder in Iran. To determine the chromosomal background of beta thalassemia mutations in Western Iran we studied β-globin gene cluster haplotypes in 314 β-thal and 70 βA chromosomes with a Kurd ethnic background from the province of Kermanshah, Iran using PCR-RFLP. β-thal mutations were analyzed using PCR-ARMS, RFLP and direct genomic sequencing. Haplotypes were constructed by analyzing the pattern of seven restriction sites through the β-globin gene cluster. Haplotype I was the most prevalent haplotype (35.7%) among β-thal chromosomes followed by haplotype III (28.6%). βA chromosomes similar to β-thal chromosomes were linked to diverse haplotypes but predominantly with haplotype I (42.9%). The predominant IVSII-1 (G → A) mutation in this population (33%) was strongly linked to haplotype III (66.1%) but was also found on chromosomes with haplotypes I, II, V, X and atypical. The second prevalent mutation was CD8/9 +G (13.5%) and showed a strong association with haplotype I (96.4%) and a weak association with haplotype V (3.6%). Haplotype background for Kurdish mutations among our studied population was similar to those among Kurdish Jews and people of Kurdistan of Iran. Identification of the most common mutations on different haplotype backgrounds can be explained by a variety of gene conversion and recombination events. © 2008 Elsevier Inc. All rights reserved.
Introduction Beta thalassemia is one of the most common genetic diseases worldwide resulting from aberrant beta-globin chain production. It is highly prevalent among people of Mediterranean, African, or Asian descent [1,2]. The presence of clinical heterogeneity among β-thalassemia patients has been paralleled by genetic heterogeneity [3]. β-thalassemia is the most common single gene disorder in Iran. More than 25,000 affected individuals have been reported. High prevalence of β-thalassemia (approximately 10%) occurs in Northern and Southern provinces, near to the Caspian Sea and Persian Gulf, respectively. The prevalence of β-thalassemia alleles in other parts of Iran has been estimated to be 4–8% [4]. In Kermanshah province the relative frequency of β-thalassemia is approximately 3% and according to the report of the Health Center of Kermanshah there are 280 β-thalassemia major registered patients.
⁎ Corresponding author. Medical Biology Research Center, Kermanshah University of Medical Sciences, Medical School, Daneshgah Avenue, Kermanshah, Iran. Fax: 0098 831 4276471. E-mail addresses: [email protected], [email protected] (Z. Rahimi). 1079-9796/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bcmd.2008.12.001
Orkin et al. [5] found strong linkage disequilibrium between polymorphic sites across the β-globin gene cluster (haplotypes) and β-thalassemia mutations. A limited number of haplotypes are found in each population, and each haplotype is usually associated with one specific type of thalassemia. When mutation analysis fails, haplotype analysis is useful in identifying a β-thal chromosome, with an accuracy of about 90% if the family is informative [6]. β-globin gene cluster haplotypes are useful in diagnosis of particular molecular defects in β-thalassemia, prenatal diagnosis of β-thalassemia, and elucidating population affinities [7]. Population of Iran consists of multiethnic groups, so each region of the country has their own distinct set of mutations. To our knowledge there is no report related to the haplotype/mutation association in Western Iran. This is the first paper to report the analysis of association of β-globin gene cluster haplotypes with β-thalassemia mutations among population of Western Iran with the Kurdish ethnic background to identify the origin and historical background of these mutations in this area. Materials and methods The patient group was comprised of 157 unrelated β-thalassemia major patients (87 males and 70 females) who referred to the Shahid
Z. Rahimi et al. / Blood Cells, Molecules, and Diseases 42 (2009) 140–143
Fahmideh Hospital of Kermanshah University of Medical Sciences for blood transfusion. The control group consisted of 35 healthy individuals with normal blood indices from the same area. Informed written consent was obtained from each individual before participation. The study was approved by Ethics Committee of Kermanshah University of Medical Sciences and was in accordance with the principles of the Declaration of Helsinki II. The levels of Hb A2 and HbF were determined by DEAE-52 microcolumn chromatography and alkaline denaturation [8,9], respectively. DNA was extracted from leukocytes of whole blood by phenolchloroform method [10]. For genotyping of variations, genomic DNA from patients was analyzed by PCR-ARMS and RFLP as previously described [11]. Unknown mutations were identified by direct genomic sequencing using an ABI 3730 DNA analyzer (Applied Biosystems, Foster City, Ca). Haplotypes were determined by analyzing the pattern of seven restriction sites through the β-globin gene cluster using primers reported by Old [12] and Lee et al. [13]. The polymorphic sites were, 5′ to ε gene by Hind II, within IVS II of the Gγ and Aγ genes by Hind III, within and 3′ to ψβ by Hind II, IVS 2 of the β gene by Ava II, and 3′ to the β gene by BamH I. To determine the heterozygous haplotypes in β-thal and βA chromosomes, it was assumed that the heterozygous individual possessed one common haplotype and a rare haplotype rather than two rare different haplotypes. The haplotype classification of Orkin [5] has been adopted.
We performed haplotype analysis using a total of 314 chromosomes and identified 17 different haplotypes. In order to define on which allele the polymorphic sites of the β-globin gene cluster is located the probands and both parents need to be analysed. Since the parents were not always available we studied 83 patients (166 chromosomes) homozygous for the same mutation. As presented in Table 1, haplotype I was the most prevalent haplotype (35.7%) among β-thal chromosomes followed by haplotypes III (28.6%), VII (10.2%), V (7.3%), IX (3.5%), A3 (3.2%), VI (2.9%), II (1.6%), IV (1%) and haplotypes VIII, X, A, and D each with 0.6%. Atypical haplotypes accounted for 3.6% of the total haplotypes.
Table 1 β-globin gene cluster haplotypes linked to β-thalassemia and βA chromosomes in Western Iran Haplotype I II III IV V VI VII VIII IX X A3 A D
II atypical Total
Table 2 Haplotypes associated with the same mutation Mutation (N)⁎
Haplotype (N)
IVSII.1 G:A (56)
III (37) V (8) Atypical (6) I (3) X (1) II atypical (1) I (27) V (1) I (12) III (2) V (1) A (1) VII (6) III (2) IX (2) I (4) Atypical (2) IX (1) A (1) I (8) VII (5) II atypical (2) D (1) I (5) VII (1) I (5) V (1) I (4) II (1) VIII (1) A3 (2) VI (2) A3 (2) III (2) IX (2) III (4)
CD 8/9 + G (28) CD 36/37-T (16)
CD 15 TGG: TAG (10)
IVSI.110 G:A (8)
IVSI.1 G:A (8) CD 39 C:T (8)
CD 44-C (6) IVSI.110 C:T (6) IVSI.5 G:C (4) CD 8-AA (4)
Results
+−−−−++ −++−+++ −+−+++− −+−++−+ +−−−−+− −++−−−+ +−−−−−+ −+−+−+− −+−++++ −+−−−−+ −++−+−+ −−−−−++ −−−−−−+ −++++++ −+−+++ −+++++− −++ + + + + − − − − − + ++ −++−−+− −++−−++ −++−++−
βA (%)
βThal (%)
30 5 4 – 10 – 3 – 1 – 1 – – 5 3 3
112 (35.7) 5 (1.6) 90 (28.6) 3 (1) 23 (7.3) 9 (2.9) 32 (10.2) 2 (0.6) 11 (3.5) 2 (0.6) 10 (3.2) 2 (0.6) 2 (0.6) 4 (1.3) – – 1 (0.3) – 3 (1) 3 (1) – 314
(42.9) (7.1) (5.7) (14.3) (4.3) (1.4) (1.4)
(7.1) (4.3) (4.3)
2 (2.9) 2 (2.9) 1 (1.4) 70 (100)
The order of the sites from 5 to 3 is as follows: Hind II 5′ to ε gene, Hind III within IVS 2 of the Gγ and Aγ genes, Hind II within and 3′ to ψβ gene, Ava II within IVS 2 of the β-globin gene, and BamH I 3′ to the β-globin gene.
141
IVSII.745 C:G (4) IVSI.108-25 (4) CD 22/23/24 del (4) N:number of chromosomes.
Distribution and frequency of β-globin gene cluster haplotypes for βA chromosomes are also depicted in Table 1. βA chromosomes similar to β-thal chromosomes were linked to diverse haplotypes but predominantly with haplotype I (42.9%). The magnitude of these frequencies, for other haplotypes including atypical, is also presented in Table 1. From the twenty different mutations that were analyzed, the mutation IVSII.1 (G → A) accounted for approximately 33% of studied thalassemic alleles followed by CD8/9 +G (13.51%). The results of scrutiny of individuals with homozygous mutations (166 chromosomes) and their linkage to corresponding haplotypes are demonstrated in Table 2. Discussion The linkage of β-thalassemia mutation with specific haplotypes has been reported previously [5]. In chromosomes of β-thalassemia patients from Mediterranean area, haplotype I of the β-thal gene was found to be the most prevalent haplotype (47%) followed by haplotypes II (17%), V (12%), III (8%), VI (6%), VII (6%), IX (3%), IV (1%) and VIII (1%) (Orkin et al., 1982). In the present study haplotype I was the most prevalent haplotype among β-thalassemia patients (35.7%), and normal individuals (42.9%). Previously we reported that haplotype I was the most prevalent among β-thalassemia minor individuals and normal controls from Southern Iran (46.2% and 43.3%, respectively) [14]. However, in the present study the second prevalent haplotype was haplotypes III (28.6%) and V (14.3%) for patients and controls, respectively. Interestingly, in Southern Iran the second prevalent haplotype among patients was haplotype V (15.4%) and in normal
142
Z. Rahimi et al. / Blood Cells, Molecules, and Diseases 42 (2009) 140–143
individuals was haplotype III (15.4%) [14]. Since haplotype I is the most prevalent haplotype in β-thal and βA chromosomes, it implies that β-thalassemia mutations might have arisen from the chromosomal background common in the population, rather than due to selection pressure or gene flow (migration). Recently, in screening 370 chromosomes from β-thalassemia patients of Kermanshah Province we identified twenty different mutations. The most common mutation was IVSII-1 (G → A) (32.97%) followed by, in order of frequency, CD8/9 +G (13.51%), IVSI-110 (C → T) (8.38%), CD 36/37-T (7.84%), CD8-AA (5.94%), CD15 (G → A) (4.86%), IVSI.1 (G → A) (4.59%), IVSI.6 (T → C) (3.78%), CD39 (C → T) (2.97%), IVSII.745 (C → G) (2.43%), CD44-C (2.43%), IVSI.5 (G → C) (2.16%), IVSI.108-25 bp (2.16%), CD82/83-G (1.08%), and CD 22/23/24 (−AAGTTGG) (1.08%). Rare mutations (b1%) were IVSII.2,3 (+ 11 − 2) (0.54%), IVSI.128 (T → G) (0.54%), CD6-A (0.27%), CD37 (G → A) (0.27%), and CD 9/10 + T (0.27%) (data not presented). The frequency of IVSII.1 (G → A) is high in Middle East (7–47%). This mutation is linked to various β haplotypes (I, III, IX or atypical) and results in a disease of variable severity [15]. The IVS II.1 (G → A) mutation has been reported as the most prevalent mutation among Iranians [16]. In the patient sample analyzed on the present study the prevalence of this mutation was 33%. This mutation was strongly linked to haplotype III (66.1%). However, it was also found on chromosomes with the haplotypes I, II, V, X and atypical. In β-thalassemia patients from the Mediterranean countries this mutation is associated with haplotypes III and V [17]. In contrast to our study Yavarian et al. [18] in small sample from South Western Provinces of Hormozgan (presumably Persian and Arab in origin) and Fars (Farsi speakers and presumably Persian in origin.) found that most of the patients with IVS II.1 (G → A) mutation (80%) had haplotype I and 20% of patients with this mutation had haplotype III. These differences could be attributed to the different ethnic backgrounds of patients in two studies. It has been suggested that a variety of gene conversion and recombination mechanisms could be responsible for heterogeneity of mutation/haplotype associations [15]. Haplotypes associated with IVSII.1 (G → A) mutation are also represented among the normal individuals indicating that the mutation arose on existing chromosomal background. The association between haplotype and mutation is not invariant, because the same haplotype, when present in two different ethnic groups, has been associated with different mutations. Also, mutation spread from one haplotype to another within a racial group has been found [3]. Very mild to moderate phenotypes are virtually found only among individuals who are homozygotes for the mutation in cis XmnI+ of haplotype III. In our study patients with haplotype background III and IX were XmnI positive. The presence of an XmnI cleavage site (+) causes a significant delayed decline of Hb F production in infant age, a high propensity of postnatal γ-globin expression and ameliorating the clinical phenotypes [19]. In Mediterranean area the polymorphism XmnI indicates linkage disequilibrium with haplotypes III, IV, and IX [20]. The CD 8/9 (+G) in Lebanon was linked to haplotype I [21] and in India to 5′ subhaplotype + −−− [22]. This mutation in our population indicates a strong association with haplotype I (96.4%) and a weak association with haplotype V (3.6%). Haplotype V has the same subhaplotype (+ − − − − +) as haplotype I. In India IVSI.5 (G → C) was linked to 5′ subhaplotype + − (found in haplotypes I, V and VII) [22] and to haplotypes VII, and X [17]. In present study this mutation was associated to haplotype I. In Lebanese population this mutation was associated with haplotype VII [20]. The frameshift 36/37 (−T) mutation that has been found in South Western province of Iran (Khuzistan) and also among Kurdish Jewish patients was found to be the fourth prevalent mutation (7.84%). In our studied patients with Kurdish ethnic origin this mutation was mostly associated with haplotype I (75%) and with lesser frequency was linked to haplotypes III, V, and A (Table 2). Among Kurdish Jews this mutation was associated with haplotype I [23].
In Mediterranean CD 39 (C → T) mutation was linked to haplotypes I, II, VII and haplotype IX [17]. In our patients, the CD 39 (C → T) was found with haplotype VII (62.5%) and also was associated with haplotypes II atypical and D. Similarly among Kurdish Jews this mutation was linked to haplotype VII. In Kurdistan of Iran 58.3% of the nonsense CD 39 mutation had haplotype background VII [23]. Conversely, the mutation was linked to haplotypes I, II in Lebanon, with haplotypes I, II, IX in Algeria and with haplotypes II, IV in Turkey [2]. It seems that the ancestral background for the frameshift mutations 36/37 (−T) and CD 39 (C → T) are haplotype I and VII, respectively. In Lebanon and among Kurdish Jews haplotype I was associated with IVS I.110 (G → A) mutation [2,23]. In the present study IVS I.110 (G → A) mutation was linked to haplotype I in 50% of patients and the remaining patients had haplotype backgrounds II, atypical, IX, and A (Table 2). In the Fars province of Iran this mutation was linked to haplotype I [14]. The IVS I.110 (G → A) mutation is tightly linked to haplotype I in Algeria (92%) and Turkey (93%). However a highest diversity of haplotypes linked to IVSI.110 (G → A) was observed in Turkey [2]. In Mediterranean area this mutation was associated with haplotypes I, II, and IX [20]. The CD 15 (G → A) mutation was associated with haplotype VII in 60% of chromosomes and the remaining 40% was linked to haplotypes III, IX. The mutation IVSII.745 (C → G) in Mediterranean area was linked to haplotype VII [17]. However, in the present study this mutation equally linked to haplotypes VI and A3 (Table 2). In Lebanese population the mutation IVSI.108-25 was found to be linked to haplotype IX [21] whereas, in our population this mutation is linked to haplotypes III and IX with equal frequency. Haplotype IX includes the same 5′ subhaplotype (− + − + + +) as haplotype III. The Mediterranean β-thalassemia patients with IVSI.1 (G → A) mutation are linked to haplotype V [17] like in the majority of the Lebanese β-thalassemia chromosomes, although, in one case linkage of chromosome to haplotype IX was detected. All β-thalassemia chromosomes of our study were in linkage disequilibrium with haplotype I. The FSC 8 (−AA) mutation in our study existed in three different haplotypes, II, VIII, and A3 while in the Mediterranean area this mutation is associated with haplotypes IV, VI and VII [20,21]. It has been demonstrated that a single mutation can be appear in association with different haplotypes. It has also been suggested that the association between specific haplotypes and a specific thalassemic mutation could be very high (85%). In fact the same haplotype could be associated with different genetic defects. In general, involvement of the same mutation to more than one haplotype is most consistent with the crossing-over of 5′ into the β-globin cluster, presumably within the region of increased recombination [17]. The linkage of CD44-C to haplotype I have been previously reported in the Kurdish population [23]. In Western Iran, 83.3% of chromosomes with this mutation that are present in individuals with Kurdish ethnic background were linked also to this haplotype. The minor association of CD44-C with haplotype VII observed in our study could be due to a meiotic recombination starting from haplotype I. CD44-C in two chromosomes from Lebanon was found on haplotypes I and on 5′ haplotype X, respectively [21]. The CD −22/23/24 deletion was in linkage disequilibrium with haplotype III. We found that several mutations can occur within a single haplotype. However, the association of one specific mutation to one haplotype is the most predominant. Conclusion Finding the most common mutations on different haplotype backgrounds can be explained by a variety of gene conversion and recombination events. However, almost all mutations have arisen on
Z. Rahimi et al. / Blood Cells, Molecules, and Diseases 42 (2009) 140–143
the same chromosomal background of normal population. Also, we demonstrated that haplotype background for Kurdish mutations among our studied population with ethnic background of Kurd is similar to those among Kurdish Jews and people of Kurdistan of Iran. References [1] M. Knott, K.M. Ramadan, G. Savage, et al., Novel and Mediterranean beta thalassemia mutations in the indigenous Northern Ireland population, Blood Cells Mol. Dis. 36 (2006) 265–268. [2] L. Zahed, J. Demont, R. Bouhass, et al., Origin and history of the IVSI.110 and codon 39 beta thalassemia mutations in the Lebanese population, Hum. Biol. 74 (2002) 837–847. [3] C. Wong, S.E. Antonarkis, S.C. Goff, et al., On the origin and spread of βthalassemia: recurrent observation of four mutations in different ethnic groups, Proc. Natl. Acad. Sci. U. S. A. 83 (1986) 6529–6532. [4] Z. Rahimi, A. Vaisi Raygani, A. Merat, et al., Thalassemic mutations in Southern Iran, Ir. J. Med. Sci. 31 (2006) 70–73. [5] S.H. Orkin, H.H. Kazazian Jr, S.E. Antonarakis, et al., Linkage of β thalassemia mutations and β globin gene polymorphism with DNA polymorphism in human β globin gene cluster, Nature 296 (1982) 627–632. [6] B.G. Forget, H.A. Pearson, Hemoglobin synthesis and the thalassemias, in: R.I. Handin, S.E. Lux, T.P. Stossel (Eds.), Blood: Principles and Practice of Hematology, J.B. Lippincott Company, Philadelphia, 1995, pp. 1525–1590. [7] R.L. Nagel, H.M. Ranney, Genetic epidemiology of structural mutations of the βglobin gene, Semin. Hematol. 27 (1990) 342–359. [8] V.F. Fairbanks, G.G. Klee, Biochemical aspects of haematology, in: C.A. Burtis, E.R. Ashwood (Eds.), Tietz Text Book of Clinical Chemistry, W.B Saunders, Philadephia, 1994, pp. 2041–2042. [9] K. Betke, H.R. Marti, I. Schlicht, Estimation of small percentages of foetal haemoglobin, Nature 184 (1959) 1877–1878.
143
[10] J.M. Old, D.R. Higgs, Gene analysis, in: D.J. Weatherall (Ed.), Methods in Hemotology. Vol:6. The Thalassemias, Livingstone, 1983, pp. 74–101. [11] J.M. Old, N.Y. Varawalla, D.J. Weatherall, Rapid detection and prenatal diagnosis of beta-thalassaemia: studies in Indian and Cypriot populations in the UK, Lancet 386 (1990) 834–837. [12] J.M. Old, Hemoglobinopathies, in: R. Elles (Ed.), Methods in Molecular Medicine: Molecular Diagnosis of Genetic Disease, Humana press Inc, Totowa, NJ, 1996, pp. 169–183. [13] Y.J. Lee, S.S. Park, J.Y. Kim, H.L. Cho, RFLP haplotypes of β-globin gene complex of β-thalassemic chromosomes in Koreans, J. Korean Med. Sci. 17 (2002) 475–478. [14] Z. Rahimi, A. Merat, M. Akhzari, et al., β-Globin gene cluster haplotypes in Iranian patients with β-Thalassemia, Int. J. Hematol. Oncol. 2 (2005) 30–34. [15] G.O. Tadmouri, N. Garguier, J. Demont, et al., History and origin of β-thalassemia in Turkey: sequence haplotype diversity of β-globin genes, Hum. Biol. 73 (2001) 661–674. [16] H. Najmabadi, R. Karimi-Nejad, S. Sahebjam, et al., The beta-thalassemia mutation spectrum in the Iranian population, Hemoglobin 25 (2001) 285–296. [17] S.E. Antonarakis, H.H. Kazazian Jr, S.H. Orkin, DNA polymorphism and molecular pathology of the human globin gene cluster, Hum. Gene 69 (1985) 1–14. [18] M. Yavarian, M. Karimi, E. bakker, et al., Response to hydroxyurea treatment in Iranian transfusion-dependent β-thalassemia patients, Haematologica 89 (2004) 1172–1178. [19] V. Viprakasit, W. Chinchang, L. Suwanthol, V.S. Tanphaichitr, Common origin of a rare beta-globin initiation codon mutation (ATG → AGG) in Asian, Clin. Lab. Haematol. 27 (2005) 409–415. [20] W. Lemsaddek, I. Picanco, F. Seuanes, et al., The β-thalassemia mutation/haplotype distribution in the Moroccan population, Hemoglobin 28 (2004) 25–37. [21] L. Zahed, M. Qatanani, M. Nabulsi, A. Taher, β-thalassemia mutations and haplotype analysis in Lebanon, Hemoglobin 24 (2000) 269–276. [22] A. Bandyopadhyay, S. Bandyopadhyay, M.D. Chowdhury, U.B. Dasgupta, Major beta-globin gene mutations in Eastern India and their associated haplotypes, Hum. Hered. 49 (1999) 232–235. [23] D. Rund, T. Cohen, D. Filon, et al., Evolution of a genetic disease in an ethnic isolate: β-thalassemia in the Jews of Kurdistan, Proc. Natl. Acad. Sci. 88 (1991) 310–314.