Hemoglobin, 2012; 36(4): 371–380Copyright © Informa Healthcare USA, Inc.ISSN: 0363-0269 print/1532-432X onlineDOI: 10.3109/03630269.2012.691147

ORIGINAL ARTICLE

THE XmnI AND BCL11A SINGLE NUCLEOTIDE POLYMORPHISMS

MAY HELP PREDICT HYDROXYUREA RESPONSE IN IRANIAN

β-THALASSEMIA PATIENTS

Mehdi Banan,1 Hadi Bayat,1 Azita Azarkeivan,2,3

Saeid Mohammadparast,1 Koorosh Kamali,4 Samaneh Farashi,1

Nooshin Bayat,5 Masumeh Hadavand Khani,3 Maryam Neishabury,1 and

Hossein Najmabadi1,5

1Genetics Research Center, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran2Blood Transfusion Organization, Tehran, Iran3The Adult β-Thalassemia Clinic, Tehran, Iran4Reproductive Biotechnology Research Center, Avicenna Research Institute, Tehran, Iran5Kariminejad-Najmabadi Pathology & Genetics Center, Tehran, Iran

� Hydroxyurea (HU), a drug which can reactivate fetal hemoglobin (Hb F) production, is frequentlyprescribed to β-thalassemia (β-thal) patients. However, transfusion requirements of only a subset ofpatients are reduced upon HU treatment. Because of its potential side-effects, targeted prescription ofHU is imperative. To identify genetic markers that correlate with drug response, we have carried out aretrospective association study of single nucleotide polymorphisms (SNPs) in three Hb F quantitativetrait loci (QTLs): the XmnI polymorphism, BCL11A, and the HBS1L-MYB intergenic region, withthe response to HU in a cohort of 81 transfusion-dependent Iranian β-thal patients. An increase inblood transfusion intervals post-therapy was used to measure drug response. Our results suggest thatpresence of theXmnI T/T genotype or the BCL11A rs766432 C allele correlates strongly with responseto HU (p <0.001). Accordingly, these markers may be used to accurately predict the HU response ofIranian β-thal patients.

Keywords β-Thalassemia (β-thal), Hydroxyurea (HU), BCL11A Single nucleotide poly-morphism (SNP), XmnI Polymorphism

Received 28 December 2011; Accepted 24 February 2012.Address correspondence to Dr. Mehdi Banan, Genetics Research Center, University of Social Welfare

and Rehabilitation Sciences, Evin, Daneshjoo Blvd., Koodakyar St., Tehran, Iran; Tel.: þ9821-22180106;Fax: þ9821-22180138; E-mail: mbbanan@yahoo.com

371

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

INTRODUCTION

Hydroxyurea (HU) is the only drug approved by the US Food and DrugAdministration (FDA) for the treatment of β-thalassemia (β-thal) patients.Hydroxyurea treatment can confer clinical benefits to β-thal major (β-TM),β-thal intermedia (β-TI) and Hb E/β0-thal patients (1). In particular, HUadministration may increase steady-state Hb F levels and consequently totalhemoglobin (Hb) levels in β-thal patients. This occurrence may, in turn,account for the reduction in transfusion requirements of a large proportionof these patients following HU treatment.

While clinically beneficial, the administration of HU in β-thal patients alsohas its drawbacks (2). Hydroxyurea is a chemotherapeutic agent with poten-tial side effects. There is concern that HU therapy may increase the risk ofleukemia, leg ulcers and spermatogenesis defects in sickle cell anemia patients(3). Hydroxyurea may also increase the frequency of acquired somatic muta-tions (4). In addition, not all β-thal patients respond favorably to HUtreatment.

Because of its potential side-effects and its efficacy in only a subset of β-thalpatients, targeted prescription of HU is of paramount importance.Unfortunately, genetic markers that accurately predict the response to HUin β-thal patients have not yet been reported. Several genes involved in theHU-mediated Hb F induction pathway have been identified (5,6). Singlenucleotide polymorphisms (SNPs) in a number of these genes correlatewith increased Hb F levels in sickle cell anemia patients. However, the associa-tion of these SNPs with HU response in β-thal patients has not beenestablished.

Recently, genome-wide association studies (GWAS) and follow-up replica-tion studies have identified three quantitative trait loci (QTL) that can influ-ence Hb F and F-cell levels in healthy persons and patients withhemoglobinopathies (7–11). In the β-globin locus, an SNP of unknown func-tional significance at position �158 (C>T) of the Gγ promoter (referred to asthe XmnI polymorphism) has been widely associated with Hb F and F-celllevels (12). Another QTL is located in intron 2 of the BCL11A gene, encodinga transcription factor that mediates γ- to β-globin switching during erythroiddevelopment (13,14). A third locus is situated in the intergenic region of theHBS1L andMYB genes. There is evidence suggesting that theMYB oncogenemay function as a γ-globin repressor (15). Notably, a small number of SNPs inthese loci account for 20–50% of the variance in Hb F levels in diversepopulations (16).

Previous studies suggest a correlation between the XmnI T/T genotypewith HU response, and the XmnI C/C genotype with a lack of response(17–19). An association between HU response and SNPs in the BCL11A andHBS1L-MYB loci, however, has not been reported. We hypothesized that such

372 M. Banan et al.

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

a correlation might exist. Firstly, these loci affect baseline Hb F levels, andsome reports suggest that higher pre therapy Hb F levels may lead to a morefavorable HU response (20,21). Furthermore these QTLs may influence theseverity of disease, and there is evidence to suggest that β-TI patients mayrespond more positively to HU treatment than β-TM patients (7,19,22–26).

In a retrospective association study, we examined correlation of the XmnIpolymorphism (rs7482144), SNPs in BCL11A intron 2 (rs766432, rs4671393),and SNPs in the HBS1L-MYB intergenic region (rs9399137, rs4895441) withthe response to HU in a cohort of 81 Iranian transfusion-dependent β-thalpatients. The selection of these SNPs was based on their strong correlationwith Hb F and F-cell levels in African, Asian, Caribbean and European popula-tions (16).

MATERIAL AND METHODS

Cohort Description

Over 400 β-thal cases referred to The Adult β-Thalassemia Clinic, Tehran,Iran were examined. A rise in steady-state Hb levels (1–2 g/dL) is frequentlyused as basis for HU response in transfusion-independent β-thal patients,while an increase in blood transfusion intervals is deployed to assess HUresponse in transfusion-dependent patients (17–21,26,27). In order to use asingle criterion for drug response, we focused on transfusion-dependentβ-thal patients. Of the >400 patients examined, 89 were transfusion-dependent, had undergone HU treatment and possessed a clinical file withperiodical (monthly or sometimes bi-monthly) entries for the doses of drugstaken (hydroxyurea, deferoxamine and folic acid); periodical platelet, whiteblood cell (WBC), and Hb levels; and dates of transfusion administrationsduring the course of treatment.

Of these patients, seven had a single �α3.7 (rightward) deletion and onewas heterozygous for the Hb Constant Spring [Hb CS, α142, Term!Gln,TAA>CAA (α2)] mutation. In sickle cell anemia patients, co-inheritance ofα-thal reduces the clinical response to HU therapy (28). The eight patientswith α-globin mutations were therefore excluded from the study. The finalcohort of 81 patients consisted of 42 males and 39 females with an average ageof 31 � 9 years. The ethnic background of the patients was heterogeneous,thus minimizing the effect of ethnicity-based stratification. There were 54β-TM and 27 β-TI patients.

Based on alterations in blood transfusion intervals after treatment, threeHU responder classes could be identified: good responders (GR), minorresponders (MR), and non responders (NR) (Supplementary Information –

available online). Good responders were defined as patients who weretransfusion-dependent but became transfusion-independent after treatment.

Hb F QTLs Associated With HU Response in β-Thal 373

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

Minor responders were defined as patients whose transfusion intervals wereincreased by >2-fold after HU treatment. Patients in this group encompasseda spectrum of responses. After treatment, some patients had attained a 2–3fold increase in their transfusion intervals, while others had shifted fromperiodical to sporadic transfusions. Non responders were classified as patientswho did not show any appreciable change (<2-fold) in blood transfusionintervals after HU administration. Using the above criteria, 37 patients werelabeled as GRs, 24 as MRs and 20 as NRs.

Genotyping

The α- and β-globinmutations, the three most common α-globin deletions[–α3.7, polyadenylation signal site (polyA2) (AATAAA>AATGAA; αpolyA2),�α4.2 (leftward)] and the XmnI polymorphism of most patients were diag-nosed prior to the study (29).Mutations in the α1-, α2- and β-globin genes weredetermined by DNA sequencing. Deletions on the α-globin gene were diag-nosed by multiplex ligation-dependent probe amplification (MLPA) (MRC-Holland, Amsterdam, The Netherlands). The study was approved by theUniversity of Social Welfare and Rehabilitation Sciences Ethics Committee,Tehran, Iran. Written consent was obtained from patients prior to bloodsampling. DNA was extracted from 5 mL whole blood in EDTA-containingcollection tubes using the salting-out method (30). DNA was dissolved in100–200 μL nuclease-free water and 1 μL (containing 100–200 ng DNA) wasused for each diagnostic procedure. The XmnI polymorphism was identifiedby polymerase chain reaction-restriction fragment length polymorphism(PCR-RFLP). Diagnosis of the other SNPs was carried out by using amplifica-tion refractory mutation system (ARMS)-PCR (rs4671393, rs4895441 andrs9399137) or tetra-primer ARMS-PCR (rs766432). The ARMS-PCR primerswere designed using a freely available algorithm (31). Samples were run on8%polyacrylamide gels and visualized by silver staining. For each SNP, controlDNA samples [MM, Mm, and mm (where M: major allele and m: minorallele)] were PCR-amplified and co-electrophoresed on each gel. The accu-racy of diagnosis was verified by sequencing 10% of the DNA samples. Primersequences are shown in Table 1. The PCR conditions will be made availableupon request.

Statistical Analysis

Statistical analyses were performed using the OpenEpi Version 2.3.1 opensource software. To simplify the analyses, patients were classified as respon-ders (GR þ MR) or non responders (NR). Subsequently, the presence of theXmnI T/T genotype (recessive model) or the BCL11A andHBS1L-MYBminoralleles (dominant model) were used as a basis to create 2 � 2 contingency

374 M. Banan et al.

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

tables. A chi-square test was performed to obtain p values and p <0.05 wasconsidered significant. Furthermore, odds ratios (OR) were estimated with95% confidence intervals (95% CI).

RESULTS

Distribution of the β-globin mutations and mutation types (β0, βþ andβþþ) in the three subgroups (GR, MR and NR) are shown in Table 2 (32). Noclear pattern in the distribution of mutation types could be detected, with theexception of a clear predominance of patients homozygous for the IVS-II-1(G>A) mutation in the responders. In particular, 70.3, 45.8 and 15.0% of GR,MR, and NR patients, respectively, were IVS-II-1/IVS-II-1. This pattern is likelydue to linkage of the IVS-II-1 mutation with the XmnI T allele in the Iranianpopulation (33,34).

Allele frequencies of the XmnI polymorphism, BCL11A and HBS1L-MYBSNPs in each of the subgroup are summarized in Table 3. The HBS1L-MYBmarkers (rs9399137, rs4895441) did not show significant association with HUresponse. Conversely, the XmnI polymorphism (rs7482144) T/T genotypeshowed significant association with the response to HU (p <0.001).Interestingly, the presence of the BCL11A SNP minor alleles (rs766432 C

TABLE 1 Primer Sequences Used For Amplification and Detection of the Indicated SNPs (and alleles)Are Shown

SNP (allele) Primer Sequences (5’>3’)

rs7482144 Forward: AAC TGT TGC TTT ATA GGA TTT TReverse: AGG AGC TTA TTG ATA ACT CAG AC

rs766432 (A) Forward: TTG TTT CGC TTT AGC TTT ATT AAG GTA CAAReverse: GAC GTG TTC TGT ATC TTG ATT TTG GT

rs766432 (C) Forward: CCA AAC AGT TTA AAG GTT ACA GAC AGA CTReverse: AAA ATG AAT GAC TTT TGT TGT ATG TAG AG

rs4671393 (A) Forward: CTG TGG ACA GCA AAG CTG CAReverse: TCT CCC CCT TGC ATT GTT GTC

rs4671393 (G) Forward: CCC CCA CTA GCT CAG AAA TGG AReverse: GGG AAT CTT AAT TTC CTG CCC C

rs4895441 (A) Forward: CTG GGG AAA GAC TCT TTG TAA AGG GAReverse: GCT GCC ATT TTT TTC TTT CTC CCT ATA A

rs4895441 (G) Forward: TTG GGA TAT AGG CCA TAG ACA AAA ATC CReverse: TCA CTT ACT CAG TTC TCT GCT CAT GGA C

rs9399137 (C) Forward: AAT GTA ATT AAC TGA ACA TAT GGT TAG TCReverse: TTT ATT GTT ACA AGG TTA ATT CAC TGC C

rs9399137 (T) Forward: GAA ATA CCA TCA CTG AGA AAA GCA TAA GReverse: CAG CAG GGT TGC TTG TGA AAA AAC TT T A

The primers were deployed for PCR-RFLP (XmnI polymorphism), ARMS-PCR (rs4671393, rs4895441 andrs9399137), or tetra-primer ARMS-PCR (rs766432) as discussed in Materials and Methods. In the ARMS-PCR primers, mismatched nucleotides are in italics (31).

Hb F QTLs Associated With HU Response in β-Thal 375

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

TABLE 3 Frequencies of the Indicated Alleles For Each Single Nucleotide Polymorphism in the Subgroups(GR, MR, and NR) Are Shown

Allele Frequency

Locus SNP AlleleCohort(n ¼ 81)

GR(n ¼ 37)

MR(n ¼ 24)

NR(n ¼ 20) p Value OR (95% CI)

β-Globin rs7482144 T 0.65 0.82 0.63 0.38 <0.001 14.9 (3.2–70.1)BCL11A rs766432 C 0.20 0.28 0.23 0.03 0.001 15.1 (1.9–120.0)

rs4671393 A 0.15 0.22 0.15 0.03 0.008 10.7 (1.3–85.6)HBS1L-MYB rs9399137 C 0.11 0.16 0.04 0.10 0.975 NA

rs4895441 G 0.22 0.20 0.27 0.20 0.381 NA

NA: not applicable.Also given are the p values, odds ratios and 95% confidence intervals that indicate differences in the

distribution of alleles in hydroxyurea responders and non responder patients (see Materials andMethods).

TABLE 2 Distribution of β-Globin Mutations in the Good Responder, Minor Responder and NonResponder Subgroups

Mutation Type GR (n ¼ 37) MR (n ¼ 24) NR (n ¼ 20)

IVS-II-1(G>A)/IVS-II-1(G>A) β0/β0 26 11 3IVS-II-1(G>A)/FSC 8/9(þG) β0/β0 3 1 0ND/ND �/� 3 0 1FSC 25/26(þT)/FSC 25/26(þT) β0/β0 1 0 0IVS-I-110(G>A)/codon 8(–AA) βþ/β0 1 0 0IVS-I-5(G>C)/IVS-I-5(G>C) β0/β0 1 0 0IVS-II-1(G>A)/FSC 74/75(–C) β0/β0 1 0 0IVS-II-1(G>A)/codon 8(–AA) β0/β0 1 0 0codon 39(C>T)/codon 39 (C>T) β0/β0 0 1 1IVS-I-6(T>C)/IVS-I-6(T>C) βþþ/βþþ 0 1 1IVS-I-110(G>A)/IVS-I-110(G>A) βþ/βþ 0 1 1IVS-II-1(G>A)/IVS-I-110(G>A) β0/βþ 0 1 1codon 22(A>C)/codon 22(A>C) β0/β0 0 1 0codon 30(G>C)/codon 22(A>C) β0/β0 0 1 0FSC 8/9(þG)/IVS-II-745(C>G) β0/βþ 0 1 0IVS-II-1(G>A)/codon 39(C>T) β0/β0 0 1 0IVS-I-110(G>A)/codon 5(–CT) βþ/β0 0 1 0IVS-I-110(G>A)/ND βþ/β– 0 1 0IVS-II-1(G>A)/IVS-I-130(G>C) β0/β0 0 1 0IVS-II-1(G>A)/IVS-I-5(G>C) β0/β0 0 1 0codon 30(G>C)/codon 30(G>C) β0/β0 0 0 2IVS-I-110(G>A)/IVS-I, 25 bp deletion βþ/β0 0 0 2IVS-II-1(G>A)/ND β0/β– 0 0 2codon 30(G>C)/FSC 36/37(–T) β0/β0 0 0 1codon 5(–CT)/IVS-I-1(G>A) β0/β0 0 0 1IVS-I-5(G>C)/IVS-I, 25 bp deletion β0/β0 0 0 1IVS-II-1(G>A)/codon 30(G>C) β0/β0 0 0 1IVS-II-1(G>A)/FSC 36/37(–T) β0/β0 0 0 1IVS-II-1(G>A)/codon 44(–C) β0/β0 0 0 1

FSC: frameshift codon(s); ND: not determined.

376 M. Banan et al.

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

and rs4671393 A) that are in strong linkage disequilibrium, also correlatedsignificantly with drug response (p ¼ 0.001 and p ¼ 0.008, respectively). Thiscorrelation was even stronger when the presence of the XmnI T/T genotypeor the BCL11A rs766432 C allele was taken into consideration [p <0.001; OR(95% CI) ¼ 37.5 (8.9–157.6)]. These results indicate that the HU responsecan, to a large extent, be predicted by the presence of either the XmnI T/Tgenotype or the BCL11A SNP minor alleles.

Consequently, a simple binary scoring system may potentially be used byclinicians to distinguish between responder (GR,MR) andnon responder (NR)patients. In particular, a score of 1 is given for the presence of either the XmnIT/T or the rs766432 C marker. The presence of the rs766432 C allele ratherthan the rs4671393 A allele is deployed because of its stronger correlation withHU response. Conversely, a score of 0 is allocated for the lack of either marker.

Using this approach, the three groups showed distinct scoring patterns(Figure 1). In particular, 97.3% (36/37) of GR patients acquired a score of 1.Conversely, only 15.0% (3/20) of NR patients obtained a score of 1 and therest (85.0% or 17/20) acquired a score of 0. In the MR group, 70.8% (17/24)of patients acquired a score of 1. Interestingly, 76.5% (13/17) of MR patientsacquiring a score of 1 had shifted from periodic to sporadic transfusions post-treatment. In contrast, 85.7% (6/7) of MR patients with a score of 0 gainedonly a 2–3 fold increase in blood transfusion intervals after HU treatment(Supplementary Information – available online).

DISCUSSION

In the present study, we have examined the possible association of threeHb F and F-cell QTLs (XmnI polymorphism, BCL11A and HBS1L-MYB) withthe response to HU in a cohort of 81 Iranian transfusion-dependent β-thalpatients. Our results suggest that the presence of the XmnI T/T genotype and

FIGURE 1 Distribution of patients according to the allocated binary score (0 or 1) in each subgroup (GR,MR, andNR) are shown. Scores are based on presence of theXmnI T/T genotype or the rs766432 C allele asdiscussed (see Results and Discussion).

Hb F QTLs Associated With HU Response in β-Thal 377

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

two BCL11A SNPs (rs766432 C and rs4671393 A) predicts the response toHU. Furthermore, using either the XmnI T/T or the rs766432 C markerappears to significantly improve the prediction of HU response.

Association of the XmnI polymorphism with the response to HU has beeninvestigated previously. In a cohort of 54 transfusion-dependent Algerianpatients (45 β-TM, nine β-TI), the XmnI C/C genotype was significantlyassociated with a lack of drug response (17). Furthermore, a study involving45 Iranian β-thal patients (36 β-TM, nine β-TI) suggested that the XmnI T/Tgenotype correlates strongly with the response to HU (19). Similar resultswere obtained in a cohort of 133 Iranian β-TM patients (18).

In the present study, we also detected a prevalence of the XmnI C/Cgenotype in NR patients (p ¼ 0.03). However, since the presence of the XmnIT allele was used as basis for HU prescription, our cohort contains relatively fewpatients with theC/Cgenotype.Ourdata strongly suggests that two copies of theXmnI T allele (T/T) are correlated with a positive HU response (p < 0.001).These results are in agreement with the latter two reports (18,19).

At present, it remains unclear whether one or two copies of the XmnI Tallele (i.e., a T/T or T/C genotype) are needed for a robust HU response.However, drawing unifying conclusions from these reports is difficult. Firstly,the investigators used different criteria for gauging the response to HU. Forexample, some considered a>2-fold increase in blood transfusion intervals as apartial response, while others defined a partial response as a shift from regularto sporadic blood transfusions. Furthermore, the inclusion criteria for patientsin these studies were different. Some of these criteria, such as β-thal subtypeand co-inheritance of α-thal, may affect the response to HU (19,26–28).

In addition, we have established that minor alleles of two BCL11A SNPs(rs766432 C and rs4671393 A) correlate strongly with the response to HU. Tothe best of our knowledge, this is the first report of such a relationship. Thetwo BCL11A SNPs that are located in intron 2, are in strong linkage disequili-brium. However, rs766432, which has a higher minor allele frequency thanrs4671393, appears to be a better predictive marker. In this regard, we are inthe process of identifying the BCL11A haplotypes correlating with HUresponse. These haplotypes may further improve prediction of responderpatients (35). In addition, they may help in identifying functional SNPs thataffect BCL11A expression, and subsequently γ-globin levels.

The effect of the XmnI T/T and BCL11A rs766432 C markers on drugresponse does not seem to be synergistic. Specifically, 75.7% (28/37) of the GRpatients in our cohort carried one of the abovemarkers, whereas only 21.6% (8/37) carried bothmarkers. Moreover, in theMR group 54.2% (13/24) of patientsbear one marker and only 16.7% (4/24) have both markers. Therefore, itappears that the above SNPs influence HU response independently.

It is possible that these markers influence pre therapy Hb F and Hb levels,which in turn determine post-therapy HU response. In support of this notion,pre therapyHb levels in our subgroupswere as follows: GR (9.7� 1.1 g/dL),MR

378 M. Banan et al.

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

(9.1� 1.4 g/dL), NR (9.0� 0.9 g/dL) (Supplementary Information – availableonline). Moreover, thesemarkers may affect the severity of β-thal, whichmay inturnplay apart inHUresponse.Wenoted that thedistributionof β-thal subtypesin the groups followed a clear trend (Supplementary Information – availableonline). In particular, 54.1% (20/37) of patients in the GR group were β-TI,whereas 45.9% (17/37) were β-TM. In the MR group, 25.0% (6/24) wereβ-TI and 75.0% (18/24) were β-TM. Of the β-TI patients in the MR group,83.3% (5/6) bear one of the markers. Conversely, only one NR patient was β-TI(5% or 1/20) and the rest (95.0% or 19/20) were β-TM.

ACKNOWLEGMENTS

We thank Dr. Sjaak Philipsen (Erasmus Medical Center, Rotterdam, TheNetherlands) for his critical comments on the manuscript. This study wassupported by the Genetics Research Center (University of Social Welfare andRehabilitation Sciences, Tehran, Iran) grant 64032 to Mehdi Banan.

Declaration of Interest: The authors report no conflicts of interest. Theauthors alone are responsible for the content and writing of this article.

REFERENCES

1. Testa U. Fetal hemoglobin chemical inducers for treatment of hemoglobinopathies. Ann Hematol.2009;88(6):505–528.

2. Saunthararajah Y, Atweh GF. Induction of fetal hemoglobin in the treatment of sickle cell disease andβ thalassemia. In: Steinberg MH, Forget BG, Higgs DR, Weatherall DJ, Eds. Disorders of Hemoglobin:Genetics, Pathophysiology, andManagement, 2nd ed.NewYork:CambridgeUniversity Press. 2009:745–754.

3. Lanzkron S, Strouse JJ, Wilson R, et al. Systematic review: hydroxyurea for the treatment of adults withsickle cell disease. Ann Intern Med. 2008;148(12):939–955.

4. Hanft VN, Fruchtman SR, Pickens CV, Rosse WF, Howard TA, Ware RE. Acquired DNA mutationsassociated with in vivo hydroxyurea exposure. Blood. 2000;95(11):3589–3593.

5. Ma Q, Wyszynski DF, Farrell JJ, et al. Fetal hemoglobin in sickle cell anemia: genetic determinants ofresponse to hydroxyurea. Pharmacogenomics J. 2007;7(6):386–394.

6. Kumkhaek C, Taylor JG, Zhu J, Hoppe C, Kato GJ, Rodgers GP. Fetal haemoglobin response tohydroxycarbamide treatment and sar1a promoter polymorphisms in sickle cell anaemia. Br JHaematol. 2008;141(2):254–259.

7. Uda M, Galanello R, Sanna S, et al. Genome-wide association study shows BCL11A associated withpersistent fetal hemoglobin and amelioration of the phenotype of β-thalassemia. Proc Natl Acad SciUSA. 2008;105(5):1620–1625.

8. Lettre G, Sankaran VG, Bezerra MAC, et al. DNA polymorphisms at the BCL11A, HBS1L-MYB, andβ-globin loci associate with fetal hemoglobin levels and pain crises in sickle cell disease. Proc Natl AcadSci USA. 2008;105(33):11869–11874.

9. Galarneau G, Palmer CD, Sankaran VG, Orkin SH, Hirschhorn JN, Lettre G. Fine-mapping at threeloci known to affect fetal hemoglobin levels explains additional genetic variation. Nat Genet. 2010;42(12):1049–1051.

10. Sedgewick AE, Timofeev N, Sebastiani P, et al. BCL11A is a major Hb F quantitative trait locus in threedifferent populations with β-hemoglobinopathies. Blood Cells Mol Dis. 2008;41(3):255–258.

11. Nguyen TK, Joly P, Bardel C, Moulsma M, Bonello-Palot N, Francina A. The XmnI Gγ polymorphisminfluences Hemoglobin F synthesis contrary to BCL11A and HBS1L-MYB SNPs in a cohort of 57β-thalassemia intermedia patients. Blood Cells Mol Dis. 2010;45(2):124–127.

Hb F QTLs Associated With HU Response in β-Thal 379

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.

12. Thein SL,Menzel S. Discovering the genetics underlying foetal haemoglobin production in adults. Br JHaematol. 2009;145(4):455–467.

13. Sankaran VG, Menne TF, Xu J, et al. Human fetal hemoglobin expression is regulated by the develop-mental stage-specific repressor BCL11A. Science. 2008;322(5909):1839–1842.

14. Sankaran VG, Xu J, Ragoczy T, et al. Developmental and species-divergent globin switching are drivenby BCL11A. Nature. 2009;460(7259):1093–1097.

15. Jiang J, Best S, Menzel S, et al. cMYB is involved in the regulation of fetal hemoglobin production inadults. Blood. 2006;108(3):1077–1083.

16. Thein SL, Menzel S, Lathrop M, Garner C. Control of fetal hemoglobin: new insights emerging fromgenomics and clinical implications. Hum Mol Genet. 2009;18(R2):R216–R223.

17. Bradai M, Pissard S, Abad MT, et al. Decreased transfusion needs associated with hydroxyurea therapyin Algerian patients with thalassemia major or intermedia. Transfusion. 2007;47(10):1830–1836.

18. Yavarian M, Karimi M, Bakker E, Harteveld CL, Giordano PC. Response to hydroxyurea treatment inIranian transfusion-dependent β-thalassemia patients. Haematologica. 2004;89(10):1172–1178.

19. Alebouyeh M, Moussavi F, Haddad-Deylami H, Vossough P. Hydroxyurea in the treatment of majorβ-thalassemia and importance of genetic screening. Ann Hematol. 2004;83(7):430–433.

20. Dixit A, Chatterjee TC, Mishra P, et al. Hydroxyurea in thalassemia intermedia—a promising therapy.Ann Hematol. 2005;84(7):441–446.

21. Singer ST, Kuypers FA, Olivieri NF, et al. Fetal haemoglobin augmentation in E/β0 thalassaemia:clinical and haematological outcome. Br J Haematol. 2005;131(3):378–388.

22. Galanello R, Sanna S, Perseu L, et al. Amelioration of Sardinian β0 thalassemia by genetic modifiers.Blood. 2009;114(18):3935–3937.

23. Badens C, Joly P, Agouti I, et al. Variants in genetic modifiers of β-thalassemia can help to predict themajor or intermedia type of the disease. Haematologica. 2011;96(11):1712–1714.

24. Danjou F, Anni F, Galanello R. β-Thalassemia: from genotype to phenotype. Haematologica. 2011;96(11):1573–1575.

25. Danjou F, Anni F, Perseu L, et al. Genetic modifiers of β-thalassemia and clinical severity as assessed byage at first transfusion. Haematologica 2012 [Epub ahead of print].

26. Italia KY, Jijina FJ, Merchant R, et al. Response to hydroxyurea in β thalassemia major and intermedia:experience in western India. Clin Chim Acta. 2009;407(1–2):10–15.

27. Koren A, Levin C, Dgany O, et al. Response to hydroxyurea therapy in β-thalassemia. Am J Hematol.2008;83(5):366–370.

28. Vasavda N, Woodley C, Allman M, et al. Effects of co-existing α-thalassaemia in sickle cell disease onhydroxycarbamide therapy and circulating nucleic acids. Br J Haematol. 2011;157(2):249–252.

29. Hadavi V, Taromchi AH, Malekpour M, et al. Elucidating the spectrum of α-thalassemia mutations inIran. Haematologica. 2007;92(7):992–993.

30. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from humannucleated cells. Nucleic Acids Res. 1988;16(3):1215.

31. Ye S, Dhillon S, Ke X, Collins AR, Day INM. An efficient procedure for genotyping single nucleotidepolymorphisms. Nucleic Acids Res. 2001;29(17):e88–e88.

32. Thein SL, Wood WG. The molecular basis of β thalassemia, δβ thalassemia, and hereditary persistenceof fetal hemoglobin. In: Steinberg MH, Forget BG, Higgs DR, Weatherall DJ, Eds. Disorders ofHemoglobin, 2nd ed. New York: Cambridge University Press. 2009:323–356.

33. Karimi M, Yarmohammadi H, Farjadian S, et al. β-Thalassemia intermedia from southern Iran: IVS-II-1(G!A) is the prevalent thalassemia intermedia allele. Hemoglobin. 2002;26(2):147–154.

34. Arab A, Karimipoor M, Rajabi A, Hamid M, Arjmandi S, Zeinali S. Molecular characterization ofβ-thalassemia intermedia: a report from Iran. Mol Biol Rep. 2010;38(7):4321–4326.

35. Lewis CM. Genetic association studies: design, analysis and interpretation. Brief Bioinform. 2002;3(2):146–153.

Supplementary Material Available Online:

Supplementary Table 1 – Good Responders (patients 1–37)Supplementary Table 2 – Minor Responders (patients 38–61)Supplementary Table 3 – Non Responders (patients 62–81)

380 M. Banan et al.

Hem

oglo

bin

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f Su

ssex

Lib

rary

on

10/1

4/12

For

pers

onal

use

onl

y.