1. ORIGINAL ARTICLE
Genetics of fetal hemoglobin in tribal Indian
patients with sickle cell anemia
APARNA A. BHANUSHALI, PRADIP K. PATRA, SMARNIKA PRADHAN, SURAJ S. KHANKA,
SUJATA SINGH, and BIBHU R. DAS
MAHARASHTRA AND CHHATTISGARH, INDIA
India tops the list of countries with sickle cell disease (SCD) with an estimated 44,000
live births in 2010 and a prevalence of 10%–33%. In the present study, the first from
India, we have investigated the effect of genetic variants in the BCL11A, the HMIP
(HBS1L-MYB intergenic polymorphism) locus, in addition to the HBB locus, which
are known to be associated with fetal hemoglobin (HbF) levels, a major modulator
of the disease phenotype. The present study was conducted on 240 individuals
with SCD and 60 with sickle cell trait. Genotyping was performed for the BCL11A
rs11886868 and rs34211119; HMIP rs9399137, rs189600565, rs7776196, rs34778774,
and rs53293029; HBG2 Xmn1 polymorphism rs7482144; and 268C . T HBD promoter
polymorphism. All the 3 quantitative trait loci were associated with HbF levels in
Indian patients with SCD. The highest difference was seen in the Xmn1 single-
nucleotide polymorphism, which accounted for 11% of the trait variance, the
BCL11A rs11886868 for 3.65%, whereas the HMIP rs9399137 for 3.8%. The present study
indicates the BCL11A, HMIP, and b-globin region to be associated with increased HbF
levels in Indian patient. Further interrogation of these genotypes with respect to pain
crisis is warranted in this population, which may help in prognostication, as also a
genome-wide association study, which may help uncover new loci controlling HbF
levels. (Translational Research 2015;-:1–8)
Abbreviations: AI haplotype ¼ Arab-Indian haplotype; ANOVA ¼ Analysis of variance; ARMS-
PCR ¼ Amplification refractory mutation system-polymerase chain reaction; BCL11A ¼ B-cell
lymphoma/leukemia 11A; CSSCD ¼ Co-operative study of sickle cell disease; DNA ¼ Deoxyribo-
nucleic acid; EDTA ¼ Ethylenediaminetetraacetic acid; GWAS ¼ Genome wide association
studies; HbA2 ¼ a-globin gene; HBB ¼ b-globin gene; HBD ¼ d-globin gene; HbF ¼ Fetal hemo-
globin; HbG2 ¼ g-globin gene; HBS ¼ Sickle hemoglobin; HBS1L ¼ HBS1-like translational GTPase;
HMIP ¼ HBS1L-MYB intergenic region; HPLC ¼ High Performance Liquid Chromatography; HWE
¼ Hardy-Weinberg Equilibrium; LD ¼ Linkage disequilibrium; MYB ¼ oncogene; OBC ¼ Other
backward class; PCR ¼ Polymerase chain reaction; QTL ¼ Quantitative trait loci; SC ¼ Sched-
uled caste; SCD ¼ Sickle cell disease; SNP ¼ Single nucleotide polymorphism; SS ¼ Sickle cell dis-
ease patients; ST ¼ Scheduled tribe
From the Research and Development, SRL Ltd, Mumbai,
Maharashtra, India; Department of Biochemistry, Pt. J.N.M.
Medical College Raipur, Raipur, Chhattisgarh, India.
Submitted for publication August 18, 2014; revision submitted
January 5, 2015; accepted for publication January 7, 2015.
Reprint requests: Bibhu R. Das, SRL Ltd, Prime Square Building, Plot
No. 1, S.V. Road, Goregaon (W), Mumbai 400 062, Maharashtra,
India; e-mail: brdas@srl.in.
1931-5244/$ - see front matter
Ó 2015 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.trsl.2015.01.002
1

2. INTRODUCTION
Sickle cell disease (SCD) a common genetic disorder
results in the annual loss of several millions of disability-
adjusted life years, particularly in the developing
world.1,2
In India, the sickle gene frequency is found to
vary from 2% to 34%.3,4
It is highly prevalent in
central India and among the tribal belts in western,
eastern, and southern India.
SCD caused because of a single amino acid substitu-
tion glutamic acid (GAG) to valine (GTG) in the
b-globin gene is a potentially devastating disease.5
The disease expression is highly variable with different
clinical outcomes such as stroke, avascular necrosis, leg
ulcers, priapism, and retinopathy. Thus, although SCD
is a monogenic disorder, at the phenotypic level it is a
multigenic disease.6
Several modifiers of the disease have been elucidated,
which include b-globin cluster haplotypes, a-globin
gene number, and fetal hemoglobin (HbF) expression.7
Of these HbF levels have received the most attention.8-10
HbF as its name implies is the primary hemoglobin
present in the fetus from mid to late gestation.11
HbF
is hereditable12-14
and is the most powerful modulator
of the clinical and hematologic features of sickle cell
anemia. Higher HbF levels were associated with a
reduced rate of acute painful episodes, fewer leg
ulcers, less osteonecrosis, less frequent acute chest
syndromes, and reduced disease severity.
Initial studies mapped a quantitative trait locus (QTL)
controlling F cells in Asian-Indian kindred with b-thal-
assemia to a 1.5 kb locus on chromosome 6q23.15
In
another study, a new F cell QTL was mapped to
BCL11A protein on chromosome 2p15. The 2p15
BCL11AQTL accounts for 15.1% of the trait variance.16
More recently genome-wide association studies and
family studies have shown that other regions outside
the b-globin gene cluster, including 2q16, 6q23, 8q,
and Xp22.2, are implicated in the regulation of HbF
levels.17,18
However, almost 45% of variations in HbF
levels are associated with the presence of 3 main
quantitative trait loci (QTLs), including the Xmn1
polymorphism on chromosome 11(11p15), the HMIP
(HBS1L-MYB intergenic polymorphism) loci on
chromosome 6 (6q23), and the BCL11A loci on
chromosome 2.10,16,19,20
Moreover, not only are
variants at BCL11A, HMIP, and b-globin gene
associated with HbF levels but also with pain crisis
rate.21
Because, in India, the prevalence of sickle cell gene is
high in both tribal and nontribal populations, it becomes
very important to evaluate genetic markers that are asso-
ciated with amelioration of SCD in the populations. It
would be imperative to first evaluate the presence and
impact of the 3 known major loci on the HbF trait. In
the present study, which is the first report from India,
we have attempted to determine the frequency of spe-
cific variants in the BCL11A, HBS1L-MYB loci, and
HBB gene in a cohort with SCD and sickle cell trait
from central India and to correlate the association of
these single-nucleotide polymorphisms (SNPs) with
HbF levels.
PATIENTS AND METHODS
Subject group. The present study was conducted on
300 individuals, which consisted of 240 patients with
SCD and 60 patients with sickle cell trait. These individ-
uals are a subset of the total cohort identified during a
screening survey of 2094 villages from the district of
Chhattisgarh in central India. The blood specimens
were collected from the general castes, schedule castes
(SC), schedule tribes (ST), and other backward classes
(OBC) under ‘‘Sickle cell project’’ was carried out
by Department of biochemistry (Pt. J.N.M Medical
College) and funded by Government of Chhattisgarh.
The nomenclature of social classes was derived from
Articles 340(1) and 340(2) of the Constitution of India.
The relative proportions of these social groups in the
State of Chhattisgarh are general castes (17%), SC
(25%), ST (8%), and OBC (50%).22
The study was approved by the local ethical commit-
tee and is performed in accordance with the Helsinki
declaration. An informed consent and detailed case re-
cord form pertaining to information on demographics,
AT A GLANCE COMMENTARY
Bhanushali AA, et al.
Background
India tops the list of countries with sickle cell dis-
ease (SCD); however, there is minimal informa-
tion about genetic factors influencing the disease
course in Indians. Fetal hemoglobin (HbF) is the
most powerful modulator of the clinical features
of SCD. The present study is the first report in
this population investigating the effect of genetic
variants affecting HbF levels.
Translational Significance
Elucidation of genetic variants responsible for the
phenotypic variability in SCD will have important
implications for genetic counseling and clinical
management. In future it may help in the develop-
ment of individual therapeutic strategies and pre-
ventive cures to produce better outcomes.
Translational Research
2 Bhanushali et al - 2015

3. medical history, and risk factors such as occurrence of
pain crisis and its frequency, and so forth was obtained
from each participant through interview and perusal of
their medical records.
SCD diagnosis and hemoglobin profiling. Briefly
finger-prick samples were taken into glass tubes
containing reagents (hyperosmolar phosphate buffer,
saponin, and reducing agent) for the solubility test,
all positive subjects had 5 mL of venous blood
taken in ethylenediaminetetraacetic acid (EDTA)
vacutainers. Alkaline hemoglobin electrophoresis on
cellulose acetate in Tris-EDTA-borate buffer at pH 8.6
was conducted on all solubility positive samples along
with known controls. Molecular analysis confirmation
was done for the entire study population by
amplification refractory mutation specific polymerase
chain reaction as previously described.23
Analysis of hemoglobin fractions and variants was
done on the Bio-Rad Variant high performance liquid
chromatography system with use of the Beta-
Thalassemia Short Program Reorder Pack (Bio-Rad
Laboratories) as per the manufacturer’s instructions.
DNA extraction and genotyping. Genomic DNA was
extracted from the collected EDTA whole blood using
QIAamp DNA extraction kit following the manufac-
turer’s instructions (Qiagen, Duesseldorf, Germany).
Genotyping was performed for the BCL11A
rs11886868 and rs34211119; HMIP rs9399137,
rs189600565, rs7776196, rs34778774, and
rs53293029; HBG2 Xmn1 polymorphism rs7482144;
andHBD268C.Tbydirectsequencingoftheproducts
obtained after amplification using the primers described
subsequently. PCR product for the rs9399137 also
covered the adjacent SNPs rs189600565, rs7776196,
rs34778774, and rs53293029, whereas PCR product
for rs11886868 covered the adjacent polymorphism
rs34211119. Sequencing was done on the Automated
ABI prism 3100 Avant Genetic Analyzer (Applied Bio-
systems Inc, Foster city, California) using ABI prism
BigDye terminator kit (version 3.1).
Xmn1 F 50
-30
: TTT TAT TCT TCA TCC CTA GC
Xmn1 R 50
-30
: GAG CTA CAG ACA AGA AGG TG
rs9399137 F 50
-30
: CAA CAT CAC CTT AAA AGG
CG
rs9399137 R 50
-30
: ATT CAC TGC CAG AAG CAC
TT
rs11886868 F 50
-30
: CAG TGT TGA GAATTC TAG
AAT
rs11886868 R 50
-30
: ATT GTA GCA CTG TTC ATA
GTG 30
HBD 268C . T F 50
-30
: AGT GGA ATG A AG GTT
CAT TTT TC0
HBD 268C . T R 50
-30
: TAT GTC AGA AGA AAG
TGT AAG C 30
Statistical analysis. Allele frequency was calculated as
the number of occurrences of the test allele in the
population divided by the total number of alleles. Any
deviation in the genotype frequencies from the Hardy-
Weinberg equilibrium was assessed by Fischer’s exact
test. Chi-square tests were used for comparison of
binary variables across groups. Analysis of variance
was performed to determine differences in the HbF
and HbA2 based on genotypes. Routine statistical
analysis was carried out with the Statistical Package
for the Social Sciences (SPSS) v 15 software (SPSS
Inc, Chicago, Illinois). SNPStat online software tool24
was used to determine the association of genotypes
with the quantitative heritable trait such as HbF and
HbA2. The association between HbF and each SNP
was tested using linear regression with a genotypic
genetic model. In all regression models, we used the
natural logarithmic transformation of HbF to satisfy
the normality assumption.
RESULTS
General characteristics. The demographic characteris-
tics of study population are indicated in Table I. Among
the social classes, the highest frequency of SCD was seen
Table I. Description of SCD and SCT cohorts
General characteristics SCD (n 5 240) SCT (n 5 60)
Age (mean 6 SD) 13.2 6 7.44 16.6 6 9.98
Sex, n (%)
Male 110 (45.8%) 39 (65%)
Female 130 (54.2%) 21 (35%)
Category
General (n 5 7) 5 (2.08%) 2 (3.3%)
OBC (n 5 204) 166 (69.2%) 38 (63.4%)
SC (n 5 44) 36 (15%) 8 (13.3%)
ST (n 5 36) 27 (11.25%) 9 (15%)
Unknown 6 (2.5%) 3 (5%)
Hematologic parameters
Hb* (g/dL) 10.49 6 2.79
RBC 3 106
* 3.92 6 1.17
MCV* 81.20 6 13.21
MCH* 22.58 6 9.93
WBC* 10.83 6 5.39
Platelet count 3 106
* 347.39 6 162.52
HbF%†
(mean 6 SD) 19.5 6 6.59 —
HbA2%†
(mean 6 SD) 3.5 6 1.92 —
Abbreviations: HbF, fetal hemoglobin; MCH, mean corpuscular he-
moglobin; MCV, mean corpuscular volume; OBC, other backward
classes; RBC, red blood corpuscles; SC, schedule castes; SCD,
sickle cell disease; SCT, sickle cell trait; SD, standard deviation; ST,
schedule tribes.
*Data on 162 individuals.
†
Data on 216 individuals.
Translational Research
Volume -, Number - Bhanushali et al 3

4. in the OBC category (69.2%), followed by SC (15%) and
ST (11.3%). SCD was also seen in general castes albeit
with a lower frequency of 2% as indicated in Table I.
The mean HbF% and HbA2% in SCD subjects were
19.5 6 6.59 and 3.5 6 1.92, respectively.
Frequency and linkage analysis. The genotype and
allele frequencies are depicted in Table II. None of the
genotypes deviated from Hardy-Weinberg equilibrium
for SCD individuals; however for SC trait individuals,
a significant deviation for the HBG2 Xmn1
polymorphism was seen. The Arab-Indian (AI)
haplotype represented by the HBG2 Xmn1
polymorphism (rs7482144A . G) was seen in almost
98.5% of the study population, with genotypic
frequency AA 70.5% and AG 28%, and only 1.5% of
the study population showed the absence of (A) allele.
Further confirmation that the cohort studied truly
represents the AI haplotype was done by additionally
genotyping the HBD 268C . T variant
(Supplementary Fig 1). All SCD samples that were
homozygous for the ‘‘A’’ allele Xmn1 (rs7482144) had
homozygous ‘‘T’’ allele of the promoter polymorphism
268C . T, SCD samples that were homozygous for
the ‘‘G’’ allele Xmn1 (rs7482144) had homozygous
‘‘C’’ allele of the promoter polymorphism, whereas
individuals who were heterozygous ‘‘AG’’ for Xmn1
were also heterozygous ‘‘CT’’ at 268 HBD promoter.
These findings indicate that the population in the
present study is truly of the AI haplotype.
HbF-associated BCL11A SNP rs11886868 allele fre-
quency was found to be 0.60 for major allele C and
0.40 for the minor allele T (Table II). In our study, we
found the rs11886868 SNP to be in complete linkage
disequilibrium (LD) with the neighboring rs34211119
(2/T), which is 73 bp downstream (Supplementary
Fig 2). The rs34211119 is a part of the same haplotype
block as rs11886868 even in Caucasian population.
For HMIP SNP rs9399137, the genotype frequencies
were 17% for CT and 83% for ‘‘TT’’ genotype, no indi-
vidual with the ‘‘CC’’ genotype was seen. Interestingly,
it was found that all individuals who were heterozygous
CT showed the presence of a mixed template on
sequencing indicating an insertion or deletion (in/del)
upstream (Supplementary Fig 3). Similar finding has
Table II. Genotype and allele frequencies of the fetal hemoglobin–related variants
SNP Genotype frequency n (%) Allele frequency HWE
HBG2 rs7482144 AA AG GG A G
Total (n 5 288) 203 (70.5) 81 (28) 4 (1.5) 0.85 0.15 0.26
SCD (n 5 232) 190 (81.9) 39 (16.8) 3 (1.3) 0.90 0.10 0.46
SCT (n 5 56) 13 (23.2) 42 (75) 1 (1.8) 0.60 0.40 ,0.0001*
BCL11A rs11886868 CC CT TT C T
Total (n 5 296) 108 (36.4) 139 (47) 49 (16.6) 0.60 0.40 0.72
SCD (n 5 238) 87 (36.4) 116 (48.7) 35 (14.7) 0.60 0.40 0.79
SCT (n 5 58) 21 (36.2) 23 (39.7) 14 (24.1) 0.56 0.44 0.18
HMIP rs9399137 TT CT CC T C
Total (n 5 299) 249 (83) 50 (17) 0.92 0.08 0.24
SCD (n 5 240) 195 (81.2) 45 (18.8) — 0.91 0.09 0.24
SCT (n 5 59) 54 (91.5) 5 (8.5) — 0.96 0.04 0.1
Abbreviations: HWE, Hardy-Weinberg equilibrium; SCD, sickle cell disease; SCT, sickle cell trait.
*Statistically significant at P , 0.005.
Table III. Analysis of variance: genotypes and HbF levels
SNP Genotypes F0
P value
HBG2 XmnI rs7482144 AA AG GG
HbF% (mean 6 SD) 20 6 6.3 17 6 8.1 8.9 6 2.8 7.857 0.0005*
BCL11A rs11886868 CC CT TT
HbF% (mean 6 SD) 20.4 6 6.19 19.6 6 6.4 16.8 6 7.6 5.461 0.0048†
HMIP rs9399137 TT CT CC
HbF% (mean 6 SD) 18.7 6 6.1 22.1 6 8.5 — 14.81 0.0002*
Abbreviations: F0
, F statistics; HbF, fetal hemoglobin; SNP, single-nucleotide polymorphism.
Analysis of variance was performed to determine if there were significant differences in the percent HbF levels based on the genotypes. The
degree of freedom d(f) between groups 5 2, and within groups d(f) 5 213. Data on 216 individuals.
Bold indicates statistically significant.
*Statistically significant at P , 0.005.
†
Statistically significant at P , 0.01.
Translational Research
4 Bhanushali et al - 2015

5. been reported by Farrell et al25
who have indicated that a
3 bp TAC deletion between Chr.6: 135,460,326 and
135,460,328 upstream of rs9399137 (135,460,710) is
in complete LD with it. Our findings suggest that this
would be the most plausible reason for the mixed tem-
plate, which was seen in CT heterozygotes. The other
SNPs in the HMIP region rs189600565C . T and
rs7776196A . G were found to be monomorphic,
whereas rs34778774 (2/C) and rs56293029A . C had
a minor allele frequency of ,1% and have not been
considered for subsequent analysis.
Association of genetic variants with HbF levels. In the
present study, the HbF% levels significantly differed be-
tween the Xmn1 SNP genotypes (Table III, Fig 1, A) as
indicated by analysis of variance (F’statistic, 7.857;
P 5 0.0005). Almost 55% decrease in HbF levels
were observed for the ‘‘GG’’ genotype compared with
the ‘‘AA,’’ whereas for heterozygotes AG the percent
decrease was 15%. Similar results were obtained in the
linear regression analysis (Table IV), where the Xmn1
genotype accounts for 11% of HbF trait variance. The
genetic model of inheritance indicated the codominant
and the recessive model to explain the association of
the rs7482144 SNP with HbF levels (Table V).
The HbF% levels as seen in the rs11886868 SNP were
20.4 6 6.19, 19.6 6 6.4, 16.8 6 7.6 for CC, CT, and TT
genotypes, respectively (Table III, Fig 1, B). For the TT
genotype the percent decrease in HbF levels was 18%,
for CT it was 4% compared with the CC genotype.
The trait variance explained by this locus is 3.65% by
linear regression analysis (Table IV), the mode of inher-
itance indicates the recessive as well as codominant
Fig 1. Fetal hemoglobin values based on genotypes: (A) rs7482144, (B) rs11886868, and (C) rs9399137.
Table IV. Association of the 3 loci with HbF levels by linear regression analysis
Parameter b coefficient
95% Wald CI
Wald chi square P valueUpper Lower
HBG2 XmnI rs7482144 AA 11.02 3.670 18.369 8.635 0.003
HBG2 XmnI rs7482144 AG 7.517 20.079 15.113 3.762 0.052
HBG2 XmnI rs7482144 GG 0*
rs9399137 CT 3.679 1.491 5.866 10.861 0.001
rs9399137 TT 0*
rs11886868 CC 3.646 0.937 6.355 6.959 0.008
rs11886868 CT 2.758 0.132 5.385 4.237 0.040
rs11886868 TT 0*
Abbreviations: CI, confidence interval; HbF, fetal hemoglobin.
*Set to zero because this parameter is redundant.
Translational Research
Volume -, Number - Bhanushali et al 5

6. model, with the presence of T allele resulting in
decreased HbF levels (OR 20.14, 95% CI 20.21 to
20.06, P 5 0.0007) (Table V).
For HMIP SNP rs9399137 the mean HbF values were
22.1 6 8.5 for CT and 18.7 6 6.1 for TT, which were
highly significant (F’statistic 14.81, P 5 0.0002) as
shown in Table III (Fig 1, C). The TT homozygotes
had almost 16% lower HbF levels compared with CT
heterozygotes. The contribution of this SNP to HbF trait
variance is 3.8% (Table IV), and the SNPstat analysis
(Table V) also indicated similar results.
DISCUSSION
HbF level is a well-characterized factor that modifies
SCD expression. Three major HbF QTLs have been
identified, which can modulate HbF levels and disease
severity in SCD and b-thalassemia,18,26-29
namely, the
BCL11A locus on chromosome 2p16,16
the HMIP on
chromosome 6q23,20
and the b-globin locus. This study
explores for the first time the frequency and association
with HbF levels of genetic variants in these QTLs in
SCD population from India.
In the present study, all the 3 QTLs were associated
with HbF levels in Indian patients with SCD. The asso-
ciation of rs7482144 or the Xmn I polymorphism with
HbF levels is a well-documented fact. The present study
population has the AI denoted by the presence of the
rs7482144 SNP in 98.5% of the subjects. Recent studies
by Asultan et al30
have indicated a polymorphism in the
promoter region of the HBD 268C . T to be exclu-
sively present only when the HBS gene was on the A1
haplotype, which is also seen in the present study indi-
cating that this SCD population is truly of the A1 haplo-
type. Although the Xmn1 SNP is present in both the AI
and Senegal haplotypes, the HbF level in Senegal haplo-
type carriers is lower than in AI haplotype carriers.
Carriers of the Benin Bantu haplotype lack the Xmn1
SNP. In Tanzanians, the prevalent HBB haplotype is
the Bantu bS
, which lacks the A allele (also referred
to as Xmn1 G
g1), this cluster thus shows a diminished
effect on HbF.31
Considerable effect of the BCL11A locus on HbF
levels was seen in Indian patients with SCD, which is
in agreement with studies on Tanzanians,31
British,31
African American, and Brazilian subjects.21
The initial
Table V. Association of variants with HbF levels based on different genetic models
Model Genotype Log HbF mean (SE) OR (95% CI) P value AIC
rs7482144 association with HbF levels
Codominant A/A 1.28 (0.01) 0.00 6 3 1024*
-169.63
A/G 1.24 (0.03) 20.05 (20.11 to 0.01)
G/G 0.94 (0.07) 20.34 (20.52 to 20.16)
Dominant A/A 1.28 (0.01) 0.00 0.016‡
2162.4
A/G-G/G 1.21 (0.03) 20.07 (20.13 to 20.01)
Recessive A/A-A/G 1.27 (0.01) 0.00 4 3 1024*
2169.1
G/G 0.94 (0.07) 20.33 (20.52 to 20.15)
Overdominant A/A-G/G 1.27 (0.01) 0.00 0.17 2158.4
A/G 1.23 (0.03) 20.04 (20.10 to 20.02)
Log-additive — 20.08 (20.13 to 20.03) 0.0017 2166.5
rs11886868 association with HbF
Codominant C/C 1.29 (0.02) 0.00 0.0026†
271.7
C/T 1.26 (0.02) 20.02 (20.08 to 20.04)
T/T 1.15 (0.06) 20.15 (20.23 to 20.06)
Dominant C/C 1.29 (0.02) 0.00 0.08 264.6
C/T-T/T 1.24 (0.02) 20.05 (20.11 to 0.01)
Recessive C/C-C/T 1.27 (0.01) 0.00 7 3 1024*
2
T/T 1.15 (0.06) 20.14 (20.21 to 20.06)
Overdominant C/C-T/T 1.25 (0.02) 0.00 0.53 261.8
C/T 1.26 (0.02) 0.02 (20.04 to 0.07)
Log-additive — 20.06 (20.10 to 20.02) 0.003†
270.5
rs9399137 association with HbF
— T/T 1.250 (0.01) 0.00 2 3 1024*
2170
C/T 1.36 (0.03) 0.11 (0.05 to 0.16)
Abbreviations: AIC, akaike information content; CI, confidence interval; HbF, fetal hemoglobin; OR, odds ratio; SE, standard error.
Logarithmic transformation of HbF has been used in the linear regression analysis.
Bold indicates statistically significant.
*Statistically significant at P , 0.001.
†
Statistically significant at P , 0.01.
‡
Statistically significant at P , 0.05.
Translational Research
6 Bhanushali et al - 2015

7. findings of Uda et al18
indicated strong association of
rs11886868 with HbF levels, subsequent genotyping of
1242 patients from the Cooperative study of sickle cell
disease (CSSCD) cohort indicated that the C allele of
rs11886868C . T accounted for 8.6% of the variance
in HbF levels.18
In the present study, the rs11886868 mi-
nor allele results in almost 18% reduction in HbF levels.
This finding also concurs with the study on an indepen-
dent cohort from Brazil21
and the CSSCD,21
where 2
SNPs rs4671393 and rs7557939 in addition to
rs11886868 at BCL11A locus were associated with
HbF levels. The fraction of HbF phenotypic variation ex-
plained by rs4671393 in the CSSCD and the Brazil
cohort was 14.1% and 9.0%. However, when the associ-
ation analysis on rs4671393 was conditioned, associa-
tion at rs11886868 and rs7557939 became
nonsignificant, suggesting that these markers tag the
same causal polymorphism. Importantly another cohort
with the AI haplotype (Eastern Saudi Arabia) also had an
association of the rs11886868 with HbF levels,32
which
is similar to our findings. The BCL11A (rs766432) was
an important HbF QTL even in Saudi patients from
Southwestern province with HBB haplotypes of African
origin.33
Frequency comparison indicates the
rs11886868 (C) allele frequency in the present study to
be much higher than the Brazilian patients (C allele
frequency 5 0.39, c2
5 24.07, P , 0.0001) or the
CSSCD patients (C allele frequency 5 0.31,
c2
5 72.46, P , 0.0001),21
but lower than in Tanzanian
patients (C allele frequency 5 0.74, c2
5 17.63,
P , 0.0001) and British population (C allele
frequency 5 0.71, c2
5 4.427, P 5 0.035).31
The role of 6q23 QTL on HbF control was explored
by Thein et al,20
who identified principal genetic vari-
ants in 3 blocks referred to as 1, 2, and 3, which are
responsible for 17.6% of the trait variance. The
rs9399137 at this locus in block 2 showed strongest as-
sociation with HbF levels (P 5 10275
). In concurrence
with the findings of Thein et al,20
the rs9399137 is
strongly associated with HBF levels in the present study.
Similar findings were also seen in the Brazilian SCD
and CSSCD cohort.21
In Tanzanian and British patients
also the largest allele effect detected at the HMIP locus
was with rs9399137 leading to mean HbF values of
8.8% for C/T vs 4.5% for T/T, where C/C homozygotes
were not seen.31
In our study, we have observed higher
mean HbF values viz 22% for C/T and 18.7% for T/T,
with no C/C homozygotes. Farrel et al25
in their study
indicated the variance in HbF level attributed to
rs9399137 to be 13.5%. Two other studies in Saudi
Arab population, however, did not find any association
of the rs9399137 with HbF levels in contrast to our
results. Interestingly one study was based on the
population from Southwestern province where the
predominant HBB haplotype is of the African origin33
and the other study by Ngo et al32
is based on the
Eastern province population with the AI haplotype.
No significant differences were observed in the allele
frequency in the present study compared with the
British,31
CSSCD cohort,21
and Southwestern Saudi pa-
tients,33
but differs from the Tanzanian population (Mi-
nor allele frequency 5 0.01, c2
5 46.812, P , 0.0001).
The rs9399137 is also important because it is in com-
plete LD with the sequence containing rs7775698 (T)
with the 3-bp deletion in the Chinese and European pop-
ulations. Functional studies indicate that this 3-bp dele-
tion changes the normal DNA binding configuration of
transcription factors and results in possible changes to
the spatial orientation for DNA-protein binding and/or
protein-protein interactions.24
Recently, it has been
indicated that the HBS1-MYB intergenic variants
notably the 3-bp deletion may affect long-range MYB
regulation and through this mechanism exert their influ-
ence on human erythroid blood parameters.34
Hence,
further studies in different populations are required to
determine the presence and the LD status of this 3 bp
deletion, as it could have therapeutic applications
whereby HbF can be induced by reducing MYB levels
via its intergenic regulatory elements.
CONCLUSIONS
The present study sheds information on the associa-
tion of genetic variants with HbF levels in Indian popu-
lation, especially ethnic tribals from central India.
Clinically, HbF level is the strongest predictor of disease
severity. Moreover, previous literature indicates that
these variants affect the HbF levels over a lifetime.
SNPs at BCL11A and HMIP also have the power to pre-
dict pain crisis and information on pain rate beyond
their effect on single measurements of steady-state
HbF levels. This suggests an interesting possibility
that clinical genotyping of these variants (and other
HbF-associated variants yet to be found) may someday
be potentially useful to stratify patients with SCD ac-
cording to severity risk and to adjust therapeutic strate-
gies accordingly.
Hence, further interrogation of these genotypes with
respect to pain crisis and quality of life aspects is war-
ranted, as well as a genome-wide association study or
high resolution association scan in this population,
where SCD is endemic may help uncover new loci
and variants controlling HbF.
ACKNOWLEDGMENTS
Conflicts of Interest: All the authors have read the
journal’s policy on conflicts of interest and have none
to declare.
Translational Research
Volume -, Number - Bhanushali et al 7

8. All the authors have read the journal’s policy on
authorship agreement.
Author Contributions: A.A.B. conceived the study,
designed the experimental work, analyzed results, and
wrote the manuscript. B.R.D. conceived the study and
reviewed the manuscript. P.K.P. participated in the se-
lection of patients and reviewed the manuscript.
S.S.K., S.P., and S.S. performed the experimental
work. All authors have read and approved the final
manuscript.
No funding for this project was received from any
external source.
Supplementary Data
Supplementary data related to this article can be
found online at http://dx.doi.org/10.1016/j.trsl.2015.
01.002.
REFERENCES
1. Weatherall DJ, Craig JB. Inherited haemoglobin disorders: an
increasing global health problem. Bull World Health Organ
2001;79:704–12.
2. Weatherall DJ. The inherited disorders of hemoglobin are an
emerging global health burden. Blood 2010;115:4331–6.
3. Lehman H, Cutbush M. Sickle cell trait in Southern India. Br Med
J 1952;1:404–5.
4. Mohanty D, Mukherjee M. Sickle cell disease in India. Curr Opin
Hematol 2002;9:117–22.
5. Ingram VM. Gene mutations in human haemoglobin: the chemi-
cal difference between normal and sickle haemoglobin. Nature
1957;180:326–8.
6. DrissA,AsareKO,HibbertJM,GeeBE,AdamkiewicTV,StilesJK.
Sickle cell disease in the post genomic era: a monogenic disease
with a polygenic phenotype. Genomics Insights 2009;2:23–48.
7. Samarah F, Athanasiou-Metaxa M, Vavatsi-Christaki N. Genetic
modulators of sickle cell disease. Aristotle Univ Med J 2007;34:
25–35.
8. Akinsheye I, Alsultan A, Solovieff N, et al. Fetal hemoglobin in
sickle cell anemia. Blood 2011;118:19–27.
9. Garner C, Tatu T, Reittie JE, et al. Genetic influences on F cells
and other hematologic variables: a twin heritability study. Blood
2000;95:342–6.
10. Garner C, Matsuda F, Farrall M, Lathrop M. Intergenic variants of
HBS1L-MYB are responsible for a major quantitative trait locus
on chromosome 6q23 influencing fetal hemoglobin levels in
adults. Proc Natl Acad Sci U S A 2007;104:11346–51.
11. Higgs DR, Wood WG. Genetic complexity in sickle cell disease.
Proc Natl Acad Sci U S A 2008;105:11595–6.
12. Gibney GT, Panhuysen CI, So JC, et al. Variation and heritability
of HbF and F cells among beta thalassemia heterozygotes in Hong
Kong. Am J Hematol 2008;83:458–64.
13. Dover GJ, Boye SH, Pembrey ME. F-cell production in sickle cell
anemia: regulation by genes linked to beta-hemoglobin locus. Sci-
ence 1981;211:1441–4.
14. 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;67:1404–10.
15. Thein SL, Craig JE. Genetics of HbF/F cell variance in adults and
heterocellular hereditary persistence of fetal hemoglobin. Hemo-
globin 1998;22:401–14.
16. Menzel S, Jiang J, Silver N, et al. The HBS1L-MYB intergenic re-
gion on chromosome 6q23.3 influences erythrocyte, platelet, and
monocyte counts in humans. Blood 2007;110:3624–6.
17. Nguyen TKT, Joly P, Bardel C, Moulsma M, Bonello-Palot N,
Francina A. The Xmn1 Gg polymorphism influences hemoglobin
F synthesis contrary to BCL11A and HBS1L-MYB SNPs in a
cohort of 57 b-thalassemia intermedia patients. Blood Cells
Mol Dis 2010;45:124–7.
18. Uda M, Galanello R, Sanna S, et al. Genome-wide association
study shows BCL11A associated with persistent fetal hemoglobin
and amelioration of the phenotype of b-thalassemia. Proc Natl
Acad Sci U S A 2008;105:1620–5.
19. Sedgewick AE, Timofeev N, Sebastiani P, et al. BCL11A is a ma-
jor HbF quantitative trait locus in three different populations with
b-hemoglobinopathies. Blood Cells Mol Dis 2008;41:255–8.
20. Thein SL, Menzel S, Peng X, et al. Intergenic variants of
HBS1LMYB are responsible for a major quantitative trait locus
on chromosome 6q23 influencing fetal haemoglobin levels in
adults. Proc Natl Acad Sci U S A 2007;104:11346–51.
21. Lettre G, Sankaran VG, Bezerra MA, et al. DNA polymorphisms
at the BCL11A, HBS1L-MYB, and -globin loci associate with
fetal hemoglobin levels and pain crises in sickle cell disease.
Proc Natl Acad Sci U S A 2008;105:11869–74.
22. Patra PK, Chauhan VS, Khodiar PK, Dalla AR, Serjeant GR.
Screening for the sickle cell gene in Chhattisgarh state, India:
an approach to a major public health problem. J Community
Genet 2011;2:147–51.
23. Wu DY, Ugozzoli L, Pal BK, Wallace RB. Allele-specific enzy-
matic amplification of beta-globin genomic DNA for diagnosis
of sickle cell anemia. Proc Natl Acad Sci U S A 1989;86:2757–60.
24. Sole X, Guino E, Valls J, Iniesta R, Moreno V. SNPStats: a web
tool for the analysis of association studies. Bioinformatics 2006;
22:1928–9.
25. Farrell JJ, Sherva RM, Chen ZY, et al. A 3-bp deletion in the
HBS1L-MYB intergenic region on chromosome 6q23 is associ-
ated with HbF expression. Blood 2011;117:4935–45.
26. Thein SL, Menzel S, Lathrop M, Garner C. Control of fetal hemo-
globin: new insights emerging from genomics and clinical impli-
cations. Hum Mol Genet 2009;18:216–23.
27. Galanello R, Sanna S, Perseu L, et al. Amelioration of Sardinian
beta0 thalassemia by genetic modifiers. Blood 2009;114:3935–7.
28. Ma Q, Abel K, Sripichai O, et al. b-Globin gene cluster polymor-
phisms are strongly associated with severity of HbE/b0-thalas-
semia. Clin Genet 2007;72:497–505.
29. Nuinoon M, Makarasara W, Mushiroda T, et al. A genome-wide
association identified the common genetic variants influence dis-
ease severity in b0 thalassemia/hemoglobin E. Hum Genet 2010;
127:303–14.
30. Asultan A, Ngo DA, Farell JJ, et al. A functional promoter poly-
morphism of the d-globin gene is a specific marker of the Arab-
Indian haplotype. Am J Hematol 2012;87:824–6.
31. Makani J, Menzel S, Nkya S, et al. Genetics of fetal hemoglobin in
Tanzanian and British patients with sickle cell anemia. Blood
2011;117:1390–2.
32. Ngo D, Bae H, Steinberg MH, et al. Fetal hemoglobin in sickle
cell anemia: genetic studies of the Arab-Indian haplotype. Blood
Cells Mol Dis 2013;51:22–6.
33. Alsultan A, Solovieff N, Aleem A, et al. Fetal hemoglobin in
sickle cell anemia: Saudi patients from the Southwestern Province
have similar HBB haplotypes but higher HbF levels than African
Americans. Am J Hematol 2011;86:612–4.
34. Stadhouders R, Aktuna S, Thongjuea S, et al. HBS1L-MYB inter-
genic variants modulate fetal hemoglobin via long-range MYB
enhancers. J Clin Invest 2014;124:1699–710.
Translational Research
8 Bhanushali et al - 2015