Impact of Genetic Polymorphisms in Modiﬁer Genes in Determining Fetal Hemoglobin Levels in Beta-Thalassemia

: Genetic polymorphisms in Quantitative Trait Loci (QTL) genes such as BCL11A, HBS1L-MYB and KLF1 have been reported to inﬂuence fetal hemoglobin (HbF) levels. This prospective study was planned to evaluate the role of genetic polymorphisms in QTL genes as determinant of HbF levels in beta thalassemia major patients. The study was carried out on 100 thalassemia major patients. Blood samples were collected in EDTA and plain vials for biochemical and molecular evaluation. The BCL11A, HBS1L-MYB and KLF1 genotypes were determined using a polymerase chain reaction (PCR)-based method. Red Blood Cell (RBC) indices and HbF levels were assessed. In silico analysis was assessed using loss-of-function tool (Lof Tool). Statistical difference and genetic comparisons between groups were evaluated by using SPSS for Windows, version 16.0 (SPSS Inc., Chicago, IL, USA). Comparisons between quantitative variables were carried out after data explored for normality using Kolmogorov–Smirnov test of normality. Logistic regression was used for computation of ORs and 95% CIs (Conﬁdence Interval). We observed association of HbF levels in thalassemia major patients with the polymorphisms in BCL11A (rs11886868 rs7557939; rs1427407 and rs766432) and HBS1L-MYB (rs9399137) gene. The results of this study indicated that the presence of polymorphisms on modiﬁer genes are strongly associated with an increase in HbF levels in thalassemia major patients. Further research with a larger sample size and with other genes of modiﬁer genes is required.


Introduction
β-thalassemia is a hereditary blood disorder caused by mutations in the beta globin gene that leads to reduced or absent synthesis of beta globin chains essential for hemoglobin (Hb) biosynthesis [1]. β-thalassemia phenotypes are quite heterogenous, ranging from the severe transfusion-dependent thalassemia major to non-transfusion-dependent thalassemia intermedia. Various factors have been found to modulate the beta-thalassemia phenotype [2,3]. The primary genetic determinants of β-thalassemia are mostly different types of mutations of beta globin genes (β0/β+/β++) that lead to decreased/absent production of β globin chains of Hb molecule; secondary loci can either ameliorate the disease phenotype by increased synthesis of γ-globin and hemoglobin F or worsen the clinical phenotype by co-inheritance of excess α globin genes. Tertiary loci are not involved in globin production but that might modify the complications of the disease in many different ways. The latter group includes the many different polymorphisms that have been co-selected with the thalassemias and that might further modify their phenotype.
Fetal hemoglobin (HBF) is one of the most important modifiers of the disease severity in individuals with β-TM disorder. Since beta chains are either absent or reduced in beta thalassemia, the γ-globin chains bind with the excess α-chains thereby decreasing the deleterious effect of intracellular precipitation of unbound β-chains [4][5][6][7]. This reduces ineffective erythropoiesis and increases production of some functional HbF, which on the other hand facilitates longer duration survival of RBC in the circulation.
B-cell lymphoma/leukemia (BCL11A) gene encodes a C2H2 type, which is a zinc finger protein, involved in the generation of B-cell progenitor [7]. BCL11A is highly expressed in several hematopoietic cell lineages, and play an important role in the switch from γ to β-globin expression during the fetal to adult erythropoiesis transition phase. It is also expressed in the brain, B-lymphocytes and the adult erythroid lineage [8]. Downregulation of BCL11A expression in adult human erythroid precursors results in induction of HbF [9][10][11].
Researchers have shown that long-range interactions between BCL11A and multiple regions throughout the β-globin gene locus mediate the silencing of the γ-globin genes [4]. A total of 5788 single-nucleotide polymorphisms (SNPs) in BCL11A gene were reported in the National Centre for Biotechnology Information (NCBI) database [12].
A study demonstrated that an intronic SNP in BCL11A, rs11886868 strongly co-relates with HbF levels in a large cohort of Sickle Cell Anemia (SCA) patients [13]. This indicates that BCL11A variants act as an important ameliorating factor of the SCA by modulating HbF levels. Genotyping of two other SNPs, rs4671393 and rs7557939 in the Brazil Sickle cell disease cohorts, have shown that these SNPs were more strongly associated with HbF level variation than rs11886868 [11,14].
The HBS1L-MYB intergenic region located on chromosome 6q23, near the oncogene MYB, was first detected through linkage in a large South Asian family [15] that regulates erythroid cell proliferation, maturation and fetal hemoglobin expression [16]. MYB encodes cMYB, an important transcriptional regulator of hematopoiesis [17]. There are multiple sub-loci situated in the intergenic interval between HBS1L, such as translational GTPase and MYB [8]. A significant effect of these sub-loci, mainly HMIP-2 (HBS1L-MYB intergenic polymorphism, block 2) and HBS1L-MYB, was demonstrated in patients with SCD [17,18].
In a study of polymorphism associated with the variability of HbF expression in non-anemic humans, variants in the intergenic region between the HBSL1 and MYB genes were identified in MYB expressing primary human erythroid cells [19][20][21]. The HBS11L-MYB SNPs especially rs9399137 had an effect on HbF production and the variation of hematological parameters in homozygous HbE subjects [22].
A recent study by Qadah et al. in 2020 confirms the strong relationship among HBS1LMYB rs9399137 and rs28384513 genotypes. However, the contributions of different SNPs of HBS1L-MYB are found to be highly variable in different populations and the association varies in different hemoglobinopathies, so it is important to study the status of these polymorphisms to understand their influence on fetal hemoglobin expression in thalassemia cases of north India.
Krüppel-like factor (KLF1) is an essential erythroid transcription factor that was first identified by Miller and Bieker in 1993 [23]. KLF1 functions in two ways: first it binds to the locus control region of the β-globin gene and regulates its expression; second, as it is a positive regulator of BCL11A, it activates the BCL11A expression which in turn silences the gamma globin gene and maintains the level of HbF [24][25][26][27]. So, KLF1 (the erythroid-specific transcription factor) is a crucial regulator of the γ to β-globin, the developmental switching between fetal and adult hemoglobin.
Genetic variations in KLF1 have been found to have an indirect role in causing high HbF levels in HPFH and Thalassemia Intermedia (TI) [28]. Therefore, it may be possible that by attenuating the activity of KLF1, HbF level can be induced which has the possibility of becoming a central approach in reducing the severity of thalassemia phenotype [29,30].
We focused to study key regulators of HbF production to demonstrate the benefit of modern human genetics in understanding clinical heterogeneity in monogenic disorder. Thus, in the second part of thesis, we have investigated the role of genetic variants present in QTLs such as BCL11A, HBS1L-MYB and KLF1 genes which ameliorate the disease severity in patients with β-thalassemia major.

Materials and Methods
A total of 100 beta thalassemia major patients aged less than 5 years from the Department of Hematology and Medical Genetics of Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow were enrolled in the study after obtaining written informed consent from the patient and their parents. We included only patients who received less than 3 transfusions for our study.
Peripheral blood was collected from each individual in an EDTA vial (2 mL). About 1 mL EDTA blood sample was used for estimation of hematological investigations. The remaining 1 mL was used for DNA isolation by phenol chloroform method. The quality and quantity of DNA was checked on agarose gel electrophoresis and then the quantity and quality were determined by measuring optical density (O.D) on Nano Drop (ND-1000).

Hematological Investigation
Complete Blood Count: Hematological parameters were obtained by using automated hematology cell counter SYSMEX KX-21.
Quantification of Hemoglobin Variant: CE-HPLC was performed with each blood sample on BIO-RAD Variant II using beta thalassemia short program pack consisting of elution buffer 1 (sodium phosphate), elution buffer 2 (sodium phosphate), whole blood primer (lyophilized human red blood cell hemolysate with preservative, HbA2/F, Calibrator (Diluent set), Analytical Cartridge sample vial and wash solution (with deionized water). The Variant II BIO-RAD is a fully automated HPLC system used to separate and determine the area percentage for hemoglobin A2 and F to provide a qualitative determination of abnormal hemoglobin. The different Hb is separated based on their ionic interaction with cartridge material (a buffer gradient of increasing ionic strength on cartridge. The separated Hb passed through two different filter photometers of absorbance at 415 nm and secondary wavelength at 690 nm to reduce the background absorbance. Different hemoglobin has a characteristic retention time with different peak in defined windows, displayed in a chromatogram of percentage absorbance versus time. The HbA2 levels ranging between 4 and 9% and between HbF 1 and 5% were diagnosed as beta thalassemia.

Genotype Analysis of Genetic Markers
Genotype studies for 11 SNPs of genetic modifier genes BCL11A, HBS1L-MYB and KLF1 were carried out by PCR-RFLP and Amplification Refractory Mutation System (ARMS PCR) ( Table 1).

Statistical Analysis
Statistical difference and genetic comparisons between groups were evaluated by using SPSS for Windows, version 16.0 (SPSS Inc., Chicago, IL, USA). Data were presented as mean ± standard deviation for continuous variables and number and percentages for discrete variables. Comparisons between quantitative variables were carried out after data were explored for normality using the Kolmogorov-Smirnov test of normality. Logistic regression was used for the computation of ORs and 95% CIs (Confidence Interval). Genotype and allele frequency distributions of the polymorphisms in the BCL11A, HBS1L-MYB and KLF1 gene were compared between patients with high (>40%) and low (<40%) HbF levels, and was analyzed by multiple linear regression to evaluate the contribution of each of the SNPs to the HbF% variation, with age and sex included as covariates. The demographic data were expressed as mean ± standard deviation (SD). The hematological parameters and HbF levels were presented as mean and standard deviation (SD). Data were found nonparametric for hematological parameters, the Mann-Whitney U test was used to see the difference of median between two groups with high (>40%) and low (<40%) HbF levels. Clinical data (hepatomegaly and transfusion therapy) from groups of patients with different HbF levels (>40% and <40%) were compared using the chi square test for categorical variables and t-test for continuous variables. Multiple linear regression was used to test for association of genetic markers with hematological parameters and HbF, considering age and sex as covariate. One-way ANOVA was used to compare the history of splenomegaly in groups of patients with high and low HbF levels. The p-values were considered to be significant at <0.05.

In Silico Analysis
The biological significance or functional impacts of the 4 noncoding single nucleotide variants (ncSNVs) of oxidative stress (genetic marker) were determined by using in silico tools (CADD phred and FATHMM pred) for variant effect prediction. For the CADD* C-Score, a cut-off value >10 indicated potential pathogenic variants. However, in FATHMM predictions were given as p-values in the range [0, 1]: values above 0.5 were predicted to be deleterious, while those below 0.5 were predicted to be neutral or benign. A p-value close to the extremes (0 or 1) was the highest confidence prediction that yielded the highest accuracy.

BCL11A rs1427407 (G>T) Polymorphism
The genotype frequencies of BCL11A rs1427407 (G>T) polymorphism in a group of patients with low HbF (<40%) and in a group of patients with high HbF (>40%) ( Table 2). We found a significant difference in genotype distributions of BCL11A gene, (rs1427407) for the heterozygous genotype (GT) with p-value = 0.034, OR = 0.321, 95% CI = 0.113 to 0.916) and homozygous mutant genotype TT with p-value = 0.039, OR = 0.329, 95% CI = 0.114 to 0.944) between patients' groups.  Figure 1 shows the gel pictures of wild type, heterozygous mutant and homozygous mutant genotypes for BCL11A gene, rs1427407 (G>T) polymorphisms. Sanger sequencing was performed to validate the results ( Figure 2). 0.644 Figure 1 shows the gel pictures of wild type, heterozygous mutant and homozygous mutant genotypes for BCL11A gene, rs1427407 (G>T) polymorphisms. Sanger sequencing was performed to validate the results ( Figure 2).

BCL11A rs766432 (A>C) Polymorphism
The genotype and allele frequencies of rs766432 (A>C) polymorphism present on BCL11A gene is shown in Table 2. A gel picture of BCL11A rs766432 (A>C) polymorphism is shown in Figure 3. Sanger sequencing was performed for the validation of the RFLP result ( Figure 4).

BCL11A rs766432 (A>C) Polymorphism
The genotype and allele frequencies of rs766432 (A>C) polymorphism present on BCL11A gene is shown in Table 2. A gel picture of BCL11A rs766432 (A>C) polymorphism is shown in Figure 3

BCL11A rs766432 (A>C) Polymorphism
The genotype and allele frequencies of rs766432 (A>C) polymor BCL11A gene is shown in Table 2. A gel picture of BCL11A rs766432 (A> is shown in Figure 3. Sanger sequencing was performed for the valid result ( Figure 4).

BCL11A rs11886868 (C>T) Polymorphism
The genotype and allele frequencies of the rs11886868 (C>T) polymorphism present on the BCL11A gene is mentioned in Table 2. A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 5. The Sanger sequencing performed for the validation of results is shown in Figure 6.

BCL11A rs11886868 (C>T) Polymorphism
The genotype and allele frequencies of the rs11886868 (C>T) polymorphism present on the BCL11A gene is mentioned in Table 2. A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 5. The Sanger sequencing performed for the validation of results is shown in Figure 6.

BCL11A rs11886868 (C>T) Polymorphism
The genotype and allele frequencies of the rs11886868 (C>T) polymorphism present on the BCL11A gene is mentioned in Table 2. A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 5. The Sanger sequencing performed for the validation of results is shown in Figure 6.

BCL11A rs4671393 (A>G) Polymorphism
The comparison of the distribution of the genotype and allele frequencies between the two groups with high and low HbF levels indicated that the genotypes AG (heterozygous) and GG (homozygous mutant) of the BCL11A gene, rs4671393 (A>G) polymorphism had no significant difference in the distribution of this variant between the two groups as p value is >0.05 (Table 2). A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 7. The Sanger sequencing performed for validation of results is shown in Figure 8.

BCL11A rs4671393 (A>G) Polymorphism
The comparison of the distribution of the genotype and allele frequencies between the two groups with high and low HbF levels indicated that the genotypes AG (heterozygous) and GG (homozygous mutant) of the BCL11A gene, rs4671393 (A>G) polymorphism had no significant difference in the distribution of this variant between the two groups as p value is >0.05 (Table 2). A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 7. The Sanger sequencing performed for validation of results is shown in Figure 8.

BCL11A rs4671393 (A>G) Polymorphism
The comparison of the distribution of the genotype and allele frequencies b the two groups with high and low HbF levels indicated that the genotypes AG (he gous) and GG (homozygous mutant) of the BCL11A gene, rs4671393 (A>G) po phism had no significant difference in the distribution of this variant between t groups as p value is >0.05 (Table 2). A gel picture of the PCR product of BCL11A rs10 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 7. The Sanger sequ performed for validation of results is shown in Figure 8.

BCL11A rs10189857 (A>G) Polymorphism
The genotype frequency distribution of the rs10189857 (A>G) BCL11A gene variant compared between patients with low and high levels of HbF (Table 2). A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 9. The Sanger sequencing performed for validation of results is shown in Figure 10.

BCL11A rs10189857 (A>G) Polymorphism
The genotype frequency distribution of the rs10189857 (A>G) BCL11A gene variant compared between patients with low and high levels of HbF (Table 2). A gel picture of the PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% agarose gel is shown in Figure 9. The Sanger sequencing performed for validation of results is shown in Figure 10.

BCL11A rs10189857 (A>G) Polymorphism
The genotype frequency distribution of the rs10189857 (A>G) BCL11A g compared between patients with low and high levels of HbF (Table 2). A gel p PCR product of BCL11A rs10189857 (A>G) polymorphisms run on 2% ag shown in Figure 9. The Sanger sequencing performed for validation of results Figure 10.

BCL11A rs7557939 (G>A) Polymorphism
We investigated the impact of the BCL11A rs7557939 (G>A) variant in patients with low and high HbF levels ( Table 2). A gel picture of PCR product of BCL11A rs7557939 (G>A) polymorphisms run on 2% agarose gel is shown in Figure 11. Sanger sequencing performed for validation of results is shown in Figure 12.

BCL11A rs7557939 (G>A) Polymorphism
We investigated the impact of the BCL11A rs7557939 (G>A) variant in patients with low and high HbF levels ( Table 2). A gel picture of PCR product of BCL11A rs7557939 (G>A) polymorphisms run on 2% agarose gel is shown in Figure 11. Sanger sequencing performed for validation of results is shown in Figure 12.

BCL11A rs7557939 (G>A) Polymorphism
We investigated the impact of the BCL11A rs7557939 (G>A) variant in p low and high HbF levels ( Table 2). A gel picture of PCR product of BCL11 (G>A) polymorphisms run on 2% agarose gel is shown in Figure 11. Sanger performed for validation of results is shown in Figure 12.

HBS1L-MYB rs9399137 (T>C) Polymorphism
The comparison of the distribution frequencies of genotype and allele between the two groups ( Table 2). A gel picture of PCR product of HBS1L-MYB rs9399137 (T>C) polymorphisms run on 2% agarose gel is shown in Figure 13.

HBS1L-MYB rs9376090 (T>C) Polymorphism
The genotype frequencies of HBS1L-MYB rs9376090 (T>C) polymorphism in the group with low HbF levels and in the group with high HbF (Table 2). On comparing the genotypes and allele frequencies between two groups, we observed no significant difference for the HBS1L-MYB rs9376090 (T>C) variant with a p > 0.05. A gel picture of PCR

HBS1L-MYB rs9399137 (T>C) Polymorphism
The comparison of the distribution frequencies of genotype and allele between the two groups ( Table 2). A gel picture of PCR product of HBS1L-MYB rs9399137 (T>C) polymorphisms run on 2% agarose gel is shown in Figure 13.

HBS1L-MYB rs9399137 (T>C) Polymorphism
The comparison of the distribution frequencies of genotype and allele betwe two groups (Table 2). A gel picture of PCR product of HBS1L-MYB rs9399137 (T>C) morphisms run on 2% agarose gel is shown in Figure 13.

HBS1L-MYB rs9376090 (T>C) Polymorphism
The genotype frequencies of HBS1L-MYB rs9376090 (T>C) polymorphism group with low HbF levels and in the group with high HbF (Table 2). On comparin genotypes and allele frequencies between two groups, we observed no significant ence for the HBS1L-MYB rs9376090 (T>C) variant with a p > 0.05. A gel picture o

HBS1L-MYB rs9376090 (T>C) Polymorphism
The genotype frequencies of HBS1L-MYB rs9376090 (T>C) polymorphism in the group with low HbF levels and in the group with high HbF (Table 2). On comparing the genotypes and allele frequencies between two groups, we observed no significant difference for the HBS1L-MYB rs9376090 (T>C) variant with a p > 0.05. A gel picture of PCR product run on 2% agarose gel is shown in Figure 14. The Sanger sequencing performed for validation of results is shown in Figure 15.
Thalass. Rep. 2023, 13, FOR PEER REVIEW product run on 2% agarose gel is shown in Figure 14. The Sanger sequencing perf for validation of results is shown in Figure 15.  Thalass. Rep. 2023, 13, FOR PEER REVIEW 15 product run on 2% agarose gel is shown in Figure 14. The Sanger sequencing performed for validation of results is shown in Figure 15.

HBS1L-MYB rs28384513 (A>C) Polymorphism
The genotype frequency distribution of HBS1L-MYB rs28384513 (A>C) polymorphism between high and low HbF groups. The statistical analysis shows that genotypes (AA, AC, CC) and allele C are not associated with fluctuations in HbF level ( Table 2). A gel picture of the PCR product run on 2% agarose gel is shown in Figure 16. The Sanger sequencing performed for validation of results is shown in Figure 17.

HBS1L-MYB rs28384513 (A>C) Polymorphism
The genotype frequency distribution of HBS1L-MYB rs28384513 (A>C) polymorphism between high and low HbF groups. The statistical analysis shows that genotypes (AA, AC, CC) and allele C are not associated with fluctuations in HbF level (Table 2). A gel picture of the PCR product run on 2% agarose gel is shown in Figure 16. The Sanger sequencing performed for validation of results is shown in Figure 17.

KLF1 rs2072597 (A>G) Polymorphism
Genotype frequency distributions for KLF1 rs2072597 (A>G) polymorphism in patients with two groups, low and high HbF levels, are presented in Table 2. The representative gel picture of KLF1 rs2072597 (A>G) polymorphism showing three genotypes is pre-

KLF1 rs2072597 (A>G) Polymorphism
Genotype frequency distributions for KLF1 rs2072597 (A>G) polymorphism in patients with two groups, low and high HbF levels, are presented in Table 2. The representative gel picture of KLF1 rs2072597 (A>G) polymorphism showing three genotypes is presented in Figure 18. The Sanger sequencing was performed to validate the results is shown in Figure 19.

KLF1 rs112631212 (T>G) Polymorphism
The genotype and allele frequencies of rs112631212 (T>G) polymorphism, pre KLF1 gene is shown in Table 2. The representative gel picture and Sanger sequenc

KLF1 rs112631212 (T>G) Polymorphism
The genotype and allele frequencies of rs112631212 (T>G) polymorphism, present on KLF1 gene is shown in Table 2. The representative gel picture and Sanger sequencing results are shown in Figures 20 and 21, respectively.

KLF1 rs112631212 (T>G) Polymorphism
The genotype and allele frequencies of rs112631212 (T>G) polymorphism, present on KLF1 gene is shown in Table 2  Evaluation of hematological parameters in two groups of TM patients with lo high HbF levels.
The hematological data of individuals in the two groups are summarized in T All patients were treated with Deferasirox (Asunra 400 mg), an orally administere chelation agent. There was no significant difference in Hb, mean corpuscular v (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin conc tion (MCHC) and HbA2 level except the HbF level between the two groups with hig low HbF levels. The mean of hemoglobin, MCV, MCH, MCHC, HbA2 and the ferriti of the 100 -thalassemia major patients were 6.9 ± 1.3, 67.7 ± 5.1, 22.6 ± 3.8, 28.6 ± 3  Evaluation of hematological parameters in two groups of TM patients with low and high HbF levels.
The hematological data of individuals in the two groups are summarized in Table 3. All patients were treated with Deferasirox (Asunra 400 mg), an orally administered iron chelation agent. There was no significant difference in Hb, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC) and HbA2 level except the HbF level between the two groups with high and low HbF levels. The mean of hemoglobin, MCV, MCH, MCHC, HbA2 and the ferritin level of the 100 -thalassemia major patients were 6.9 ± 1.3, 67.7 ± 5.1, 22.6 ± 3.8, 28.6 ± 3.6 and 546.1 ± 379.2, respectively. The range of HbF in patients was 3.1-95.0%. Only twenty patients were found to have an HbF level below 10%. All other patients in the low HbF group had a value of HbF between 3.1 and 40.1%. From the total patient group, the HbF Evaluation of hematological parameters in two groups of TM patients with low and high HbF levels.
The hematological data of individuals in the two groups are summarized in Table 3. All patients were treated with Deferasirox (Asunra 400 mg), an orally administered iron chelation agent. There was no significant difference in Hb, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC) and HbA2 level except the HbF level between the two groups with high and low HbF levels. The mean of hemoglobin, MCV, MCH, MCHC, HbA2 and the ferritin level of the 100 β-thalassemia major patients were 6.9 ± 1.3, 67.7 ± 5.1, 22.6 ± 3.8, 28.6 ± 3.6 and 546.1 ± 379.2, respectively. The range of HbF in patients was 3.1-95.0%. Only twenty patients were found to have an HbF level below 10%. All other patients in the low HbF group had a value of HbF between 3.1 and 40.1%. From the total patient group, the HbF pretransfusion level was >40% in 62 patients and <40% in 38 patients. We found a significant variation for HbF and HbA levels between two groups with p-value < 0.05.

Comparison of Clinical Parameter between Two Groups of TM Patients
Clinical parameters such as hepatomegaly, splenomegaly and transfusion history were compared between the groups of patients with high and low HbF levels (Table 4). From the study groups, 14 patients had hepatomegaly and 86 patients had splenomegaly. There was no association of HbF levels with hepatomegaly p value = 0.849 and splenomegaly p value = 0.113 between TM major patients of two groups with low and high HbF levels.
We obtained the details of time intervals between two consecutive transfusions in the TM patients. In this study, we found that patients who have shown low HbF levels need transfusions at shorter intervals of time (i.e., at 3 weeks) than patients with high HbF levels (transfused at 4 weeks) and this relationship was statistically significant with a p-value = 0.013.

Influence of Genetic Modifier Variants on HbF and on Hematological Parameters
HbF levels were strongly influenced by two variants on the BCL11A gene, (rs11886868, rs1427407) out of eleven tested modifier gene variants. The HbF-promoting alleles at BCL11A (rs11886868, rs1427407) had the strongest effect on HbF levels with a Beta value-0.200, p value-0.038 and a Beta value-0.215, p value-0.028, respectively ( Table 5). The HbF-promoting polymorphism of BCL11A gene (rs4671393) among all genotyped genetic markers of modifier genes for increased HbF was positively associated with increased Hb levels, but no change were observed in MCV, MCH and MCHC values (p > 0.05). However, another variant on the KLF1 gene, (rs2072597) had a significant effect on HbA levels (Beta value-0.266, p value-0.016). All the values for each hematological parameter were adjusted for HbF, and age and gender were included as covariates. Table 5. Regression coefficient (Beta) estimates and significance p value (in brackets). Age and gender were included as covariates. For the genetic data, Beta serves as a measure of the effect of an allele change from low-HbF to high-HbF allele. Nominally significant effects are in bold font.

Age of Onset of the Disease and Age at First Blood Transfusion among TM Patients with Low and High HbF Levels
In the group of patients with low HbF levels (<40%) 16 (30.2%), patients were diagnosed at less than 1 year of age whereas 37 (69.8%) were diagnosed at an age of more than 1 year. However, 25 (53.2%) patients were diagnosed at the >1 year age in the group with high (>40%) HbF whereas 22 (46.8%) were those who were diagnosed at <1 year of age. On comparing the p-value for the age of diagnosis before and after 1 year between TM patients with high and low HbF levels we have not found a significant p-value 0.066. A total of 16 (32.0%) TM patients received their first blood transfusion before one year of age and 34 (68.0%) after one year of age in the low (<40%) HbF levels group. However, in 28 (56.0%) patients, transfusion was required after 1 year age whereas 22 (44.0%) were those who were transfused at <1 year of age in group among the high (>40%) HbF levels group. For the age at first transfusion the p value (0.151) was not found to be significant.
A correlation was also evaluated between patient groups who were diagnosed and first transfused before or after 1 year of age with selected genetic markers of modifier genes (Tables 6 and 7) among TM patients with high and low HbF levels. Among the 11 selected genetic markers of the modifier gene, none of the polymorphism of the BCL11A, HBS1L-MYB, KLF1 gene, showed a significant p-value <0.05 among TM patients with high and low HbF levels group (Tables 6 and 7). This shows that the presence of polymorphism on the modifiers' gene is not associated with variation in age of onset and need for blood transfusions in TM patients.

In Silico Analysis for All Variants of Modifier Genes
The biological significance or functional impacts of the 10 noncoding single nucleotide variants (ncSNVs) and 1 missense variant of modifier genes (genetic marker) was determined using in silico tools (CADD phred and FATHMM pred) for variant effect prediction. For CADD C-Score, a Cut-off value >10 indicates potential pathogenic variants. However, in FATHMM, predictions are given as p-values in the range [0, 1]: values above 0.5 are predicted to be deleterious, while those below 0.5 are predicted to be neutral or benign. p-values close to the extremes (0 or 1) are the highest confidence predictions that yield the highest accuracy. We have found that out of 11 ncSNV, 1 ncSNV on BCL11A, 1 ncSNV on the HBS1L-MYB gene and 1 missense variant on the KLF1 gene, (rs112631212) were predicted to be pathogenic using the CADD and FATHMM tool (Table 8). Moreover, in the present study we found an association of BCL11A rs1427407 (G>T) polymorphism with increased HbF levels in beta thalassemia major patients. Thus, evaluation of the status of genetic marker of modifier genes and prediction of their pathogenic effect are useful in TM patients for better management of the disease.

Discussion
Beta-thalassemia is one of the most common hereditary disorders caused by a defective synthesis of β-globin chain or an imbalanced α/β chain ratio [31].
The clinical manifestations of β-thalassemia are extremely heterogeneous, ranging from the transfusion-dependent state to asymptomatic carrier condition [9]. It is now known that residual HbF synthesis in the adult stage is inheritable as a quantitative trait. Modifying factors that reduce the imbalance in αvs. non α-hemoglobin chains are the genetic determinants of the clinical severity of β-thalassemia as well as SCA. The reduced globin chain imbalance allows the selective survival of the erythroid precursors and thereby reduces the unproductive erythropoiesis [32,33]. Moreover, the presence of genetic polymorphism in different modifier genes plays an important role in elucidating and ameliorating the pathogenesis of diseases. Thus, the present study was designed to evaluate the contribution of BCL11A, the HBS1L-MYB intergenic region and KLF1 gene polymorphisms on the variation in HBF levels in β-TM patients.
BCL11A gene that encodes zinc finger protein plays a central role in the silencing of HbF expression in humans, through binding the key regions in β-globin gene cluster and forming a HbF repressor complex (nucleosome remodeling and histone deacetylase, NuRD) to silence γ-globin gene expression [11,34,35].
It was previously reported that several polymorphisms within intron 2 of the BCL11A gene have been strongly associated with Hb F levels in different ethnic populations, reflecting the heterogeneity of the allele frequencies among different ethnic groups: rs766432, rs4671393, rs1427407 and rs11886868 [11,13].
The association of different polymorphisms present on the BCL11A gene, with variation in HbF levels, has been widely investigated, but interestingly, the conclusions reported in these studies varied significantly among different ethnic populations, as some authors have described the presence of associations for rs766432, rs4671393, rs1427407, rs11886868, rs10189857 and rs7557939 [11,22,[36][37][38][39] variants and others have described the absence of associations for rs766432, rs4671393, rs1427407, rs11886868, rs10189857 [39,40] and rs7557939 [33,38] polymorphisms with variation in HbF levels in thalassemia major disease.
In this study, we examined the status of selected SNPs present on the BCL11A gene in β-thalassemia major patients to determine their association with variations in HbF levels. The selection of the SNPs was based on recently published studies in which these genetic variants were most strongly associated with increased HbF levels in TM patients [33,[36][37][38][39][40]. We found that four (rs1427407, rs766432, rs11886868, rs7557939) out of the six selected SNP variants in the BCL11A gene showed a strong association with increased HBF levels in TM patients. On comparing individuals who have shown an absence of the above mentioned risk genotypes with those TM patients who carry risk genotypes of SNPs (rs11886868 rs7557939; rs1427407 and rs766432) on BCL11A gene have shown a gradual increase in HbF levels.
These findings were in concordance with the previous study reported by Rashiya et al., 2020; Chamouine et al., 2020 where the increase in HbF levels was observed for the BCL11A gene polymorphism with rs11886868 rs7557939; rs1427407 and rs766432 [38,39]. In contrast, they differed from the results of Bhanushali et al., 2016; who reported no association between the intronic region BCL11A gene polymorphism and variation in HbF levels [40]. Moreover, on in silico analysis by using CADD and FATHMM, only one (rs1427407) variant out of the six noncoding variants of the BCL11A gene was predicted to be pathogenic and deleterious in nature.
On comparing the hematological parameters among TM patients, we found a significant variation for HbF and HbA levels in group of patients with high HbF (p-value < 0.05). Moreover, an analysis of status of clinical parameters among TM patients with low and high HbF levels was performed and we found that patients who have high HbF levels need transfusion in a shorter duration of time (i.e., at 3 weeks) than patients with low HbF levels (transfused at 4 weeks) and this relationship was found to be statistically significant. Furthermore, the polymorphism of BCL11A gene (rs4671393) among all genotyped genetic markers of modifier genes was positively associated with increased Hb levels, but no change were observed in (red cell indices) MCV, MCH and MCHC values (p-value > 0.05). However, other variants on HB1L-MYB and KLF1 genes have not shown any significant effect on HbA levels. We believe that the significant effect of the modifier loci on general blood traits (Hb) shown represent a combination of both, disease amelioration through HbF modification and pleiotropic effects, and that the mechanisms underlying both phenomena are diverse and gene-specific.
Among the 11 selected genetic markers of modifier genes, none of the polymorphisms of the BCL11A, HBS1L-MYB, KLF1 gene, showed a significant p-value for age of onset of disease and age at first blood transfusion among TM patients in the two groups. This shows that the presence of polymorphism on the modifiers' gene was not associated with factors such as age of onset and need for blood transfusions in TM patients. This has also been noted in other studies [41][42][43], which state that the age at diagnosis and age at first transfusion may vary according to the awareness of the family and economic status. Therefore, it is expected that there are some other genetic factors which also affect the clinical phenotype of thalassemia and the role of which needs to be studied further.
Thus, to conclude, some of BCL11A gene variants contribute to an increase in HbF levels by affecting BCL11A gene expression in TM patients. Hence, it was important to explore the therapeutic targets of BCL11A gene for reactivation of HbF in patients with β-hemoglobin disorders.
The next modifier of HbF levels is the HBS1L-MYB gene which is mainly expressed in the erythroid precursor cells [16]. Low levels of MYB are associated with reduced cell expansion and accelerated erythroid differentiation, suggesting that variation in the intrinsic levels of MYB might affect HbF by its effect on the cell cycle [6]. Polymorphisms in this region accounted for 17.6% variance in normal Europeans and 3 to 7% of the trait variance in African-American and Brazilian patients with sickle cell anemia [4,44].
In this study, we genotyped three SNPs (block 1; rs9399137, rs28384513, rs9376090), out of which only one SNP, i.e., rs9399137, was significantly associated with increased HbF levels in TM patients. For SNP HBS1L-MYB (rs9399137), our results showed that the frequency of CC (homozygous mutant) genotype was significantly higher than that of the genotype TT (wild type). Additionally, the C allele frequency of this SNP was significantly higher in group with high HbF levels when compared to the T allele. Our results are consistent with those reported by previous studies [11,26,40,45]. In addition, the frequency of HBS1L-MYB for the rs28384513 and rs9376090 polymorphism were not associated with variations in HbF level in TM patients and these findings are consistent Thus, we conclude that the HBS1L-MYB gene variant (rs9399137) has a contribution in the increase in HbF levels in TM patients. Conversely. numerous studies confirmed the effect of rs9399137 on HbF levels; however, its role on molecular mechanisms regulating HbF levels is still not clear and needs to be explored. KLF1 (formerly known as EKLF Erythroid Krüppel-like factor) is an erythroid-specific transcription factor critical for erythropoiesis and hemoglobin switching [46]. It was previously reported that reduced expression of KLF1 results in downregulation of BCL11A and an increased γto β-globin ratio [10,47]. Any polymorphisms in the KLF1 gene causing haploinsufficiency for the KLF1 gene is associated with increased HbF synthesis and the amelioration of severity of homozygous β0-thalassemia, hence manipulating KLF1 expression is possibly another approach for activating HbF in β-thalassemias [29,48,49].
In the Leiden Open Variation Database (LOVD), till the present date > 60 genomic variations in the KLF1 gene have been reported and depending on the location of the variation, distinct hematological phenotypes have been explained [49]. Moreover, the NCBI dbSNP database showed that the c.304T>C mutation corresponded to a previously reported SNP (rs2072597) and was unlikely pathogenic; this was consistent with Borg et al. who reported the identification of a defective KLF1 allele in a Maltese family which caused the persistence of high levels of fetal hemoglobin in adults-a benign and usually asymptomatic condition known to alleviate the severity of β-thalassemia and sickle cell disease [50]. In our study, SNP (rs2072597) was not associated with alteration in HBF levels and our results were consistent with previously published studies by Muszlak  On performing in silico analysis by using the CADD and FATHMM tool, the missense variant on KLF1 gene, (rs112631212) was predicted to be pathogenic. Thus, evaluation of the status of genetic marker of modifier genes and prediction of their pathogenic effect are useful in TM patients for better management of the disease. The detailed functional mechanism among these genetic markers of modifiers genes and HbF levels need be further identified and studied.

Conclusions
In β-thalassemia disease there is a clinical range from mildly affected such as nontransfusion dependent thalassemia (NTDT) to severely affected such as Cooley's anemia. The heterogeneity of HbF levels in β-thalassemia is associated with the disease severity. So, induction of HbF synthesis can ameliorate the clinical severity of β-thalassemia by reducing the degree of imbalance in αto non-α-globin chains. The level of HbF in βthalassemias is controlled by genetic modifiers. Therefore, in our present study, we analyzed genetic association including 11 SNPs at 3 main loci (BCL11A, HBS1L-MYB and KLF1) affecting HbF levels and demonstrated that 5 SNPs were independently associated with higher HbF levels (rs1427407, rs766432, rs11886868, rs7557939 in BCL11A, and rs9399137 in HBS1L-MYB) with p-value <0.05. Our study demonstrates the importance of examining polymorphism in genetic determinants that play important roles in ameliorating the severity of clinical symptoms and complications of thalassemia by targeting genes involved in the upregulation of HbF. Stimulation of HbF production can be used as an option to treat thalassemia patients. The present study validates the relationship between HbF levels and SNPs in a BCL11A erythroid-specific enhancer, HBS1L-MYB and KLF1 among thalassemia patients.
More research in β-thalassemia is needed to understand the regulatory mechanisms underlying the variation in clinical outcomes, HbF production and response to HbF induction therapy to improve the management of patients. So, future studies should be designed to evaluate in vivo novel strategies with multitarget effects on erythropoiesis with the final goal to understand the status of anemia in beta-thalassemia major patients.