Fetal hemoglobin (HbF) is a tetrameric molecule composed of two alpha-globin and two gamma-globin chains ($\alpha$2$\gamma$2) typically expressed during fetal development. It starts to decrease just before birth when it is replaced by adult hemoglobin A (HbA; $\alpha$2$\beta$2), which has a lower affinity for oxygen [1]. In healthy adults, red blood cells predominantly contain HbA, along with 2.5% to 3.5% HbA2 ($\alpha$2$\delta$2) and less than 1% HbF, which is unevenly distributed among erythrocytes. However, about 10% to 15% of normal adults have elevated levels of HbF that can reach up to 5.0%. This condition, known as hereditary persistence of fetal hemoglobin (HPFH), can be caused by deletions within the HBB gene cluster on chromosome 11, point mutations in the promoters of the HbF genes HBG1 and HBG2, or variations in major repressors of the HbF genes, such as BCL11A and MYB [2, 3].

In adults of African descent with sickle cell disease (SCD), HbF typically comprises 4% to 10% of the total hemoglobin levels, although it rarely exceeds 30% [2]. In contrast, African Americans with sickle cell trait (HbAS) have a mean HbF level of 1.4% [4]. While the increase in HbF levels has no consequences in healthy adults, it can provide clinical benefits in patients with SCD and $\beta$-thalassemia [5]. Therefore, understanding and managing the biological pathways associated with HbF expression could lead to clinical interventions for individuals with severe $\beta$-hemoglobinopathies by genetic modification of HbF expression [6, 7].

Research has shown that variations in HbF levels in adults are significantly influenced by genetic factors [5, 8]. One of the most studied variants is the XmnI restriction site polymorphism located in the promoter region of the HBG2 gene, upstream of the transcription start site (–158C$>$T) (rs7482144). This polymorphism is known to impact HbF levels, particularly under conditions of increased erythropoietic stress, such as in patients with $\beta$-thalassemia and SCD patients [7, 8]. Some patients with SCD have exceptionally high levels of HbF associated with the Senegal and Saudi-Indian haplotypes, both characterized by the presence of the rs7482144 [T] allele [5]. It is known that only the HBG2 expression is affected by the –158 polymorphism; however, the transcription factors binding to the –158 HBG2 motif remain uncharacterized [8].

Additionally, minor alleles of several single nucleotide polymorphisms (SNPs) in two other quantitative trait loci (QTL), the BCL11A gene on chromosome 2 and the HBS1L-MYB (HMIP) intergenic region on chromosome 6, are also strongly associated with HbF regulation [5]. Both genes BCL11A and MYB encode repressors of $\gamma$-globin genes [8]. The transcription factor BCL11A binds TGACCA motifs in the HBB gene cluster, preferentially to position –115 in the HBG2 and HBG1 gene promoters [8]. Functional variants of the BCL11A gene are marked by the intron 2 SNP rs1427407 which marks an erythroid-specific enhancer [8, 9]. The HMIP intergenic variants affect regulatory elements that are occupied by key erythroid transcription factors within this region, affecting long-range interactions with MYB and in consequence the MYB expression levels [10]. The most significant functional motif accounting for HMIP modulation of HbF is rs66650371, a 3-bp deletion polymorphism, which is in complete LD with the SNP rs9399137 shown by several Genome-Wide Association Studies to be most significantly associated with HbF in Europeans, Asians, and African Americans [4, 9].

Variants in these three QTLs have been reported to contribute up to 50% of HbF variation [11]. The increase of HbF remaining unexplained could be a result of rare alleles or polymorphic variations in other genes coding for protein factors involved in the regulation of HbF production, such as KLF1 that activates expression of the BCL11A gene [12] or other still uncharacterized HbF repressors. Moreover, the impact of these variants can be different from one population to another, depending on how common a risk allele is in a population. For example, in European populations, a strong HbF influence is known for variants at HMIP, which are rare in the Sub-Saharan populations and hence have a low impact on HbF levels in such populations [11].

The high prevalence of SCD in Africa justifies genetic studies linked to HbF regulation in African populations. Although many HbF association studies were implemented in SCD patients of African ancestry [13], a large proportion of the HbF variance remains unexplained. Additional studies in healthy populations could highlight different genetic factors or pathophysiological pathways associated with HbF regulation. Therefore, the present study examines for the first time an adult population sample of the São Tomé e Príncipe archipelago in Central Africa for the potential association with HbF levels of several SNPs in the HBB gene cluster, BCL11A gene, and HMIP intergenic region.

Characterizing genetic variants associated with HbF regulation is a paramount topic because it has been established that the induction of higher levels of HbF by genome-editing strategies or pharmacological inducer agents can improve the clinical and hematological features of severe hemoglobinopathies such as SCD and $\beta$-thalassemia [6].

In the present study, we replicated in a population sample of HbAA women from São Tomé e Príncipe the significant associations with HbF levels of the two well-known BCL11A variants rs11886868 T$>$C and rs1427407 G$>$T. The study of a population group of 98 HbAA women in a case-control framework using logistic regression revealed that the minor allele frequencies of the two BCL11A polymorphisms rs11886868 [C] and rs1427407 [T] were significantly higher in the group with higher levels of HbF. In concordance, homozygous individuals for the minor allele present significantly higher HbF levels compared with heterozygous and homozygous for the major allele. Despite the higher levels of HbF in two homozygous individuals for the minor allele in the HbAS women, no significant association was found between the minor allele of the two BCL11A SNPs and high HbF levels using linear regression in the additive, dominant, or recessive models.

SNPs in the intronic BCL11A enhancer were implicated in HbF regulation by many genome-wide and candidate association studies, addressing mainly SCD and $\beta$-thalassemia patients [18, 19, 20], but also $\beta$-thalassemia carriers [16, 21, 22] or even normal healthy Caucasian individuals with common forms of hereditary persistence of fetal hemoglobin (HPFH) [21, 23]. The transcription factor BCL11A is a major repressor of $\gamma$-globin gene expression and switching from fetal to adult hemoglobin in erythroid cells of mammals by directly binding the fetal globin (HBG1/2) gene promoters and other sites in the $\beta$-globin gene cluster [24, 25]. Increased levels of HbF are present with the downregulation of BCL11A in adult erythroblasts and cell lines expressing adult hemoglobin [8]. Moreover, the BCL11A binding site disruption in the $\gamma$-globin promoters increases $\gamma$-globin gene expression and HbF production [3]. Because of these results, the down-regulation or blockage of BCL11A has been a guide for new targeted HbF induction therapies [7, 26].

The HBG2-XmnI polymorphism (–158C$>$T) (rs7482144) in HbAA women, showed significantly higher values of HbF in the homozygous genotypes for the minor allele T in comparison with heterozygous and homozygous for the major allele. However, as only three homozygous TT individuals were found in this group, this result should be interpreted with caution. No signals of association between rs7482144 and HbF was observed in women with HbAS genotypes. It is well established that the HBG2-XmnI polymorphism is strongly associated with increased levels of HbF and can improve the phenotypes in response to stress erythropoiesis in conditions like $\beta$-thalassemia and sickle cell anemia [5]. The effect on HbF levels linked to the XmnI polymorphism is minimal or undetectable in healthy adults of Caucasian ancestry [21, 23]. Consequently, the significant or near-significant association of the XmnI minor allele [T] with elevated HbF levels observed in HbAA individuals in São Tomé and Príncipe may indicate some form of erythropoietic stress.

In the association tests for the BGLT3 SNP rs7924684 C$>$T, significant and near-significant results were found linking the major [C] allele to higher levels of HbF in women with the HbAA genotype. These findings align with previous studies that reported significant associations between the ancestral [C] allele of SNP rs7924684 and elevated levels of HbF among Portuguese $\beta$-thalassemia carriers [27] and Angolan children with SCD [28]. The BGLT3 gene, which encodes a long non-coding RNA (lncRNA), serves as a positive regulator of $\gamma$-globin genes expression [29]. This regulation may occur through various mechanisms, including looping to the $\gamma$-globin genes and recruiting coregulators via the BGLT3 transcript [29].

For the two remaining SNPs here analyzed, rs66650371 and rs4895441, located in the HMIP intergenic region, no significant associations were observed with HbF levels in the HbAA or HbAS groups. Several SNPs linked to increased HbF levels have been identified in this genomic region on chromosome 6q23, particularly within the contexts of $\beta$-thalassemia and SCD [5, 30]. Variants in these loci have been reported to contribute to 20–50% of the variability in HbF levels among non-African populations; however, the impact of these variants varies from one population to another. As the HMIP variants, such as rs66650371 and rs4895441, are less frequent in Sub-Saharan African populations, as was also observed herein for the study sample of São Tomé e Príncipe showing lower MAFs of 0.036 and 0.046, respectively, in HbAA women, and 0.032 and 0.064, respectively, in the HbAS women, this can explain the limited impact on HbF levels in that populations.

In conclusion, we successfully replicated in HbAA women from São Tomé e Príncipe (Central Africa), the known associations with HbF variation of the two BCL11A variants rs11886868 and rs1427407. No evidence of association was found between levels of HbF and the two HMIP variants rs66650371 and rs4895441, which can be explained by the lower MAF from these variants in the study population. For the variants located at the HBB gene cluster, the BGLT3 SNP rs7924684 C$>$T showed signs of an association between the major C allele and higher levels of HbF in women with the HbAA genotype. A signal of association with increased levels of HbF was observed for the minor allele of the XmnI polymorphism rs7482144 in HbAA women. In the group of women with HbAS, no significant associations were found between HbF levels and the six analyzed SNPs in BCL11A, HMIP, and HBB gene cluster. This lack of association may be attributed, in part, to the small sample size of only 47 individuals, in particular given the low frequency of homozygous individuals for the minor alleles of the two BCL11A variants rs11886868 and rs1427407.

The authors confirm that the findings of this study are available within the article and its supplementary materials. The data supporting the findings of this study are available from the first author upon reasonable request.

CB, GQ and CM collected samples, performed blood analyses, and interpreted the hematological data. LR performed laboratory blood analysis. AMM, SMA and IS performed molecular and statistical analysis. LM designed the research study, interpreted the molecular and statistical data, wrote, and edited the manuscript. All authors read and approved the final manuscript. All authors contributed to editorial changes in the manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The study was approved by the Ethics Committee of São Tomé e Príncipe, identified as CESIC Case PC022_2022. All participants signed an informed consent form before participating in the study. All data were anonymized with respect to confidentiality and processed according to the ethical principles outlined in the Declaration of Helsinki.

This work was supported by the Fundação para a Ciência e a Tecnologia (FCT) under the institutional grant UIDB/00283/2020 and by Forum Hematologico (CHUC).

The authors declare no conflict of interest. Given his role as the Editorial Board member, Licínio Manco had no involvement in the peer-review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Gustavo Caetano-Anollés.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/FBS38388.