INTRODUCTION — Fetal hemoglobin (hemoglobin F, HbF, alpha2gamma2) is the major hemoglobin present during gestation; it constitutes approximately 60 to 80 percent of total hemoglobin in the full-term newborn. In individuals without hemoglobinopathies, it is almost completely replaced by adult hemoglobin (hemoglobin A, HbA, alpha2beta2) by approximately 6 to 12 months of age, and it amounts to less than 1 percent of total hemoglobin in adulthood.
As a minor hemoglobin following the first year of life, HbF has little in the way of clinical relevance in normal physiology. However, it is assuming ever greater importance in certain of the hemoglobinopathies, in which congenital, acquired, and drug-induced increases in HbF have been shown to improve the clinical features of affected individuals with sickle cell disease and beta thalassemia. This important subject is discussed in depth separately. (See "Hydroxyurea use in sickle cell disease".)
The biology of HbF in health and disease will be discussed here. Reviews of the other normal hemoglobins (eg, hemoglobin A, hemoglobin A2) and the most common variants affecting globin genes are presented separately. (See "Structure and function of normal hemoglobins" and "Overview of the clinical manifestations of sickle cell disease" and "Diagnosis of thalassemia (adults and children)" and "Hemoglobin variants including Hb C, Hb D, and Hb E".)
BIOLOGY OF FETAL HEMOGLOBIN
Evolution — Hemoglobin evolved from ancient hemoproteins by gene duplication, gene conversion (non-reciprocal exchange of genetic material between two linked homologous genes), translocation to different chromosomes, and mutations that caused changes in the primary structure and properties of globin and their various genetic regulatory regions [1].
●Hemoglobin alpha (HBA1 and HBA2) and beta (HBB) gene clusters diverged from their predecessors approximately 450 million years ago, with modern adult hemoglobin becoming a heterotetramer of the products of these two genes (alpha2beta2).
●The gamma-globin genes (HBG1 and HBG2), which produce gamma-globin chains used in HbF and the delta-globin gene (HBD), which produce delta globin used in HbA2 (alpha2delta2), were the products of duplications of HBB.
Structure and function of globin genes — Human globin genes are present in two separate clusters on chromosomes 11 and 16 (figure 1).
●The alpha-globin genes and their related embryonic zeta-globin genes (HBZ) are near the telomere of the short arm of chromosome 16 (16p13.3).
●The embryonic epsilon-globin gene (HBE), the two gamma-globin genes (HBG1 and HBG2), HBD, and HBB are all found on the short arm of chromosome 11 (11p15.5). The closely related and linked gamma-globin genes are 5' HBG2 (encoding G-gamma-globin protein) and 3' HBG1 (encoding A-gamma-globin).
●The two gamma-globin genes HBG2 and HBG1 have the intron-exon structure of all human globins, and their exons are nearly identical except at codon 136 where HBG2 codes for glycine, while HBG1 codes for alanine [2]. This striking similarity in protein sequence and gene structure is due to gene duplication and gene conversion [3].
Gamma-globin chains differ from beta-globin chains in either 39 or 40 amino acid residues, depending on whether a glycine or alanine is present at the gamma codon 136 position [3]. Most of the differences lie at the surface of the tetramer and therefore have little functional importance. Two differences in the alpha1/gamma1 interface (gamma 112 threonine, which is a cysteine residue in HBB, and gamma 130 tryptophan, which is a tyrosine in HBB) might contribute to the resistance of HbF to alkali compared with HbA, and provide the structural basis for measuring HbF by the historical method of alkali denaturation.
●A common polymorphism is found in HBG1, in which threonine replaces isoleucine at codon gamma 75 [4]. The HBG1 variant with isoleucine at codon gamma 75 is known as HbF-Sardinia.
●The proteins encoded by HBG2 and HBG1 are identical in their function and in their capacity to inhibit sickle hemoglobin (HbS) polymerization. (See 'F cells in sickle cell anemia' below.)
●Structural differences at amino acid positions 1, 5, and 7 in the amino-terminal A helix of gamma- compared with beta-globin chains account for the greater strength of the HbF tetramer compared with HbA [5]. This difference might explain, in part, the protective effect of fetal red cells and all erythrocytes with increased HbF against malaria, as a stronger hemoglobin tetramer less readily dissociates into digestible dimers [6]. Many other mechanisms of malaria resistance conferred by HbF might also be operative [7]. (See "Protection against malaria in the hemoglobinopathies", section on 'Neonatal red cells and hemoglobin F'.)
●The alpha2/gamma2 interface is identical to the alpha2/beta2 interface, accounting for the similar cooperativity of HbA and HbF. (See "Structure and function of normal hemoglobins", section on 'Cooperativity'.)
Crystallographic studies of HbF at 2.5 angstrom resolution show almost complete isomorphism between HbA and HbF, with the sole difference being located at the N-terminus [8]. This change, which increases the distance between 2,3-diphosphoglycerate (2,3-DPG) inserted in the central cavity and the gamma2 histidine residue, is possibly a secondary contributor to the absence of 2,3-DPG effect exhibited by HbF. (See "Structure and function of normal hemoglobins", section on '2,3-bisphosphoglycerate'.)
Another aspect of HbF not readily understood in terms of mechanism is its strong globin-heme interaction and lower rate of dimerization compared with HbA [9]. Approximately 20 percent of HbF in the developing fetus has a post-translational modification where the N terminus of gamma-globin is acetylated [10]. While variant hemoglobins with N-terminal residue substitutions can be acetylated, no other normal human globin subunits are acetylated.
Oxygen affinity — Fetal blood has higher oxygen affinity than adult blood in nearly all mammals, suggesting its importance in transporting oxygen from the maternal circulation to that of the fetus. The P50, or partial pressure at which the hemoglobin molecule is half saturated with O2, is approximately 19 mmHg in cells with primarily HbF compared with approximately 27 mmHg for HbA-containing cells (figure 2).
Solutions of HbF in which organic phosphates have been removed have a P50 identical to that of solutions of HbA. The differences in oxygen affinity between adult and fetal blood is primarily a result of the failure of HbF to interact with 2,3-DPG due to gamma 143 serine and the acetylation of NA1 valine. The Bohr effect of HbF, through which an increase in blood acidity results in increased release of oxygen from hemoglobin, is 20 percent higher than that of HbA because of gamma 143 serine rather than beta 143 histidine. This helps maximize oxygen transport to the fetus, and contributes approximately half of the oxygen transport between mother and fetus. (See "Structure and function of normal hemoglobins", section on 'pH'.)
Differing expression of the two gamma globin genes — Although structurally almost identical, HBG2 and HBG1 are expressed at different levels in the fetus and the adult. In the newborn, approximately two-thirds of gamma chains have glycine at position 136 (G-gamma), while the remaining third have alanine (A gamma). This ratio falls during the switch from gamma- to beta-chain production and is nearly reversed in adult erythroid cells.
Although the promoters of HBG2 and HBG1, which contain binding sites for key transcription factors, differ [11], the mechanisms underlying the "switch" in expression of HBG2 and HBG1 during maturation are unknown. In individuals with the 5' HBG -158 C-T single nucleotide polymorphism (SNP) (rs7482144), sometimes called the Xmn1 G-gamma site, the HBG2 to HBG1 "switch" does not occur [12]. It was suggested that rs7482144 was in linkage disequilibrium (LD) with the functional motif that enhanced HBG2 expression, but studies have failed to find HbF repressive elements in this region between HBD and HHBBP1. Nevertheless, transcription of this pseudogene appears to have a roles in erythropoiesis and fetal to adult hemoglobin switching [13-16].
Variants of the two gamma-globin genes — The Globin Gene Server reports more than 140 variants affecting the structure or the expression of the gamma-globin genes. Another resource is located at ithanet.eu. As with structural variants of adult globin genes, these gamma-globin variants can be unstable, have high or low O2 affinity, or be subject to oxidation and cause methemoglobinemia and neonatal "cyanosis." (See "Hemoglobin variants including Hb C, Hb D, and Hb E".)
HbF variants are expressed during fetal and neonatal life and nearly disappear after age 1 year. The levels of the mutant HbF can differ depending on which gamma-globin gene contains the mutation or if there is variation in the numbers of gamma-globin genes as occurs with their rare duplication or triplication.
●Levels of HBG2 variants can range from 5 to 30 percent of total hemoglobin present in neonates, with most accounting for approximately 25 percent of all hemoglobin.
●Levels of most HBG1 gene variants are approximately half this amount. These differences reflect levels of expression of the respective genes, as discussed above.
Detection of HbF and measurement of HbF levels — HbF can be detected by high performance liquid chromatography (HPLC) (figure 3), alkaline and acid hemoglobin electrophoresis (figure 4), isoelectric focusing (figure 5), or capillary electrophoresis (figure 6). HPLC and capillary electrophoresis are the most commonly used methods to quantitatively measure HbF levels. When HbS is also present, capillary electrophoresis gives a small but statistically significant higher HbF level than HPLC [17].
Using HPLC in adults without a hemoglobinopathy, HbF concentrations were between 0.1 and 0.4 percent [18].
Sex, age, variation within the beta-globin gene cluster, gamma-globin gene duplications, gamma-globin gene promoter mutations, variants in the major trans-acting quantitative trait loci (QTL) associated with HbF (see 'Major genetic modifiers' below), and stress erythropoiesis can all increase HbF and also account for major differences in HbF among people with sickle cell anemia (homozygosity for the HbS gene) and beta thalassemia (table 1). (See 'Evaluating increased HbF in adults' below.)
F cells (HbF-containing cells) — HbF in the blood is normally restricted to a small number of erythrocytes called F cells. Determining the number of F cells is another method of estimating levels of HbF. Usually, there is a high correlation (R2 = 0.97) between the number of F cells and the percent HbF in the hemolysate [19,20]. F cells contain both HbF and HbA; in individuals who are homozygous for the sickle hemoglobin variant, they contain both HbS and HbF.
The origins, genetics, and physiology of F cells have been studied [21,22]. F cells do not resemble fetal erythrocytes, except for their high levels of HbF [23]. F cells are typical adult erythrocytes with adult MCV of approximately 80 fL, in contrast to fetal erythrocytes, which are much larger with MCV of approximately 120 fL. F cells also display the adult pattern of erythrocyte membrane antigens and enzymes [24].
The amount of HbF per F cell remains stable as nucleated red cells transit to reticulocytes that mature to erythrocytes, whereas HbA concentration increases [25]. Flow cytometry with fluorescence activated cell sorting (FACS), the standard method of enumerating F cells, can detect cells containing approximately 4 to 6 picograms (pg) of HbF. Single-cell analyses can detect cells with approximately 3 to 5 pg of HbF, but such methods are not used clinically [21,26,27]. A new FACS-based method has been developed to quantitatively measure the distribution of HbF amongst erythrocytes. This method claimed to be able to detect at least 2 pg of HbF per F cell.
Reticulocytes — HbF production can be estimated by the number of reticulocytes that contain measurable HbF, or F-reticulocytes. In sickle cell anemia and beta thalassemia, F-reticulocytes provide a more accurate measure of the expression of gamma-globin genes than F cells. Enrichment of F cells compared with F-reticulocytes suggests that F cells have a survival advantage in these hemoglobinopathies [28]. Using biotin-labeled sickle cells, F cells survived six to eight weeks compared with approximately two weeks for non-F cells [29].
In individuals without a hemoglobinopathy, F-reticulocytes comprise 1.0±0.8 percent of all circulating red cells; the corresponding percentages are 1.8±1.7 percent in sickle cell trait (HbAS) and 10.6±7.0 percent in sickle cell anemia [30].
HEMOGLOBIN SWITCHING: GENETIC BASIS OF HbF EXPRESSION — After the eighth week of gestation, HbF becomes the predominant hemoglobin of the fetus, its levels increasing until midway through gestation. The concentration of HbF then decreases with increasing gestational age. As an example, the concentration of HbF in an infant born at 28 weeks gestation is approximately 90 percent, decreasing to approximately 60 percent 10 weeks after birth, a value similar to that of full-term infant born at 38 weeks. (See "Anemia of prematurity (AOP)", section on 'Oxygen delivery'.)
●After birth, HbF is gradually replaced by HbA such that after approximately six months of age, HbF comprises less than 1 percent of the total hemoglobin. This switch from fetal to adult hemoglobin involves repression of HBG followed by upregulation of HBB expression; it is not complete in that residual amounts of HbF continue to be synthesized throughout adult life. This provides the basis for the therapeutic induction of HbF as treatment for the beta hemoglobinopathies.
●As noted above, during maturation there is also a switch in expression of HBG2 and HBG1. (See 'Differing expression of the two gamma globin genes' above.)
Population surveys show that the levels of HbF and F cells vary >20-fold in healthy adults; the distribution is continuous and positively skewed [31-33]. Approximately 10 percent of the populations at the upper tail of this distribution have >0.8 percent HbF, levels that correspond to >4.0 percent F cells. Historically, such individuals were said to have heterocellular hereditary persistence of HbF (HPFH). Most of these people have variants in the three HbF quantitative trait loci (QTL) discussed below that modulate HbF gene expression.
A genetic network regulating the switch from HbF to HbA production has emerged with multiple layers of genetic and epigenetic regulation, with interacting transcription factors and repressive complexes, all contributing essential functions to HbF control [34-36]. The nucleosome remodeling and deacetylase (NuRD) complex acts as the gateway underlying this epigenetic regulation in the developmental silencing of the gamma-globin genes. In addition to binding to their cognate recognition sites in HBG1 and HBG2 promoters, BCL11A and ZBTB7A need to physically interact with the NuRD co-repressor complex to effect their repressive activity [37,38]. These factors are illustrated in the figure (figure 7).
Further support for the important role for the NuRD complex is provided by a novel fetal Hb repressor, ZNF410, which does not bind directly to the gamma-globin promoter but acts via its highly specific regulation of CHD4, a protein subunit of the NuRD complex [39]. Contributions of lineage-specific factors such as KLF1 and MYB to the switch are indirect, through modulating expression of direct repressors such as BCL11A. Modulation of gamma-globin gene expression thus involves interaction of these transcription factors with each other and with GATA1 and other co-repressor complexes that involve chromatin-modeling and epigenetic modifiers [40].
Chromosomal looping from the locus control region (LCR) to the promoters of the genes of the beta-globin gene cluster, including the gamma-globin genes, plays a critical role in the switching process. In addition to the LCR, other hypersensitive sites in the beta-globin gene cluster might have roles in hemoglobin gene switching [41]. The silenced gamma-globin genes can be reactivated in adult erythroid cells by forcing looping of the LCR to the promoters of these genes causing increased gene expression, HbF production, and a decrease in the amount of adult hemoglobin [42-44]. Our modern understanding of HbF gene regulation has provided satisfying explanations to explain why HbF in some adults is higher than expected, as discussed below. The emerging network of HbF regulation has provided new insights and leads for therapeutic HbF reactivation [34,36,45]. (See 'Increased HbF in adults' below.)
INCREASED HbF IN ADULTS — Increased HbF levels in adults can be inherited or secondary to other conditions (table 1 and table 2 and figure 8). With our enhanced understanding of the mechanisms of hemoglobin switching and the genetic basis of the modulation of gamma-globin gene expression, the causes of higher-than-normal HbF in adulthood can be more precisely delineated.
Hereditary persistence of fetal hemoglobin — The designation of hereditary persistence of fetal hemoglobin (HPFH) might be restricted to deletions that remove delta- and beta-globin genes but preserve one or both gamma-globin genes and to variants in the gamma-globin gene promoters that alter binding motifs for the various transcription factors. Delta- and beta-globin gene deletions that are generally smaller than HPFH deletions are associated with the phenotype of delta-beta thalassemia that overlaps that of HPFH but where the increase in gamma-globin chain synthesis does not fully compensate for the loss of beta globin. The customary HbF increases in sickle cell anemia and beta thalassemia result from secondary responses to the stress erythropoiesis and hemolysis characteristic of these disorders and are modulated by co-inherited HbF boosting genetic variants. (See 'Secondary increases in HbF' below and 'HbF in the thalassemias and hereditary persistence of fetal hemoglobin' below and 'Sickle cell disease' below.)
●Some of the HbF increase in sickle cell anemia is related to polymorphisms in regulatory regions of HBB that are linked to the HBB gene cluster and marked by the haplotypes of these genes. In addition, some patients have co-inherited high HbF determinants in BCL11A and the HBS1L-MYB intergenic interval not linked to the beta S-globin gene haplotype.
●HbF response is also variable in beta thalassemia. Although some of this can be explained by the specific beta thalassemia variant itself and the HBB haplotype background, as in patients with sickle cell anemia, HbF-boosting determinants in BCL11A and the HBS1L-MYB intergenic interval contribute to the HbF response. Variants in the KLF1 gene have also been reported to influence HbF levels and disease severity in a Chinese population [46].
HPFH variants can be present as simple heterozygotes and homozygotes without associated hematologic disease. It is not uncommon for HPFH to coexist with secondary causes of increased HbF, such as sickle cell anemia. Historically, HPFH was classified as either "pancellular" or "heterocellular." (See 'HbF in the thalassemias and hereditary persistence of fetal hemoglobin' below.)
●Pancellular HPFH – HPFH inherited in a Mendelian fashion is caused either by large deletions in the HBB cluster (deletion HPFH) or point mutations and small deletions in the gamma-globin gene promoters (non-deletion HPFH). These are rare variants, and individuals heterozygous for these HPFH variants have elevations of HbF ranging from 10 to 40 percent. While HbF is detectable in all erythrocytes (pancellularly), the concentration of HbF among F cells is not uniform, whether estimated by Betke staining or by fluorescence intensity. Although the increases in HbF reflect the genotype of the different HPFH variants, a range of HbF levels has been noted in individuals heterozygous for the same variant that is likely to be related to co-inheritance of the common HbF quantitative trait loci (QTLs). (See 'Major genetic modifiers' below.)
●Heterocellular HPFH – In contrast to pancellular HPFH, increases in HbF in heterocellular HPFH are usually more modest, and some erythrocytes do not have detectable HbF. This might be a result of epigenetic influences on gene expression leading to position-effect variegation; it might also reflect the insensitivity of HbF measurement in individual erythrocytes. Some cells are likely to have HbF below the limits of detection. Pancellularity of HbF becomes obvious with HbF levels of approximately 30 percent. When HbF levels are this high, most red cells can have an average of 10 pg of HbF, which is above the limits of fluorescence activated cell sorting (FACS) detection. Family studies show that heterocellular HPFH tends to be inherited, but the inheritance pattern is not always clear; in some families, the high HbF determinant is not linked to the HBB complex. These observations led to the detection of the first trans-acting QTL that affected HBG expression on chromosome 6q. (See 'Major genetic modifiers' below.)
The variable increases in HbF, whether primary due to HPFH variants or as a secondary response to disease, occur on an underlying background of common HbF variation, as observed in healthy adults, which is due in part to variations in the major HbF-associated QTL. Higher-than-normal HbF, often a result of inheritance of HbF-boosting variant alleles of HbF QTL, has often been referred to as heterocellular HPFH. Twin studies show that the HbF in adults is predominantly genetically controlled; genetic factors account for 89 percent of the variability, with the remaining 11 percent accounted for by age, sex (2 percent), and unknown environmental factors [47]. The conglomeration of multiple factors, genetic and environmental, leads to the complex inheritance of the measured HbF values in families and explains the lack of clear Mendelian inheritance patterns in some patients with increased HbF [48,49].
Drugs that increase HbF production are of interest because they have the potential to ameliorate disease manifestations of the beta hemoglobinopathies.
●Increased HbF is thought to be the major mechanisms by which hydroxyurea reduces vaso-occlusive complications in sickle cell disease. (See "Hydroxyurea use in sickle cell disease".)
●The thalidomide analog pomalidomide has been shown to increase HbF production; this effect appears to be mediated by derepression of HbF expression via effects on transcriptional regulators such as KLF1 and BCL11A, which are discussed below, and others [50]. (See 'Kruppel-like factor 1 (KLF1)' below and 'BCL11A' below.)
●Many other drugs and cell-based therapies are being studied as a means of increasing HbF [51,52].
Major genetic modifiers — Xmn1-HBG2 on chromosome 11p, HBS1L-MYB intergenic region (HMIP) on chromosome 6q23, and BCL11A on chromosome 2p16 are considered to be QTL for HbF. They have been identified in genetic studies, validated in mechanistic studies, and contribute to the complex inheritance of some forms of high HbF.
Observations of variable HbF with different beta S–globin gene haplotypes first suggested that the HBB cluster is a prime location for a HbF determinant, represented by the C>T single nucleotide polymorphism (SNP) (rs7482144) at position -158 5' of the G-gamma-globin gene promoter, also referred to as Xmn1-HBG2 polymorphism [53].
There is emerging evidence that rs7482144 marks a functional domain. While it is associated with reduced methylation in six CpG sites flanking the transcription start site of HBG2 and alters a putative binding motif for ZBTB7A (LRF), a silencer of HbF gene expression, studies of BCL11A binding in the HBG2/1 promoters (see 'BCL11A' below) did not support a mechanistic role for rs7482144. CRISPR-Cas targeted disruption of the -158 motif in sickle CD34+ cells led to approximately 25 percent gamma globin and 55±5 percent F cells, compared with 30 to 40 percent gamma globin and 75 to 80 percent F cells after disruption of BCL11A and ZBTB7A binding sites at positions -115 and -200, respectively [54]. This implied that the region surrounding the -158 5’ HBG2 site binds an uncharacterized transcription factor. It also comports with the observation that variants within transcription factor binding sites in gamma-globin gene promoters are responsible for the phenotype of persistent HbF caused by promoter variants. The functional basis for the association of the Benin, Bantu, and Cameroon HbS gene haplotypes with characteristic HbF levels is unclear. The Arab-Indian haplotype, associated with very high HbF, is discussed below. (See 'Sickle cell disease' below.)
The first indication that chromosome 6q23 could be the location of a HbF QTL came from linkage association studies of an extended family with beta thalassemia and HPFH [55]. In this family, segregation analysis showed that the genetic determinant for HPFH was inherited independently from the HBB cluster [56]. The Xmn1 site and 6q23 sites were "rediscovered" in two genome-wide association studies (GWAS), which also identified a new HbF/F cell locus in intron 2 of the BCL11A gene on chromosome 2p16 [57,58]. BCL11A (B cell lymphoma/leukemia 11A) was known as an oncogene involved in leukemogenesis [59-61], but its relevance to HbF and erythropoiesis was previously unsuspected.
In GWAS, variants in the HBB gene cluster (rs7482144), chromosome 6q (rs66650371), and BCL11A (rs766432) loci account for 10 to 50 percent of the variation in HbF levels in healthy adults, or in those with sickle cell anemia or beta thalassemia, depending on the population studied [57,62-67]. The remaining variation ("missing heritability") is likely to be accounted for by many loci with relatively small effects, and/or rare variants with significant quantitative effects on gammaglobin gene expression that are typically missed by GWAS population studies [68]. (See "Regulation of erythropoiesis", section on 'Transcription factors'.)
Kruppel-like factor 1 (KLF1) — The association of KLF1 with HbF levels was identified through genetic studies in a Maltese family with beta thalassemia and HPFH that segregated independently of the HBB locus [69]. Linkage studies identified a locus on chromosome 19p13 that encompassed KLF1, and expression profiling of erythroid progenitor cells confirmed KLF1 as the gamma-globin gene modifier in this family. Family members with HPFH were heterozygous for the nonsense K288X mutation in KLF1 that disrupted its DNA-binding domain, a key erythroid gene regulator. Numerous reports of different mutations in KLF1 associated with increases in HbF soon followed [70,71].
●The HbF increases secondary to KLF1 mutations occurred as a primary phenotype [69] or in association with red blood cell disorders such as abnormal red cell membranes along with expression of the rare In(Lu) blood group [72], congenital dyserythropoietic anemia [73,74], congenital non-spherocytic hemolytic anemia with or without reduced expression of pyruvate kinase [75,76], beta thalassemia [46], other hemolytic anemias [77], and sickle cell anemia [78].
●Several GWAS of HbF, including ones in sickle cell anemia patients of African descent, have failed to identify common variants in KLF1, and it seems unlikely that polymorphisms of this gene play an important role in HbF regulation in sickle cell anemia [64,79,80].
●KLF1 mutations were overrepresented in a southern Chinese population with beta thalassemia. Two mutations were also associated with a thalassemia intermedia phenotype in beta thalassemia homozygotes, perhaps because the KLF1 mutation leads to suppression of BCL11A and increased expression of HbF genes [46].
KLF1, a direct activator of BCL11A and ZBTB7A (see 'BCL11A' below and 'LRF/ZBTB7A' below), is also essential for the activation of HBB expression (figure 7) [81-84]. A microRNA, miR-326, suppresses KLF1 expression directly by targeting its 3' untranslated region. Overexpression in CD34+ cells reduced KLF protein levels and increased expression of gamma-globin [85]. Collectively, studies suggest that KLF1 is key in the switch from HBG to HBB expression; it not only activates HBB directly, providing a competitive edge, but also silences the gamma-globin genes indirectly via activation of BCL11A [86] and the other major fetal globin repressor ZBTB7A/LRF [84]. KLF1 may also play a role in the silencing of embryonic globin gene expression [75]. In the light of these findings, KLF1 has now emerged as a major erythroid transcription factor with pleiotropic roles underlying many of the previously uncharacterized anemias [71].
BCL11A — BCL11A, a repressor of HbF expression, is a key regulator of hemoglobin switching (figure 7). BCL11A levels are developmentally regulated; its levels are low or absent when HbF is highly expressed, and it is present at high levels in adult erythroblasts where HbF levels are low [87]. Variants in intron 2 of this gene are associated with altered HbF levels [87]. Fine-mapping demonstrated that these HbF-associated variants, in particular rs1427407 and rs7606173, were localized to an enhancer that is erythroid-specific and not functional in lymphoid cells [88]. BCL11A interacts with several co-repressor complexes occupying discrete regions in the HBB complex leading to reconfiguration of the locus [89,90].
As it had not been possible to demonstrate direct binding of BCL11A to gamma-globin promoters, the mechanistic basis for the increased HBG expression was thought to be indirectly via long range interactions of the upstream locus control region (LCR) with the HBG promoter [90]. In 2018, key studies by two groups showed that BCL11A and ZBTB7A each bind to a cognate recognition site, around -200 and -115 bp, respectively, within the gamma-globin promoter [37,38]. These binding sites are the positions of the various naturally occurring HPFH-associated variants in the HBG promoter.
The overexpression of insulin-like growth factor 2 mRNA-binding protein 1 (IGF2BP1) caused a nearly complete and pancellular reversal of HbA expression with an increase in HbF to nearly 70 percent. These changes were partially mediated by reduced protein expression of BCL11A due to post-transcriptional loss [91].
●BCL11A is the major target of LIN28B-mediated HbF induction. An RNA-binding protein, LIN28B, is expressed reciprocally to that of BCL11A. BCL11A mRNA translation is directly suppressed by LIN28B. LIN28B directly binds BCL11A mRNA preventing its effective translation. The absence of LIN28B expression in adult cells allows effective BCL11A synthesis repressing gamma-globin gene expression [92]. Additional translational regulation might involve arginine methyltransferases (PRMTs) that suppress gamma-globin synthesis at the translational level [93].
●The composite BCL11A erythroid-specific enhancer has three elements, marked by DNase hypersensitivity at +55, +58, and +62 kilobases from the transcription start-site of this gene.
●The features of the enhancer elements that have the greatest effect on HBG gene expression and the least effect on erythropoiesis have been defined at near nucleotide resolution [94].
●BCL11A binds TGACCA motifs at 35 sites within the HBB gene cluster, two of which are in gamma-globin gene promoters. BCL11A binds preferentially to the -118 to -113 motif, repressing this promoter while favoring LCR interactions with the beta-globin gene promoter [37,38]. The favored BCL11A binding site at position -113 to -118 contains a (distal) CCAAT box; another (proximal) CCAAT box is 24 bp downstream of the distal site and binds the activator NF-Y. Occupancy by NF-Y is rapidly established following BCL11A depletion, preceding LCR contacts with gamma–globin promoters and gamma-globin expression. Both BCL11A and ZBTB7A need to physically interact with NuRD to achieve its repressive effect on HBG expression and it was hypothesized that concerted recruitment of NuRD is required for stable repression [16]. BCL11A also competes with the activator NF-Y for the gamma-globin gene promoters and that loss of either factor increases chromatin accessibility, thereby allowing the activator NF-Y to gain partial competitive advantage [16]. Fetal to adult hemoglobin switching is, in large part, a result of competition of BCL11A, the stage-selective repressor, and NF-Y, the ubiquitous activator, binding to these two CCAAT boxes.
●A G to A point mutation at position -113 causing HPFH does not disrupt BCL11A binding but rather creates a de novo binding site for the transcriptional activator GATA1 thereby increasing HbF [95].
●The erythroid specificity of the BCL11A enhancer has led to targeting this gene for therapeutic purposes in the beta hemoglobinopathies [96]. A single biallelic base edit at the BCL11A erythroid enhancer at +58 GATA1 binding site in CD34+ sickle erythroblasts increased HbF to about 30 percent while preventing sickling; globin chain synthesis balance was restored in beta-thalassemia erythroblasts [97].
●BCL11A microdeletions have been associated with neurological deficits and HbF of about 5 to about 30 percent, with normal hematological findings [98,99].
Trials in sickle cell anemia and beta thalassemia that either disrupt the BCL11A erythroid-specific enhancer using CRISPR-Cas9 or interfere with enhancer function by its targeting with an shRNA in isologous CD34+ cells have returned dramatic initial results, leading to almost 100 percent F cells and eliminating disease complications [100,101]. These approaches are discussed in detail separately. (See "Investigational therapies for sickle cell disease" and "Management of thalassemia" and "Overview of gene therapy, gene editing, and gene silencing", section on 'Gene editing'.)
The HBS1L-MYB intergenic region (HMIP) — High-resolution genetic mapping has refined the 6q QTL to a group of variants in tight linkage dysequilibrium (LD) in a 24-kb block referred to as HMIP-2 [62]. The causal single nucleotide polymorphisms (SNPs) were likely to reside in two clusters within the block, at -84 and -71 kb respectively, upstream of MYB [102]. Functional studies in transgenic mice and primary human erythroid cells provide overwhelming evidence that the SNPs at these two regions disrupt binding of key erythroid enhancers affecting long-range interactions with MYB and MYB expression [102-104], providing a functional explanation for the genetic association of the 6q HBS1L-MYB intergenic region with HbF and F cell levels.
A three-base pair (3-bp) deletion (rs66650371) in HMIP-2 is one putative functional element in the MYB enhancers and is associated with increased HbF expression in individuals who have the sentinel SNP rs9399137. This SNP is common in European and Asian populations, although less frequent in African-derived populations [105]. The DNA fragment encompassing the 3-bp deletion had enhancer-like activity that was augmented by the introduction of the 3-bp deletion. Saturating-mutagenesis with Cas9 nucleases identified the -71 site as containing functional sequence but suggested enhancer activity at -83 as opposed to the -84 site and also showed that the -36 site and possibly the -7 and -126 elements were also active [106]. The 3-bp deletion resided within the -84 site. A 1283 bp long noncoding RNA is transcribed from this enhancer and its downregulation increased gamma-globin gene mRNA 200-fold with a heterocellular increase in gamma globin in an adult-like erythroid cell line expressing HbA [107].
The MYB transcription factor is a key regulator of hematopoiesis and erythropoiesis [108,109], and modulates HbF expression via two mechanisms:
●Indirectly through alteration of the kinetics of erythroid differentiation: low MYB levels accelerate erythroid differentiation leading to release of early erythroid progenitor cells that are still synthesizing predominantly HbF [110].
●Directly via activation of KLF1 and other repressors (eg, nuclear receptors TR2/TR4) of gamma-globin genes [82,104,111,112].
Modulation of MYB expression also provides a functional explanation for the pleiotropic effect of the HMIP2 SNPs with other erythroid traits such as red cell count, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), HbA2 levels, and also with platelet and monocyte counts [113-118]. MYB expression is also reduced by GATA1 [119] and miRNA-15a and -16-1 [120]. Elevated levels of the latter have been proposed as the mechanism for the elevated HbF levels in infants with trisomy 13. (See 'Trisomy 13' below.)
The HBS1L-MYB intergenic enhancers do not appear to affect expression of HBS1L, the other flanking gene [102]. Further, one study also excluded HBS1L as having a role in the regulation of HbF and erythropoiesis. In whole-exome sequencing of rare uncharacterized disorders, mutations in the HBS1L gene leading to a loss of function in the gene were identified in a female child [121]. The child had normal blood counts and normal HbF levels. Thus, HMIP-2 is likely to affect HbF and hemopoietic traits via regulation of MYB [102].
LRF/ZBTB7A — ZBTB7A (chr 19p13.3), a Kruppel-like transcription factor, is a potent repressor of HBG expression. HbF increased 49 to 70 percent on knockdown of ZBTB7A in CD34+ cells [122]. When the gene was knocked down by CRISP/Cas9 editing in an immortalized human cell line that expressed adult HbA, HBG expression was induced without changes in BCL11A expression. Further, in the cells where both BCL11A and ZBTB7A were knocked out, HbF comprised >90 percent of total hemoglobin, suggesting that the silencing pathway of ZBTB7A is independent of BCL11A and that the two factors account for the majority of HbF silencing. Binding was observed at the LCR and the HBB and HBG loci but ZBTB7A depletion opened chromatin only at the HBG locus. Occupancy of ZBTB7A at the HBG locus coincides with sites of naturally occurring point mutations of HPFH in the promoters, and disruption of these residues disrupted binding of ZBTB7A [38]. ZBTB7A does not appear to be polymorphic or a common QTL for HbF, in contrast to the variants in BCL11A [123].
The 5' HBB region — Within the HBB cluster, apart from Xmn1-HBG2, are numerous polymorphisms associated with HbF variability:
●(AT)x (T)y polymorphism upstream of the HBB promoter in the olfactory gene
●(AT)x N12 (AT) y
●Polymorphism within the 5′ HS3 and 5′ HS4 elements of the LCR [124,125]
Functional evidence for how these alleles affect gamma-globin gene expression has not been provided for any of these genetic variants. High resolution mapping suggests that these variants are markers in LD with causal element(s) that remain to be discovered [13,126].
X-linked elements — HbF concentrations are lower in males than females, a difference attributed to an F cell production locus in Xp22.2 [127]. In 98 men compared with 124 women with sickle cell disease, F cells were associated with rs16998911 in FRMPD4. It is not clear how this locus or genes in LD with this locus might impact the sexual dimorphism of HbF levels [128].
MicroRNAs
●MicroRNAs-15a and 16-1, which contribute to MYB expression, could have a potential role in regulating HbF persistence in adults. (See 'Trisomy 13' below.)
●Additionally, targeting let-7 microRNA increased HbF to more than 30 percent of the total in primary erythroid progenitors [129], and expression of a high-mobility group protein (HMGA2), a downstream target of let-7 microRNA, increased HBG expression [130].
Other potential HbF regulatory factors
●The emerging network of HbF regulation also includes the orphan nuclear receptors TR2/TR4 (part of the DRED complex) [131-133], the protein arginine methyltransferase PRMT5, involving DNA methylation and histone deacetylases 1 and 2 epigenetic modifiers [134], a speckle-type POZ protein (SPOP), a substrate adaptor of the CUL3 ubiquitin ligase complex that is repressor of HbF gene expression, SIRT1, a protein deacetylase, an inducer of HbF gene expression [135,136].
●Knockdown of POGZ, which encodes a zinc-finger protein, in primary human CD34+ progenitor cell-derived erythroblasts, reduced BCL11A expression and increased HbF [137].
●FOXO3 might be a positive regulator of HbF gene expression. Its silencing reduced gamma-globin mRNA and HbF levels in erythroblasts whereas overexpression increased HbF [138].
●In addition to multiple mechanisms of transcriptional control, HbF synthesis was induced at the level of translation by phosphorylation of the initiation factor eIF2alpha (HRI) [139]. EIF2AK1 phosphorylates eIF2alpha. In contrast to the above cited study, depletion of EIF2AK1, an erythroid-specific kinase also known as the heme-regulated inhibitor, increased HbF. This effect was largely due to the decreased expression of BCL11A [140]. ATF4, a transcription factor and HRI-regulated protein, directly stimulates BCL11A transcription by binding to its enhancer and increasing enhancer-promoter contacts [141].
●Epigenetic modulation includes the nucleosome remodeling and deacetylase or NuRD complexes of multiple proteins involved in chromatin remodeling, histone deacetylation and gamma-globin gene silencing. One gene MBD2, a member of the methyl-CpG binding domain, when a member of the NuRD complex (MBD2-NuRD), is a strong silencer of gamma-globin genes [142]. ERF encodes the ETS2 repressor factor and represses gamma-globin expression. Promoter hypermethylation and mRNA downregulation of ETS2 was associated with increased HbF in human CD34+ erythroid progenitor cells, HUDEP-2 cell lines, and a mouse model [143].
●Targeting elements of the NuRD complex using rationally designed small molecules could be an important approach to derepressing HbF gene expression [144]. For example, ZNF410, a DNA-binding protein, activates only CHD4, a NuRD component. Loss of ZNF410 in adult erythroid cells diminishes CHD4 levels derepressing HbF genes [39,145].
●Chromatin structure also effects HbF gene regulation. Deletion of the HBBP1 region in adult cells increases LCR-gamma-globin contacts and gamma-globin transcription [14]. HBBP1 is essential for erythropoiesis, indirectly upregulating TAL1, a key regulator of erythropoiesis. Higher levels of HBBP1 and TAL1 transcripts were associated with increased HbF in beta thalassemia and in cultured erythroid cells [15].
SECONDARY INCREASES IN HbF — Increased HbF levels in adults have been found in various acquired and genetic disorders other than the hemoglobinopathies. In the majority of cases, the increases appear to be secondary to perturbation of erythropoiesis.
Increased HbF levels have also been recorded in sporadic reports of thyrotoxicosis and pernicious anemia [146]. Review of a larger number of such cases showed that increased HbF is not a consistent feature. It is likely that the isolated increases are in individuals who have co-inherited the HbF-boosting quantitative trait loci. (See 'Major genetic modifiers' above.)
Prematurity, diabetes, and pregnancy — A number of gestational issues may affect HbF levels, as follows:
●Prematurity – The switch from fetal to adult hemoglobin production proceeds on a set developmental clock and is not affected by the gestational age of the infant [147]. HbF remains the major hemoglobin synthesized up to the 37 weeks gestation. In normal early preterm newborns, the rate of transition from HbF to HbA synthesis postnatally resembles that in utero, such that at the 38 weeks gestation, the relative amounts of HbA and HbF correspond to that in the full-term newborn group.
●Infants of diabetic mothers – Full-term infants (36 to 38 weeks of gestation) of diabetic mothers have a delayed HbF to HbA switch, synthesizing more HbF than is expected for their gestational age [148,149]. Although the exact mechanisms are not clear, this could be related to elevated levels of alpha-amino-butyric acid. This supposition was tested in sheep; butyrate infusion in the ovine fetus delayed the developmental clock for the HbF to HbA switch [150]. This initial study led to a series of clinical trials of butyrate and its analogues for therapeutic HbF reactivation in patients with sickle cell anemia and beta thalassemia, but this approach has not proven therapeutically useful [151-153].
●Pregnancy – Increases in HbF during pregnancy are very characteristic and reach a peak in the second trimester. Increased production of F cells is the physiologic result of erythroid expansion during this period [154].
Trisomy 13 — A delayed HbF to HbA switch, along with persistently elevated HbF levels, is one of the unique features in infants with trisomy 13 [155,156]. One study provided compelling evidence that the elevated HbF levels relate to the increased expression of microRNAs 15a and 16-1 produced from the triplicated chromosome 13 [120]. The increased HbF effect is mediated, at least in part, through down-modulation of MYB via targeting of its 3′ UTR by microRNAs 15a and 16-1. (See "Congenital cytogenetic abnormalities", section on 'Trisomy 13 syndrome'.)
Bone marrow regeneration — A number of conditions or experimental settings associated with increased or "stressed" erythropoiesis have been associated with increases in HbF [157]. These include the following:
●Bone marrow regeneration and acute expansion in erythropoietic activity have been proposed to underlie increases in F cells and HbF in many leukemia patients following chemotherapy [158], after hematopoietic stem cell transplantation [159], following acute blood loss, and after treatment with iron in severe untreated iron deficiency [160].
●Acute hemolysis also results in increased F cell production. The mechanistic basis for this is likely to be related to the acute compensatory expansion in erythropoietic activity. Chronic non-hemoglobinopathic hemolytic anemias are rarely associated with increased HbF. In some studies, increases in HbF were highly variable, HbF was heterogeneously distributed, and F cell numbers were increased [159,160]. HbF levels also rose with an increase in reticulocyte count. Co-inheritance of heterocellular HPFH [161] and co-inheritance of HbF-boosting quantitative trait loci might explain this variability. (See 'Major genetic modifiers' above.)
●Transient erythroblastopenia of childhood (TEC) is characterized by a transient arrest in erythropoiesis. Spontaneous recovery is typically associated with increased numbers of F cells and HbF concentration [160,161]. (See "Overview of causes of anemia in children due to decreased red blood cell production", section on 'Transient erythroblastopenia of childhood'.)
●Some of the increased HbF production under stress conditions may also result from enhanced translation of gamma-globin messenger RNA, rather than enhanced transcription [139,162]. Increased levels of HbF following the experimental use of butyrate in patients with sickle cell disease may be an example of this effect [163].
Bone marrow failure syndromes — Patients with inherited bone marrow failure syndromes (Diamond-Blackfan anemia, dyskeratosis congenita, Fanconi anemia, Shwachman-Diamond syndrome) frequently have increased HbF as part of their "stressed" hematopoiesis that also includes macrocytosis and erythropoietin levels higher than predicted by their degree of anemia [164]. Again, a wide range in the increases in HbF has been observed; one study has shown that the increased HbF levels were associated with young age, male gender, anemia, high erythropoietin levels, and the minor allele (T) of the Xmn1-HBG2 quantitative trait locus (QTL) [165]. (See "Diamond-Blackfan anemia", section on 'Laboratory findings'.)
Leukemia — Increases in HbF levels have been observed in hematologic malignancies [166], among which juvenile myelomonocytic leukemia (JMML; juvenile chronic myeloid leukemia, chronic myelomonocytic leukemia of childhood) is a prime example [167]. JMML is a rare, aggressive myeloproliferative disorder of early childhood. In a retrospective review of 100 children with chronic myelomonocytic leukemia of childhood under the age of 16 years, 64 percent of the patients had a normal karyotype, 25 percent had monosomy 7, and 10 percent had chromosomal abnormalities other than monosomy 7 [168]. A characteristic feature of those with a normal karyotype and other karyotypic abnormalities is the sustained and marked increase in HbF of up to 90 percent of the total hemoglobin. This is in contrast to those with monosomy 7 where HbF levels were within normal limits. The clinical features were similar in both groups.
The increased HbF levels in JMML are accompanied by red cell changes typical of normal fetal red cells such as decreased carbonic anhydrase and increased levels of the i antigen and absence of I antigen [166,169]. The G-gamma:A-gamma ratio of HbF is also that of the fetal type. Initially, it was thought that the condition represented emergence of a clone of hematopoietic cells that had reverted to fetal erythropoiesis, but the uncoordinated expression of the various "fetal" characteristics suggests that the condition is due to grossly distorted regulation of gene expression [170]. Nonetheless, HbF levels at presentation appeared to be an independent risk factor of survival in the non-transplanted group of patients [168].
In patients with the myelodysplastic syndrome, karyotypic abnormalities have been shown to be associated with increased HbF [171,172]. HbF >10 percent represented a poorer prognosis in one study [171]. Elevated HbF levels have also been observed in acute myeloblastic leukemia, erythroleukemia, lymphoblastic leukemia, and chronic myeloid leukemia [166]. A 2017 study suggested that HbF could be a predictor of outcome in myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) patients receiving decitabine, an agent also used experimentally for therapeutic reactivation of HbF in sickle cell disease [173]. Given what is known about the pleiotropic influence of the HbF QTLs on hematologic traits, it seems that HbF may be acting as a marker for the presence of the genetic variants that influence the hematologic response in MDS/AML [173].
Solid tumors — Rarely, increased HbF levels have been observed in solid tumors including choriocarcinoma, adenocarcinoma of the lung, and hepatoma [174-176]. The increased HbF could be related to paraneoplastic phenomenon that involves inappropriate overproduction of erythropoietin [177], a key cytokine that induces erythropoiesis, thus influencing F cell production.
HbF IN THE THALASSEMIAS AND HEREDITARY PERSISTENCE OF FETAL HEMOGLOBIN
Beta thalassemia — Beta thalassemia shows remarkable phenotypic diversity, ranging from life-threatening anemia to an extremely mild condition identified only by chance [178-181]. This disorder is caused by a quantitative deficiency of beta-globin chains, leading to globin chain imbalance and redundant alpha-globin chains. The free alpha-globin chains are highly unstable; they form intracellular inclusions, interfering with red cell maturation, causing premature death of these erythroid precursor cells, or ineffective erythropoiesis. Affected infants present with severe anemia and failure to thrive within 6 to 12 months of age when the switch from HbF to HbA is almost complete. (See "Pathophysiology of thalassemia".)
While the severity of beta thalassemia is primarily determined by the degree of beta chain deficiency, for any given beta thalassemia allele the severity of the disease can be alleviated by co-inheritance of alpha thalassemia or by co-inheritance of factors that increase gamma-globin chain production and HbF levels. In the latter case, gamma-globin chains combine with the excess alpha globin to form HbF; cells that contain a relatively higher percentage of HbF are protected against the deleterious effect of alpha-globin chain precipitation and premature death and have selective survival. Thus, all individuals with beta thalassemia have variable increases in HbF due to survival of these F cells.
Certain beta thalassemia mutations, notably those that involve small deletions or mutations of the promoter sequence of the HBB gene, are associated with much higher levels of HbF production than mutations affecting other regions of HBB [182]. (See "Molecular genetics of the thalassemia syndromes".)
This might reflect the competition between the HBG and HBB promoters for interaction with the upstream beta locus control region (LCR) and limited transcription factors. Heterozygotes for such types of beta-thalassemia mutations have unusually high HbA2, and although the increases in HbF levels are variable, the increase in HbF production in homozygotes is often adequate to compensate for the complete absence of HbA [183].
●HbF levels are normal or slightly elevated in beta-thalassemia heterozygotes [180]. Higher HbF (and HbA2) levels are found with mutations that involve promoter of the HBB gene, but variations in HbF levels also reflect the genetic background of the individual.
●In homozygous beta thalassemia, the proportion of HbF ranges from 10 percent in those with the milder alleles to almost 100 percent in homozygotes or compound heterozygotes with beta0 thalassemia. In the most severe cases, the absolute amount of HbF is approximately 3 to 5 g/dL, produced as a result of extreme erythroid hyperplasia, selective survival of F cells, and some increase in HBG transcription.
●Non-transfusion-dependent beta0 thalassemia intermedia with hemoglobin levels of 8 to 11 g/dL and 100 percent HbF has been observed. In some cases, the increase in HbF production reflects the type of beta-thalassemia allele, but in others, co-inheritance of quantitative trait loci (QTL) associated with increased HBG expression might explain their more benign clinical features [184,185]. Unusually, two KLF1 variants were associated with a milder phenotype in beta thalassemia homozygotes, perhaps because the KLF1 variant leads to suppression of BCL11A and increased expression of HbF genes [46].
●In Chinese individuals with beta thalassemia, rs368698783 in the promoter of HBG1 predicted clinical severity and was associated with increased HbF. The minor allele of this single nucleotide polymorphism (SNP) had its effect through demethylation of HBG promoter sites [186].
Effect of the quantitative trait loci — While selection of F cells provides an explanation for the increases in HbF in beta thalassemia, the mechanism does not explain the wide variation in the amount produced. Much of this variability is genetically determined, in part from the co-inheritance of one or more of HbF-boosting alleles of the Xmn1-HBG2, HMIP-2 and BCL11A HbF QTLs. The Xmn1-HBG2 QTL is a common sequence variation in all population groups, present at a frequency of approximately 0.35 [47,187]. Although increases in HbF and F cells associated with Xmn1-HBG2 are minimal or undetectable in healthy adults, clinical studies have shown that under conditions of stress erythropoiesis, as in homozygous beta thalassemia and sickle cell disease, the presence of Xmn1-HBG2 leads to a much higher HbF response. This could explain why the same mutation on different beta chromosomal backgrounds, some with and others without the Xmn1-HBG2 variant, are associated with different clinical severity. (See 'Major genetic modifiers' above.)
The three HbF QTLs are associated with HbF and severity of beta thalassemia in diverse groups including Sardinian, French, Chinese, and Thai populations [58,63,188-190]. More than 95 percent of Sardinian beta-thalassemia patients are homozygous for the same codon 39 beta0-thalassemia mutation but have extremely variable clinical severity. Co-inheritance of variants in BCL11A and HMIP-2 and alpha thalassemia accounts for 75 percent of the differences in disease severity [191]. In France, a combination of the beta-thalassemia genotype, Xmn1-HBG2 and SNPs in BCL11A and HMIP-2, can predict up to 80 percent of disease severity [188]. In a cohort of 316 beta0-thalassemia patients, delayed or absent transfusion requirements correlated with status of the three HbF QTLs and the alpha-globin genotype [192]. Using these genetic markers and the patient's age at the initiation of transfusion therapy, the hematologic severity could be predicted and scored [193].
A striking example of the possible clinical impact in carriers of the minor alleles of the HbF QTL was the report of dizygotic twins homozygous for the HBB codon 8 frame-shift beta0-thalassemia mutation (FSC8; HBB:c25_26delAA), a well-known cause of severe transfusion-dependent beta0 thalassemia. Nevertheless, the twins were asymptomatic, their hemoglobin concentration was 12 to 13 g/dL, and their HbF was 98 percent. Both were homozygous for rs7482144, homozygous for the 3-bp deletion HBS1L-MYB intergenic polymorphism at rs66650371, and heterozygous for the BCL11A intron 2 polymorphism at rs766432. In 22 patients with this same beta-thalassemia mutation, four individuals with the thalassemia intermedia phenotype had some of the HbF QTL, and 18 with transfusion dependency lacked both rs7482144 and rs6665037 [194]. Further studies suggested that a variant Sp1 (R218Q) transcription factor found in both twins, but not in the 22 other severely affected cases, might play a role in upregulating HBG expression [195].
Delta beta thalassemia — Large deletions in the HBB gene cluster that remove the beta- and delta-globin genes, sparing both gamma-globin genes or HBG2, alone cause delta beta thalassemia (figure 9). These deletions range from <0.8 to >100 kb, but most are 10 to 30 kb. The mechanism of increased HbF in these conditions is unclear; several hypotheses have been advanced:
●Removal of regulatory regions between HBG1 and HBD that silence gamma-globin gene expression; the area of the Corfu deletion is especially germane for this hypothesis [196-199].
●Removal of competition between HBG promoters with promoters of HBD and HBB for transcriptional machinery, including the LCR enhancers.
●Translocation of 3' distal enhancers into the proximity of HBG [200,201].
As a group, these deletions cause G-gamma/A-gamma (delta beta0) thalassemia or G-gamma (delta beta0) thalassemia and several sub-varieties of each have been described [124,202]. Most are rare, but the Mediterranean deletion of some 13 kb that includes HBD and extends 3' of HBB is common in its cognate areas.
●Heterozygotes for G-gamma/A-gamma (delta beta0) thalassemia resemble a mild beta-thalassemia trait with the exception of elevated HbF and normal HbA2 levels. They have between 4 and 24 percent HbF, with most averaging approximately 10 to 12 percent. Microcytosis is mild, and the mean corpuscular volume (MCV) might even be normal. HbF distribution among erythrocytes is heterocellular.
●G-gamma/A-gamma (delta beta0) HPFH heterozygotes are characterized by normal HbA2 levels and HbF levels of 15 to 30 percent. There are only minor reductions in the red cell indices, which are frequently within the normal range. Homozygotes have been described, and their hemoglobin consists entirely of HbF. These individuals are clinically unaffected and have normal or high hemoglobin levels (15 to 18 g/dL), presumably as a result of the higher oxygen affinity of HbF. They have mildly microcytic, hypochromic red cells; globin chain synthesis imbalance similar to that of beta-thalassemia trait is present, indicating that the output of gamma-globin chains does not fully compensate for the lack of beta-globin chains.
●Homozygosity for some of these deletions causes mild thalassemia intermedia with hemoglobin concentrations of 10 to 13 g/dL and only mild hepatosplenomegaly. They may develop more severe anemia during infections. Red cell morphology is more abnormal than in beta-thalassemia trait, and the hemoglobin consists of 100 percent HbF, containing both G-gamma- and A-gamma-globin chains. Too few cases have been studied to know whether the variability in clinical severity among delta beta thalassemia homozygotes is directly related to the underlying molecular defect. HbF is heterogeneously distributed, and cells with the most gamma-globin and the least chain imbalance survive longest. Even in heterozygotes, cells with higher amounts of HbF survive longer.
●The G-gamma (delta beta0) thalassemias are similar to the G-gamma/A-gamma (delta beta0) thalassemias, but only HBG2 is expressed such that their HbF has only G-gamma-globin chains. Homozygotes might be more severely affected, as only a single gamma-globin gene can be expressed.
Rare cases of co-inheritance of delta beta thalassemia with beta thalassemia may result in transfusion-dependent thalassemia; in other cases, milder thalassemia can occur. (See "Diagnosis of thalassemia (adults and children)", section on 'Overview of subtypes and disease severity'.)
A gamma+ thalassemia phenotype was proposed to result from an IVS II-115 A-G substitution in both the A-gamma- and G-gamma-globin genes and the deletion of an A at position -6 relative to the G-gamma-globin gene polyadenylation site. Yet, the IVS II substitutions appear to have gene frequencies of 0.73 for the A-gamma-globin gene and 0.86 for the G-gamma-globin gene, suggesting that they are common SNPs. One gamma thalassemia resulted from an unequal crossover between the G-gamma and A-gamma-globin genes, deleting one gamma-globin gene from the affected chromosome, and leaving a hybrid gamma-globin gene, akin to Hb Lepore-type genes. Found in heterozygotes and in two homozygotes, newborn homozygotes for this deletion had 50 percent HbF, all A-gamma.
Gene deletion hereditary persistence of HbF (HPFH) — Deletional HPFH is caused by 13 to 106 kb deletions in the HBB gene cluster that remove parts or all of the HBD and HBB genes and are associated with a variable but often nearly complete compensatory increase in gamma-globin gene expression and HbF levels (figure 9). These deletions are characteristically associated with a pancellular or homogenous distribution of HbF among erythrocytes. It seems likely that each red cell does not have identical concentrations of HbF; nevertheless, all cells have HbF.
Gene deletion HPFH 1 and HPFH 2 are the most common types in people of African descent. They are characterized by >80 kb deletions including HBD and HBB that are staggered by approximately 5 kb at the 5' and 3' ends. Heterozygotes for both deletions have HbF levels of 20 to 30 percent and mild microcytosis. They differ in the ratio of G-gamma to A-gamma chains that is approximately 50:50 in HPFH 1 and 30:70 in HPFH 2. Other HPFH deletions have similar hematologic findings with minor differences in the ratios of G-gamma to A-gamma chains and levels of HbA2 when HBD is not deleted. Homozygotes have 100 percent HbF and no HbA2.
Both delta beta thalassemia and deletional HPFH have a range of HbF levels. A careful study of trans-acting quantitative trait loci (QTL) polymorphisms has not been done in these cases, and it is possible that, as with normal individuals and patients with hemoglobinopathies and thalassemia, variations in co-inheritance of the HbF QTLs (and possibly others) that modulate HBG expression might account for some of this variation.
Non-gene deletion HPFH — The non-deletional forms of HPFH can be divided into those that are inherited in a Mendelian fashion caused by point mutations in either HBG2 or HBG1 promoters, or those that are inherited as a complex trait caused by inheritance of a conglomerate of QTLs (Xmn1-HBG2, BCL11A, and HMIP-2; historically referred to as heterocellular HPFH).
Mutations in the Mendelian non-deletion HPFH include single base substitutions and minor deletions clustered in two regions in the gamma-globin gene promoters: (1) positions -114 to -117 and (2) from positions -195 to -202 [124,203]. The regions in the gamma-globin gene promoter that harbor the mutations coincide with binding sites for ubiquitous and erythroid-specific transcription factors, and it seems probable that the increased gamma-globin gene and HbF expression involves altered binding of these transcription factors: altered binding of the ubiquitous Sp-1 elements and ZBTB7A in positions -195 to -202; altered binding of GATA1, NFE-3, BCL11A, and NF-gamma to the duplicated CCAAT boxes around position -117. Some small deletions of critical regions of the HBG promoters, for example a 13 nt deletion of the HBG1 promoter that included the CCAAT box, were associated with HbF of about 30 percent that was distributed pancellularly [204]. By recreating this 13 nt deletion first described in the HBG1 promoter, it was possible to increase HbF to levels likely to be therapeutic in sickle cell disease [205]. Engineering an HPFH deletion in erythroid progenitors also led to increased HBG expression [206]. A 13 bp deletion and a point mutation associated with the phenotype HPFH are found at positions -118 to -113. This is the critical locus of BCL11A binding. The major HbF repressors BCL11A and ZBTB7A bind to the sites at -115 and -200 bp, respectively, and the introduction of these naturally occurring HPFH mutations into erythroid cells disrupted this binding and increased HbF expression [38]. These loci are being targeted therapeutically by CRISPR/Cas editing of BCL11A to relieve the repression of the gamma-globin genes by BCL11A.
Heterozygotes for these Mendelian non-deletion HPFHs have HbF levels varying from 5 to 40 percent; mutations at position -175 are associated with the highest HbF increases. HbF shows a pancellular distribution, and the composition of the gamma-globin chains is predominantly G-gamma or A-gamma, reflecting whether the mutation is in the HBG2 or HBG1 genes. Heterozygotes have normal hematologic findings including normal red cell indices; synthesis of alpha and non-alpha globin chains may be balanced or slightly imbalanced with an excess of alpha-globin chains.
Compound heterozygosity with mutations affecting the HBB allele in trans (eg, sickle cell anemia, beta0 thalassemia) suggest that the combined output of the gamma- and beta-globin chains from the non-deletion HPFH chromosome closely approximate the output from a normal HBB gene. For example, combined output (46.5 percent) of HbA and HbF at 30.7 and 15.8 percent, respectively, is comparable to HbS at 45.4 percent in a heterozygote for -197 C>T HBG1 [207]. This reciprocity in cis supports the notion that activation of the gamma- and beta-globin genes depends on competition to access to the cis-acting upstream LCR and direct physical contact. In reported cases, homozygotes have HbF levels approximately twice that of heterozygotes with normal hemoglobin levels and normal red blood cell indices.
Small increases in HbF levels have also been described in adults in which the condition appears familial but clearly not inherited in a Mendelian fashion; in other cases, the condition behaves as if allelic to the HBB cluster. These modest increases are frequently lumped together as heterocellular HPFH due to the uneven distribution of HbF. The "unlinked" causes include SNPs associated with increased HBG expression in BCL11A and HBS1L-MYB QTLs, and the possible "linked" forms include Xmn1-HBG2 and gene rearrangements such as Atlanta HPFH and triplicated and quadruplicated gamma-globin genes [208,209].
The high HbF associated with naturally occurring HPFH mutations of several different varieties has prompted the therapeutic exploitation of these variants using genome editing. The T-C SNP in position -198 of the A-gamma-globin gene promoter (British type HPFH) is typically associated with HbF levels of 8 to 10 percent. When this mutation was introduced into an erythroid cell line expressing HbA, HbF levels rose substantially. The mutation created a binding site for KLF1 [210].
SICKLE CELL DISEASE — Newborns with sickle cell anemia have few symptoms because their high HbF prevents the polymerization of deoxy-HbS [211] that drives the pathophysiology of disease. HbF and its mixed hybrid tetramer (alpha2betaSgamma) are excluded from the deoxy-HbS polymer phase. The latter property is unique to HbF, whereas most other hemoglobins such as HbA and HbC simply dilute intracellular HbS. The primary effect of HbF remans dilutional [212]. HbS polymerization is highly sensitive and dependent on intracellular HbS concentration [213]; thus, even a small decrease in HbS concentration is therapeutic because more cells can escape the small vessels before sickling occurs. Reduction of HbS intracellular concentration, such as by increasing HbF or the red cell volume, increases the delay time to sickling, while strategies that reduce adherence of sickle cells to endothelium and shorten transit time should be therapeutic. The phenotype of sickle cell disease becomes manifest within six months to two years of age as HbF levels decline. (See "Pathophysiology of sickle cell disease", section on 'Genetics'.)
●Individuals with sickle cell trait (HbAS) usually have normal levels of HbF. Minor increases can be associated with HbF-boosting alleles of the HbF quantitative trait loci (QTL). Rarely, people with HbAS can have HbF of 15 to 40 percent with certain promoter mutations responsible for HPFH [214].
●The average HbF level among people with sickle cell anemia of African descent is between 5 and 8 percent with a range of less than 5 to more than 20 percent [52]. Part of this variation resides in regions linked to the HBB complex and is associated, at least in part, with the HBB haplotype.
●Based on restriction fragment length polymorphisms (RFLPs), five common beta S-haplotypes are found, named for their sites of origin in Africa, the Middle East, and the Indian subcontinent [215-217]. Varying levels of HbF are observed in adults homozygous for the different haplotypes: approximately 5 percent for Bantu beta S-haplotype, 6 to 8 percent for the Benin and Cameroon, approximately 10 percent for the Senegal, and approximately 20 percent HbF for Arab-Indian (AI) beta S-haplotype [53,218,219]. Each beta S-haplotype has considerable variance in HbF levels, suggesting the importance of co-inheritance of the aforementioned trans-acting QTLs modulating HBG expression.
●Only the Senegal and AI beta S-haplotypes have the Xmn1 C-T restriction site polymorphism (rs7482144) [53,217,220].
●The variance in HbF among different populations with sickle cell anemia is modulated by the three major known quantitative trait loci (QTL); however, the contribution by each of these loci is not uniform because of the different frequencies of the minor alleles amongst the different populations. In Indians with the AI haplotype, the BCL11A sentinel SNP 1427407 explained approximately 23 percent of HbF variance, an amount similar to that found in studies of African Americans, but rs69334904 (marking HMIP) was not associated with HbF [13,63,65-67,79,80,126,221]. In Saudi Arabs with the AI haplotype, BCL11A and HMIP explained 8.8 percent of HbF variance. In 780 patients from the Eastern Province of Saudi Arabia, the minor allele frequency of rs9399137 in MYB and rs766432 BCL11A were 0.08 and 0.28 percent, respectively; together they accounted for 4.2 percent of HbF variance [222]. In African populations, the variance in HbF explained by polymorphisms in BCL11A ranged from 2 to 8 percent [223,224].
Modulation of the phenotype of sickle cell anemia by HbF — HbF is the most powerful modulator of the clinical and hematologic features of sickle cell anemia. In large population-based studies, any increment in HbF had a beneficial effect on mortality [225,226]. Such observations are consistent with the modeling of the distribution of HbF/F cell that suggests that very few F cells that are fully protected from HbS polymer-induced damage are present when HbF levels are approximately 5 percent, larger numbers of "protected" cells are possible when HbF reaches levels of 10 percent, and levels of 20 percent or more can be associated with a large fraction of "protected" cells. F cells in patients with sickle cell disease survive longer than non-F cells, depending on the amount of HbF/F cell [29,227]. (See 'F cells in sickle cell anemia' below.)
An ensemble of genetic risk models predicted 23.4 percent of HbF variability with an accuracy of 0.28 to 0.44, a number that compared very favorably with some other genetic risk scores [228]. Another genetic prediction model used four SNPs in the HbF QTLs (rs6545816, rs1427407, rs66650371, and rs7482144) that accounted for 21.8 and 27.5 percent of HbF variability in HbSS and HbSC patients, respectively [229]. The comparatively larger value in HbSC patients suggests that the genetic component has a larger impact in HbSC disease, in which the anemia and stress erythropoiesis is relatively less compared with HbSS disease.
The beneficial effect of HbF on sickle-related complications is variable, and studies of the association of HbF with the same complications have resulted in disparate findings [230]. Events most closely linked epidemiologically to sickle vaso-occlusion like acute painful episodes and acute chest syndrome were related to HbF concentration, whereas events linked to hemolysis like priapism, chronic kidney disease, cerebrovascular disease, and tricuspid regurgitant velocity were often not [230-232]. However, HbF levels were lower in patients with the highest quartile of hemolysis compared with the lowest quartile, high HbF was associated with fewer leg ulcers and longevity, and low HbF was associated with white matter lesions on MRI [233,234]. A study evaluating the overall health-related quality of life in children showed that HbF was the most powerful predictor [235]. In a longitudinal study of infants recruited at age 4 months and followed for 20 months, HbF level, was an independent predictor of severe disease complications [236].
Co-inheritance of HPFH — In HbS-gene deletion HPFH with HbF levels of about 30 percent, the concentration of HbF is homogeneously distributed in every cell and can be calculated to be approximately 10 pg per cell. This level of HbF protects all cells from HbS polymer-induced damage, and the affected individuals have nearly normal hematology, and clinical events associated with sickle cell anemia appear to be rare. In 30 individuals with HbS-HPFH whose mutation was unequivocally ascertained by DNA-based diagnostics, HbF level was 50 to 90 percent during infancy and stabilized between ages three and five years at approximately 30 percent. Mean HbF of individuals aged five or older was 31±2 percent; average hemoglobin concentration was 12.7 g/dL; mean corpuscular volume (MCV) 75 fL; and reticulocytes were 1.4 percent (range 0.5 to 3 percent). Patients were healthy without complications attributable to sickle cell disease [237]. However, some patients with proven HPFH deletions appear to have had acute sickle cell disease complications [214].
In contrast, individuals with compound heterozygous HbS/Black (A-gamma delta beta)0 thalassemia, despite the relatively mild anemia and HbF of 20 to 30 percent, experience clinical complications typically associated with sickle cell anemia [238].
F cells in sickle cell anemia — The distribution of concentrations of HbF per F cell (HbF/F cell ratio) amongst patients is likely to be a critical determinant of the protective effect of HbF in sickle cell disease and perhaps also plays a role in the pathophysiology of beta thalassemia.
●Individuals with sickle cell anemia have individually characteristic distributions of HbF/F cell regardless of their total HbF level [239]. As an example, in 46 African Americans with sickle cell anemia, F cells ranged between 2 and 80 percent of erythrocytes, and the average HbF/F cell was 6.4±1.6 pg [21]. However, calculating the concentration of HbF/F cell using HbF levels and the number of F cells falsely assumes that each F cell contains the same amount of HbF [240].
●Studies suggest that deoxyHbS polymerization is prevented at physiologic venous and capillary O2 saturations of 40 to 70 percent when HbF/F cell is in the range of 9 to 12 pg [240].
●From a "HbF-centric" perspective, the reason that some patients with sickle cell anemia and very high levels of HbF can have severe disease is because many of these F cells contain insufficient concentrations of HbF to inhibit HbS polymerization-based injury. Individuals with compound heterozygous HbS/Black (A-gamma delta beta)0 thalassemia, despite the relatively mild anemia and HbF of 20 to 30 percent, suffer clinical complications typically associated with sickle cell anemia when compared with individuals with HbS-HPFH with similar HbF levels because of their heterocellular HbF distribution [238]. (See 'Co-inheritance of HPFH' above.)
When the distribution of HbF/F cell was modeled in patient groups with mean HbF levels of 5, 10, 20, and 30 percent, the distribution of HbF/F cell was shown to greatly vary, even when the mean was constant. As examples:
●At HbF levels of 20 percent, as few as 1 percent and as many as 24 percent of cells can have "protective" or polymer-inhibiting levels of HbF; with lower HbF, few or no "protected" cells might be present.
●Only when the total HbF concentration was near 30 percent was it possible for the number of "protected" F cells to approach 70 percent.
Imaging flow cytometric studies of sickle cell anemia found that some cells with detectable HbF were capable of sickling supporting the notion of a threshold of HbF/F-cell required to prevent the sickling process. However, some cells not containing HbF were resistant to sickling [241].
Accordingly, rather than knowing the total number of F cells or the concentration of HbF in the hemolysate, it has been proposed that the amount of HbF per F cell and the proportion of F cells that have enough HbF to thwart HbS polymerization are the most critical predictors of the likelihood of severe sickle cell disease [242,243]. In sickle cell anemia, when high levels of HbF are successfully induced with hydroxyurea, HbF is distributed heterocellularly with a calculated mean HbF/F cell of 8 pg, a value that is still less than optimal (ie, less than the optimal levels of 9 to 12 pg of HbF per F cell). Methods for determining HbF/F cell are being developed but are not yet clinically available [27].
The variable amounts of HbF per F cell might permit HbS polymerization and sickling to occur in cells with low HbF, leading to intravascular hemolysis and sufficient nitric oxide (NO) scavenging by plasma hemoglobin to provoke hemolysis-related complications (figure 10) [244,245]. (See "Pathophysiology of sickle cell disease", section on 'Vasoconstriction'.)
Sickle cell disease and unusually high HbF — Rarely, African Americans with typical African beta S-haplotypes have been reported to have unusually high levels of HbF [246]. Twenty of these individuals with HbF levels of 17.2±4.8 percent were compared with 30 controls with HbF levels of 5.0±2.5 percent. The frequencies of the HbF QTL alleles BCL11A (rs766432) and HBS1L-MYB intergenic region (rs9399137) were higher in those with high HbF compared with the controls. There were frequency differences in high and lower HbF cases of single nucleotide polymorphisms (SNPs) in the 14.1 kb DNA fragment between HBG1 and HBD that included the site of the 7.2 kb Corfu deletion; nevertheless, a definitive explanation for the high HbF in these patients was not found.
●In Saudi Arabia, sickle cell anemia is concentrated in the Southwestern and Eastern Provinces. Southwestern Province patients have typical African-derived beta S-haplotypes, usually the Benin. Southwestern Province patients differ phenotypically from African American patients, with fewer episodes of stroke, priapism, and leg ulcers, and a higher prevalence of splenomegaly that might be related to differences in HbF levels or co-inherited alpha thalassemia [247-250]. In 77 patients whose mean HbF was 11.9±7.1 percent, a level almost twice that of African Americans, BCL11A was the sole QTL associated with HbF [251]. Despite having African-derived haplotypes, the genetic population structure of these Southwestern Province cases was similar to that of all Arabs. New candidate genes that might explain this observation have not been found [252].
●In the Eastern Province of Saudi Arabia and in parts of India, sickle cell anemia is commonly associated with the AI beta S-haplotype [250,253-258]. The phenotype of Indian and Saudi patients seems similar. Although AI sickle cell disease had been considered "benign" and attributed to high HbF, these studies were largely done in children [250,258-260]. Studies in adults have suggested that as HbF falls with maturation, the phenotype of disease changes [261-263]. In 104 adults with average age of approximately 27 years, 96 percent had painful episodes; 47 percent had acute chest syndrome; 18 percent had osteonecrosis; 17 percent had priapism; 6 percent had overt stroke; 66 percent had cholelithiasis; and none had leg ulcers. This change in phenotype might be due to the continued decline in HbF level from 30 percent in childhood to approximately 15 to 20 percent in adults [222]. As their HbF is distributed heterocellularly, insufficient numbers of "protected" F cells could be the cause of this phenotypic switch. In another study of 137 AI haplotype patients, all were homozygous for AI beta S-haplotype-specific elements cis to HBB including core elements of the LCR, HBG promoters, the BP1 binding site and indels; known SNPs in BCL11A and HBS1L-MYB explained less than 10 percent of the variation of HbF; KLF1 SNPs associated with high HbF were not present [79]. Sixty-one Saudi AI beta S-haplotype patients with HbS-beta0 thalassemia (mean HbF 17.7 percent) were similarly studied. While no associations with HbF were found, the small sample size precluded definitive findings [264].
●Homozygosity for minor alleles at rs16912979, rs7119428, and rs7482144, forming a T/A/T sub-haplotype, was exclusive to the AI haplotype. It was hypothesized that this sub-haplotype might represent a functional cis-acting domain modulating HBG2 expression. Patients with the Senegal haplotype have the minor allele of rs7482144 but not rs16912979. Perhaps this divergence accounts in part for the differences in HbF between Senegal and AI haplotypes as 16912979 is in HS-4 of the locus control region (LCR) that has strong binding signals for GATA1, GATA2, and POLR2A [265]. Although it is likely that cis-action regulation plays a major role in HbF gene expression in the AI haplotype, the mechanism of this association has not been defined experimentally except for the increase in HbF when the motif containing rs7482144 is disrupted [54].
●Two studies totaling about 900 Saudi adult HbS homozygotes with the AI haplotype found an association between variants in ANTXR1, an anthrax toxin receptor, and HbF [266,267]. There was no association of ANTXR1 variants with HbF in 970 HbS homozygotes with other HBB haplotypes. ANTXR1 variants explained 10 percent of HbF variability compared with 8 percent for BCL11A. The two genes had independent additive effects on HbF, together accounting for 15 percent of variability. Studies in erythroid progenitors suggested that ANTXR1 was a repressor of HbF expression. A mechanism explaining this genetic association has not been defined experimentally.
EVALUATING INCREASED HbF IN ADULTS — In adulthood, the level of HbF is generally <1 percent, and HbF will only be present in <4 percent of red cells (F cells). (See 'Detection of HbF and measurement of HbF levels' above and 'F cells (HbF-containing cells)' above.)
A conceptual scheme for classifying increased levels of HbF in adults is shown in the figure (figure 8). Increased HbF in adults should be evaluated using the following criteria:
●Level of HbF
●Whether increased HbF is an isolated finding or is secondary to other acquired or genetic conditions
●The accompanying hematologic profile (eg, level of hemoglobin, red cell indices, reticulocyte count)
●Presence of abnormal hemoglobins (eg, HbS, Hb Lepore), altered levels of HbA2
HbF level assay is appropriate in diagnostics and in monitoring therapy for HbF induction (eg, hydroxyurea in the treatment of sickle cell disease and beta thalassemia). In the context of interpretation of the hemoglobin disorders, the level of HbF together with red cell indices and HbA2 level helps to differentiate beta thalassemia trait from delta beta thalassemias and hereditary persistence of fetal hemoglobin (HPFH). In sickle cell disease and beta thalassemia, increased HbF is clinically important. When found serendipitously on high performance liquid chromatography (HPLC) analysis during diagnostic studies for other anemias, increased HbF rarely has clinical significance. Over interpreting this finding should be avoided.
Specific examples — The following are examples of the most common conditions that result in increased levels of HbF (table 1):
●Individuals with sickle cell anemia have average HbF levels of 5 to 8 percent before treatment with hydroxyurea; such treatment may increase HbF levels markedly. (See "Hydroxyurea use in sickle cell disease".)
●Slightly increased HbF levels (ie, >1 percent) along with hypochromic microcytic red blood cells and elevated HbA2 suggests the presence of heterozygous beta thalassemia. If HbF is >3 percent, this suggests co-inheritance of variants of the major known HbF quantitative trait loci (QTLs). (See 'Beta thalassemia' above.)
●It can be difficult to determine if high HbF (5 to 12 percent), with normal red cell indices and normal HbA2 during pregnancy is due to co-inheritance of HbF QTL variants boosted by the physiologic erythropoietic stress of pregnancy or is due to primary HPFH. The two conditions can be differentiated by DNA analysis or by re-testing the patient following delivery. (See 'Prematurity, diabetes, and pregnancy' above.)
●A HbF level >10 percent, when accompanied by hypochromic microcytic red blood cell indices and normal HbA2, is likely to be due to the presence of delta beta thalassemia. (See 'Delta beta thalassemia' above.)
●An increase of HbF to more than 10 percent of total hemoglobin, along with normal or slightly reduced hemoglobin and minimally affected red cell indices (ie, mean corpuscular volume [MCV] and mean corpuscular hemoglobin [MCH]) suggests the presence of a primary HPFH, usually non-deletional. (See 'HbF in the thalassemias and hereditary persistence of fetal hemoglobin' above.)
●Extremely high levels of HbF (ie, 30 to 100 percent) in association with severe anemia and hypochromic microcytic red cell indices is a typical finding in severe beta thalassemia. HbF of 100 percent with only minor degrees of microcytosis and little or no anemia suggests homozygosity for deletional HPFH. (See 'Beta thalassemia' above.)
Resources for testing — Making a specific diagnosis in some of the above examples (eg, sickle cell anemia, beta thalassemia) can be accomplished in most clinical laboratories. However, some (eg, HPFH, delta beta thalassemia) may require the use of molecular methods. (See "Methods for hemoglobin analysis and hemoglobinopathy testing", section on 'Referral to a specialized laboratory'.)
The decision concerning which reference laboratory to send the sample to depends upon several factors. Many hospitals and medical centers have contractual agreements with one of the national reference laboratories (eg, LabCorp, Quest, ARUP, Mayo Clinic) and forward the samples to these laboratories for further analysis. Some institutions prefer to forward samples to one of the few laboratories that specialize in globin abnormalities. This choice may depend upon the familiarity of the referring clinician. In addition, there are three laboratories in the United States that specialize in globin abnormalities and are based in academic institutions. These include:
●Titus HJ Huisman Hemoglobinopathy Laboratory at Augusta University, Augusta, GA (https://www.augusta.edu/centers/blood-disorders/hemoglobinopathy/index.php)
●Hemoglobin Diagnostic Reference Laboratory at Boston University, Boston, MA (www.bu.edu/sicklecell)
●Hemoglobinopathy Laboratory at UCSF Benioff Children's Hospital, Oakland, CA (https://www.childrenshospitaloakland.org/main/departments-services/hemoglobinopathy-laboratory-123.aspx)
SUMMARY
●HbF levels – Fetal hemoglobin (hemoglobin F, HbF) predominates during fetal development but is a minor component of hemoglobin in the adult (eg, <1 percent of total hemoglobin), and is generally of little pathophysiologic importance. Major regulators of the switch from HbF to adult hemoglobin (HbA) are illustrated in the figure (figure 7) and discussed above. (See 'Biology of fetal hemoglobin' above and 'Hemoglobin switching: genetic basis of HbF expression' above.)
●Beta thalassemia – Individuals with beta thalassemia major are generally asymptomatic from birth until the age of 6 to 12 months, during which time HbF is their major hemoglobin. Following the switch from fetal to adult hemoglobin, those who are able to produce the most HbF have the mildest disease. (See 'Beta thalassemia' above.)
●Sickle cell disease – Individuals with sickle cell anemia can be asymptomatic until the age of 6 to 12 months. Thereafter, HbF is variably increased in most individuals. Children and adults with sickle cell anemia with the highest HbF levels usually have the fewest complications of disease. This may be due to inheritance of variants in quantitative trait loci (QTL) modulating HbF, hereditary persistence of fetal hemoglobin (HPFH), or treatment with agents such as hydroxyurea. (See 'Sickle cell disease' above and 'Sickle cell disease and unusually high HbF' above.)
●Other conditions – HbF levels are increased in other acquired and genetic conditions, but the clinical impact is far less than in the hemoglobinopathies such as sickle cell disease and beta thalassemia. (See 'Secondary increases in HbF' above.)
●Manipulating HbF levels – Because of its disease-modifying effects, especially in beta thalassemia and sickle cell disease, there has been a strong focus on understanding (and potentially reversing) the switch from fetal to adult hemoglobin. (See 'Hemoglobin switching: genetic basis of HbF expression' above and "Hydroxyurea use in sickle cell disease" and "Investigational therapies for sickle cell disease".)
88 : An erythroid enhancer of BCL11A subject to genetic variation determines fetal hemoglobin level.
120 : MicroRNA-15a and -16-1 act via MYB to elevate fetal hemoglobin expression in human trisomy 13.
160 : Production of erythrocytes that contain fetal hemoglobin in anemia. Transient in vivo changes.
222 : Sickle cell disease in the Eastern Province of Saudi Arabia: Clinical and laboratory features.
239 : Estimation of fetal hemoglobin levels in individual red cells via fluorescence image cytometry
