Targeting Regulators of Hemoglobin F

Breakthroughs have initiated largely from human genetic studies.¹  Residual HbF expression in adulthood is a heritable quantitative trait. Genome-wide association studies (GWAS) demonstrate that polymorphisms in three loci account for about half of the heritable variation in HbF level. These loci include the β-globin cluster itself, an intergenic interval between the HBS1L and MYB genes, and the BCL11A gene. Subsequent studies have demonstrated that BCL11A is a zinc-finger transcription factor that functions as a developmental stage-specific repressor of HbF expression.^2,3  It cooperates with other transcription factors to directly silence HbF expression in human erythroid precursors and controls globin switching in mice and humans by binding discrete regions of the β-globin cluster.⁴ 

A recent linkage study of a family with hereditary persistence of fetal hemoglobin identified mutations in the KLF1 (previously known as EKLF) gene associated with increased HbF.⁵  Additional studies have demonstrated that a variety of KLF1 mutations lead to derepressed fetal/embryonic globin gene expression in humans and mice.^6-8  Interestingly, KLF1 is a DNA-binding transcription factor that activates BCL11A expression by associating with the BCL11A promoter, suggesting a dual role of KLF1 in globin gene regulation by both functioning as a direct activator of adult-stage β-globin and indirect repressor of fetal-stage γ-globin.

Molecular analysis of the globin switch has also revealed the crucial roles of epigenetic modifiers in regulating HbF expression. The γ-globin genes become methylated in the adult stage and are occupied by multiprotein repressive chromatin complexes including DNA methyltransferases and histone deacetylases. Studies focusing on the biochemical characterization of the globin genes have identified a repertoire of trans-acting regulators of HbF expression (Figure), although few of these nuclear factors have yet been fully validated as robust regulators of HbF in vivo.

The identification of a new set of regulators has therapeutic implications for directed reactivation of HbF. A crucial task will be to place these factors into a hierarchy to prioritize the most promising leads for further target-based clinical development. To define the “optimal” characteristics of a target, it is important to consider a number of features including quantitative effect on globin expression, non-globin effect on erythropoiesis, impact outside of the erythroid lineage, and feasibility of therapeutic modulation. Of note, reactivation of HbF expression would be required for the lifetime of the patient to achieve desired benefit; therefore, any cumulative toxicity would be undesirable.

While some of the HbF regulators are ubiquitously expressed, others are more restricted to the erythroid lineage. Many of the factors influencing globin gene transcription have general importance in hematopoietic or erythroid differentiation, thereby complicating their use as possible targets for therapeutic manipulation.

Inhibition of epigenetic activities has been a longstanding focus for HbF induction and has been reinvigorated by chemical screens “rediscovering” their potential efficacy.⁹  DNA hypomethylating agents, such as 5-azacytidine and decitabine, have been used successfully to induce HbF expression in animal models and patients.^10-11  Inhibitors of histone deacetylases, such as butyrate and its derivates, have also been shown to increase HbF synthesis and continue in clinical development.¹²  Increased understanding of the regulatory pathways in HbF silencing may reveal additional epigenetic modifiers as potential targets. Furthermore, next-generation epigenetic modulators with increased specificity for particular isozymes or protein-protein interfaces may minimize “off-target” effects. However, the broad role of many chromatin regulators both within and beyond the erythroid lineage raises concern about the achievability of an adequate therapeutic window.

Recently identified transcription factors participating in HbF regulation also serve as attractive targets for future therapeutics (Table). BCL11A controls HbF expression in a robust and dose-dependent manner, has few erythroid non-globin targets, and is dispensable for erythropoiesis.

One concern is that BCL11A is expressed in the developing central nervous system and is essential for normal B-lymphocyte development, suggesting inhibition of BCL11A might have an impact on other cell lineages. In contrast, KLF1 expression is highly erythroid-restricted, and the dual role of KLF1 in globin regulation makes it an appealing target. However, this “master regulator” has numerous target genes and coordinates multiple aspects of terminal erythroid differentiation. Of note, a variety of erythroid phenotypes have been observed in patients carrying KLF1 mutations. An improved understanding of these genotypephenotype correlations could help predict how KLF1 could be targeted to specifically result in HbF reactivation.

Despite intensive investigation in the past two decades, hydroxyurea, a ribonucleotide reductase inhibitor and HbF-inducing agent, remains the sole FDA-approved medication for SCD. While partially effective for many patients, its utility is limited by predictable myelosuppression, unpredictable induction of HbF, and marginal benefit for patients with β-thalassemia. There remains an unabated need for more effective therapies. Strategies for potential targeting of HbF regulators include interfering with their expression and/or function. For example, shRNA-mediated inhibition of BCL11A expression induces HbF in primary human erythroid precursors and merits further exploration with emerging in vivo RNAi technologies. Small molecule inhibition of key regulators of the globin switch may be an alternative approach. Traditionally, DNA-binding transcription factors have been considered “undruggable,” but recent progress in chemical biology indicates that this important protein class may be therapeutically tractable.¹³ 

The complexity of the mechanisms of HbF regulation suggests that combination therapy consisting of two or more modulators, each with a different mechanism of action, may be the most effective strategy for the induction of very high levels of HbF while limiting adverse effects. There is actually historical precedence for this; combined therapy with hydroxyurea and recombinant erythropoietin elevates HbF levels more so than hydroxyurea alone in SCD patients.¹⁴  By gaining an increased mechanistic understanding of globin gene regulation, it is anticipated that targeted, mechanism-based therapeutic approaches for efficient HbF induction can be developed.