Fetal Globin Gene Inducers: Novel Agents & New Potential

Abstract

Inducing expression of endogenous fetal globin (γ-globin) gene expression to 60-70% of alpha globin synthesis produces β-thalassemia trait globin synthetic ratios and can reduce anemia to a mild level. Several classes of therapeutics have induced γ-globin expression in beta thalassemia patients and subsequently raised total hemoglobin levels, demonstrating proof-of-concept of the approach. Butyrate treatment eliminated transfusion requirements in formerly transfusion-dependent patients with treatment for as long as 7 years. However, prior generations were not readily applicable for widespread use. Currently, a novel oral dual-action therapeutic sodium 2,2-dimethylbutyrate is in clinical trials, an oral decitabine formulation is under development, and agents with complementary mechanisms of action can be applied in combined regimens. Identification of 3 major genetic trait loci which modulate clinical severity provides avenues for developing tailored regimens. These refinements offer renewed potential to apply fetal globin induction as a treatment approach in patient-friendly regimens that can be used world-wide.

Introduction

β-thalassemia syndromes are the most common monogenic disorders worldwide, characterized by molecular mutations of the β-globin chain of adult hemoglobin (HbA; α₂β₂ with resulting defective or deficient β-globin chains and an excess of unmatched α-globin chains.^1-8 The excess α-globin damages the red blood cell membrane and causes apoptosis of developing erythroblasts and intramedullary hemolysis.^1-8 In patients with co-inheritance of β-thalassemia and hereditary persistence of fetal hemoglobin (HbF; α₂γ_β) mutations, high-level expression of fetal globin (γ-globin) ameliorates the clinical severity of thalassemia, including the degree of anemia¹. Many patients with higher γ-globin levels than their counterparts with the same mutations often do not require regular transfusions, or do not require transfusions on a regular basis as early in life as others. These observations, and clinical trials of fetal globin inducers, have clearly established that patients with β-thalassemia benefit from persistence of, or pharmacologic induction of γ-globin.^1-20 Inducing γ-globin expression by even small increments is an established therapeutic approach and is most likely to be widely feasible to apply worldwide, as the γ globin genes are universally present and integrated in hematopoietic stem cells^1-2.

Lessons from prior trials and factors for tailored regimens

Proof-of-principle has been demonstrated in clinical trials in which pharmacological reactivation of γ-globin expression have reduced anemia and eliminated transfusion requirements in patients with thalassemia. Fetal globin re-induction was accomplished with chemotherapeutic agents, particularly 5-azacytidine, and 5-aza-2-deoxy-cytidine (decitabine),^14-20 and with short-chain fatty acids (SCFAs), such as arginine butyrate (AB) and sodium phenylbutyrate ^{2,9-10,12-13,37}. SCFAs would be preferable for a long-term therapy compared to chemotherapeutics, which typically are cytotoxic and have carcinogenicity risks long-term. However, while SCFAs have not been found to be mutagenic, many prototype SCFAs also have limitations, including rapid metabolism and requirements for high doses, and those which are also global HDAC inhibitors cause inhibition of erythropoiesis.² Erythropoiesis stimulating agents are limiting for long term application, also due to parenteral administration and high costs.^19,21-24 However, these 3 classes of therapeutics have already rendered some thalassemia patients transfusion-independent in small trials. Agents which are more tolerable would allow for broader application, particularly where thalassemia is common globally and transfusions carry particularly high risks of infections².

In the SCFA clinical trials, several observations were highly informative regarding magnitude of responses and patterns of thalassemia mutations which appear to respond differently. Several γ-globin inducers have produced significant hematologic responses with increases in total hemoglobin of 2-5 g/dL or more above baseline, and although the clinical trials have been small, patients with diverse thalassemia syndromes have had significant responses, including transfusion-independence^2,10-19. Clinical trials of 5-azacytidine, phenylbutyrate, arginine butyrate, and EPO preparations have all demonstrated responses of this magnitude in some thalassemia patients. Collins and colleagues found that sodium phenylbutyrate could increase total Hgb by 2 grams/dl above baseline, and that responses occurred more frequently in patients with EPO levels > 130 mU/ml.⁹ Increases in total Hgb levels of 1-5 g/dL above baseline were achieved when these agents were administered for at least 3-6 months^9-10,38,. This is remarkable, since thalassemic cells survive for only a few days^4-8, compared to the normal red cell survival of 120 days.

Of the chemotherapeutic agents, hydroxyurea (HU) treatment has increased total Hgb by 0.6-1.0 g/dL in HbE/β-thalassemia patients, and although not as great an effect in magnitude, was still significant.^2,16-17 Hajjar and Pearson reported that γ-globin increased rapidly with HU treatment, with a 6-week treatment time-frame required for a peak response, but was followed by a decline in total Hgb, suggesting cellular growth inhibition.¹⁵ 5-Azacytidine administration has increased total Hgb levels by an average of 2.5 g/dL (range 1-4 g/dL), even in end-stage patients with life-threatening severe anemia.^1-2,18-19

Of the SCFAs and histone deacetylase (HDAC) inhibitors, arginine butyrate (AB), administered first intensively and then intermittently, increased total Hgb levels by 1-5 g/dL (mean 2.9 g/dL) when administered frequently for at least three months, as shown in Figure 2.¹⁰ AB treatment has made four patients transfusion-independent for several years with home therapy, given initially for 5 days/week for 4 weeks, followed by intermittent or pulse therapy for 4 nights every other week to optimize γ-globin induction and simultaneously avoid the anti-proliferative effects common to HDAC inhibitors. AB has been safe in long-term use, with no butyrate-related adverse events in more than 16 patient-years of home administration provided by parents. A representative profile of rises in total hemoglobin levels in a formerly transfusion-dependent patient is shown in Figure 2C. Isobutyramide has also increased fetal globin within 28 days of treatment and has reduced transfusion requirements.^12-13

Figure 2.

(A) Representative responses in fetal hemoglobin and (B) Total hemoglobin responses to arginine butyrate (AB) +/− erythropoietin (EPO) in a patient with β⁺-thalassemia. The combination of the two therapeutics produced the highest rise in total hemoglobin which required both agents; while only Butyrate treatment produced an increase in fetal globin (C) AB treatment rendered a formerly transfusion-dependent a β^0/+-thalassemia patient transfusion-independent for 7 years. An initial frequent dose regimen (solid line) and a maintenance pulse regimen, administered 4 nights, twice per month (shown by the dotted line). Addition of EPO to the Butyrate regimen did not produce any added therapeutic benefit in this type of thalassemia patient.

EPO preparations have increased Hgb levels by 1-3 g/dL above baseline in thalassemia intermedia patients, with highest responses in children, and have decreased transfusion requirements in thalassemia major.^{1-2, 21-24}

The studies generally found that HbF did not increase, so that thalassemic red blood cell production increased with EPO. All of these trials have already shown proof-of-principle of the utility of therapeutic induction of γ-globin +/− enhancement of erythropoiesis in β-thalassemia patients.

Studies with HbF Inducers with differing mechanisms of action

For a long term therapeutic strategy, γ-globin inducers with high potency or improved pharmacologic profiles are desirable, so that smaller, more tolerable doses than those required for SPB or AB, and combination of agents with complementary actions can be administered, as required in most medical conditions. Agents with differing mechanisms of action, such as agents with global epigenetic actions and targeted activity, could be combined sequentially (Supplement Table 1). Chemotherapeutic agents are not suitable for simultaneously dosed combinations, as this would result in additive cytotoxicity and greater degrees of erythroid cell apoptosis. Combinations of butyrate and 5-azacytidine in the baboon were shown by Stamatoyannopoulos and colleagues to produce a synergistic 3-fold increase in γ-globin expression, above the significant levels induced by each drug alone.² Higher responses than with single agents alone, with both additive and synergistic effects, have been found in cell culture studies and in animal models using hydroxyurea, decitabine, or MS-275 with a novel SCFAD which is not a global HDAC inhibitor , sodium 2,2 dimethylbutyrate⁴⁰.

Clinical trials of AB +/− EPO have shown that β⁺-thalassemia patients often require EPO in addition to a SCFA to achieve substantial increase in total Hgb levels. Hgb can be increased on average by 3 g/dL above baseline with AB +/− EPO.² We studied AB and EPO in combination and found that a subset of β⁺-thalassemia patients had relatively low HbF (<30%) and low baseline EPO levels (<130 mU/mL), as shown in Figure 1. The β+ thalassemia group responded to the therapeutic combination with higher rises in total hemoglobin to a greater degree than with they did with either agent administered alone.² Examples of HbF and total Hgb responses in representative β-thalassemia patients with higher vs lower baseline EPO levels to treatment with Arginine Butyrate (AB) +/− EPO are shown in Figure 2A-B. Patients with baseline EPO levels <80 mU/mL required AB and EPO in combination to elicit a high hematologic response (a rise in total hemoglobin of 3 g/dL above baseline); neither agent alone was as effective as the two agents together. When butyrate was discontinued and EPO administered alone, HbF and total Hgb gradually declined back to baseline. AB and EPO in combination were therefore more effective than either agent alone in β⁺-thalassemia. In contrast, patients with at least one β⁰-thalassemia mutation and baseline EPO levels >130 mU/mL have responded to the Arginine Butyrate alone increasing total Hgb levels by 2-4 g/dL, without any further rise in total Hgb when EPO was added, as shown in Figure 2C in a formerly transfusion-dependent patient who became transfusion-independent with home therapy with Arginine Butyrate administered 4 nights, twice per month, for 7 years. These clinical findings indicated that combinations of agents with complementary, yet distinct, molecular mechanisms of actions can produce significantly higher responses than single agents. Based on these collective observations, our group hypothesized that prolonging the survival of erythroid precursor cells for a sufficient interval may allow a γ-globin inducer to then reduce the pro-apoptotic α:β globin chain imbalance. Identifying a therapeutic with these dual actions became a therapeutic goal.

Figure 1.

Basal levels of fetal hemoglobin (HbF) and endogenous erythropoietin (EPO) in patients with β⁺- and β⁰-thalassemia. Values represent the mean baseline levels in the two groups enrolled in a trial of arginine butyrate + erythropoietin.

A new generation oral fetal globin inducer: sodium 2,2-dimethylbutyrate

Based on the beneficial therapeutic effects of sodium phenylbutyrate, arginine butyrate, and isobutyramide, an orally active SCFAD which requires more tolerable doses would be of benefit for long-term treatment in beta thalassemia. An oral butyrate derivative, sodium 2,2-dimethylbutyrate (SDMB), has entered clinical trials. This therapeutic stimulates γ-globin production in erythroid cell cultures, anemic baboons, and transgenic mice.³⁴ The agent also enhances thalassemic erythroid cell survival through Bcl-family pro-survival proteins.³⁵ SDMB has had excellent safety profile in long-term animal studies in two species and tested negative in mutagenicity testing. In normal volunteers, the therapeutic was found to have a half-life of 9-11 hours at tolerable doses from 2 to 20 mg/kg/dose, allowing for once per day dosing (shown in Figure 3).³⁹ These actions, and the pharmacologic profile, defines SDMB as an excellent candidate for single-use therapy and for combined modality therapy in thalassemia, as it can be used with therapeutics which are cytotoxic or suppress erythropoiesis, such as the global HDAC inhibitors, decitabine, or hydroxyurea. SDMB is now undergoing initial evaluation in short-term studies in β-thalassemia intermedia patients in dose-escalation trials. With beneficial dual actions on fetal globin induction and in prolonging erythroid cell survival, the agent offers potential for providing a potential mainstay or maintenance therapeutic ^33-35 to which the cytotoxic agents might be intermittently added or “pulsed”.

Figure 3.

Pharmacokinetic profiles of sodium 2,2-dimethyl butyrate (SDMB) in healthy human volunteers. This therapeutic was administered in single oral doses at 2 mg/kg (○), 5 mg/kg (•), 10 mg/kg (□), and 20 mg/kg (■). The drug was detectable in the plasma after single oral doses with a t_1/2 of 9-11 hours.

The influence of quantitative trait loci on basal and induced levels of HbF

HbF and proportions of F-cells can vary by 10-fold in different normal subjects and in patients with the same molecular mutations; it has been difficult to predict whether clinical courses will become thalassemia intermedia or major with the same globin mutations. The influence of specific genetic modifiers, including single nucleotide polymorphisms (SNPs) and quantitative trait loci (QTL), which alter basal HbF levels in both normal and hemoglobinopathy subjects and ameliorate clinical severity has been recognized for many years, but is now considered to be influenced by the presence of genetic modifiers, such as a T-allele present at promoter nucleotide (nt) −158 5′ upstream of the HbG (^Gγ-globin gene) on chromosome 11p15 (rs7482144), associated with elevated HbF levels particularly during stress erythropoiesis ^25,29. More recently, this and two other QTL have been found to account for nearly 50% of the variation in basal HbF/ F-cell levels.^25-31 Genome-wide SNP association studies by Thein and colleagues yielded the unexpected finding that BCL11A is a major QTL for HbF and F-cell production in normal individuals and in patients with β-thalassemia, which subsequently was confirmed in sickle cell disease; Bcl11A was found to have profound effects in preventing the fetal to adult globin switch when absent.^25-31 Thein and colleagues also showed that the HBS1LMYB intergenic polymorphism (HMIP) on chromosome 6q23 exerts significant negative effects^25,27-28. Other groups have confirmed that BCL11A on chromosome 2p16 is a major HbF QTL in populations both with or without β-hemoglobinopathies. Further, Chui and colleagues reported that BCL11A is a transcriptional repressor of HbG (^Gγ-globin) proximal promoter activity, which is abolished by butyrate in vitro³⁰. (Modifiers of fetal globin expression are listed in Supplemental Table 1). BCL11A binds to the GGCCGG motif in nucleotide −56 to −51 on the gamma globin proximal promoter, a region in which our group previously identified alterations in protein binding in the nucleated erythroid cells of patients who responded to butyrate therapy, including disappearance of a putative repressor from the region.³⁷ Those, and our more recent findings demonstrated that with butyrate and a newer agent, RB7, identified through molecular modeling, displacement of a repressor complex (potentially BCL11A) occurs in the promoter, leading to histone acetylation and γ globin transcription, highly targeted effects.^32,36-38

Differences in QTLs occur commonly. In a recent analysis of random sickle cell and thalassemia patients enrolled in clinical trials, more than 50% of the sickle cell subjects were found to have a high basal HbF genotype at the Bcl11A locus and 70% of Thai HbE-beta thalassemia patients had a high HbF genotype including the XmnI polymorphism and a Bcl11A QTL. Three QTLs in the Thai population which produce low or high F genotypes in genome wide screening studies have been found to profoundly alter thalassemia severity, producing an average 1 gram/dl difference in total hemoglobin levels with the same globin mutations.²⁶ It is entirely possible that the presence or absence of genetic modifiers (QTL) will also affect inducibility of responses to different HbF-inducing therapeutic agents. Determination of whether the presence of QTL polymorphisms are associated with differential responses to epigenetic modifiers, such as HDAC inhibitors or demethylating agents, or with targeted SCFA derivatives (SCFADs) which are not global HDAC inhibitors, will allow design of clinical trial protocols tailored for a greater likelihood of success in patients with different QTLs.

Summary

Proof-of-concept of fetal globin induction as an approach to improve globin chain balance and reduce the anemia in β-thalassemia has been established with 3 different therapeutic classes of γ-globin inducers. Chemotherapeutic agents such as hydroxyurea or decitabine and global HDAC inhibitors which inhibit cellular proliferation probably should be used intermittently in beta thalassemia, as cell growth arrest can aggravate apoptosis. A novel oral HbF inducer, SDMB, which is not cytotoxic, can be used in combination regimens with the epigenetic and cytotoxic agents. An oral formulation of decitabine is being developed, and higher potency oral inducers have been identified. Targeting clinical trials now for specific agents directed to low F vs high F QTL may identify regimens predictive of responses to different agents. These new options for HbF induction should benefit many thalassemia patients globally, particularly many who chronically live with very low hemoglobin levels (< 7 grams/dl) without transfusions.

Supplementary Material

01

Supplemental Table 1. HbF-inducing gents with complementary mechanisms of action

Table 1.

Genetic modifiers of fetal globin gene expression

Quantitative trait locus	Activity
Bcl-11A	Repressor of γ-globin
−158 C→T Gγ promoter	Increases Gγ-globin
6p22.3-23.2 QTL, myb	Decreases F-cell proportions
Xp22 QTL	Increases F-cells proportions
PDE 7B	Reduces γ-globin
MAPK5	Increases γ-globin through distant upstream site
PEX7	Binds factor that binds a negative regulator
TOX	Not determined