Fetal hemoglobin in β hemoglobinopathies: Is enough too much?

Abstract

Cell-based therapies inducing about 40% fetal hemoglobin (HbF), or a HbF-like hemoglobin in most erythrocytes, can—at least in the short-term—effect a cure or near-cure of β hemoglobinopathies, which are humankind's most common Mendelian diseases.1-3 In sickle cell disease, a point mutation in the normal β-globin gene (HBB) directs the synthesis of sickle hemoglobin (HbS), which polymerizes on deoxygenation, damaging the red cell and triggering a complex pathophysiology. HbF exerts a powerful anti-polymerization effect because the chances of both the HbF tetramer (α2γ2) and the hybrid tetramer α2γβS entering the polymer phase are nearly nil.4 Hundreds of different HBB mutations cause β thalassemia where insufficient β-globin synthesis allows free unpaired α globin to accumulate, leading to ineffective erythropoiesis and severe anemia. γ-Globin pairs with α globin forming HbF and preventing, at its earliest stage, the pathophysiology of the β thalassemia phenotype. Hemoglobin isoforms change during progression from embryo to fetus to adult. The switch from HbF to HbA expression starts midway through gestation and normally is complete within months after birth. HbF in normal adults constitutes <1% of total hemoglobin. Sickle cell disease adults average about 5% HbF; homozygotes for β0 thalassemia have 100% HbF, albeit at concentrations insufficient to sustain normal life. The controls of the switching process are partly understood and depend largely on the activation of HbF repressors like the gene BCL11A.5-7 The successes of gene therapy of β hemoglobinopathies2, 3 have prompted the question: can blood with high concentrations of HbF cause physiological impairment? HbF evolved to potentiate the transfer of O2 from maternal blood to fetal tissues: a goal achieved by the higher O2 affinity of HbF compared with HbA because of the insensitivity of HbF to 2,3 BPG, the major modulator of hemoglobin-O2 affinity.8 Hemoglobin-O2 affinity can be estimated by measuring the P50, or partial pressure of O2 where hemoglobin is half saturated. P50 of HbF is about19 torr while HbA is 26–30 torr9 (Figure 1A,B). An individual with 12% HbF and 75% HbA had normal P50 because the tetramer species included only 5% α2γ2 and >90% HbA and α2γβ hybrids. Similarly, α2γβS hybrids predominate in sickle cell anemia patients with very high HbF levels who have near-normal P50 (Figure 1C). P50 is convenient to measure in the laboratory but does not precisely estimate tissue oxygenation in sickle cell disease. Factors that affect the kinetics of HbS polymerization, like microvascular flow rates, tissue O2 gradients, hemoglobin isoform distribution, and MCHC, all have important roles.9 (modified from Reference 10) Currently, there is little evidence to suggest that the level of HbF needed to “cure” β hemoglobinopathies is harmful developmentally or physiologically. Fetal development could be retarded if mothers have increased HbF that reduces O2 transport from maternal to fetal blood. Homozygous for large gene deletions that remove HBB but preserve the HbF genes have 100% HbF. These individuals with hereditary persistence of HbF (HPFH) seem clinically well.11 Two women, 1 with 100% and 1 with 80% HbF (reduced to <50% by exchange transfusion), delivered normal healthy children.12 These observations should be cautiously interpreted as data on these rare individuals are limited. Carbon monoxide (CO) binds hemoglobin forming carboxyhemoglobin that has a low P50. Infants of mothers who smoked, and therefore had higher than normal levels of CO, were compared with infants of mothers who did not smoke. P50 of infants of smokers had an inverse relationship to maternal CO. HbF was 72.6% in infants of smokers compared with 69.2% in controls. While statistically significant, these changes in dynamics of oxygen transport were deemed trivial by the authors.13 The effects of P50 on fetal development were studied in patients with high O2 affinity hemoglobin variants. These rare individuals have erythrocytosis, a defining feature of high O2 affinity variants and commonly have thrombosis. Among 12 patients with high O2 affinity variants with P50 between 14 and 25 torr, 23 pregnancies resulted in 78% live births.14 Three additional women with high affinity variants with P50 9.1–12 torr had normal pregnancies and fetuses.15 Seven women whose blood P50 of 27.3 ± 0.4 mm Hg was at the lower bound of the normal range had smaller babies than women with P50 of 28.5 ± 0.3 torr.16 Among 21 normal women with third-trimester fetal growth retardation, hemoglobin-O2 affinity was linearly related to birth weight.17, 18 Twenty-two women with 11%–100% HbF had 43 pregnancies with 33 live births. Four women, all with severe forms of β thalassemia and HbF of 70%–100%, had growth-restricted or small for gestational age fetuses; fetal growth was normal in 17 heterozygotes with different genotypes of β thalassemia who had HbF 11%–29%.19 Women with sickle cell disease and HbF >15% had superior pregnancy outcomes and larger fetuses than patients with lower HbF.20 This is consistent with the observation that HbF benefits many of the complications of sickle cell disease.21 Is HbF with its reduced P50 detrimental at altitude? Studies in two patients with a high O2 affinity variant and P50 of 17 torr showed minimal increases in erythropoietin, normal maximal oxygen consumption, and normal 2,3-BPG response when studied at ~3000 m. The investigators concluded that low P50 might provide some preadaptation to altitude.22 A patient with ~23% HbF and the remainder largely HbA had 34% hybrid α2γβΑ tetramers with only 5% α2γ2; P50 was 26.3 torr.9 The calculated P50 and O2 delivery for sickle cell disease and 35% HbF, at a rate of oxygen pressure decrease of 40 mm Hg/s, was normal. Hybrid α2βSγ tetramers should account for about half of total hemoglobin in sickle cell anemia patients with 40% HbF. Their remaining hemoglobin is HbS, which has high P50 and HbF that has low P50. Oxygen delivery by unpolymerized HbS is like that of HbA. Whole blood P50 in sickle cell anemia patients with successful gene therapy should approximate normal as it is a composite of the P50 of hybrid tetramers, HbS and HbF. (Fig, 1C) Measuring P50 in patients with sickle cell disease with a range of HbF levels following gene or other HbF-inducing therapy should be informative. Additive gene therapy that introduces a fetal-like HBB (HbAT87Q) achieves about 40% HbAT87Q with excellent clinical results.1 HbAT87Q in mice had an O2 affinity like that of HbA.23 In two patients with sickle cell disease treated with LentiGlobin gene therapy, and achieving HbAT87Q levels of about 46%, P50 was 27–29 torr compared with 32 torr in untreated sickle cell disease controls.24 Although it is unlikely that high HbF levels, even 100% that occurs after CRISPR-Cas9 gene editing the BCL11A enhancer in β thalassemia, will be physiologically detrimental, LentiGlobin treatment should not incur the same issues of hemoglobin-O2 affinity raised with HbF induction. The drug voxelotor binds the amino-terminal of α globin locking the HbS tetramer in its non-polymerizing oxy configuration. At therapeutic doses, about 30% of hemoglobin is modified, but the P50 in sickle cell disease patients is nearly normal because the usual right-shift in hemoglobin-O2 equilibrium is caused by increased 2,3 BPG and the molecular crowding of HbS-containing erythrocytes is offset by the left-shift caused by voxelotor.25 Any clinical benefits of low P50 and reduced HbS polymerization are theoretically offset by reduced O2 delivery to tissues.9, 10 In clinical trials, voxelotor reduced hemolytic anemia and appeared to decrease acute vasoocclusive events, at least in the best responders to this treatment.26, 27 There was no evidence for tissue hypoxia. In normal volunteers given voxelotor P50 fell from 28 to 26 torr (Figure 1E). Arterial O2 saturation increased, perhaps as a result of hyperventilation, but there was little effect on cardiopulmonary responses during normoxic and hypoxic low-intensity exercise.28 Clinical trial results of voxelotor do not yet confirm the theoretical adverse effects of increasing hemoglobin-O2 affinity (summarized by Howard et al27). A second therapeutic approach to reducing P50 in β hemoglobinopathies uses inducers of PKLR, the gene for erythrocyte pyruvate kinase, to drive glycolysis down the Embden–Meyerhof pathway, bypassing the Rapoport–Luebering shunt, decreasing the level of 2,3 BPG and P50. Early-phase clinical trials have shown reduced hemolysis in sickle cell anemia.28, 29 The hemoglobin-O2 dissociation curve in PKLR-agonist treated patients, while left-shifted is normally sigmoidal compared with biphasic curve seen with voxelotor perhaps benefitting tissue O2 delivery.9 Compound heterozygotes for HbS and HPFH with HbF about 30% are usually well.30 Available evidence, albeit limited, strongly suggests that induction of HbF to levels >40% in sickle cell disease will not impact adversely fetal development or normal physiology. Although data are inconclusive and the critical studies have not been done, limited observations on maternal P50 and the fetus suggest that HbF levels sufficient to normalize hemoglobin concentration in β thalassemia might affect fetal growth. Additional studies of O2 transport and hemoglobin function in β hemoglobinopathy patients with different HbF levels, in addition to long-term studies of gene therapies inducing very high levels of HbF in most erythrocytes, will fill the gaps in our knowledge of the effects of HbF in human physiology. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.

Topics & Concepts