HO-1 and CO: fighters vs sickle cell disease?

In this issue of Blood, Belcher et al report on the use of pegylated hemoglobin saturated with carbon monoxide (CO) to inhibit vaso-occlusive events in mouse models of sickle cell disease (SCD).¹ 

It has been more than 60 years since sickle cell anemia (SCA) was first characterized at the molecular level by Linus Pauling. SCA occurs when thymine is substituted for adenosine in the 6th codon of the betaglobin gene, resulting in a substitution of the hydrophilic aminoacid glutamic acid by the hydrophobic amino acid valine (Hb S). SCD includes SCA and the compound heterozygous sickle hemoglobinopathies. It is one of the most common monogenetic diseases worldwide, responsible for over 80% of the significant hemoglobinopathies in the world. Remarkable progress has been achieved in the care of individuals with SCD, including the use of hydroxycarbamide (hydroxyurea), yet the only definite therapy remains limited to stem cell transplantation, and the overall life expectancy is still quite low.²  Development of new therapies is therefore required.

The work of Belcher et al in this issue¹  explores the therapeutic potential of CO in 2 mouse models of SCD, first shown to reduce the degree of sickling of erythrocytes in a human subject in 1963. The investigators had previously shown that inhaled CO reduced vascular stasis in transgenic sickle mice expressing β^s hemoglobin.³  They now extend their previous findings by using a pegylated form of hemoglobin (polyethylene glycol-conjugated human Hb) saturated with CO gas, termed MP4CO. Upon administration, MP4CO releases CO that rapidly equilibrates with erythrocyte Hb. MP4CO was shown to be efficacious in limiting vascular stasis induced by hypoxia in NY1DD transgenic sickle mice and in preventing cardiorespiratory collapse induced by hemin in heterozygous HbAS-Townes mouse, models for SCD and SCD trait, respectively.

This work is quite important, not only because it describes a novel therapeutic modality in experimental mice that could have the potential for clinical applications in humans but also because of new insights into the mechanism(s) of action, which may open the window to explore additional therapeutic strategies aimed at different targets. The pathogenesis of SCD is primarily determined by polymerization of Hb S during its deoxygenation, resulting in sickling of the erythrocytes that become rigid, irregularly shaped, and unstable, with subsequent intravascular hemolysis, blood cell adhesion, vaso-occlusion, and hypoxia-reoxygenation (H/R) injury. Reactive oxygen species can be generated at various points, especially during the reoxygenation phase when the reintroduction of oxygen to the tissues leads to a marked increase in the concentration of reactive oxygen species,⁴  activation of NF-κB⁵  and inflammatory responses. The investigators found that administration of MP4CO resulted in decreased activation of NF-κB together with decreased expression of P-selectin and von Willebrand Factor in vessels of the lungs and the liver. It appears that most, if not all, of these effects were due to the release of CO, because the administration of the pegylated hemoglobin saturated with oxygen instead of CO (MP4OX), 24 hours before hypoxia, did not alter the degree of vascular stasis, and although MP4OX somewhat inhibited vascular stasis when administered 30 minutes after 1 hour of hypoxia and 1 hour of reoxygenation, it did so to a lesser extent than did MP4CO.

CO can directly mediate these effects as it has been reported that it exerts antiinflammatory and possibly antioxidant actions in the vasculature.⁶  At the same time, CO can trigger upregulation of Heme oxygenase-1 (HO-1),⁷  an important antioxidant and antiinflammatory enzyme, responsible for the catalytic cleavage of heme groups to generate equimolar amounts of biliverdin, CO, and Fe²⁺⁶. Indeed, the investigators determined that administration of MP4CO led to increased expression of HO-1 in vessels of the lungs and liver. Treatment with tin-protoporphyrin (Sn-PP), which binds to the catalytic site of the HO-1 molecule in an irreversible fashion and inhibits its enzymatic activity, resulted in complete abrogation of the MP4CO-mediated reduction in vascular stasis, strongly suggesting that MP4CO beneficial effects were mediated by HO-1 upregulation. These findings are consistent with previous reports from the same group in which HO-1 delivery by the Sleeping Beauty transposon system⁸  or administration of bilirubin or CO, HO-1 enzymatic byproducts, resulted in decreased vascular stasis in sickle cell transgenic mice.³  It is possible, however, that Sn-PP could have induced nonspecific pharmacological effects, collateral to HO-1 inhibition. Therefore, full demonstration that MP4CO-mediated beneficial effects are HO-1 mediated will require the use of HO-1 null mice.

The investigative team also showed that HO-1 induction was likely mediated by activation of the Nrf2 transcription factor because there were increased levels of Nrf2 protein in the nucleus. Nrf2 is regarded as a master regulator of the antioxidant defense, comprising a large number of antioxidant genes and phase 2 detoxifying enzymes, in addition to HO-1.⁶  It is possible that Nrf2 activation may have led to upregulation of various antioxidant genes, in addition to HO-1, that could have contributed to the MP4CO-induced beneficial effects. Although the findings shown by the authors are suggestive of an involvement of Nrf2, future work with Nrf2 null mice is required to determine whether MP4CO effects and HO-1 induction were entirely Nrf2 dependent. Fully dissecting the involvement and concerted actions of HO-1 and Nrf2 on MP4CO-induced effects in the vasculature are needed because it has been shown that they could exert opposite actions, beneficial vs deleterious, in other vascular inflammatory processes such as atherosclerosis.^6,9 

The present work invites the continued exploration of the therapeutic potential of MP4CO, which will need to be carefully titrated to avoid CO toxicity and the dreaded reduction in the capacity of erythrocytes to deliver oxygen to the tissues.¹⁰  In addition, the identification of convergent beneficial actions from Nrf2, HO-1, and CO into a synchronized axis that feeds onto itself, as the end byproduct (CO) is capable of inducing Nrf2 activation and HO-1 upregulation, offers an opportunity to design additional therapeutic interventions that can be aimed at the different levels of this axis.