In this issue of Blood, Allali et al report that free hemoglobin S (HbS) released by intravascular hemolysis leads to monocyte activation and inflammatory cytokine production in patients with sickle cell disease (SCD).1 Strikingly, the effects of HbS are reported to be significantly and profoundly greater than the effects of hemoglobin A (HbA), despite the difference of only one amino acid between the hemoglobins.
Allali et al address an old and perplexing issue: Why do patients with SCD have elevated baseline inflammatory cell activation and elevated levels of nearly all proinflammatory cytokines, compared with patients with β-thalassemia or patients without SCD with intravascular hemolysis? In patients with SCD, proinflammatory cytokines induced by NF-κB signaling (ie, tumor necrosis factor [TNF], interleukin 6 [IL-6], and IL-1β) are elevated at baseline and further increase during vaso-occlusive crises, during which they activate inflammatory cells, promote endothelial cell damage, and ultimately potentiate vaso-occlusion.2 In addition, elevated levels of type 1 interferons (IFN-α/β) have been recently identified in patients with SCD,3 with these inflammatory interferons having implications during red blood cell (RBC) transfusion and viral infection.4 However, triggers for cytokine production in SCD represent a gap in knowledge in the field.
One described mechanism is that patients with hemolytic anemia have elevated levels of free heme, which has been shown to activate endothelial cells during vaso-occlusive crises.5 Hemin, the oxidized form of heme, binds a complex of Toll-like receptor 4 and myeloid differentiation 2 (TLR4/MD-2), resulting in downstream production of NF-κB–induced cytokines. Although activated monocytes producing these proinflammatory cytokines are elevated in patients with SCD,2 hemin fails to induce IL-6 production by monocytes of patients with SCD, despite elevated expression of TLR4.6
Allali et al hypothesized that HbS may contribute to the enhanced inflammation in SCD, compared with patients without SCD who had hemolytic anemia. They illustrated that although hemin and HbA had minimal effects, addition of HbS from patients with SCD to monocytes from healthy controls profoundly stimulated IL-6 and other NF-κB–induced cytokine production (see figure). They also demonstrated that HbS induced expression of IFN-α/β–stimulated genes to a lesser degree. HbS, but not HbA, promoted chemotaxis of monocytes. These effects were abrogated when adding a TLR4 inhibitor or when studying monocytes lacking TLR4. The authors also used surface plasma resonance to illustrate that HbS has strong affinity for TLR4/MD-2, whereas HbA has significantly lower affinity. Finally, treatment of humanized sickle cell mice with HbS led to increased serum TNF and IL-6, whereas HbA had a minimal effect. Interestingly, treatment of control mice with either hemoglobin had no effect on cytokine production, which may be due to hemoglobin binding to haptoglobin, which is absent in sickle cell mice.
One caveat to these findings is that lipopolysaccharide (LPS; ie, endotoxin) also binds the TLR4/MD-2 complex on monocytes and strongly induces NF-κB–induced cytokines and IFN-α/β.7 The authors excluded hemoglobin preparations containing detectable endotoxin, and by utilizing an antibiotic that inactivates LPS, they showed that LPS inactivation had no effect on HbS-induced TNF and IL-6 production. Although this indicates that HbS preparations did not contain LPS, HbS content, measured by an oxidizing reagent, ranged from 56% to 88% of total protein in HbS preparations. Thus, the possibility that undetectable levels of LPS or other components in the hemoglobin preparations influenced cytokine production cannot fully be excluded.
Nevertheless, the elevated responses to HbS beg the question: Why is the affinity of HbS for TLR4/MD-2 so much greater than that of HbA? This question requires further investigation. LPS and other endogenous hydrophobic molecules bind a large hydrophobic pocket within MD-2.7 Given that HbS has a hydrophobic valine instead of hydrophilic glutamic acid in HbA, the authors postulate that the hydrophobicity of HbS may lead to increased affinity for TLR4/MD-2. The testing of this hypothesis is eagerly awaited.
Patients with SCD have high levels of proinflammatory cytokines, which can potentiate SCD sequalae. The work by Allali et al demonstrates that cell-free HbS induces high levels of proinflammatory cytokine production by monocytes. The authors further elucidate the pathway by showing that HbS, although differing by only one amino acid from HbA, has high affinity for the TLR4/MD-2 complex. On ligation of HbS to the TLR4/MD-2 complex, downstream signaling induces NF-κB activation and subsequent proinflammatory cytokine production.
Patients with SCD have high levels of proinflammatory cytokines, which can potentiate SCD sequalae. The work by Allali et al demonstrates that cell-free HbS induces high levels of proinflammatory cytokine production by monocytes. The authors further elucidate the pathway by showing that HbS, although differing by only one amino acid from HbA, has high affinity for the TLR4/MD-2 complex. On ligation of HbS to the TLR4/MD-2 complex, downstream signaling induces NF-κB activation and subsequent proinflammatory cytokine production.
The induction of IFN-α/β by HbS may also have clinical implications. IFN-α/β were discovered for their role in antiviral immunity, and some reports have indicated patients with SCD are more susceptible to severe complications of viral infection.8 In addition, IFN-α/β promote deleterious anti-RBC antibody production in murine transfusion models.9 Given that patients with SCD have a high prevalence rate of anti-RBC antibodies, one wonders whether HbS-induced IFN-α/β production may contribute to alloantibody formation.
As the identification of elevated levels of IFN-α/β in patients with SCD is a relatively new finding, recent studies have addressed mechanisms underlying IFN-α/β production. Mitochondria from SCD RBCs were reported to induce IFN-α/β in neutrophils.10 Intravascular hemolysis in mice with SCD induces IFN-α/β production by hepatic monocytes, which enhances erythrophagocytosis of transfused RBCs. Interestingly, hemolysis-induced IFN-α/β production was shown to be independent of TLR4 activation.3 Thus, multiple mechanisms, including HbS/TLR4 interactions, contribute to IFN-α/β production in SCD, and more are likely to emerge.
In conclusion, the authors are congratulated on significant progress toward identifying mechanisms that promote inflammation in patients with SCD. The striking difference in HbS- and HbA-induced inflammatory cytokine production and monocyte activation is highly novel and will lead to studies that shed light on an underlying mechanism. Although other sources also contribute to inflammation in SCD, identification of HbS-induced inflammation by monocytes, and possibly other inflammatory cells, has the potential to lead to interventions that combat severe sequelae of inflammation in patients with SCD.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
