Targeted Disruption of Nrf2 in Non-Hematopoietic Tissues Exacerbates Hemolysis and Accelerates Chronic Lung Disease in Transgenic Sickle Cell Anemia Mice

Nuclear factor erythroid-2 related factor 2 (Nrf2) is a transcription factor that regulates the cellular defense mechanism by mediating a coordinate induction of cytoprotective antioxidant responsive element-driven genes. Nrf2 agonists augment fetal hemoglobin expression in hematopoietic progenitors and have emerged as a new class of drugs for therapeutic induction of fetal hemoglobin in sickle cell anemia (SCA). However, the cytoprotective effect of Nrf2 on the pathobiology of SCA has not been previously defined. To investigate the role and mechanism of Nrf2 in SCA independent of globin gene modulation, we generated chimeric mice with disruption of Nrf2 in non-hematopoietic tissues. A total of twenty-six Nrf2^-/- mice were transplanted with bone marrow harvested from the Berkeley sickle and hemizygous (Hemi) mice and the Townes sickle (SS) and control (AA) mice. All sickle/Nrf2-null chimeras (SS^NHNrf2-/-; n=13) developed classical hematological and intravascular hemolysis features of SCA after eight weeks of transplantation. Despite the presence of erythrocyte Nrf2, SS^NHNrf2-/- chimeras developed more severe intravascular hemolysis than SS wild-type (SS^WT) donor-litter-mates. The concentration of plasma cell-free Hb (p<0.01), plasma heme (p<0.05) and lactate dehydrogenase (p<0.05) were 1.5 to 2.0-fold higher in the chimeras than in the SS^WT mice. The chimeras had markedly higher concentrations of soluble vascular cell adhesion molecule-1 (p<0.01) and soluble P-selectin in their plasma (p<0.05), and more oxidizing extracellular redox status (p<0.05). The latter finding was corroborated by markedly lower reduced-glutathione in the SS^NHNrf2-/- mouse lungs (19.07±0.32 vs 26.5 ±1.5 nmol/mg protein, p<0.01). The worsening of hemolysis, vascular inflammation and oxidative stress in the SS^NHNrf2-/- chimeras was unrelated to the transplantation procedure since none of these effects were seen in the AA^NHNrf2-/- and Hemi^NHNrf2-/- controls. Next, we found that the SS^NHNrf2-/- chimeras had hypoxemia at baseline (oxygen saturation, 95.8±1.7%) while age-matched SS^WT mice had normal oxygen saturation (99.2±0.2%, p<0.05). Consistent with these disparate pulmonary phenotypes, the SS^NHNrf2-/- but not the SS^WT mice developed pulmonary edema at a relatively young age, which was confirmed by histological evidence of alveolar edema. Remarkably, the severely damaged SS^NHNrf2-/- mouse lungs contained a similar quantity of heme oxygenase-1 (HO-1) mRNA and protein as the relatively less inflamed SS^WT lungs. In addition, the amount of HO-1 in the plasma was also similar in these animals. These results indicate that other transcription factors modulate HO-1 induction, and that HO-1 may play a lesser role in mitigating the chronic effects of SCA. To test the latter idea, we performed a longitudinal study in a cohort of thirty-two SS and AA mice of both sexes. At 8 weeks of age, plasma HO-1 was 3-fold higher in both male and female SS mice than in the AA mice (p<0.001). The HO-1 values did not change at 12, 24 and 40 weeks in any group, however intravascular hemolysis, vascular inflammation and endothelial barrier integrity deteriorated progressively only in the SS^WT mice. Overall, our results demonstrate that the elevated HO-1 expression typical of SCA is insufficient to blunt many of the chronic effects of the disease in mice. However, other Nrf2 target genes in non-hematopoietic tissues can slow down this process. Such genes offer an attractive prophylactic therapeutic strategy to preserve organ function in adults with SCA.