Nrf2 Activation Attenuates Hemolysis but Promotes Pulmonary Vascular Dysfunction in Sickle Mice

Nuclear factor erythroid 2-related factor 2 (Nrf2) is the major transcription factor that coordinates the body's antioxidant and cytoprotective defense against a variety of toxins. Several Nrf2 activators can reactivate gamma globin gene expression and augment fetal hemoglobin production. More recently, genetic and pharmacologic evidence have shown that Nrf2 activation can specifically mitigate the severity of hemolytic anemia, and systemic and local inflammation in transgenic sickle cell disease (SCD) mice. Based on these encouraging results Nrf2 activation has emerged as an attractive therapeutic strategy in SCD. However, the BEACON trial of the Nrf2 activator CDDO-Methyl Ester (CDDO-Me) showed that this therapeutic approach can cause adverse cardiovascular events in patients with chronic kidney disease with comorbid diabetes. Hitherto, the efficacy-toxicity profile generated by individual Nrf2 activating drugs has not been investigated in SCD. There are hundreds of synthetic and naturally occurring Nrf2 activating compounds, and each class of Nrf2 activating compound has a unique pharmacokinetic, pharmacodynamic, toxicokinetic and toxicodynamic profile. We have recently demonstrated that intravascular hemolysis deteriorates with aging in transgenic sickle (SS) mice in a process that can be mitigated by the Nrf2 activator 3H-1,2-dithiole-3-thione (D3T) (Ghosh et al., JCI Insight, 2016). In this study, an in vitro screen of five Nrf2 activating compounds revealed CDDO-Me to be the most potent inducer of cytoprotective enzymes in human pulmonary microvascular endothelial cells. Thus, we performed a long-term prophylactic CDDO-Me treatment of SS mice and examined the effect of the drug on intravascular hemolysis and vascular dysfunction. A cohort of newly weaned SS mice aged ~4 weeks were randomly assigned to receive CDDO-Me (20µmoles/kg/TIW, n=6) or Vehicle (DMSO/TIW, n=10) by oral gavage for 4 months. After the treatment, the total hemoglobin increased by 10% in the CDDO-Me group while it decreased by 5% in the vehicle-treated group (p<0.05). Plasma concentration of VCAM-1 increased in both groups compared to their respective baseline (vehicle p<0.001; CDDO-Me p<0.05). The lack of impact on VCAM-1 by CDDO-Me was contrary to our results with D3T (Ghosh et al., JCI Insight, 2016). To examine this further, we performed vascular reactivity studies on isolated pulmonary arteries. Second order pulmonary arteries were isolated, cut into 2mm rings, and mounted using 40 μm wire. Rings were set at a resting tension equivalent to 100 mmHg transmural pressure. Cumulative dose response curves of endothelin-1 (100 pM-10 nM) and acetylcholine (10nM-100 μM) were used to study vasoconstriction and vasodilation, respectively. Pulmonary arteries isolated from the CDDO-Me treated SS mice showed a 15% reduction in acetylcholine stimulated vasodilation compared to the arteries isolated from the vehicle treated littermates (vehicle n=5, CDDO-Me n=5; p<0.001). In vitro studies of rat smooth muscle cells revealed significant up-regulation of endothelin receptor subtype A (ET_A) mRNA by CDDO-Me (p < 0.01, n=6). We obtained similar results of increased ET_A expression in rat smooth muscle cells treated with another Nrf2 activator, Dimethyl Fumarate (DMF) (p < 0.01, n=6). Upregulation of endothelin or the ET_A is frequently found in pulmonary hypertension. The hypertensive phenotype that we report here in SS mice treated with CDDO-Me is consistent with the adverse cardiovascular events found in the BEACON trial. Together our results show that while CDDO-Me can stabilize and potentially improve hemolytic anemia in SCD, this drug and potentially other Nrf2 activators may promote the development of pulmonary hypertension, which is a major adult complication of SCD.