Endothelial Hemoglobin Alpha Mediates Iron Regulation of Endothelial Nitric Oxide

Iron deficiency is associated with heart failure, chronic kidney disease, and chronic inflammation. All of these pathologies are associated with endothelial dysfunction and impaired nitric oxide (NO) signaling. Previous clinical studies have found iron deficiency anemia (IDA) increases endothelial NO signaling and attributed this change to the reduction in circulating hemoglobin. However, while free hemoglobin is a potent scavenger of NO, it has been shown in ex vivo preparations that endothelial NO signaling is unaffected by the presence of erythrocytes suggesting changes in circulating hemoglobin do not alter NO signaling. For these reasons we focused on endothelial iron regulation. Using single cell RNA sequencing data of isolated mesenteric and adipose endothelial cells generated by our lab, we have observed arterial endothelial cells have higher transferrin receptor expression compared to capillary, veinous, and lymphatic endothelial cells. This suggests that the arterial endothelium, which controls blood flow and is especially susceptible to endothelial dysfunction, may be particularly sensitive to changes in iron status. Additionally, our lab has previously shown the alpha chain of hemoglobin (Hbα) is expressed in the endothelium of small arteries and scavenges NO. We therefore hypothesized endothelial Hbα, rather than blood hemoglobin is responsible for the increased NO signaling in IDA. We first tested this hypothesis in a mouse model of IDA in which we were able to replete vascular iron stores. C57BL/6 mice were fed an iron deficient (2-6 ppm Fe) or nutrient matched control diet (48 ppm Fe) beginning at weaning. IDA mice underwent phlebotomies (10% blood volume) at 9 and 11 weeks old to aid in the progression of anemia. To restore vascular iron stores, a subset of IDA mice received a single i.p. injection of iron dextran (FeDex; 20 mg/kg) at 12 weeks. One week later complete blood counts were assessed, tissues were collected, and blood flow studies were performed. As expected, IDA mice had low blood hemoglobin (Con: 15.54 ± 0.43 vs. IDA: 6.04 ± 0.35 g/dL; p < 0.0001). FeDex did not rescue the anemia (8.55 ± 0.74 g/dL; p < 0.0001 vs. Con), but did rescue vascular iron stores as measured by ferritin light chain protein. To investigate NO signaling, we used laser speckle contrast imaging to measure changes in in vivo blood flow in response to the NO synthase inhibitor L-NAME. The magnitude of the response to L-NAME is an indicator of NO signaling. IDA mice exhibited an exaggerated response to L-NAME indicating increased NO signaling. Rescuing vascular iron stores with FeDex restored the L-NAME response back to control levels suggesting vascular iron stores, rather than the anemia itself increased NO signaling. In order to investigate the role of endothelial Hbα, we repeated these studies in endothelial specific Hbα knockout mice (Hba1 ^fl/fl-Cdh5-CreER ^T2+) developed by us. Lack of endothelial Hbα did not prevent the increased response to L-NAME in IDA mice, but did prevent the rescue by FeDex. These data suggest a model in which endothelial Hbα participates in regulating NO in response to iron, but not the increased NO in IDA.