Mechanisms of Hypoxic Erythrovasodilation.

Abstract Red blood cells (RBC) have the unique ability to relax blood vessels under low oxygen conditions (Pawloski et al. Nature409:622, 2001), an activity now termed hypoxic erythrovasodilation (HEV). With human erythrocytes, HEV appears to be mediated, in part, by the export of nitric oxide (NO) bioequivalents formed in the RBC membrane by nitric oxide (NO) group transfer from S-nitrosohemoglobin (SNO-Hb) to cysteine thiols in the cytoplasmic domain of the anion exchanger-1 (AE1 or band 3) protein. Alternatively, HEV has been proposed to be mediated via the nitrite reductase activity of deoxy Hb, resulting in the formation of heme-bound or nitrosyl Hb (HbNO), with subsequent export of NO from the RBC. For both paradigms, downstream export mechanisms, and a precise chemical identity of the discharged specie(s), have yet to be resolved. Interactions between RBCs and vascular tissue (i.e. endothelium and smooth muscle) involved in the HEV response remain undefined, and are the focus of the current study. Rabbit aortic rings were suspended in jacketed organ chambers filled with Krebs-bicarbonate buffer at 37°C, and bubbled continuously with either 21% O2/5% CO2 or 95% argon/5% CO2. Basal tension was maintained at 2 grams, and active tension induced with phenylephrine (PE) or prostaglandin F2a (PGF2a). Fresh (<4 days) or old (>60 days) RBCs were washed and resuspended in phosphate-buffered saline at 50% hematocrit prior to use. Following induction of active tension with PE or PGF2a, changing the aeration gas from O2 to argon resulted in a brisk increase in vascular tone (hypoxic vasoconstriction, HVC). At the peak of this contractile response, RBCs were added to the tissue chambers, causing a dose-dependent relaxation of the rings. There was no correlation between the magnitude of the HVC and that of the ensuing HEV. Following physical removal of the endothelium (confirmed by loss of relaxation to acetylcholine), HVC was unaffected, but HEV was reduced about 20%. In contrast, pretreating the rings with the NO scavenger Hb (100 uM), the NOS inhibitor L-NAME (10 uM), or the soluble guanylate cyclase inhibitor ODQ (10 uM), virtually abolished HVC, while augmenting the HEV response. Deoxygenating RBCs prior to use reduced HEV ~ 50% versus oxyRBCs, while treating the RBCs with carbon monoxide (CO-RBC) abolished HEV. Adding a molar equivalent of purified oxyHb (10 uM), instead of RBCs, at the peak of HVC, resulted in a comparable relaxation response. HEV activity was markedly attenuated using old RBCs, but could be fully restored by pretreating the old RBCs with aqueous NO (1:250 NO:heme) or 1 mM N-acetylcysteine (NAC). NO and NAC treatment did not enhance the HEV of fresh RBCs. In contrast, treating old RBCs with nitrite (100 or 500 nM), or adding nitrite to the tissue bath (100 or 500 nM) had no effect on the HEV response. These results demonstrate that hypoxic erythrovasodilation is a multi-mechanistic biologic response involving the vascular endothelium, oxygen transfer to substrate-limited smooth muscle inducible NOS (iNOS), and possibly a direct vasorelaxant effect on the vascular smooth muscle. Further, we have shown that hypoxic vasoconstriction can occur in non-pulmonary vascular tissue, an effect likely mediated by increased iNOS expression.