Restoring Endogenous Nitric Oxide in Sickle or Transfused Red Cells Ameliorates Adhesion and Vaso-Occlusion in Vivo

Abstract 3251

Sickle erythrocytes (SSRBCs) are adherent to the endothelium, activate leukocyte adhesion, are deficient in nitric oxide (NO) content, and are inefficient in inducing vasodilation in response to hypoxia, all of which promote vascular occlusion, the hallmark of sickle cell disease (SCD). Furthermore, normal RBCs stored for transfusion are frequently used in SCD but are also NO-deficient. We therefore tested the hypothesis that repletion of intracellular SS or normal RBC NO could improve the abnormal circulatory characteristics of SSRBCs, using previously described in vitro (Zennadi et al. 2004 & 2008) and in vivo systems (Zennadi et al. 2007, Zhu et al. 2011). While exposure of SSRBCs to epinephrine (epi) up-regulated SSRBC adhesion to cultured HUVECs by 2.7±0.2-fold over baseline (p=0.0019), as previously described, loading SSRBCs with NO under hypoxic conditions (favoring SNO-Hb formation and bioactive NO availability), followed by reoxygenation prior to epi treatment, significantly decreased epi-induced SSRBC adhesion by 89±6.4% (p<0.01). In contrast, sham-loading of SSRBCs by deoxygenation/reoxygenation failed to inhibit epi-activated SSRBC adhesion, suggesting that increasing NO content rather than deoxygenation/reoxygenation reduced SSRBC adhesion. NO-loading of SSRBCs also significantly reduced the ability of epi-activated SSRBCs to activate mononuclear leukocyte adhesion to HUVECs to a level below adhesion induced by non-activated SSRBCs (11±5.8% vs 21±5.6% adherent leukocytes, p=0.02). The effect of NO-loading of SSRBCs on RBC adhesion was not due to NO release by SSRBCs, with subsequent effect of cell-free NO on HUVECs, since addition of free Hb to NO-loaded SSRBC suspensions failed to prevent NO from decreasing epi-activated SSRBC adhesion to HUVECs. We then studied the ability of NO loading of red fluorescent SSRBCs to affect in vivo adhesion and vaso-occlusion visible in implanted window chambers (Zennadi et al. 2007). While sham-treated SSRBCs adhered minimally to endothelium, epi-stimulated human SSRBCs showed strong adhesion, with frequent postcapillary obstruction, as previously shown. However, NO-loading of SSRBCs reduced epi-stimulated SSRBC adhesion to vessel walls and vaso-obstruction in vivo and resulted in improved RBC circulatory behavior. When SSRBC adhesion and vaso-occlusion were quantified by red fluorescence intensity (RFI) in vessels, using images obtained 30 minutes after RBC infusion, mice infused with epi-activated SSRBCs showed substantially greater RFI in vessels compared to animals infused with sham-treated SSRBCs (1343±395 vs 5±3.5 RFI, p<0.05). However, RFI in vessels of animals infused with NO-loaded, epi-activated SSRBCs was significantly decreased by 93.5% compared to RFI in vessels of mice infused with sham-loaded, epi-activated SSRBCs (p =0.00001). We also studied the effect of loading SSRBCs with the low-mass S-nitrosothiol (SNO) S-nitroso-L-cysteine (SNO-Cys) on the ability of SSRBCs to affect pulmonary artery pressures in intact mice. SNO-Cys attenuated the exaggerated pulmonary hypertension seen basally and during hypoxia in SSRBC-perfused lungs (p<0.05), suggesting that SSRBC NO/SNO deficiency may also contribute to SCD pulmonary vascular hypertension. Finally, we studied the effect of stored normal RBCs, with and without NO repletion, on the circulatory behavior of SSRBCs by admixing fresh or banked normal RBCs stored for 14 or 30 days with epi-activated SSRBCs (1:4 ratio of normal:SS) prior to infusion into mice. Fresh or 14 day old normal RBCs reduced epi-activated SSRBC adhesion to vascular endothelium (90% and 99% inhibition, respectively), preventing visible vascular blockade. In contrast, normal RBCs stored for 30 days failed to decrease adherence of epi-activated SSRBC to endothelium or vascular occlusion. However, when normal RBCs that had been stored for 30 days were NO-loaded prior to admixture with activated SSRBCs, we observed significant abrogation of adhesion of epi-stimulated SSRBCs to postcapillary venules (97% inhibition) and markedly reduced small diameter vessel obstruction (p < 0.001). Taken together, our results suggest that abnormal SSRBC NO content and NO/SNO-dependent vasoactivity contribute to the vaso-occlusive pathophysiology of SCD and are potentially amenable to correction by transfusion of NO-repleted normal RBCs.