Differential Efficacy of Anti-Sickling and Anti-Inflammatory Mechanisms in a Fluorescent Intravital Microscopy Dorsal Skinfold Vaso-Occlusion Model in Sickle Cell Disease Townes Mice

Sickle cell disease (SCD) is characterized by acute and repetitive vaso-occlusive crises (VOC). These crises have been hypothesized to occur when blood flow is reduced following obstruction of sickle-shaped red blood cells in the vasculature. However, it is now well established that inflammation, oxidative stress, endothelial activation and pro-coagulation in sickle cell disease patients also contribute to the formation of heterocellular aggregates that can lead to VOC (Vercellotti and Belcher, 2014).

Transgenic SCD mice recapitulate the pathology of human disease in response to stimuli such as heme injection and hypoxia/reoxygenation. SCD SS Townes mice, which express human α and sickle γ^Aβ^S globins, AA Townes mice expressing normal human α and normal γ^Aβ^A globins and heterozygous AS mice which express only one allele of the γ^Aβ^S sickle gene were used. To characterize vaso-occlusion in these mice and evaluate the efficacy of different pharmacological mechanisms, we modified the skinfold vaso-occlusion model (Kalambur et al, 2004) using fluorescent intravital microscopy to visualize blood flow occlusion following hemin injection or hypoxia/reoxygenation challenge. Dorsal skinfold chambers were implanted and24 hours post-surgery mice were injected with FITC-dextran for visualization of flowing blood vessels. Skinfold bearing mice were then subjected to hemin treatment (16 μmoles/kg) or hypoxia (7%; 1 hour)/reoxygenation (1 hour) followed by the injection of Alexa fluor 647-labeled albumin to allow quantification of occluded vessels through dual fluorescent image analysis.

Following hemin injection, SS mice showed significant ~30% vaso-occlusion in comparison to AA mice with ~8%, whereas the AS mice showed an intermediate phenotype with ~20% vaso-occlusion. Hypoxia/reoxygenation challenge also resulted in significant vaso-occlusion for SS mice (~25%) whereas only 5% was observed in AA mice. Interestingly, AS mice also showed a significant amount of vaso-occlusion (~25%) similar to SS mice when challenged with hypoxia/reoxygenation. Although no sickling can be observed in an ex vivo sickling assay using AS red blood cells, an intermediate amount of free Hemoglobin (Hb) can be detected in the plasma of these mice and rolling can be observed. This suggests that these vaso-occlusive models relate more on the inflammatory and endothelial activation state and are independent of the sickling potential of the red blood cell.

We then used our model with hypoxia/reoxygenation challenge to evaluate the effects of dimethyl fumarate (DMF, 15 mpk BID), an anti-P-Selectin antibody (150ug/mouse) and the covalent hemoglobin oxygen affinity modulator GBT-440 (300 mpk). As anti-inflammatory agents, DMF and Anti-P-Selectin significantly reduced vaso-occlusion in SS mice by ~60% compared to the vehicle treated mice, but GBT-440 did not inhibit vaso-occlusion at a dose where a significant reduction in p50 was observed.

In conclusion, our data have shown that obstruction of blood flow in the skinfold vaso-occlusion model in SCD Townes mice reflects the vascular inflammatory state of the disease and is independent of the ex vivo capacity of red blood cell to sickle.