Role of Protein Disulfide Isomerase In Thrombus Formation In a Collagen-Induced Pathway of Thrombus Formation

Abstract 345

We have demonstrated that platelet activation in vivo can take place via two pathways, one initiated by the generation of thrombin and the other initiated by the exposure of collagen on the injured vessel wall (Furie & Furie, 2008). In our laser-induced model of thrombosis, in which the endothelium is activated but intact, platelet activation by thrombin dominates and collagen is not required. In the widely used ferric chloride model of thrombosis the endothelium is denuded exposing collagen which leads to initial platelet activation. Using our laser-induced thrombosis model we previously demonstrated that protein disulfide isomerase is expressed on the vessel wall and within the platelet thrombus at the site of injury. Both bacitracin A, a non-specific inhibitor of thiol isomerases, and an inhibitory antibody specific for protein disulfide isomerase (RL90) block platelet thrombus formation and fibrin generation. Here we extend our study of the role of protein disulfide isomerase in thrombus formation to the ferric chloride model of thrombosis. We used intravital fluorescence microscopy in mouse arterioles exposed to filter paper saturated with 10% ferric chloride for 3 minutes. Protein disulfide isomerase, detected with a non-inhibitory polyclonal anti-protein disulfide isomerase antibody, accumulated in ferric chloride-induced platelet thrombi in cremaster arterioles. Bacitracin A (5 mg/mouse) delayed initiation of thrombus formation in mesenteric arterioles. Median time to initial platelet accumulation increased from 1 min in the absence of inhibitor to 4 min in the presence of inhibitor. In 4 out of 8 mice treated with 7 mg of bacitracin A platelet accumulation was completely inhibited. Similarly, bacitracin A prolonged the time to occlusion of ferric chloride-injured arterioles. Less than 50% of injured arterioles in mice treated with 5 mg of bacitracin A and only 25% of injured arterioles in mice treated with 7 mg of bacitracin A occluded after 30 minutes compared to a 100% of arterioles occluded in control saline treated mice. Pretreatment of mice with RL90 at 0.3, 1 or 3 μg/g mouse delayed the appearance of the first aggregates of platelets. Median time to initial platelet accumulation was prolonged from 0.6 min in the presence of isotype-matched control antibody (1 ug/g mouse) to 1.5 min in the presence of 0.3 mg/g mouse of RL90. Platelet accumulation was not observed in 1 out of 7 animals treated with RL90 at 1 mg/g mouse and in 3 out of 7 animals treated with RL90 at 3 mg/g mouse. RL90 also inhibited fibrin deposition after ferric chloride injury. Minimal fibrin was detected in the presence of RL90 at 1 μg/g mouse while fibrin appeared rapidly in mice treated with a control antibody. These data indicate that PDI is a component of a general regulatory pathway for initiation of thrombus formation.