Alteration of the Structure and Dynamics of Venous Clot Formation in Human and Murine Sickle Cell Disease

Sickle cell disease (SCD) is associated with chronic activation of coagulation and an increased risk of venous thromboembolism. Traditionally, it is believed that during venous thrombosis, red blood cells (RBC) are simply trapped within fibrin-rich thrombi and do not actively affect thrombosis. However, a study from our group showed that factor XIII (FXIII) activity is critical for the retention of RBC within clots and directly affects thrombus size. Others reported that during clot contraction polyhedral shaped RBC formed a densely packed core and that SCD alters the formation of polyhedrocytes which may affect clot stability (Strauss et al, 2015, ASH abstract). We further investigated if SCD affects the structure and the dynamics of clot formation. Ex vivo clot retraction was performed using blood from sickle cell patients and Townes mice (a model of SCD). Citrated blood was added to siliconized wells of 96-well plates containing CaCl2 (10mM) and tissue factor (1pM) and incubated for 2 hours at 37°C. The number of RBC extruded from the clot was counted in serum by Hemavet™ and expressed as a percentage of initial RBC number in the anticoagulated blood. Morphology of the clots was evaluated using scanning and transmission electron microscopy (EM). Thrombosis in Townes SS (sickle) and AA (control) mice was studied using femoral vein thrombosis induced by electrolytic injury and inferior vena cava stenosis models. The number of mouse SS RBC in the serum extruded during clot contraction was dramatically reduced compared to the number of AA RBC (0.8±0.8% vs. 19.4±0.8%, n=3, p<0.0001, Fig. A). A similar result was observed for human RBC. Since SS mice and sickle patients have a lower hematocrit (HCT), we investigated if the number of RBC affects the extrusion of these cells during clot retraction. Indeed, lowering HCT in AA mouse blood reduced RBC extrusion from the clots. However, increasing HCT in SS mouse blood to that of AA blood did not increase the number of SS RBC extruded from the clot. Furthermore, inhibition of FXIIIa activity with T101 (10µM) increased the release of AA (by 64%, n=6, p<0.05) but not SS RBC from mouse clots. These data indicate that the entrapment of SS RBCs within the clot is not simply caused by lower RBC number, and is FXIII-independent. Mixing the platelet poor plasma (PPP) and cellular fraction of AA and SS mouse blood revealed that the entrapment of SS RBC is not mediated by PPP (AA blood recombined = 6.9±3.6%; SS blood recombined = 0.4±0.4%; AA cells/SS PPP = 6.7±6.4%; SS cells/AA PPP = 0.1±0.2%, n=3 per group). Clots formed ex vivo from AA blood had a gel-like, soft structure, whereas SS clots were more firm and stiff. EM demonstrated that RBC within AA clots had polyhedral shapes and were tightly packed in the central part of the clot. In contrast, most of the mouse SS RBC did not have polyhedral shapes, underwent sickling and were not compacted within the clot. They also formed long "spicule-like" processes that intertwined with fibrin fibers (Fig. C). Similar results were observed in blood of sickle patients; however, the sickled RBC phenotype was less prominent. Importantly, sickling of RBC was observed in clots formed in the inferior vena cava of SS mice, 2 hours after vessel stenosis (Fig. D). We also performed a tPA challenge assay on clots formed ex vivo from human blood and showed that SS clots challenged with low tPA concentration (0.6nM) were more resistant to fibrinolysis compared to AA clots (clot lysis time, 714±6 vs. 388.3±120.7 minutes, n=6, p=0.08). The electrolytic injury model of venous thrombosis was used to investigate the dynamics of clot formation in SS mice in vivo. Mice were infused with fluorescently labeled antibodies for fibrin (green) and platelets (red). Electrolytic injury was applied to the femoral vein; a relative intensity of fibrin and platelet accumulation was assessed by fluorescence microscopy for one hour at 10-minute intervals. SS mice have increased platelet and fibrin accumulation compared to AA mice (~ 2 fold, n=5-7, p<0.05 for 40, 50 and 60 minute time points). Interestingly, in the AA clots, platelets were mostly localized on the surface, in contrast to their widespread distribution throughout the clot in SS mice (Fig. E, yellow color). Our data demonstrated that SCD alters the structure and dynamics of venous clot formation. Experiments investigating the consequence of these observations in mouse models of stroke and pulmonary embolism are currently ongoing. Figure Figure. Disclosures No relevant conflicts of interest to declare.

Thrombostatin Inhibits Induced Canine Coronary Thrombosis

SummaryThrombostatin (RPPGF), an angiotensin converting enzyme metabolite of bradykinin, is an inhibitor of α-thrombin’s ability to activate platelets. We examined the in vivo pharmacokinetics and pharmacodynamics of thrombostatin in rabbits and its ability to inhibit coronary thrombosis induced by electrolytic injury in dogs. Plasma half-life of thrombostatin had a t1/2α of 2.6 min and a t1/2β of 24 min in rabbits. Ligating the renal arteries did not prolong clearance (t1/2α = 2.4 min; t1/2β = 12 min). Thrombostatin produced a prolonged in vivo antiplatelet effect. At 30 min after a single intravenous administration in rabbits, thrombostatin’s plasma concentration was <8.7 μM (5 μg/ml). However, ex vivo 20 and 40 nM γ-thrombin-induced platelet aggregation of these rabbits’ platelets was inhibited 40% for 2.75 and 1 h, respectively. In vitro, flow cytometry studies revealed that thrombostatin specifically bound to human platelets and washed human platelets treated with thrombostatin were less responsive to γ-thrombin than control platelets. Using electrolytic injury to induce coronary artery thrombosis, dogs treated with thrombostatin, aspirin, or combined thrombostatin and aspirin occluded in 62 ± 25 (mean ± SD), 62 ± 36, or 89 ± 32 min versus untreated animals which occluded at 39 ± 27 min, (p <0.01, p <0.01 and p <0.001, respectively). These studies show that thrombostatin binds to platelets and can delay coronary occlusion in vivo. Abbreviations: RPPGF: thrombostatin; PAR1: protease activated receptor 1, the first cloned thrombin receptor; PRP: platelet-rich plasma; PPP: plateletpoor plasma; LCX: left circumflex coronary artery; APTT: activated partial thromboplastin time; PT: prothrombin time

Primary Prevention of Coronary Arterial Thrombosis with the Factor Xa Inhibitor rTAP in a Canine Electrolytic Injury Model

SummaryThe antithrombotic efficacies of the coagulation factor Xa inhibitor recombinant tick anticoagulant peptide (rTAP) and heparin were compared in a canine model of left circumflex (LCX) coronary artery electrolytic lesion. Intravenous infusions of saline (controls), rTAP (50 μg/kg/min continuous infusion) or heparin (200 U/kg bolus followed by 2 U/kg/min continuous infusion) were started 60 min prior to the initiation of LCX coronary artery electrolytic injury (150 μA continuous anodal current). All 6/6 saline-treated control animals developed occlusive thrombi at 49.8 ± 13.6 min after the initiation of vessel injury, and possessed a residual thrombus mass of 20.7 ± 3.3 mg. In the rTAP treatment group, 4/6 preparations developed occlusive thrombi, but with times to thrombosis delayed significantly compared to both the saline control as well as to the heparin treatment group (202.7 ± 28.9 min; p <0.01 to both saline and heparin groups). The remaining 2 rTAP-treated preparations remained patent despite the continued electrical stimulation of the coronary vessel for 5 h. Residual thrombus mass in the rTAP treatment group was reduced markedly compared to the saline control group (4.4 ± 1.0 mg; p <0.01). Heparin infusion resulted in a modest but statistically insignificant delay in occlusive LCX coronary artery thrombosis compared to saline controls, with all 6/6 heparin-treated preparations occluding at 79.7 ± 16.5 min after the initiation of vessel injury. Residual thrombus mass in heparin-treated animals, however, was reduced compared to saline controls (9.4 ± 1.4 mg; p <0.01). These results support a pivotal role for fXa in the process of arterial thrombosis, as well as the feasibility of inhibiting fXa as an effective strategy for the primary prevention of arterial thrombosis.

Effects of Antithrombotic Drugs in a Rat Model of Aspirin-Insensitive Arterial Thrombosis

SummaryThese studies describe experimental conditions where aspirin is less effective than other antiplatelet and anticoagulant drugs in inhibiting acute arterial thrombosis. External electrolytic injury of the rat carotid artery was used to induce occlusive thrombi in 97% of vehicle-treated rats. Thrombi were revealed by light and electron microscopy to be comprised primarily of platelets enmeshed in a fibrin network. The thrombin inhibitor D-phenylalanyl-L-prolyl-L-arginyl chloromethy ketone (PPACK; 6 mg/kg, i. v.) decreased thrombus weight by 90%. Aspirin alone (1, 10 and 30 mg/kg, i. v.), dipyridamole alone (5 mg/kg i. v.) and aspirin (1 and 10 mg/kg, i. v.) in combination with dipyridamole (5 mg/kg, i. v.) did not inhibit thrombosis. The platelet-activating factor (PAF) antagonist, WEB 2086, (1 mg/kg i. v.) was also ineffective. Other drugs had intermediate activity. Thrombi were decreased 56% by the thromboxane receptor antagonist, BMS 180,291, either alone (5.8 mg/kg i.v.) or in combination with aspirin (10 mg/kg, i.v.). Heparin (900 U/kg, i.v.), warfarin (0.25 mg/kg, p.o. once daily for 3 days) and ticlopidine (200 mg/ kg, p.o. once daily for 3 days) reduced thrombus weight by 63, 73 and 43% respectively. Reductions in thrombus weight were always associated with improvements in either average blood flow or vessel patency.