Cocaine and Certain of Its Long-Acting Metabolites Stimulate Endothelial Cell Von Willebrand Factor Secretion

Abstract 1148

Cocaine ingestion is associated with increased risk of myocardial infarction, stroke, and deep vein thrombosis. Elevated risk for these thrombotic events persists for 7–10 days after cocaine ingestion, despite the fact that cocaine is rapidly cleared from the blood (with a half-life in circulation of less than 1 hour) being rapidly converted into numerous metabolites that circulate for 1–2 weeks. Thus, it is likely that these longer-lived cocaine metabolites account for the long-term thrombotic risk with cocaine use. Cocaine ingestion has been associated with increased plasma VWF concentration, early atherosclerosis, and platelet-rich arterial thrombi. VWF is the major adhesive ligand that attaches platelets to the vessel wall, either to the subendothelium at sites of injury where the endothelium has been denuded or, during inflammation, to intact endothelium, from which it is secreted. VWF is secreted by endothelial cells from intracellular storage granules (Weibel-Palade bodies) in a large, hyperadhesive multimeric form (ultra-large VWF, or ULVWF) that either remains tethered to the endothelial surface or is released into bulk flow. Platelet-VWF adhesion, and subsequent thrombus formation, may be augmented further by cocaine-induced vaso-constriction, increasing shear stress. Thus, we hypothesized that cocaine and/or its metabolites would stimulate endothelial VWF secretion as a mechanism of thrombotic risk. To test this possibility, we exposed cultured human endothelial cells from umbilical vein (HUVEC), brain microvasculature (BMVEC), or coronary artery (CAEC) to cocaine or one of its four major metabolites at concentration ranges reported to occur in plasma following cocaine use. The cocaine metabolites we tested were benzoylecgonine (BE), cocaethylene (CE), norcocaine (NC), and ecgonine methyl ester (EME). We assayed VWF release by platelet-VWF string formation in a parallel-plate flow chamber (2.5 dyne/cm²) and by measuring the concentration of VWF released into the supernatant. Cocaine concentrations as low as 0.1 μg/ml induced VWF release from HUVEC; 1–2 μg/ml cocaine was as effective in releasing VWF as 25 μM histamine or 4 mg/ml dDAVP. Of the cocaine metabolites, only BE and CE induced VWF release from endothelial cells. BMVEC were 10-fold more sensitive to cocaine and metabolites BE and CE than HUVEC in both platelet-string formation and VWF antigen assays. In CAEC, VWF release was slightly reduced compared to HUVEC in response to cocaine, BE, or CE. Consistent with this pattern, staining for intracellular VWF revealed the following hierarchy of intracellular VWF: BMVEC > HUVEC >> CAEC. In BMVEC and HUVEC, VWF staining was restricted to Weibel-Palade bodies, with CAEC also demonstrating a diffuse cytoplasmic pattern. We tested whether intracellular cAMP levels increased after cocaine or cocaine metabolite exposure to explore whether VWF release was dependent on Protein Kinase A, as is the case with dDAVP. Intracellular cAMP did not increase following exposure to cocaine or its metabolites in any of the endothelial cell lines. These results suggest that cocaine contributes to thrombosis by activating endothelial cells to secrete ULVWF via a mechanism that does not involve increased intracellular cAMP. The prolonged thrombotic risk after cocaine ingestion likely relates to the continued action of cocaine metabolites BE and/or CE on endothelial ULVWF secretion. VWF secretion is likely to vary between vascular beds, with brain endothelial cells being particularly sensitive. Furthermore, these results suggest that clinical management of cocaine-induced ischemia or vaso-occlusion may benefit from therapies aimed at disrupting the ULVWF–platelet interaction.