Constriction of carotid arteries by urokinase-type plasminogen activator requires catalytic activity and is independent of NH(2)-terminal domains.

Urokinase-type plasminogen activator (uPA) is expressed at increased levels in stenotic, atherosclerotic human arteries. However, the biological roles of uPA in the artery wall are poorly understood. Previous studies associate uPA with both acute vasoconstriction and chronic vascular remodeling and attribute uPA-mediated vasoconstriction to the kringle - not the catalytic - domain of uPA. We used an in-vivo uPA overexpression model to test the hypothesis that uPA-induced vasoconstriction is a reversible vasomotor process that can be prevented - and uPA fibrinolytic activity preserved - by: 1) removing the growth factor and kringle domains; or 2) anchoring uPA to the endothelial surface. To test this hypothesis we constructed adenoviral vectors that express: wild-type rabbit uPA (AduPA); a uPA mutant lacking the NH(2)-terminal growth-factor and kringle domains (AduPAdel); a mutant lacking catalytic activity (AduPAS-->A), and a cell-surface anchored mutant (AdTMuPA). uPA mutants were expressed and characterised in vitro and in carotid arteries in vivo. uPAS-->A had no plasminogen activator activity. Activity was similar for uPA and uPAdel, whereas AdTMuPA had only cell-associated activity. AduPAS-->A arteries were not constricted. AduPA, AduPAdel, and AdTM-uPA arteries were constricted (approximately 30% smaller lumens; p< or =0.008 vs. AdNull arteries). Papaverine reversed constriction of AduPA arteries. uPA-mediated arterial constriction is a vasomotor process that is mediated by uPA catalytic activity, not by the NH(2)-terminal domains. Anchoring uPA to the endothelial surface does not prevent vasoconstriction. uPA catalytic activity, generated by artery wall cells, may contribute to lumen loss in human arteries. Elimination of uPA vasoconstrictor activity requires concomitant loss of fibrinolytic activity.

Improved vascular gene transfer with a helper-dependent adenoviral vector.

BACKGROUND: Adenoviral vectors are the most widely used agents for vascular gene transfer. However, the utility of adenoviral vectors for vascular gene transfer is limited by brevity of expression and by the induction of a significant host inflammatory response. Third-generation or "helper-dependent" adenoviral vectors have achieved prolonged recombinant gene expression in liver and muscle with minimal associated inflammation; however, they have never been tested for vascular gene transfer. METHODS AND RESULTS: We constructed a helper-dependent adenoviral vector expressing rabbit urokinase plasminogen activator (HD-AduPA). HD-AduPA was compared, in a rabbit model of carotid gene transfer, with a first-generation adenovirus, also expressing rabbit uPA (FG-AduPA). uPA expression and vector DNA were measured in arteries harvested from 3 to 56 days after gene transfer. Vector-specific mRNA, vascular inflammation, and neointimal formation were assessed 14 days after gene transfer. uPA expression was lost, and vector DNA declined rapidly in arteries infused with FG-AduPA. In contrast, uPA expression and vector DNA persisted in HD-AduPA arteries for > or =56 days, with stable expression from 14 to 56 days. Increased uPA expression in HD-AduPA arteries was accompanied by high levels of vector-specific uPA mRNA. Moreover, HD-AduPA arteries had significantly less inflammation and neointimal formation than FG-AduPA arteries. CONCLUSIONS: Helper-dependent adenoviral vectors can stably express a therapeutic gene in the vascular wall for > or =8 weeks, with minimal associated inflammation. Helper-dependent adenoviral vectors will be useful agents for vascular gene transfer and gene therapy.