Involvement of endothelial thromboxane A2 in the vasoconstrictor response induced by 15-E2t-isoprostane in isolated human umbilical vein

The present study was undertaken to evaluate the contractile response of several E- and F-ring isoprostanes (IsoP) in human umbilical vein (HUV) and to investigate the role of the endothelium on the effect of 15-E2t-IsoP, the most potent vasoconstrictor isoprostane, in human vessels. HUV rings with or without endothelium were suspended in an organ bath for recording the isometric tension in response to different agonists. The inhibitors to be evaluated were applied 30 min before the addition of the agonist. All of the compounds tested produced concentration-dependent contractions when tested on HUV rings with endothelium. Although these compounds were equieffective, significant differences were observed in their potency, with U46619 being the most potent followed by 15-E2t-IsoP > 15-E1t-IsoP = 15-F2t-IsoP > 15-F1t-IsoP = 9-epi-15-F2t-IsoP in descending rank order of potency. 15-E2t-IsoP was the most potent of the isoprostanes evaluated and, therefore, the one employed in the present study. When intact endothelium HUV rings were used, 15-E2t-IsoP-induced contraction was unaffected by the endothelin-converting enzyme inhibitor, phosphoramidon (10 microM), suggesting that short-term endothelin-1 release is not involved in this response. However, the non-selective cyclooxygenase (COX) inhibitor, indomethacin (10 and 30 microM), and the COX-2 selective inhibitor, NS-398 (3, 10 and 30 microM) produced inhibitory effects on 15-E2t-IsoP-induced contraction of HUV rings with endothelium. These results indicate that COX-derived contractile prostanoids are involved in this effect. Furthermore, the apparent pKb values estimated for indomethacin (5.5) and NS-398 (5.4) suggest that the prostanoids involved are derived from the COX-2 isoenzyme pathway. On HUV rings with endothelium, the phospholipase A2 inhibitor, oleyloxyethyl phosphorylcholine (30 and 100 microM), induced an inhibitory effect on 15-E2t-IsoP-induced contraction, suggesting that the phospholipase A2 pathway is also involved in this effect. In addition, the thromboxane A2 synthase inhibitor furegrelate (10 and 30 microM) also inhibited 15-E2t-IsoP-induced contraction of HUV rings with endothelium, indicating that thromboxane A2 is one of the contractile prostanoids involved in this response. Endothelium denudation clearly diminished the vasoconstrictor potency of 15-E2t-IsoP, demonstrating that the endothelium releases a vasoconstrictor factor in response to 15-E2t-IsoP. The absence of an inhibitory effect at the highest concentration of furegrelate (30 microM) on 15-E2t-IsoP-induced contraction of HUV rings without endothelium suggested that endothelium is the source of thromboxane A2. We conclude that prostanoids derived from the COX-2 isoenzyme pathway participate in 15-E2t-IsoP-induced vasoconstriction of isolated HUV rings. Our results also indicate that endothelial thromboxane A2 is one of the prostanoids involved in this effect.