MIDAZOLAM PRODUCES VASODILATION BY MIXED ENDOTHELIUM-DEPENDENT AND ENDOTHELIUM-INDEPENDENT MECHANISMS

Aortic rings were obtained from rat thoracic aorta and studied in vitro with and without functionally intact endothelium to determine whether ''the mechanism'' requires endothelium [or endothelium-derived relaxing factor (EDRF)]. In aortic rings precontracted with either phenylephrine (PE, 3 x 10(-8)-3 x 10(-7) mol/L) or KCl (40 mmol/L), midazolam produced concentration-dependent relaxation, with and without endothelium. Rings without endothelium demonstrated significantly less relaxation than those with endothelium regardless of whether they were precontracted with PE or KCl. With intact endothelium, midazolam produced greater relaxation in PE-contracted aortic rings than in KCl-contracted aortic rings; the midazolam concentration producing 50% relaxation from the contracted state (RC50) was 8.8 +/-3.6 x 10(-7) mol/L for PE-contracted rings and 3.3 +/-1.1 X 10(-6) mol/L for KCl-contracted rings (P < 0.05). In aortic rings with intact endothelium pretreated with N(G)-monomethyl-L-arginine (L-NMMA, 10(-4) mol/L), an inhibitor of nitric oxide (NO) synthesis, midazolam produced relaxation of similar magnitude to that seen in the denuded aortic rings except at the highest concentration (1 x 10(-5) mol/L). Pretreatment with the cyclooxygenase inhibitor, indomethacin (2.5 x 10(-5) mol/L), did not change the midazolam-induced relaxation in rings with intact endothelium as compared to untreated control aortic rings. In contrast to the intact endothelium state, when endothelium was removed, midazolam produced greater relaxation in the KCl-contracted aortic rings than in PE-contracted rings (RC50, 1.2 +/- 0.3 x 10(-5) mol/L vs 2.3 +/- 0.4 x 10(-5) mol/L, P < 0.05). Midazolam inhibited contractions produced by incremental addition of Ca2+ to aortic rings exposed to Ca2+-free, K+-depolarizing solution. In addition, midazolam inhibited contractions elicited by 1,4-dihydro-2,5-dimethyl-5-4-[2-(trifluoro-methyl)-phenyl]-3-pyridine carboxylic acid methyl ester (BAY K 8644), an activator of voltage-gated Ca2+ channels. These effects were similar to those produced by the Ca2+ channel blocker verapamil. Midazolam had minimal effect on intracellular Ca2+ release elicited by PE (10(-5) mol/L), indicated by the observation that initial phasic contractions elicited by PE in a Ca2+-free media were similar in the presence and absence of midazolam. We conclude that midazolam produces vasodilation by endothelium-dependent and -independent mechanisms. Endothelium-dependent vasodilation produced by midazolam probably is mediated through the release of NO from endothelium. Endothelium-independent vasodilation seems to be linked to inhibition of voltage-gated Ca2+ channels.