We investigated the effect of the potassium channel opener pinacidil on ATP-dependent K+ channels (KATP) in the relaxation of porcine and human coronary arteries by means of isometric contraction experiments in arterial rings. We also measured whole cell currents in freshly isolated porcine and human coronary artery vascular smooth muscle cells with patch clamp. We first characterized serotonin-induced precontractions in our vessels and proved that the contractions were mediated by Ca2+ influx through voltage-dependent Ca2+ channels. Similarly, we observed that serotonin-induced contractions were strongly enhanced by small K(+)-induced depolarizations. Pinacidil completely relaxed rings preconstricted with 5 microM serotonin and produced dose-dependent relaxations of 5 microM serotonin-preconstricted rings, with an IC50 of 1.26 microM. Similar results were observed (IC50 = 1.15 microM) when the endothelium was removed. The KATP blocker glibenclamide (3 microM), inhibited pinacidil-induced relaxations (5-10 microM) by approximately 25% although the KATP blocker tetrapentylammonium (10 microM), inhibited pinacidil-induced (5-10 microM) relaxations completely. Pinacidil 10 microM had only a minimal effect on rings precontracted with a 50 mM external K+ concentration (IC50 = 60 microM). Porcine and human arterial rings did not differ qualitatively in their responses. Moreover, in the patch clamp experiments pinacidil (1 microM and 20 microM) induced a large, nonrectifying, outward current in both human and porcine cells. The reversal potential was close to the K+ equilibrium potential, suggesting an induction of pinacidil-activated K+ current. The pinacidil-induced (1 microM) current was strongly inhibited by glibenclamide (3 microM). These data show that the relaxation of porcine and human coronary arteries by pinacidil is primarily induced by an opening of KATP in smooth muscle cells. Furthermore, the vasorelaxant effect of pinacidil is not endothelium dependent.