The effect of high concentrations of melatonin, and related indole-based and naphthalene-based derivatives, has been examined in the porcine coronary artery, pulmonary artery and the marginal artery of the colon. In addition, we have pharmacologically examined the role of cyclic GMP in the relaxatory action of these agents. Cumulative addition of melatonin (3-300 microM) caused a slowly developing relaxation in all three vascular preparations pre-contracted with 9,11-dideoxy-9a,11a-methanoepoxy prostaglandin F2alpha (U46619), a thromboxane mimetic agent. The estimated pIC50 values were 4.10-3.70 (coronary artery), 3.89 (pulmonary artery) and 3.96 (marginal artery). All melatonin analogues examined also produced concentration-dependent inhibition of U46619-induced contractions of the coronary and marginal arteries in a qualitatively similar manner to melatonin. The rank order of potency (based on the pIC50 values) of these compounds in both vascular tissues was N-[2-(3-ethyl-7-methoxynaphthyl) ethyl]-acetamide (S21634) >2-iodomelatonin = N-[2-(7-methoxynaphth-1-yl)-ethyl]-acetamide (S20098) = N-[2-naphth-1-yl-ethyl]-cyclobutyl carboxamide (S20928) >melatonin >N-acetyl-5-HT. Finally, the pharmacological characteristics of melatonin and S21634 as phosphodiesterase inhibitors were compared to those of zaprinast, a known cyclic GMP-specific phosphodiesterase inhibitor. Zaprinast also caused concentration-dependent inhibition of U46619-induced tone. All three compounds, zaprinast (10 microM), melatonin (300 microM) and S21634 (30 microM), significantly enhanced sodium nitroprusside-induced relaxations. The inhibitory action of zaprinast per se was greater in the presence of the endothelium and significantly attenuated by 3 microM 1H-[1,2,4]oxadiazolo[4,3-a]quinoxaline-1-one (ODQ), a selective inhibitor of guanylyl cyclase. In marked contrast, the vasorelaxant action of melatonin and S21634 was not affected by the removal of the endothelium or the addition of ODQ. In summary, we have shown that porcine arterial smooth muscle relaxes in response to high concentrations of melatonin and other related melatonin receptor ligands. However, it appears that the receptive site is pharmacologically different from the melatonin receptors currently known and does not involve inhibition of cyclic GMP-specific phosphodiesterase.