This study investigated the mechanism of ergonovine-induced hyperconstriction of coronary artery in conscious dogs that had undergone endothelial denudation one month earlier. The diameter of the large epicardial coronary artery was continuously measured by a sonomicrometer in 12 dogs in which two pairs of 10-MHz piezoelectric crystals had been surgically implanted at the denuded and nondenuded sites of coronary arteries. A month after the endothelial denudation, intravenous ergonovine (0.01, 0.1, 0.3, and 1.0 mg) produced transient dilation followed by dose-dependent constriction. The degrees of dilation were comparable between the denuded and nondenuded sites. The magnitudes of constriction induced by ergonovine were significantly larger in the denuded site than in the nondenuded site: the percent reductions in diameter evoked with 0.3 mg ergonovine were 14.4 +/- 2.3% and 3.8 +/- 0.8% (p less than 0.01) at the denuded and nondenuded sites, respectively. The magnitudes of constriction induced by intravenous phenylephrine (0.02, 0.06, and 0.2 mg) were comparable in the denuded and nondenuded sites. Methysergide (a nonselective serotonergic blocker) in a dose of 0.5 mg/kg significantly inhibited vasoconstriction induced by ergonovine (0.3 mg) from 13.1 +/- 1.1% to 2.7 +/- 1.0% (p less than 0.01) at the denuded site and from 4.2 +/- 0.6% to 0.8 +/- 0.3% (p less than 0.05) at the nondenuded site. Diltiazem (1.0 mg/kg) selectively inhibited the ergonovine-induced hyperconstriction. Ketanserin (0.5 mg/kg), prazosin (1.0 mg/kg), or indomethacin (5.0 mg/kg) did not prevent the ergonovine-induced hyperconstriction. Histological study revealed intimal thickening and regenerated endothelium in the denuded site.(ABSTRACT TRUNCATED AT 250 WORDS)