Various pulmonary artery preparations in vitro demonstrate sustained endothelium-dependent contractions upon hypoxia. To determine whether endothelin-1 could mediate this phenomenon, we examined the effect of bosentan, a new antagonist of both the ETA and ETB subtypes of the endothelin receptor. Small (300 pm) pulmonary arteries from rats were mounted on a myograph, precontracted with prostaglandin F2 alpha and exposed to hypoxia (PO2, 10 to 15 mm Hg, measured on-line) for 45 min. Endothelium-intact control rings exhibited a biphasic response, with a transient initial vasoconstriction (phase 1) followed by a second slowly developing sustained contraction (phase 2). Expressed in percent of the maximal response to 80 mmol/L KCl, the amplitudes of phase 1 (peak tension) and 2 (tension after 45 min of hypoxia) averaged 37 +/- 12% and 17 +/- 14%, respectively (n = 11). In endothelium-denuded rings, phase 1 persisted while the amplitude of phase 2 was reduced to 2 +/- 12% (p < 0.05, n = 8), showing the endothelium dependence of this contraction. Neither phase was significantly decreased in rings treated with 10(-5) mmol/L bosentan (38 +/- 15% and 17 +/- 12%, respectively, n = 6). The PO2 threshold for onset of hypoxic contraction was not significantly different among these three groups and averaged 32 +/- 24 mm Hg. In a separate experiment, we assessed the inhibitory effect of 10(-5) mol/L bosentan on the response to 10(-8) mol/L endothelin-I. Rings treated for 45 min with 10(-8) mol/L endothelin-1 alone exhibited a maximal contraction of 75 +/- 27% (n = 6). This was reduced to 4 +/- 17% (p < 0.01, n = 6) in rings treated with both 10(-8) mol/L endothelin-1 and 10(-5) mol/L bosentan. We conclude that complete blockade of all endothelin receptor subtypes has no effect on either endothelium-dependent or -independent hypoxic contractions in this preparation. This suggests that endothelial factors other than endothelin-I mediate the acute hypoxic contractions of small pulmonary arteries in the rat.