Lipopolysaccharide treated rats (25 mg/kg i.v.) were killed after 60 min and rat aortic rings were mounted in an isolated organ bath for measurement of isometric contractions in response to phenylephrine (0.01-10 microM) or potassium chloride (10 mM). Aortic rings from lipopolysaccharide-treated rats showed reduced contractility to phenylephrine and potassium chloride when compared to those from saline-treated rats. Indomethacin 10 microM, added in vitro further impaired phenylephrine-induced contraction of aortic rings from lipopolysaccharide ex vivo treated rats but was ineffective on aortic rings from saline treated rats. A similar pattern was observed when potassium chloride was used. Administration in vitro of thromboxane A2 receptor antagonist SQ29,548 gave a similar effect to indomethacin. Aortic rings collected from rat treated in vivo with dexamethasone (10 mg/kg) showed a reduction in phenylephrine induced contractions that was not further reduced by in vitro treatment with indomethacin (10 microM). Similarly, when rat aortic rings were incubated in vitro (60 min) with lipopolysaccharide (0.4 mg/ml) a reduction of phenylephrine- and potassium chloride-induced contraction was observed, but addition of either indomethacin or SQ29,548 did not further reduce contraction. Our results suggest that under these experimental conditions, in the early phase of endotoxin shock, synthesis of cyclooxygenase products (such as endoperoxides or thromboxane A2) occurs probably as a compensatory mechanism to lipopolysaccharide induced hypocontractility from the interaction, in vivo, between lipopolysaccharide, endothelium, circulating cells and vascular smooth muscles.