1 In this study the mechanisms of the acute vasodilator action of bacterial lipopolysaccharide (LPS) were investigated in the rat Langendor perfused heart. 2 Infusion of LPS (5 mg ml
71) caused a rapid and sustained fall in coronary perfusion pressure (PP) of 59+4 mmHg (n=12) and a biphasic increase in NO levels determined in the coronary euent by chemiluminescent detection. Both the fall in PP and the increase in NO release were completely abolished (n=3) by pretreatment of hearts with the NO synthase inhibitor L-NAME (50 mM). 3 LPS-induced vasodilatation was markedly attenuated to 5+4 mmHg (n=3) by pretreatment of hearts with the B 2 kinin receptor antagonist Hoe-140 (100 nM). 4 Vasodilator responses to LPS were also blocked by brief pretreatment with mepacrine (0.5 mM, n=3) or nordihydroguaiaretic acid (0.1 mM, n=4) and markedly attenuated by WEB 2086 (3 mM, n=4). 5 Thirty minutes pretreatment of hearts with dexamethasone (1 nM), but not progesterone (1 mM), signi®cantly modi®ed responses to LPS. The action of dexamethasone was time-dependent, having no eect when applied either simultaneously with or pre-perfused for 5 min before the administration of LPS but inhibiting the response to LPS by 91+1% (n=4) when pre-perfused for 15 min. The inhibition caused by dexamethasone was blocked by 15 min pretreatment with the glucocorticoid receptor antagonist RU-486 (100 nM) or by 2 min pre-perfusion of a 1 : 200 dilution of LCPS1, a selective antilipocortin 1 (LC1) neutralizing antibody. 6 Treatment with the protein synthesis inhibitor, cycloheximide (10 mM, for 15 min) selectively blunted LPS-induced vasodilatation, reducing the latter to 3+5 mmHg (n=3), while having no eect on vasodilator responses to either bradykinin or sodium nitroprusside. 7 These results indicate that LPS-induced vasodilatation in the rat heart is dependent on activation of kinin B 2 receptors and synthesis of NO. In addition, phospholipase A 2 (PLA 2 ) is activated by LPS resulting in the release of platelet-activating factor (PAF) and lipoxygenase but not cyclo-oxygenase products. These eects are dependent on de novo synthesis of an intermediate protein which remains to be identi®ed.