1 We have investigated whether glibenclamide, an inhibitor of ATP-sensitive potassium channels, influences the induction of the calcium-independent isoform of nitric oxide synthase (iNOS) in cultured J774.2 macrophages activated by bacterial endotoxin (E.coli lipopolysaccharide; LPS), as well as in the lung and aorta of rats with endotoxic shock.2 Pretreatment of J774.2 macrophages with glibenclamide (10-' to 10-1 M for 30 min) dosedependently inhibited the accumulation of nitrite caused by LPS (1 pg ml-'). In contrast, pretreatment of macrophages with tetraethylammonium (10-4 to 10-2M for 30min), a non-selective inhibitor of potassium channels, did not affect the rise in nitrite caused by LPS. At the highest concentration (10-5 M) used, cromakalim, an opener of ATP-sensitive potassium channels, caused a small, but significant inhibition of nitrite formation in macrophages activated with LPS, while lower concentrations (10-7 to 3 x 10-6 M) were without effect.3 The inhibition by glibenclamide (3 JM) of the increase in nitrite induced by LPS in J774.2 macrophages was weaker when glibenclamide was given several hours after LPS, indicating that glibenclamide inhibits the induction, but not the activity, of iNOS. In contrast, the degree of inhibition of nitrite formation caused by the nitric oxide synthase (NOS) inhibitor N-nitro-L-arginine methyl ester (L-NAME) was similar when this agent was given up to 10 h after LPS. 4 In anaesthetized rats, LPS caused a fall in mean arterial blood pressure (MAP) from 120 ± 4 (time 0) to 98 ± 4 mmHg at 180 min (P<0.05, n = 6). Treatment of LPS-rats with glibenclamide (1 mg kg-', i.v.at 60 min after LPS) caused a rapid and sustained rise in MAP (e.g. MAP at 180 min after LPS: 122±4 mmHg; n =6, P <0.05 when compared to LPS-rats). The maximum of the rise in MAP produced by glibenclamide (1 mg kg-, i.v.) was similar when the drug was given either at 60 or 180 min after LPS. However, the duration of the pressor response was significantly longer when glibenclamide was given at 60 min, rather than at 180 min after LPS. 5 LPS-treatment caused a significant reduction of the pressor responses elicited by noradrenaline (NA, 1 jig kg-', i.v.) from 35 ± 2 to 19 ± 1 mmHg at 60 min and 20 ± 2 mmHg at 180 min (P<0.05).Treatment of LPS-rats with glibenclamide (1 mg kg-', i.v. at 60 min) caused a significant restoration of the pressor responses elicited by NA from 19 ± 1 mmHg at 60 min (prior to glibenclamide injection) to 29± 3 mmHg at 180 min (P<0.05). 6 Endotoxaemia for 180 min resulted in a significant increase in a calcium-independent NOS activity (which was taken to represent iNOS activity) in the lung from 0.17 ± 0.1 (control, n =4) to 6.21 ± 0.48 pmol mg-' min-' (n =6, P<0.05). Injection of glibenclamide (1 mg kg ', i.v.) at 60 min after LPS attenuated the increase in iNOS activity caused by endotoxaemia in the lung by 43 ± 7% (n = 6, P <0.05). In contrast, injection of glibenclamide at 180 min after LPS did not result in a significant inhibition of iNOS activity (n = 6, P <0.05). 7 Thor...