Here, we demonstrate that the metabolism of glyceryl trinitrate (GTN) to nitric oxide (NO) occurs not only in bovine aortic smooth muscle cells (SMCs) but also in endothelial cells (ECs) and that this biotransformation is enhanced by pretreatment with Escherichia coli lipopolysaccharide (LPS). Two bioassay systems were used: inhibition of platelet aggregation and measurement of cGMP after stimulation by NO of guanylate cyclase in SMCs or ECs. In addition, NO produced from GTN by cells was measured as nitrite (NO2-), one of its breakdown products. Indomethacin (10 microM)-treated SMCs or ECs enhanced the platelet inhibitory activity of GTN. This effect was abrogated by coincubation with oxyhemoglobin (oxyHb; 10 microM), indicating release of NO from GTN. LPS (0.5 microgram/ml; 18 h) enhanced at least 2- to 3-fold the capacity of SMCs or ECs to form NO from GTN, and this enhancement was attenuated when cycloheximide (10 micrograms/ml) was incubated together with LPS. Furthermore, when incubated with GTN (200 microM) SMCs or ECs treated with LPS (0.5 microgram/ml; 18 h) released more NO from GTN than nontreated cells as indicated by a much higher (8- to 9-fold) increase in the levels of cGMP. Exposure of SMCs to GTN (600 microM) for 30 min led to an increase in the levels of NO2- dependent on cell numbers, which was enhanced when SMCs were treated with LPS. Incubation of nontreated or LPS-treated cells with NG-monomethyl-L-arginine (300 microM; 60 min) did not influence the metabolism of GTN to NO. SMCs failed to enhance the antiplatelet activity of sodium nitroprusside. Anesthetized rats treated with an intraperitoneal injection of LPS (20 mg/kg) 18 h beforehand showed enhanced hypotensive responses to GTN (0.25-1 mg/kg). These effects were blocked by methylene blue (10 mg/kg) but not by indomethacin (3 mg/kg). LPS did not alter the hypotensive responses induced by phentolamine, verapamil, or SIN-1. Thus, both in vitro and in vivo, LPS induces the enzyme(s) metabolizing GTN to NO.