1. The pulmonary vasculature is constantly exposed to oxygen and reactive oxygen species such as nitric oxide (NO) and superoxide anions which can combine at a near diffusion limited rate, to form the powerful, oxidant, peroxynitrite (ONOO-). When formed in large amounts, ONOO- is thought to contribute to tissue injury and vascular dysfunction seen in diseases such as the acute respiratory distress syndrome (ARDS) and septic shock. Recent studies have shown that ONOO- can cause vasodilatation and at higher concentrations can activate poly (adenosine 5'-diphosphoribose) synthase (PARS) leading to consumption of nicotinamide adenine dinucleotide (NAD+) and adenosine 5'-triphosphate (ATP). As the lung represents a prime site for ONOO- formation, we characterized its effects on pulmonary vascular tone and on endothelial function. In addition, we have assessed the role of PARS in producing the vasoactive properties of ONOO- on pulmonary artery rings. 2. Isolated pulmonary artery rings from rats were mounted in organ baths containing warmed and gassed (95% O2: 5% CO2) Krebs buffer. Force was measured with isometric force transducers. After equilibration, ONOO- (10 nM-100 microM) was added in a cumulative manner. In separate experiments designed to assess any vasodilator properties of ONOO-, tissues were pre-contracted with the thromboxane mimetic U46619 (1 microM). Once a stable base-line was achieved, ONOO- was added in a cumulative fashion. ONOO- had no significant effect on resting pulmonary artery tone but caused concentration-dependent relaxations of pre-contracted vessels in the range 1 microM to 100 microM. In some experiments the effects of freshly prepared ONOO- solutions were compared with those allowed to decay at 4 degrees C for 2 days. 3. In some experiments either vehicle or ONOO- (1, 10 or 100 microM) was added for 15 min before U46619 (1 microM). Concentration-response curves to the endothelium-dependent vasodilator, acetylcholine (10 nM-100 microM) were then constructed. In these experiments, ONOO- (1 microM or 10 microM) had no effect on the actions of acetylcholine. However, at the highest concentration tested (100 microM), ONOO- increased acetylcholine-induced relaxations. 4. The vasodilator actions of ONOO- were unaffected by the NO synthase inhibitor, NG-nitro-L-arginine methyl ester (L-NAME; 100 microM) or by removal of superoxide anions with superoxide dismutase (SOD) (30 units ml-1). However, the relaxations induced by ONOO- were significantly inhibited by the PARS inhibitor, 3-aminobenzamide (10 microM). In contrast to its effects on ONOO-, 3-aminobenzamide had no effect on the relaxation caused by acetylcholine or sodium nitrite, but actually increased that induced by sodium nitroprusside. 5. These data show that ONOO- causes vasodilatation of rat pulmonary arteries, probably via activation of PARS. Moreover, at concentrations where relaxation was achieved, ONOO- did not affect the ability of pulmonary artery rings to relax to acetylcholine. We propose that ONOO-, but not endothelially derived NO, activates PARS resulting in the rapid depletion of ATP and a consequent reduction in contraction as well as other active processes of vascular smooth muscle. The finding that 3-aminobenzamide inhibited the actions of ONOO- but not acetylcholine, suggests that NO and ONOO- cause relaxation by independent mechanisms. It has been suggested that ONOO- is responsible for the vascular hyporesponsiveness to constrictor agents seen in experimental sepsis. This observation together with our current finding, that 3-aminobenzamide inhibits the relaxation induced by ONOO- but not by acetylcholine, suggests that inhibitors of PARS may reduce the persistent hypotension seen in sepsis without affecting the actions of endothelium-derived NO. Thus, the use of PARS inhibitors may represent a novel therapeutic approach to the treatment of septic shock.