The traditional evaluation of the endothelium-derived nitric oxide (EDNO) pathway involves isolated aortic rings with attached strain gauges. This model is nonphysiologic and does not permit studies lasting longer than several hours. Our objective was to overcome the limitations of these "traditional" methods utilizing a physiologic, whole vessel model as a reproducible assay of EDNO. Canine carotid arteries (n = 4) were removed (maintaining in vivo arterial geometry), mounted in a specially designed, continuous-flow circuit, and perfused at 100 ml/min, 80 mm Hg with Medium-199/10% canine serum. Physiologic pH, pCO2, pO2, and temperature were precisely regulated. A non-contacting, helium-neon laser micrometer was interfaced with the current system to provide continuous measurement of vessel external diameter and to quantitate changes in vessel wall geometry in response to epinephrine (EPI; 2 x 10(-5) to 2 x 10(-3) mg/ml) and acetylcholine (ACh; 0.1 to 100 microM) challenge. Further characterization of the perfusion system included the use of a competitive inhibitor to EDNO production, NG-monomethyl-L-arginine (L-NMMA), and the effect of this compound on ACh-induced vasodilation. The reversibility of this blockade was verified via the sequential addition of L-arginine (L-ARG; 0 to 3 mM). Data are expressed as the ratio of steady-state vessel cross-sectional area (CSA) following administration of vasoactive substance to the CSA prior to vasoactive challenge. Our results indicate that EPI and ACh produced significant dose-dependent vasoconstrictive and vasodilatory responses, respectively (P less than 0.001, ANOVA).(ABSTRACT TRUNCATED AT 250 WORDS)