Inhibition of endothelium-dependent vasorelaxation by arginine analogues: a pharmacological analysis of agonist and tissue dependence.

1. Isolated rings of rabbit external jugular vein (RbJV) and rat thoracic aorta (RA) were used to study the effect of the NO synthase inhibitor L-NG-nitroarginine methyl ester (L-NAME) on muscarinic and 5-hydroxytryptamine (5-HT) receptor-stimulated, endothelium-dependent vascular relaxations. 2. In RbJV relaxations produced by the endothelial 5-HT receptor agonist alpha-methyl-5-HT were potently and non-surmountably inhibited by L-NAME (10 microM), whereas acetylcholine relaxations in this tissue were unaffected by this concentration of inhibitor. By contrast, acetylcholine relaxations in RA were virtually abolished by 10 microM L-NAME. In each case an equivalent concentration of D-NAME was without effect on agonist-induced relaxations. 3. The different effect of L-NAME on acetylcholine relaxations in RbJV and RA was not due to muscarinic receptor differences. Affinity estimates for acetylcholine (pKA = 6.12 +/- 0.09; 6.09 +/- 0.08 respectively) and for 4-diphenyl-acetoxy-N-methylpiperidine methobromide (4-DAMP, pKB = 9.01 +2- 0.012; 9.24 +/- 0.16 respectively) indicated that the receptors in both tissues belong to the same M3 class. Tissue differences resulting from the release of a cyclo-oxygenase product or a glibenclamide-sensitive K(+)-channel-linked hyperpolarizing factor were also ruled out by selective inhibition of these pathways. 4. When phenoxybenzamine was used to reduce the efficacy of acetylcholine in RbJV so that it behaved as a partial agonist in this tissue, L-NAME (10 microM) now produced non-surmountable inhibition of relaxation responses. In untreated tissues the same concentration of L-NAME also profoundly inhibited responses produced by butyrylcholine and pilocarpine, both of which behave as partial agonists at the M3 receptor in RbJV. 5. A simple model was developed which describes the theoretical behaviour of receptor-stimulated synthesis and release of NO. The model predicts that competitive inhibition of NO formation results in parallel displacements of the agonist response curve in the case of high efficacy agonist, but right-shift with concomitant depression of the curve maximum in the case of low efficacy agonists. Simulations based on the model showed reasonable agreement with the experimental data. 6. It is concluded that analogues of L-arginine demonstrate tissue- and agonist-dependence in terms of their ability to inhibit receptor-mediated events involving the liberation of NO. This behaviour can reflect differences in agonist efficacy in the receptor systems being studied, a possibility that should be ruled out before apparent resistance to inhibition is taken as evidence for the involvement of heterogeneous endothelium-derived relaxing factors (EDRFs).