In the course of our studies on non-peptide bradykinin (BK) B(2) receptor ligands, it was suggested that the 4-substituent of the quinoline ring may play a critical role in determining binding affinities for human and guinea pig B(2) receptors, as well as agonist/antagonist properties. We carried out an extensive investigation to elucidate the structure-activity relationships (SAR) for this key pharmacophore. Introduction of lower alkoxy groups to the 4-position of the quinoline ring of 3 led to the identification of 4-ethoxy derivative 22b as a unique partial agonist. This compound significantly stimulated inositol phosphates (IPs) formation in Chinese hamster ovary cells expressing the cloned human B(2) receptor at concentrations greater than 10 nM and displayed one-tenth of the intrinsic activity of BK. The agonist activity of 22b was selective for the B(2) receptor and was inhibited by selective peptide and non-peptide B(2) antagonists. On the other hand, 22b strongly suppressed BK-induced IPs formation through the cloned human B(2) receptor. Further studies on the key pharmacophore led to identification of a 2-picolyloxy moiety as a powerful agonist switch, leading to the discovery of a potent and efficacious non-peptide B(2) agonist, 19a. Successive optimization of the acyl side chain afforded 38, which exhibited full agonist activity on stimulation of IPs formation. Furthermore, this strategy could be applied successfully to the benzimidazole series. The representative 1-(2-picolyl)benzimidazole derivative 47c increased PGE(2) production at a 1 microM concentration to the same level as the maximum effect of BK. Thus, we have established the medicinal chemistry modifications required to convert our highly potent non-peptide B(2) antagonists to agonists with potent efficacy.