Rationally designed non-peptides: variously substituted piperazine libraries for the discovery of bradykinin antagonists and other G-protein-coupled receptor ligands.

Molecular modeling studies of potent decapeptide bradykinin antagonists suggested the de novo design of peptide mimetics based on a 1,2,3,4-tetrasubstituted 1,4-piperazin-6-one scaffold. These de novo-designed antagonists exhibited only modest potency (IC50 approximately 55 microM) on a cloned human B2 receptor and antagonist activity in an in vitro human-cell functional assay. The success of these structures led to the creation of prototype libraries based on variously substituted 1,4-piperazine scaffolds, which allowed a rapid and general search of pharmacophores attached to a piperazine scaffold. The parent piperazinedione structures and fully reduced piperazine libraries differ from recently reported diketopiperazine libraries in the use of diverse nonnatural amino acids, on-resin-submonomer synthesis to provide more diverse N-substituted structures, and the adaptation of simultaneous ring closure and resin cleavage to drive the formation of highly hindered amide bonds. Using this chemistry, a rationally directed non-peptide library of approximately 2500 N,N'-disubstituted piperazines and piperazinediones was synthesized and screened for ligand affinity on bradykinin, neurokinin, and opioid receptors. A number of lead structures were identified. Notably, a bradykinin antagonist lead, CP-2458, with good receptor selectivity and antagonist activity in human-cell assays was identified and is undergoing optimization by traditional and combinatorial methods.