Design and synthesis of cyclic peptide antagonists intended to block coactivator binding to steroid nuclear receptors

Estrogen Receptor (ER), a member of the nuclear receptor superfamily, is involved in various diseases, including breast cancer. A widely used estrogen antagonist, tamoxifen binds to ER in the steroid binding pocket in place of estrogen. It induces a conformational change within the receptor, such that coactivators can no longer bind to ER, thereby potentially blocking transcription.; In contrast, normal ligand binding causes a displacement of ER helix 12, exposing a region on ER that can bind a consensus LXXLL pentapeptide sequence. This motif, known as the NR box, is found on approximately 20 coactivators isolated to date. Following the discovery of the NR box, genetic experiments, phage display libraries, and X-ray studies have revealed further insights into the mechanism of LXXLL binding and suggest a new approach for drug design by targeting the coactivator/ER interaction directly.; To mimic the coactivator structure, the NR box peptides have to be cc-helical, and short peptides generally are not. To increase their α-helicity, we considered using either disulfide constraints between the side chains of two cysteines; at the i and i+3 positions, or lactam bridges between i and i+4 residues. Next, molecular dynamics was performed using MacroModel version 7.0 on an SGI workstation in order to evaluate the ability of the peptides to mimic the coactivator structure. The calculations were carried out at 300°K, with the AMBER* Force Field and the GB/SA solvation model. Each structure was superimposed with a truncated version of the coactivator. The best fit came from one of the disulfide bridged peptides, H-Lys-cyclo(D-Cys-Ile-Leu-Cys)-Arg-Leu-Leu-Gln-NH ₂.; In a time resolved fluorescence competition assay vs. a linear 14-mer, this cyclic peptide gave a Ki value of 390 nM with ERβ and 25 nM with ERα.