Abstract
In recent years, G protein vs. β-arrestin biased agonism at opioid receptors has been proposed as an opportunity to produce antinociception with reduced adverse effects. However, at present this approach is highly debated, a reason why more information about biased ligands is required. While the practical relevance of bias in the case of µ-opioid receptors (MOP) still needs to be validated, it remains important to understand the basis of this bias of MOP (and other GPCRs). Recently, we reported two cyclopeptides with high affinity for MOP, the G protein biased Dmt-c[d-Lys-Phe-pCF(3)-Phe-Asp]NH(2) (F-81), and the β-arrestin 2 biased Dmt-c[d-Lys-Phe-Asp]NH(2) (C-33), as determined by calcium mobilization assay and bioluminescence resonance energy transfer-based assay. The biased character of F-81 and C-33 has been further analyzed in the [(35)S]GTPγS binding assay in human MOP-expressing cells, and the PathHunter enzyme complementation assay, used to measure β-arrestin 2 recruitment. To investigate the structural features of peptide-MOP complexes, we performed conformational analysis by NMR spectroscopy, molecular docking, and molecular dynamics simulation. These studies predicted that the two ligands form alternative complexes with MOP, engaging specific ligand-receptor contacts. This would induce different displays of the cytosolic side of the seven-helices bundle, in particular by stabilizing different angulations of helix 6, that could favor intracellular coupling to either G protein or β-arrestin.