Structural insights into the opioid message–address concept from cryo-EM data and AlphaFold3 predictions

Opioid peptides activate the μ-, δ-, and κ-opioid receptors (MOR, DOR, and KOR), three closely related class A G-protein coupled receptors (GPCRs) whose subtle structural differences underlie their distinct pharmacological profiles and adverse-effect liabilities. The classical “message–address” model provides a conceptual framework for understanding peptide opioid structure–activity relationships, assigning receptor activation to an N-terminal “message” motif and selectivity to a C-terminal “address” region. Here, we reevaluate and refine this paradigm through a comprehensive structural analysis of 24 peptide–receptor complexes, combining available cryo-electron microscopy structures with high-confidence AlphaFold3 models. Predicted complexes were refined using AMBER energy minimization and prioritized based on binding free energy estimates. Benchmarking against experimentally resolved structures [Endomorphin-1–MOR, DAMGO–MOR, Deltorphin-II–DOR, and Dynorphin A(1–8)–KOR] demonstrates that the workflow accurately reproduces peptide binding modes and key receptor interactions. Comparative analysis across receptor subtypes reveals conserved activation determinants within the message region, alongside previously underappreciated interaction patterns in the address domain that contribute to subtype selectivity. These findings extend the classical message–address model by defining atomic-level interaction motifs and receptor-specific hotspots that govern ligand recognition. Collectively, this work establishes a generalizable structural framework for the rational design of receptor-selective opioid peptides and highlights the utility of integrative computational approaches for elucidating GPCR–ligand interactions.