Buprenorphine Restricts the Conformational Landscape of the μ-Opioid Receptor

Biased signaling and partial agonism in GPCRs are considered promising strategies to develop safer and more effective drugs. Various mechanisms have been proposed to explain these phenomena, including the ability of biased and partial ligands to promote specific receptor conformations able to engage distinct intracellular partners. Recent cryo-electron microscopy structures of the {micro}-opioid receptor (MOR) bound to both G-proteins and {beta}-arrestins have shown clear structural differences shedding light into the putative signaling states. However, the accessibility of such conformations in the absence of the intracellular partner and the mechanisms by which ligands modulate their populations remain poorly understood. In the present manuscript, we employ extensive enhanced-sampling simulations to explore the free-energy landscapes of MOR bound to a biased partial agonist (buprenorphine) and a full unbiased agonist (endomorphin-1) showing that the ligands indeed stabilize distinct ensembles of active-like states. In particular, we observe that endomorphin-1, the natural ligand, explores two different ensembles of active-like conformations, including states compatible with either G-protein or {beta}-arrestin1 recruitment, whereas buprenorphine restricts access to these conformations and favors a narrower subset of active states more prone to G-protein binding. Moreover, we analyze the allosteric pathways connecting the ligands to the different active-like states. Together, our findings provide a thermodynamic framework for understanding biased signaling at MOR and confirm that in MOR ligand efficacy and signaling preference emerge from the selective redistribution of receptor conformational populations even in the absence of the effector proteins. This perspective may guide the rational design of opioid therapeutics with improved signaling profiles and reduced adverse effects.