Three-Dimensional Structural Insights Have Revealed the Distinct Binding Interactions of Agonists, Partial Agonists, and Antagonists with the µ Opioid Receptor

The paper addresses the urgent public health context of the U.S. opioid crisis, noting a ~14-fold increase in opioid-related deaths from 6,984 (1999) to 94,371 (2020). It emphasizes that while opioids are indispensable for pain management, they carry high risks of addiction and adverse effects. The µ-opioid receptor (MOR), a class A GPCR with seven transmembrane helices, mediates the primary analgesic effects of opioids via G protein–dependent signaling and other pathways (β-arrestin–dependent and Gαi–β-arrestin complex signaling). The purpose of the review is to synthesize recent structural insights from X-ray crystallography and cryo-EM MOR structures to elucidate ligand–receptor interactions that underlie agonism, partial agonism, antagonism, and biased signaling, thereby informing the design of safer, more effective analgesics.

The review summarizes foundational biology of opioid receptors (MOR, DOR, KOR, and NOR/ORL-1), their endogenous ligands, distribution in pain-modulating pathways, and GPCR signaling cascades (Gi/Go coupling, AC inhibition, cAMP/PKA modulation, VGCC inhibition, and GIRK activation). It details functional selectivity/ligand bias at MOR involving G protein and β-arrestin pathways, including roles in desensitization, internalization, and MAPK/p38/ERK signaling. It compiles available MOR structures since 2012 (antagonist-bound: β-FNA, alvimopan; full agonist-bound: BU72, DAMGO, fentanyl, morphine, lofentanil, mitragynine pseudoindoxyl; partial agonist-bound: PZM21, FH210, oliceridine/TRV130, SR17018; plus a bitopic ligand structure). The review integrates MD simulation literature on ligand binding, activation mechanisms, internal water networks, sodium and magnesium ion allostery, cholesterol interactions, and small-molecule allosteric modulators (e.g., BMS-986121/122), highlighting probe dependence and predicted allosteric sites (e.g., near TM2/TM7 and extracellular grooves). Tables summarize endogenous ligands by receptor subtype (Table 1), MOR structures and conditions (Table 2), binding-pocket residue composition (Table 4), and binding-site SASA/SAV metrics (Table 5).

This is a comparative structural review and in silico analysis. The authors: (1) Collected atomic-resolution MOR structures from the PDB spanning antagonist-, partial agonist-, and full agonist-bound states (X-ray and cryo-EM). (2) Aligned structures to compare orthosteric binding poses and binding-site architectures (visualized with UCSF ChimeraX 1.4; binding pose figures generated with Maestro). (3) Cataloged ligand–receptor contacts, focusing on conserved interactions (e.g., D1473.32 salt bridge, H2976.52 water-mediated H-bond, Y3267.43 interactions). (4) Quantified residue composition within 5 Å of ligands and classified residues within 4 Å as shared or unique to ligand classes (agonist, partial agonist, antagonist). (5) Assessed water molecules explicitly resolved in structures. (6) Calculated binding-site solvent-accessible surface area (SASA) and solvent-accessible volume (SAV) using CASTp 3.0 to compare pocket openness across ligand classes. (7) Integrated and summarized findings from molecular dynamics simulations in the literature regarding ligand interactions, water networks (including NPxxY region dynamics), and ion/cholesterol allostery. The bitopic ligand structure (PDB 7U2L) was excluded from certain comparative analyses due to undetermined pharmacology.

- Conserved orthosteric interactions across morphinan-like ligands: a water-mediated hydrogen bond between the phenolic hydroxyl and H2976.52, and a salt bridge/ionic interaction between the ligand’s tertiary amine and D1473.32; D1473.32 also hydrogen bonds with Y3267.43. DAMGO’s N-terminus similarly forms a salt bridge to D1473.32 and H-bonds to Y3267.43. - Antagonist vs agonist pose similarity: β-FNA (antagonist) and BU72 (full agonist) display highly similar morphinan scaffold orientations and key interactions, despite opposite functional outcomes. - Fentanyl-specific recognition: In the fentanyl-bound MOR, π–π/stacking interactions occur between fentanyl’s benzene ring and W2936.48 and Y3267.43; a hydrophobic phenylethyl moiety engages a minor pocket between TM2–TM3 (involving Q1242.60, W13323.50, I1443.27). Alanine mutations in this pocket reduce fentanyl potency more than morphine, rationalizing fentanyl’s higher potency. - Partial agonists (e.g., PZM21, FH210): Retain D1473.32 and H2976.52 interactions but show strong complementarity to a lipophilic vestibule formed by the extracellular faces of TM2, TM3, and ECL1 (e.g., thiolphenylalkyl of PZM21, naphthyl of FH210), correlating with partial efficacy and increased G protein bias over β-arrestin recruitment. - Binding-pocket residue mapping (≤4 Å): 11 residues are common across structures; subsets are shared between specific ligand classes, and several residues are unique to each class (e.g., Y75 1.39, T218 45.51, H319 7.35, G325 7.41 unique to full agonists; T120 2.56, K233 5.39 to partial agonists; L219 5.52, E229 5.35 to antagonists). - Pocket openness by SAV: Full agonist-bound structures generally exhibit smaller solvent-accessible volumes than antagonist- or partial agonist-bound structures (Table 5). Examples: BU72 SASA 1212 Å2, SAV 619 Å3; DAMGO (6DDF) SASA 631 Å2, SAV 588 Å3; antagonists: alvimopan SASA 1016 Å2, SAV 1296 Å3; partial agonist PZM21 (7SBF) SASA 1341 Å2, SAV 2126 Å3. - Water molecules: Many cryo-EM structures lack resolved waters; where present (e.g., BU72-bound 5C1M), waters are observed near the NPxxY motif, consistent with activation-associated internal water pathways. MD studies indicate agonist-bound MOR hosts more internal water and dynamic exchange with extracellular water than antagonist-bound MOR. - Ion allostery: Na+ acts as a NAM via an allosteric site (coordinated by waters and N3.35/S3.39/D2.50, seen in inactive DOR), stabilizing inactive or intermediate states; Mg2+ preferentially binds extracellular regions (near S214ECL2, D216ECL2, E310ECL3) and acts as a PAM, promoting opening of the G protein cleft. - Cholesterol modulation: Cholesterol co-crystallizes in both active and inactive MOR at a groove between TM2–TM3–TM4/TM6, suggesting a conserved allosteric site influencing receptor function. - Small-molecule PAMs: BMS-986121/986122 enhance MOR signaling with probe dependence (e.g., increasing DAMGO/methadone potency/affinity, enhancing efficacy for morphine/nalbuphine). Docking/MD suggest allosteric sites near TM2/TM7 or extracellular TM2 region. - MD corroboration: Simulations recapitulate key contacts (e.g., PZM21 with D1473.32/Y1483.33/Y3267.43; oliceridine with D1473.32 and Y3267.43 stabilizing W2936.48; β-FNA maintaining D1473.32 salt bridge) and reveal mechanistic nuances underlying partial agonism and bias.

By systematically comparing high-resolution MOR structures and integrating MD literature, the review clarifies why chemically distinct ligands (full agonists, partial agonists, antagonists) can share a common orthosteric site yet produce different signaling outcomes. Conserved D1473.32–amine and H2976.52–phenol interactions anchor binding across classes, while additional contacts (e.g., fentanyl’s engagement of a TM2–TM3 minor pocket; partial agonists’ occupancy of a lipophilic vestibule at TM2/TM3/ECL1) correlate with potency, efficacy, and bias. Structural metrics (SAV) and water-network behavior support a model in which agonist-bound MOR adopts a more compact, water-permeable pocket conducive to activation, whereas antagonists expand the pocket and limit internal water dynamics. Allosteric modulators, including physiological ions (Na+, Mg2+) and cholesterol, act at distinct sites to stabilize specific conformational ensembles, offering routes to modulate signaling with potential safety advantages. Together, these insights address the central question of how ligand chemotypes drive distinct MOR signaling profiles and inform strategies for designing safer opioid therapeutics with improved benefit–risk profiles.

The review consolidates 3D structural and computational evidence to define conserved and class-specific MOR–ligand interactions, relate pocket geometry and water dynamics to activation, and map endogenous and small-molecule allosteric sites. Key contributions include: (1) identification of shared orthosteric anchors and class-distinct contacts (e.g., TM2–TM3 minor pocket for fentanyl; lipophilic vestibule for partial agonists) linked to potency and biased signaling; (2) quantification of pocket openness (SAV) differentiating ligand classes; (3) articulation of the roles of internal waters and ions (Na+, Mg2+) in activation and allostery; and (4) highlighting actionable allosteric sites (including cholesterol and TM2/TM7-adjacent regions) for modulator design. Future research should leverage these structures for structure-based design, explore novel allosteric sites, apply large-scale virtual screening to diversify chemotypes beyond morphinan scaffolds, and pursue ligands with tailored bias and improved safety profiles.

- Many cryo-EM structures lack resolved water molecules, limiting direct interpretation of water-mediated interactions despite their inferred importance from MD and select X-ray structures. - The bitopic ligand structure (7U2L) was excluded from certain analyses due to undetermined pharmacology, narrowing comparative scope. - SASA/SAV and residue-proximity analyses depend on static structures; dynamic conformational ensembles and probe dependence may not be fully captured without extensive simulations. - Evidence linking β-arrestin signaling to adverse effects (e.g., respiratory depression) remains controversial; translational implications of biased signaling require further validation. - Mutational and functional data are more comprehensive for some ligands (e.g., fentanyl) than others, potentially biasing mechanistic interpretations.