Structural insights into distinct signaling profiles of the μOR activated by diverse agonists

Qianhui Qu, Weijiao Huang, Deniz Aydın, Joseph M. Paggi, Alpay B. Seven, Haoqing Wang, Soumen Chakraborty, Tao Che, Jeffrey F. DiBerto, Michael J. Robertson, Asuka Inoue, Bryan L. Roth, Susruta Majumdar, Ron O. Dror, Brian K. Kobilka, Georgios SkiniotisbioRxiv 2021 · DOI

TL;DR

This study reveals how different opioid drugs, even those targeting the same receptor (μOR), can trigger distinct downstream signaling pathways. By combining cryo-electron microscopy (cryo-EM) structures and molecular dynamics simulations, researchers uncovered how subtle differences in how drugs bind to the μOR lead to vastly different cellular responses, offering a molecular basis for their varying safety profiles and potential for developing safer analgesics.

What Problem Does This Paper Address?

Opioid drugs, particularly those targeting the mu-opioid receptor (μOR), are highly effective for pain relief but come with a significant risk of fatal respiratory depression. While some opioid agonists are considered "safer" due to reduced respiratory side effects, the underlying molecular mechanisms for these differences have remained unclear. This paper aims to elucidate the structural and mechanistic basis for these distinct signaling profiles, specifically comparing a potent and dangerous opioid, lofentanil (LFT), with a kratom-derived compound, mitragynine pseudoindoxyl (MP), which exhibits reduced respiratory depression. Understanding these differences is crucial for designing new analgesics that retain pain-relieving efficacy while minimizing dangerous side effects.

Methods

The researchers employed a multi-pronged approach:

  • Bioluminescence Resonance Energy Transfer (BRET) and NanoBiT Assays: These techniques were used to measure the activation of various G protein subtypes (Gi1, Gi2, Gi3, GoA, GoB, Gz) and the recruitment of β-arrestins in response to different ligands. This allowed for a quantitative comparison of the signaling "fingerprints" of LFT and MP.
  • Cryo-Electron Microscopy (Cryo-EM): High-resolution cryo-EM structures of the μOR complexed with Gi1, and either MP or LFT, were determined. This provided detailed 3D snapshots of how the ligands interact with the receptor and influence its interaction with the G protein.
  • Molecular Dynamics (MD) Simulations: Atomistic simulations were used to explore the dynamic behavior of the μOR-ligand complexes, both with and without the G protein. This allowed researchers to observe how ligand binding influences the receptor's conformational landscape and how these changes propagate to the intracellular side where signaling occurs.

Datasets

Not explicitly discussed as "datasets" in the context of machine learning training. However, the study relies on:

  • Existing μOR structures: The study builds upon previously determined cryo-EM structures of the μOR bound to other ligands (e.g., DAMGO, BU72).
  • Biochemical and cellular assay data: Data from TRUPATH BRET assays and NanoBiT assays were used to characterize the signaling profiles of the ligands.
  • Atomic coordinates: The newly determined cryo-EM structures of MP-μOR-Gi1 and LFT-μOR-Gi1 complexes have been deposited in the Protein Data Bank (PDB IDs: 7T2G and 7T2H) and EMDB (accession codes: EMD-25612 and EMD-25613).

Key Insights and Ideas

  • Ligand-Specific Binding Pockets: LFT and MP, despite binding to the same orthosteric site on the μOR, occupy distinct sub-pockets. LFT utilizes a more hydrophobic pocket, while MP engages a novel sub-pocket. This suggests that the shape and chemical properties of the ligand dictate its precise interaction within the binding site.
  • Conformational Plasticity of the μOR: The μOR is not a static molecule. Molecular dynamics simulations revealed that the receptor can adopt at least two distinct intracellular conformations ("canonical active" and "alternative").
  • Ligand-Driven Conformational Selection: Crucially, MP favors the "canonical active" conformation, which is similar to what is observed when the G protein is bound. In contrast, LFT favors an "alternative" conformation, characterized by a counterclockwise rotation of TM7. This suggests that ligands can bias the receptor towards specific signaling-competent states.
  • TM7 Rotation as a Key Allosteric Mechanism: The rotation of transmembrane helix 7 (TM7) appears to be a critical allosteric mechanism that dictates the receptor's intracellular conformation and, consequently, its interaction with downstream signaling partners like G proteins and β-arrestins.
  • Implications for Biased Agonism: The distinct signaling profiles observed for LFT and MP, with MP showing a preference for G protein activation and minimal β-arrestin recruitment, while LFT engages both pathways more broadly, provides a molecular explanation for biased agonism. This concept is key to developing safer opioids.

Key Results

  • Differential G Protein and β-arrestin Signaling: LFT exhibited higher efficacy across all G protein subtypes and robust recruitment of β-arrestins, while MP acted as a partial agonist with significantly lower efficacy and minimal β-arrestin recruitment.
  • Cryo-EM Structures Reveal Distinct Binding Modes: The 2.5Å structure of the MP-μOR-Gi1 complex and the 3.2Å structure of the LFT-μOR-Gi1-scFv complex clearly showed how MP and LFT occupy different sub-pockets within the orthosteric binding site.
  • MD Simulations Show Ligand-Dependent Intracellular Conformations: Simulations without the G protein revealed that MP favors the canonical active state, LFT favors an alternative state, and DAMGO (a reference agonist) samples an equilibrium between the two. This difference is linked to the rotation of TM7.
  • Y3267.43 Mutation Affects Signaling: A mutation at Y3267.43, which is involved in hydrogen bonding with MP and LFT, disproportionately reduced β-arrestin recruitment for MP and DAMGO, highlighting its importance in mediating arrestin signaling.

Limitations and Future Work

  • In Vitro vs. In Vivo: While the study provides detailed molecular insights, the direct translation of these in vitro findings to in vivo physiological responses, particularly respiratory depression, requires further investigation. The authors acknowledge that the precise role of specific Gi/o/z subtypes in respiratory depression remains complex and agonist-specific.
  • Limited Ligand Set: The study focuses on two specific ligands (LFT and MP) and a reference agonist (DAMGO). Exploring a broader range of agonists with varying pharmacological profiles could provide a more comprehensive understanding of μOR signaling.
  • Future Directions: The authors emphasize the need for a more complete functional characterization of μOR agonists, considering all relevant G protein subtypes and β-arrestin recruitment. This comprehensive approach is essential for identifying and designing novel opioid analgesics with improved safety profiles. Further research could also explore how these ligand-induced conformational changes at the intracellular interface influence the precise interactions with different G protein subtypes.

Summary updated

Abstract

Abstract Drugs targeting the G protein-coupled μ-opioid receptor (μOR) are the most effective analgesics available but are also associated with fatal respiratory depression. While some partial opioid agonists appear to be safer than full agonists, the signaling pathways responsible for respiratory depression have yet to be elucidated. Here we investigated the structural and mechanistic basis of action of lofentanil (LFT) and mitragynine pseudoindoxyl (MP), two μOR agonists with different safety profiles. LFT, one of the most potent and lethal opioids, and MP, a derivative from the kratom plant with reduced respiratory depression in animal studies at equianalgesic doses, exhibited markedly different signaling efficacy profiles for G protein subtype activation and recruitment of β-arrestins. Cryo-EM structures of the μOR-Gi1 complex with MP (2.5Å) and LFT (3.2Å) revealed that the two ligands engage distinct sub-pockets, and molecular dynamics (MD) simulations showed additional differences in the binding site that propagate to the intracellular side of the receptor where G proteins and β-arrestins bind. While MP favors the precise G protein-bound active state observed in the cryo-EM structures, LFT favors a distinct active state. These results highlight how drugs engaging different parts of the μOR orthosteric pocket can lead to distinct signaling outcomes.
BibTeX
@misc{qu2021structural,
  author = {Qianhui Qu and Weijiao Huang and Deniz Aydın and Joseph M. Paggi and Alpay B. Seven and Haoqing Wang and Soumen Chakraborty and Tao Che and Jeffrey F. DiBerto and Michael J. Robertson and Asuka Inoue and Bryan L. Roth and Susruta Majumdar and Ron O. Dror and Brian K. Kobilka and Georgios Skiniotis},
  title = {{Structural insights into distinct signaling profiles of the μOR activated by diverse agonists}},
  year = {2021},
  doi = {10.1101/2021.12.07.471645},
  howpublished = {bioRxiv},
}
Figure