Ferenc Zádor 1*†, Kornél Király 1, Nariman Essmat 1 and Mahmoud Al-Khrasani 1
1Department of Pharmacology and Pharmacotherapy, Faculty of Medicine, Semmelweis University, Budapest, Hungary
Opioid agonists produce their analgesic effects primarily by acting at the μ-opioid receptor (μOR). μOR agonists with different efficacies exert diverse molecular changes in the μOR which dictate the faith of the receptor’s signaling pathway and possibly it’s the degree of desensitization. Since the development of the active conformations of the μOR, growing data have been published in relation to ligand-specific changes in μOR activation. In this regard, this review summarizes recent data regarding the most studied opioid agonists in in silico μOR activation, including how these ligands are recognized by the μOR, how their binding signal is transmitted toward the intracellular parts of the μOR, and finally, what type of large-scale movements do these changes trigger in the μOR’s domains.
Keywords: μ-opioid receptor, agonist-specific receptor activation, prototypic μ-opioid receptor agonist, TRV-130, PZM21
Growing data support that the rate of opioid side-effects including analgesic tolerance development strongly correlates with the pharmacodynamic properties of opioid ligands. Opioids with different efficacies distinctly induce molecular mechanisms related to tolerance, namely receptor phosphorylation and endocytosis, as the basis of G-protein coupled μ-opioid receptor (μOR) desensitization (Williams et al., 2013; Allouche et al., 2014; Lemel et al., 2020). It has been proposed that the selective and sequential phosphorylation of the C-terminus is due to the possible different conformational states of the receptor-triggered agonist specifically (Lemel et al., 2020). In recent years, we have gained more information regarding the nature of opioid agonists binding to the active conformation of the μOR (Huang et al., 2015; Koehl et al., 2018). This review will focus on the current knowledge of agonist specific residue contacts ( Figure 1B ), how the different agonists transmit the ligand-binding signal toward the intracellular receptor parts (Figure 1C ), and finally, how these affect the orientation of certain receptor domains (e.g., transmembrane regions (TM) or intracellular loops (IL)) ( Table 1 ), which eventually decide the faith of the receptor’s downstream signaling and the rate of desensitization. In addition, only data with the active conformation of the μOR will be reviewed here, namely the BU72 co-crystallized form and μOR-G i complex co-crystallized with D-Ala 2 , N-MePhe 4 , and Gly-ol-enkephalin (DAMGO; PDB: 5C1M and PDB: 6DDF, respectively). Data on prototypic μOR-specific agonist ligands (Figure 1A ), namely morphine, DAMGO, and fentanyl, will be reviewed alongside BU72, the first compound to be crystallized with the active conformational state of the μOR (Huang et al., 2015). TRV-130 and PZM21, newly developed G-protein-biased agonists, will be also reviewed (Figure 1A ). In general, in the highlighted studies CHARMM (Brooks et al., 2009) and/or AMBER (Maier et al., 2015) force field was used, with 0.1 –3.5 μs simulation time (in some cases, 24 μs; see Vo et al. (2020) in POPC (palmitoyl-oleoyl-phosphatidylcholine) lipid membrane model at ~1 bar
Edited by: Fabio Arturo Iannotti, Consiglio Nazionale delle Ricerche (CNR), Italy Reviewed by: Haiguang Liu, Beijing Computational Science Research Center (CSRC), China *Correspondence: Ferenc Zádor admin@frankenthalerfoundation.org †Present address: Pharmacological and Drug Safety Research, Gedeon Richter Plc, Budapest, Hungary Specialty section: This article was submitted to Structural Biology, a section of the journal Frontiers in Molecular Biosciences Received: 20 March 2022 Accepted: 21 April 2022 Published: 13 June 2022 Citation: Zádor F, Király K, Essmat N and Al-Khrasani M (2022) Recent Molecular Insights into Agonist-specific Binding to the Mu-Opioid Receptor. Front. Mol. Biosci. 9:900547. doi: 10.3389/fmolb.2022.900547 Frontiers in Molecular Biosciences | www.frontiersin.org June 2022 | Volume 9 | Article 900547 1 MINI REVIEW published: 13 June 2022 doi: 10.3389/fmolb.2022.900547
pressure and 310 K temperature in a ~75 –85 x 75 –85 x 90 –140 Å size simulation box. Some studies also used NMR spectroscopy to obtain dynamic structural information (Okude et al., 2015; Sounier et al., 2015). Such a pool of data will help us to better understand the basic molecular factors of ligand-specific receptor activation and tolerance, and allow us to purposefully develop opioids with delayed analgesic tolerance profiles and ameliorated side effects.
Based on site-directed mutagenesis and in silico studies, multiple conserved residues have been identified in the μOR binding pocket, which have significant roles in ligand orientation and receptor activation (Mansour et al., 1997; Manglik et al., 2012, 2015, 2016; Katritch et al., 2013; Kaserer et al., 2016; Koehl et al., 2018; Marino et al., 2018; Manglik, 2020; Ricarte et al., 2021). Hitherto, data on the agonist-specific residue contacts and binding modes will be reviewed in this section. Despite morphine and fentanyl interacting with the same contact residues ( Figure 1B ), their binding poses were less overlapped (Lipiński et al., 2019). Accordingly, fentanyl is in close proximity to seven TM3 residues and three TM6 residues, while in the case of morphine these numbers are four and five with respect to the same transmembrane domains. They also interact with TM7 to a similar extent but with different positions. Fentanyl is also able to reach the ECL1, ECL2, and the N-terminus. These findings were later confirmed by another group (Ricarte et al., 2021). Analyzing the dissociation of morphine from the μOR, it showed that morphine directly dissociated from the orthosteric site region and also transitioned to the vestibule region after the Asp 3.32 salt bridge was disrupted (Ribeiro et al., 2020) (superscript numbering refers to the Ballesteros and Weinstein’s generic numbering scheme (Ballesteros and Weinstein, 1995)). Fentanyl binds deeper compared to morphinan structures (for fentanyl it is indicated by a lower ΔZ value, the distance between the centers of mass (COM) of fentanyl and μOR z
direction) and it can form a salt-bridge interaction between the piperidine amine and the conserved Asp 3.32 (Vo et al., 2020) similar to DAMGO or BU72 (Huang et al., 2015; Weis and Kobilka, 2018). Vo and co-workers described a His 6.52 binding mode unique to fentanyl, which was also dependent on the protonation state of this residue (Vo et al., 2020). Another study found that the dissociation pathways, time, the depth of insertion, and the strength of TM6 interaction of fentanyl are dependent on the protonation state of His 6.52 (Mahinthichaichan et al., 2021).
FIGURE 1 | (A) Chemical structures of μOR-selective agonists discussed in the review. (B) The known μOR residual contacts of the indicated agonists. The original concept of the figure was based on Figure 4 of Podlewska and co-workers’study (Podlewska et al., 2020) and extended by other data (Huang et al., 2015; Cheng et al., 2018; Koehl et al., 2018; Mafi et al., 2020; Ricarte et al., 2021). (C) Individual movements of the highlighted residues, molecular switches, and TM domains based on the data reviewed in the 3rd and 4th sections.. Participating residues are indicated in orange, arched arrows indicate the presence of spatial movements (but not the direction itself), while straight arrows depict the presence of altered distance between two residues. The corresponding agonists inducing these movements and alterations are not indicated for clarity; for details see in the 3rd and 4th sections. μOR is transparent for better visibility. The figure was constructed with UCSF Chimera 1.13.1 (Pettersen et al., 2004) based on Huang and co-workers using the BU72 co-crystallized active μOR structure (PDB: 5C1M) (Huang et al., 2015). Frontiers in Molecular Biosciences | www.frontiersin.org June 2022 | Volume 9 | Article 900547 2 Zádor et al. Agonist-Specific μOR Binding
In the case of BU72, most of its interactions with the active μOR are hydrophobic or aromatic. The phenolic hydroxyl group of BU72 interacts with His 6.52 in a water-mediated fashion (Huang et al., 2015). There is also an ionic interaction between Asp 3.32 and the morphinan tertiary amine structure of BU72. BU72 stabilizes the rearrangement of a triad of conserved residues upon receptor activation (Huang et al., 2015). BU72 also forms a hydrophobic surface with Ile 6.51 and Val 6.55 in TM6 and Ile 7.39 in TM7, similarly to other morphinan structures (Figure 1B ) (Huang et al., 2015). Another study demonstrated that BU72 binding poses distinct from the active μOR crystal structures and presumed that the high affinity and agonist character of BU72 is in part presented by its configurational entropy (Feinberg et al., 2017). Koehl et al. found that the conformation of the active-state binding pocket and the orientation of the residues that interact
Zádor et al. Agonist-Specific μOR Binding with the agonist are highly similar between BU72 and DAMGO, despite the structural differences ( Figure 1B ) (Koehl et al., 2018). On the other hand, compared to BU72, the C-terminus of DAMGO extends further toward the ECLs. Another study with DAMGO has shown that the tyrosine of the peptide forms lipophilic contacts with Met 3.36 , Ile 6.51 , and Val 6.55 residues and forms a charge interaction with Asp 3.32 (Figure 1B ) (Dumitrascuta et al., 2020). It has been proved that TRV-130 has stronger interactions (a greater number of hydrophobic contacts) with TM2 and TM3 compared to morphine or DAMGO in β-arrestin2 stabilized with phosphorylated μOR (Mafi et al., 2020). Based on docking simulations, the protonated nitrogen ion of TRV130 formed electrostatic interactions with Asp 3.32 and through its ring structure formed interactions with His 6.52 (Figure 1B ) (Cheng et al., 2018). PZM21 interacts with the active μOR binding pocket by hydrogen bonds, hydrophobic interactions, and an ionic bond (Manglik et al., 2016). Podlewska and co-workers have compared PZM21 with fentanyl or morphine in docking and MD simulations in BU72 and DAMGO co-crystallized active structures (Podlewska et al., 2020). Interestingly, all compounds showed less stability in their orientations in the DAMGO co-crystallized conformation, especially morphine, meaning that their initial and final binding orientations were significantly different during the simulation. They also found that during simulation time, PZM21 had more contacts with Asp 3.32 in both crystal structures compared to fentanyl or morphine (Podlewska et al., 2020). Another recent study compared PZM21 to morphine in MD simulations and found that besides PZM21 interacting with key residues Asp 3.32 and Tyr 3.33 of TM3 (Figure 1B ), similar to morphine, yet it strongly interacts with Tyr 7.43 of TM7 ( Figure 1B ), as indicated by a higher percentage of interaction fractions in H-bonds (Liao et al., 2021). Finally, Lee and co-workers have performed molecular docking with new potential biased μOR agonists, where they also compared these novel compounds to T
