Opioids are the most effective drugs for the treatment of acute and chronic pain. As representative opioid alkaloids and synthetic opioids, respectively, Morphine and Fentanyl are used for cancer pain treatment, narcotic analgesia, prophylactic analgesia, and postoperative multimodal analgesia. Although opioids are effective pain relievers, they can cause serious side effects such as respiratory depression (deaths from respiratory depression have sparked a widespread "opioid crisis" ^[1,2], especially in North America), addiction and constipation, thus limiting their clinical application. According to a report published in 2019, more than 70% of deaths in the "opioid crisis" were caused by overdoses of synthetic opioids, mainly fentanyl and its derivatives. These side effects limit the clinical application of opioids.

Functions of opioids are mediated by a family of four G protein-coupled receptors (GPCRs), namely μ、κ、δ, and nociceptive peptide receptors (NOPR). Among these opioid receptors (ORs), μOR was revealed as the main receptor for both the analgesic and adverse effects of Morphine. It has been proposed that opioid-induced analgesia is attributed to Gi signaling of μOR (Fig. 1). However, which signaling pathway produces the unwanted side effects (such as respiratory suppression and constipation) is currently controversial: one view is that it is caused by β-arrestin signal transduction, and the other view suggests that the respiratory depression by opioids is caused due to G protein-gated inwardly rectifying potassium (GIRK) channel signaling.

This year in November, Cell published a research paper entitled Molecular recognition of morphine and fentanyl by the human μ-opioid receptor online. This article illustrates how fentanyl, morphine, and other μOR agonists bind to the μOR and reveals key differences in how they bind to the receptor. The study also reveals molecular details of the structural factors important for arrestin activity of μOR and provide structural templates for the design of potent and potentially safer analgesics.

The study first characterized the six opioids with diverse chemical scaffolds—fentanyl, morphine, SR17018, TRV130, PZM21, and DAMGO (a peptidomimetic agonist that activates the μOR-Gi complex, as a control)—the signaling profiles.

The study first characterized the signaling profiles of the six opioids with diverse chemical scaffolds used in this study: DAMGO, fentanyl, morphine, SR17018, TRV130, and PZM21.

As shown in Figure 2a, all five ligands were able to activate μOR to inhibit the production of cAMP, with fentanyl, morphine, and PZM21 acting as full agonists, whereas SR17018 and TRV130 being partial agonists when using DAMGO as a reference in the cAMP assay.

No response signals were detected for PZM21 in β-arrestin recruitment assays, while only weak signals were produced for SR17018 and TRV130. In contrast, both morphine and fentanyl induced robust β-arrestin recruitments (Figure 2b). All in all, fentanyl and morphine not only fully activated the μOR, but also induced robust β-arrestin recruitment.

Morphine and fentanyl are different from each other in chemical scaffolds. Therefore, the authors investigated their specific binding modes to μOR separately. Studies have shown that, the fentanyl molecule occupies a "Y"-shaped conformation in the orthosteric pocket, and contacts mainly with residues from TM2, TM3, TM6, and TM7 of transmembrane domain (TMD); morphine adopts an elliptical ‘‘O’’ configuration, interacting with hydrophobic residues from TM3, TM6, and TM7 (Fig. 3).

The study found that D149^3.32 formed polar interactions with nearby residues Q126^2.60 and Y328^7.43 (DQY polar motif, hereafter) in both structures (Fig. 3a-b), and mutations in the DQY polar motif largely decreased potencies of both G protein and β-arrestin signaling of fentanyl and morphine, indicating that the DQY polar motif was crucial for μOR activation.

In addition, mutations of residues around the ligand-binding pockets of fentanyl and morphine, including Y150^3.33, M153^3.36, V238^5.42, I298^6.51, and H299^6.52, resulted in attenuated activities toward G protein and β-arrestin signaling of both ligands (Figure 4a-b).

The authors mutated residues in the orthosteric binding pocket of μOR and tested the abilities of those mutants to activate G protein signaling and β-arrestin recruitment by fentanyl, morphine, and DAMGO, respectively.

The results showed that mutations of residues near TM6 and TM7 had more significant effects on β-arrestin signaling than those in TM2 and TM3 sides. For instance, mutations W295A from TM6 and W320A from TM7 had only minimum or partial effects on G protein signaling in cAMP inhibition and G protein recruitment, whereas they nearly abolished the β-arrestin recruitment induced by fentanyl, morphine, and DAMGO (Figure 5b). This suggested that the interaction of the ligand with TM6/7 was critical for β-arrestin signaling.

To further validate the above discovery, the authors designed two structurally similar fentanyl derivatives FBD1 (partial agonist) and FBD3 (full agonist), with the aim to obtain μOR agonists with reduced/abolished β-arrestin signaling but relatively intact G protein activity (Figure 6a). That is, the interaction between FBD1 and FBD3 and TM6/7 of μOR was weakened.

Results of this experiment showed that both FBD1 and FBD3 showed significant reduction in β-arrestin recruitment activities in reference to fentanyl. Especially, FBD3 showed nearly identical potency and efficiency as fentanyl in either cAMP inhibition or Gi recruitment assays but had very limited β-arrestin recruitment activity (Figure 6b-c).

The study further illustrated that TM6/7 was critical in eliciting arrestin signaling by μOR and that reducing this interaction may result in ligand signaling preferentially via the G protein pathway (Figure 7).

This article reveals the molecular recognition of morphine and fentanyl by the human μ-opioid receptors, and proposes an analog design based on the structure of fentanyl to weaken the β-arrestin activity of μOR. This research helps to rationally design the next-generation analgesics, and is expected to reduce the side effects of opioids without affecting their analgesic effect, or even enhance the analgesic effects.