Biased agonism at the µ-opioid receptor

The μ opioid receptor (MOP) is the main therapeutic target for the most clinically useful class of analgesics for treating severe acute and chronic pain, despite the numerous associated side effects that limit their use. The property of G protein-coupled receptors (GPCRs) where different ligands stabilise the receptor into unique active conformations, which can result in differential activation of cell signalling pathways and, eventually, in different physiological outcomes is known as biased agonism. Biased agonism is a natural phenomenon that has been observed at other neuropeptide receptors that, like the opioid system, have multiple endogenous ligands targeting the same receptor. This property of GPCRs can be exploited to design drugs that selectively activate signalling pathways that lead to the desired physiological effects whilst minimising side effects that are elicited by activation of other signalling pathways. Quantification of biased agonism of endogenous opioids at MOP across multiple different signalling pathways using the transduction coefficient ratio [ΔΔlog(τ/KA)] method was performed in a common cellular background. This has revealed that this family of peptides possess a diverse range of signalling profiles. Met-enk-RF, endomorphin-1 and α-neoendorphin in particular showed unique bias profiles when compared to the synthetic ligand DAMGO, whereas most other endogenous peptides showed bias profiles more similar to DAMGO. This diversity in bias among endogenous opioids, may enable the endogenous opioid system to have an unprecedented level of control to fine-tune MOP mediated physiology. There are a number of factors that need to be taken into consideration when quantifying biased agonism, such as the spatiotemporal kinetics of the signalling pathways. The spatiotemporal kinetics of extracellular signal regulated kinase (ERK) activation by morphine and DAMGO was examined in dorsal root ganglia neurons. This showed that morphine stimulates sustained activation of cytosolic ERK via a PKC dependent pathway, whereas DAMGO stimulates transient activation of cytosolic and nuclear ERK. Similarly, examination of the kinetics of cAMP inhibition in CHO-MOP cells revealed that the cAMP kinetic profile of Met-enk-RF is distinct from that of DAMGO and other endogenous ligands, which results in a significant change in bias of Met-enk-RF depending on the time point chosen to measure cAMP inhibition. Additionally, bias between activation of different G protein subtypes was quantified at different time points, which showed that the bias of Met-enk-RF and α-neoendorphin changed over time. Overall, the spatiotemporal characteristics of the signalling pathways can have a significant impact on the observed bias of a ligand. The impact of the cellular background on the quantification of bias was also investigated by quantifying biased agonism of the same ligands in different cell backgrounds. This revealed that even altering the expression level of a single signalling protein can change the observed bias of a ligand, where overexpression of GRK2 in CHO-MOP cells altered the bias of endomorphin-1 between inhibition of cAMP and β-arrestin recruitment. However, when biased agonism was quantified across multiple signalling pathways in different cell types, CHO-MOP and AtT20-MOP, despite the fact the overall bias profiles changed significantly between cell lines, ligands with unique bias profiles retained uniqueness. Thus suggesting that these pluridimensional bias profiles can be used to predict in vivo bias indirectly via ligand clustering.