Mu Opioids Induce Biased Signaling at the Full-Length Seven Transmembrane C-Terminal Splice Variants of the mu Opioid Receptor Gene, Oprm1

Abstract

The biased signaling has been extensively studied in the original mu opioid receptor (MOR-1), particularly through G protein and β-arrestin2 signaling pathways. The concept that the G protein pathway is often linked to the therapeutic effect of the drug, while the β-arrestin pathway is associated to the side effects has been proposed to develop biased analgesic compounds with limited side-effects associated with traditional opiates. The mu opioid receptor gene, OPRM1, undergoes extensive alternative pre-mRNA splicing, generating multiple splice variants or isoforms that are conserved from rodent to human. One type of the Oprm1 splice variants are the full-length 7 transmembrane (7TM) C-terminal splice variants, which have identical receptor structures including entire binding pocket, but contain a different intracellular C-terminal tail resulted from 3' alternative splicing. Increasing evidence suggest that these full-length 7TM C-terminal variants play important roles in mu opioid pharmacology, raising questions regarding biased signaling at these multiple C-terminal variants. In the present study, we investigated the effect of different C-terminal variants on mu agonist-induced G protein coupling, β-arrestin2 recruitment, and ultimately, signaling bias. We found that mu agonists produced marked differences in G protein activation and β-arrestin2 recruitment among various C-terminal variants, leading to biased signaling at various level. Particularly, MOR-1O, an exon 7-associated variant, showed greater β-arrestin2 bias for most mu agonists than MOR-1, an exon 4-associated variant. Biased signaling of G protein-coupled receptors has been defined by evidences that different agonists can produce divergent signaling transduction pathways through a single receptor. Our findings that a single mu agonist can induce differential signaling through multiple 7TM splice variants provide a new perspective on biased signaling at least for Oprm1, which perhaps is important for our understanding of the complex mu opioid actions in vivo where all the 7TM splice variants co-exist.

Keywords: Beta-arrestin; Biased signaling; G-protein; Mu opioid receptor; Splicing.

Fig. 1

Schematic structure of Oprm1 full-length 7TM C-terminal splice variants. The top panel is a drawing that shows structures of the mu opioid receptor encoded by exons 1–3. The transmembrane domains are indicated by cylinders. The right of the bottle panel lists the amino acid sequences encoded by alternative exons downstream of exon 3 whose composition is shown by corresponding colors on the left. The predicted phosphorylation sites are indicated by red letters. The underlined sequences in MOR-1C and MOR-1O are the phosphorylation code for high-affinity β-arrestin binding based on crystal structures of several GPCRs (Zhou et al. 2017)

Fig. 2

Mu agonist-induced β-arrestin2 recruitment and G protein activation at Oprm1 full-length 7TM C-terminal splice variants. A PathHunter® β-arrestin2 assay (Eurofins DiscoverX) and [³⁵S]GTPγS binding assay were performed using CHO cell lines stably expressing β-arrestin2-EA and indicated PK-tagged 7TM C-terminal variants, as described in Materials and Methods. In both PathHunter® β-arrestin2 and [³⁵S]GTPγS binding assays, 10 µM DAMGO was included in each dose–response curve as a reference so all the E_max values were normalized with that of 10 µM DAMGO. The ED₅₀ values and % E_max values were analyzed by non-linear regression (Prism 8) and listed on Table 3. Results are from three independent experiments

Fig. 2

Mu agonist-induced β-arrestin2 recruitment and G protein activation at Oprm1 full-length 7TM C-terminal splice variants. A PathHunter® β-arrestin2 assay (Eurofins DiscoverX) and [³⁵S]GTPγS binding assay were performed using CHO cell lines stably expressing β-arrestin2-EA and indicated PK-tagged 7TM C-terminal variants, as described in Materials and Methods. In both PathHunter® β-arrestin2 and [³⁵S]GTPγS binding assays, 10 µM DAMGO was included in each dose–response curve as a reference so all the E_max values were normalized with that of 10 µM DAMGO. The ED₅₀ values and % E_max values were analyzed by non-linear regression (Prism 8) and listed on Table 3. Results are from three independent experiments

Fig. 3

Heatmap of biased factors. Biased factors were calculated using the Black and Leff Operational Model by using different normalization methods, as described in Materials and Methods. a Normalized with respect to DAMGO at MOR-1 for a comparison between drugs and variants. b Normalized with respect to each drug at mMOR-1 for a comparison across variants. c Normalized with respect to DAMGO at each variant for a comparison across drugs. The negative (blue) values indicate β-arrestin2 bias whereas the positive bias (red) values indicate G protein bias

Fig. 4

Mu agonist induced ERK1/2 activation in HEK293 cells expressing MOR-1 and MOR-1O receptor. Western blot. Whole cell lysate from HEK293 cells expressing indicated variants with treatment of mu agonists (1 µM, 5 min) were used in Western blot analysis, as described in Materials and Methods. Lanes 3 & 5: No drug control; Lanes 1 & 6: DAMGO; Lanes 2 & 7: buprenorphine; Lanes 4 & 8: fentanyl. pERK1/2: phosphorylated ERK1/2; tERK: total ERK1/2. One representative from three independent experiments is shown. Quantification of Western blot. Band intensities on the images obtained in ChemiDoc MP (Bio-Rad) were quantified using ImageLab 6.2. The ratios of pERK1/2 and tERK1/2 were normalized with No drug control so No drug is always as 1. Results are the means ± S.D. of three independent determinations. Significant differences were analyzed by two-way ANOVA with Bonferroni’s post hoc test. * & #: p < 0.05; ** & ##: p < 0.01; ***: p < 0.001