Development of Mutant M₃ Muscarinic Receptors Biased for G Protein Activation or Recruitment of β-Arrestins

Abstract

The M₃ muscarinic acetylcholine (ACh) receptor (M3R), a prototypic class A biogenic amine G protein-coupled receptor (GPCR), regulates various important functions of the CNS and many of the effects of ACh released from peripheral parasympathetic nerves. Agonist-activated M3Rs preferentially couple to G proteins of the G_q family but are also able to bind β-arrestins. The recruitment of β-arrestins by the activated M3R disrupts receptor/G protein coupling and promotes M3R internalization, at least in certain cell types. Of note, numerous studies have shown that GPCR-recruited β-arrestins can act as signaling molecules in their own right. These findings raise the question to which extent the recruitment of β-arrestins by activated M3Rs contributes to the physiological responses triggered by M3R activation. A better understanding of the potential physiological relevance of M3R-mediated β-arrestin recruitment may guide the development of so-called biased M3R ligands that preferentially promote either G protein signaling or β-arrestin recruitment. For this reason, the present study aimed to develop two biased mutant M3Rs with opposing functional properties. One of the mutant M3Rs selectively activates G proteins, while the other one preferentially recruits β-arrestins. Here, we report an initial functional characterization of this pair of mutant M3Rs in cultured cells. The future use of these new molecular tools for generating M3R knockin mice or for structural studies may pave the way for the clinical use of biased M3R ligands endowed with high therapeutic efficacy and a favorable side effect profile.

Graphical Abstract

INTRODUCTION

Muscarinic acetylcholine (ACh) receptors (mAChRs) are prototypic members of the superfamily of class A G protein-coupled receptors (GPCRs)^{1, 2}. The mAChR family consists of five molecularly distinct subtypes (M1R-M5R) that differ in their pattern of distribution and functional properties¹. While the odd-numbered mAChR subtypes preferentially activate G proteins of the G_q/11 family, the even-numbered mAChRs are preferentially linked to G_i/o-type G proteins¹.

During the past three decades, we used the M3R subtype as a prototypic class A GPCR to identify the molecular basis underlying ligand binding, G protein coupling selectivity, GPCR dimerization, and several other GPCR functions^3–6. More recently, several high-resolution structures of the M3R in complex with chemically diverse muscarinic ligands have been reported^7–10. These structural data have created new opportunities for medicinal chemists to develop drugs that selectively target the M3R¹¹.

Various lines of evidence, including work with M3R knockout (KO) mice, suggest that M3Rs mediate many of the important physiological actions of ACh released from peripheral parasympathetic nerves¹. For example, ACh-induced activation of peripheral M3Rs triggers the contraction of smooth muscle tissues and stimulates secretion from exocrine and endocrine glands¹. In a series of papers, we demonstrated that activation of M3Rs expressed by pancreatic β-cells strongly stimulates insulin secretion, resulting in greatly improved glucose homeostasis in various mouse models of diabetes¹². The most commonly used M3R drugs in the clinic are muscarinic antagonists that are highly effective in the treatment of chronic obstructive pulmonary disease and asthma by lowering M3R-dependent contraction of bronchial smooth muscle and reducing mucus production¹³.

As is the case with most other GPCRs, agonist stimulation of the M3R recruits cytoplasmic proteins of the β-arrestin family (β-arrestin-1 and −2; alternative names: arrestin-2 and −3, respectively) to the activated receptor^14–16. In most cases, the binding of β-arrestins to the activated M3R disrupts receptor/G protein interactions, leading to receptor desensitization. In addition, M3R/β-arrestin complexes can undergo removal from the cell surface (receptor internalization) via clathrin-dependent processes, at least in certain cell types¹⁶.

β-arrestins recruited by activated GPCRs can act as signaling molecules in their own right^17–21. In many cases, β-arrestins act as cytoplasmic scaffolds allowing more efficient signaling via members of the MAPK family and other signaling cascades^{19, 22, 23}. However, β-arrestins can also directly activate cytoplasmic signaling proteins (see, for example,^{24, 25}).

On the basis of these findings, we explored the extent to which the recruitment of β-arrestins by activated M3Rs contributes to the final cellular responses triggered by M3R stimulation. Such information may provide a rational basis for the development of biased M3R ligands capable of selectively activating G proteins without recruiting β-arrestins or, vice versa, of M3R ligands that can recruit β-arrestins but are unable to stimulate G protein signaling. Studies with many other GPCRs have shown that this approach may prove useful for the development of novel classes of drugs endowed with enhanced therapeutic efficacy and/or a more favorable side effect profile^{26, 27}. For example, a G protein-biased μ-opioid receptor agonist, oliceridine, has recently been approved by the FDA for the treatment of moderate to severe acute pain²⁸, and a G protein-biased 5-HT_2C receptor agonist²⁹ is currently being evaluated in clinical trials for anti-epileptic efficacy (NCT06401538).

To gain a better understanding of the relative functional importance of M3R-mediated G proteins activation vs. M3R-dependent β-arrestin recruitment, we generated two biased mutant M3Rs. One of the receptors selectively activates G proteins and does not recruit β-arrestins, whereas the other one recruits β-arrestins, but does not activate G proteins. In the present study, we functionally characterized these two mutant receptors in cultured cells.

MATERIALS AND METHODS

Drugs.

Carbachol (carbamoylcholine chloride) was obtained from Sigma-Aldrich (# C4382).

The source of all other reagents used is indicated in the subsequent paragraphs.

Generation of Biased Mutant M3Rs.

To generate a mutant M3R capable of coupling to G_q-type G proteins but lacking the ability to recruit β-arrestins, we used the sequence of a previously published G protein-biased DREADD (M3D-Gq) (DREADD=Designer Receptor Exclusively Activated by a Designer Drug) as a template³⁰. To restore the ACh binding site in the M3D-Gq construct, we introduced the C148Y and G238A point mutations (rat M3R sequence) using the QuikChange^™ site-directed mutagenesis kit (Stratagene, La Jolla, CA) following the manufacturer’s instructions. For simplicity, we refer to this newly generated M3R as “M3R-Gq.” In a similar fashion, we introduced the C148Y and G238A point mutations into a β-arrestin-biased DREADD (Rq(R165L)) (rat M3R sequence), first described by Nakajima et al.³¹. Throughout the manuscript, we refer to this β-arrestin-biased M3R as “M3R-Arr”. All M3R constructs used in this study carried an N-terminal HA epitope tag and were inserted into the pcDNA3.1 mammalian expression vector using standard cloning techniques.

Transient Expression of M3R Constructs in Cultured Cells.

COS-7 cells were seeded in 100 mm dishes at a density of 1 × 10⁶ cells/dish and grown in Dulbecco’s Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (FBS). On the next day, cells were transiently transfected with M3R constructs using Lipofectamine^™ 2000 according to the manufacturer’s protocol (Thermo Fisher Scientific). Cells were harvested 24 h after transfection and replated in collagen-coated 96-well plates (black-walled, clear bottom; Thermo Scientific) at a density of 25,000 cells/well. About 48 h after transfection, cells were used for calcium mobilization assays (see next paragraph).

Calcium Mobilization Assay.

About 48 h after transfection, carbachol-induced changes in intracellular calcium levels ([Ca²⁺]_i) were measured via FLIPR (Molecular Devices, Sunnyvale, CA), as described previously³⁰. Briefly, after a washing step with 200 μl of Hanks’ balanced salt solution containing 20 mM HEPES (pH 7.4), 30 μl of loading dye (FLIPR Calcium 3 Assay Kit; Molecular Devices), supplemented with 2.5 mM probenecid to enhance dye retention, was added to each well, followed by a 45-min incubation at 37°C. Cells were then exposed to increasing concentrations of carbachol, and changes in cell fluorescence were measured at 22°C using the FLIPR TETRA System (excitation: 470–495 nm; emission: 515–575 nm). For each carbachol concentration, increases in [Ca²⁺]_i were measured as peak fluorescence activity divided by basal fluorescence activity prior to the addition of carbachol. Curve fitting was performed using the non-linear regression “log(agonist) vs. response (three parameters)” equation in GraphPad Prism 10. Carbachol E_max values were determined from concentration-response curves, und data were normalized to E_max values obtained with the M3R-Gq construct. EC₅₀ and log EC₅₀ values were derived from normalized concentration-response curves. Three independent experiments were performed, each in quintuplicate.

Measurement of Cell Surface Receptor Expression Levels.

About 48 h after transfection of COS-7 cells with different M3R constructs, receptor cell surface expression levels were measured by using a whole-cell ELISA strategy³². In brief, transfected COS-7 cells grown in 96-well plates (cat # 713011, Nest Biotechnology) were washed once with 100 μl PBS and then fixed with 4% paraformaldehyde for 20 min at room temperature (RT). Fixed cells were washed twice with PBS and then blocked with 5% goat serum in PBS (blocking buffer) for 1 h at RT. The extracellular HA-tag present at the N-terminus of all M3R constructs was detected by incubation with an anti-HA-biotin antibody (Millipore Sigma, B9183, 1:1,000 dilution) in blocking buffer for 1 h. After several washes with PBS, cells were incubated with horseradish peroxidase-conjugated streptavidin (Thermo Scientific, #21130, 1:2,000 in blocking buffer) for 30 min at RT. Following three washes with PBS, 100 μl of one-step TMB-ELISA solution (Thermo Scientific, #34028) were added to the cells. After 5 min, the reaction was terminated by the addition of 100 μl of 2 M sulfuric acid, and color development was determined by measuring absorbance at 450 nm with a microplate reader. Cells that had been transfected with empty vector were processed in the same fashion for control purposes.

Bioluminescence Resonance Energy Transfer (BRET) Studies.

The ability of the various M3R constructs to recruit β-arrestin-1 or −2 was assessed using BRET assays. COS-7 cells were plated at 75,000 cells/ml in 100-mm dishes containing DMEM and grown to 90–100% confluency. Cells were then co-transfected in serum-free media with a fixed amount (0.1–0.2 μg) of different M3R constructs carrying a C-terminal luciferase (Luc) tag and either β-arrestin-1 or −2 (20 μg) constructs harboring an N-terminal Venus sequence, using a 1:2.5 DNA-to-Lipofectamine^™ 2000 (Thermo Fisher Scientific) ratio. Five h post-transfection, serum-free media was replaced with DMEM supplemented with 10% FBS. After 48 h, cells were gently detached from the plates using Dulbecco’s phosphate-buffered saline (DPBS, Thermo Fisher Scientific) containing 1 mM EDTA, and 100 μl/well of 50,000 cells/ml were transferred to a 96-well plate. Increasing concentrations of carbachol (10 nM to 1 mM, 100 μl) dissolved in DPBS were added to the cells, in addition to 100 μl of 5 μM coelenterazine-h substrate. Luminescence was measured at 485 nm, and fluorescence at 535 nm after 25 min using a 96-well plate reader. Changes in BRET signals were calculated by subtracting the BRET ratio of unstimulated cells from that of carbachol-treated cells. Expression of the Venus-β-arrestin constructs was measured at 535 nm prior to carbachol treatment. Each experiment was performed with eight technical replicates, for a total of three independent experiments. Curve fitting was performed using the non-linear regression “log(agonist) vs. response (three parameters)” equation in GraphPad Prism 10. Carbachol E_max values were determined from concentration-response curves, und data were normalized to E_max values obtained with the M3R-WT receptor. EC₅₀ and log EC₅₀ values were derived from normalized concentration-response curves.

M3R Constructs Used for Internalization Assays.

To visualize the internalization of the WT M3R and the two biased mutant M3Rs (M3R-Gq and M3R-Arr), a sequence coding for the EYFP fluorophore was fused to the C-terminus of the receptors (template: human M3R) using standard molecular cloning techniques. In brief, the receptor sequences were amplified by PCR with a forward primer containing an XhoI cleavage site and a reverse primer harboring a BamHI restriction site. The pEYFP-N1-expression vector (Clontech) containing the EYFP sequence was digested with XhoI and BamHI, followed by the insertion of the various receptor sequences. All constructs were verified by sequencing.

Live-cell Imaging of M3R Internalization Using Fluorescence Microscopy.

HEK293 cells were plated onto sterile 8-well μ-slides (Ibidi GmbH, Martinsried/Munich, Germany). When cells were ~70% confluent, they were co-transfected by using Lipofectamine^™ 2000 (Life Technologies, Carlsbad, CA, USA) with 0.9 μg DNA encoding M3R constructs C-terminally tagged with EYFP and 0.1 μg of plasmid DNA encoding β-arrestin-2 fused to mCherry at its N-terminus. Live-cell imaging was performed 24 h after transfection using a Zeiss Axio Observer microscope with an ApoTome Imaging System, equipped with a P-Lab-Tec S1 heated insert unit. Fluorescence images of living cells were captured at 37°C using AxioVision Rel. 4.6 software. Prior to imaging studies, cells were starved for 30 min in serum-free Opti-MEM medium (Thermo Fisher Scientific) and stained with Hoechst 33342 (5 μg/μl) for 10 min to label nuclei. To study the internalization of WT and mutant M3Rs, cells were incubated with carbachol for 45 min at 37°C. Images were obtained prior to and after carbachol treatment.

Statistics.

Data are expressed as means ± SEM for the indicated number of independent experiments. For comparisons between two groups, the unpaired Student’s t test (two-tailed) was used. Data from concentration-response curves were analyzed using 2-way ANOVA. Tukey’s multiple comparison posthoc tests were performed for statistical comparison at each carbachol concentration. A p value of <0.05 was considered statistically significant.

RESULTS

Generation of Biased Mutant M3R Constructs.

We previously described the generation and initial functional analysis of two M3R-based designer GPCRs (DREADDs) that either preferentially activate G proteins or primarily recruit and activate β-arrestins^{30, 31}. One of these DREADDs, referred to as M3D-Gq by Hu et al.³⁰, was capable of stimulating G_q/11 signaling in an agonist-dependent fashion, but was severely impaired in its ability to recruit β-arrestins. The second DREADD, referred to as Rq(R165L) (rat M3R sequence)³¹, was able to recruit β-arrestins but was deficient in coupling to G_q/11.

Like other mAChR-based DREADDs, these two biased DREADDs contained two single point mutations in the transmembrane (TM) core of the receptor, one in TM helix 3 (TM3) and the other one in TM5 (Y148C and A238G in the rat M3R sequence; Ballesteros-Weinstein amino acid positions: 3.33 and 5.46, respectively)^{33, 34}.

To generate G protein- and β-arrestin-biased mutant M3Rs, we used site-directed mutagenesis to restore the ACh binding site in M3D-Gq and Rq(R165L) by replacing C148 and G238 with the corresponding WT residues (Y148 and A238, respectively). For the sake of simplicity, we refer to the resulting mutant M3Rs as M3R-Gq and M3R-Arr, respectively (Figure 1).

Figure 1.

Schematics of biased M3Rs analyzed in this study. Amino acid numbers refer to amino acid positions in the rat M3R. i3 loop, third intracellular loop of the rat M3R.

Receptor-mediated activation of G_q/11 signaling.

To study the ability of M3R-Gq and M3R-Arr to stimulate G proteins in an agonist-dependent fashion, we transiently transfected COS-7 cells with the two mutant M3R constructs. The WT M3R preferentially activates G proteins of the G_q/11 family³⁵, resulting in elevated intracellular calcium levels ([Ca²⁺]_i)³⁶. For example, Brands et al.³⁶ recently showed that M3R-mediated increases in [Ca²⁺]_i in cultured cells can be effectively blocked by FR900359, a highly selective G_q/11 inhibitor³⁷. In the present study, we measured the ability of carbachol, a stable ACh analog that is structurally closely related to ACh, to promote receptor-mediated increases in [Ca²⁺]_i by using FLIPR technology. Carbachol treatment of M3R-Gq-expressing cells caused robust, concentration-dependent increases in [Ca²⁺]_i (EC₅₀ = 14.5 nM; log EC₅₀ = −7.84 ± 0.14; n=3), indicating that agonist-occupied M3R-Gq receptors couple to G_q/11 with high efficiency (Figure 2A). In contrast, carbachol exposure of M3R-Arr-expressing cells only led to a marginal increase in [Ca²⁺]_i at rather high carbachol concentrations (>10 μM; n=3) (Figure 2A). Figure 2B shows that the M3R-WT receptor mediated elevations in [Ca²⁺]_i with similar carbachol potency and efficacy (E_max) as the M3R-Gq construct.

Figure 2.

Receptor-mediated increases in intracellular calcium levels ([Ca²⁺]_i) and receptor cell surface expression levels. (A, B) COS-7 cells were transiently transfected with the indicated M3R constructs. Carbachol-induced increases in [Ca²⁺]_i were measured using FLIPR technology, as described under “Materials and Methods”. Panel (B) compares M3R-WT and M3R-Gq-mediated [Ca²⁺]_i responses. (C) Relative cell surface expression levels of the indicated M3R constructs transiently expressed in COS-7 cells. A whole-cell ELISA strategy was employed to detect the presence of the extracellular HA-tag present at the N-terminus of all M3R constructs (see “Materials and Methods” for details). Background absorbance at 450 nm measured with vector-transfected cells was set equal to 1. (D) Total cellular protein content per well in samples used to obtain the ELISA data shown in (C). Data are given as means ± SEM from three independent experiments, each performed in triplicate. ***p < 0.001 vs. M3R-Gq (2-way ANOVA followed by Tukey’s multiple comparison test).

To confirm that all receptor constructs were expressed at similar levels on the cell surface, we employed a whole-cell ELISA strategy that can selectively detect the extracellular HA-tag fused to the N-terminus of all receptor constructs^{32, 38}. We found that the M3R-Gq and M3R-Arr mutant receptors were present at similar levels as the M3R-WT receptor on the cell surface (Figure 2C). We confirmed that the samples used for the ELISA studies contained similar amounts of total protein (Figure 2D).

Receptor-mediated β-Arrestin Recruitment.

We next used BRET to examine the ability of M3R-Gq and M3R-Arr to recruit β-arrestins. COS-7 cells were co-transfected with the WT M3R (M3R-WT) and the two mutant M3R constructs containing a C-terminal Renilla luciferase 8 sequence (BRET donor) and β-arrestin-1 or −2 constructs (abbreviated as βarr1 and βarr2, respectively, hereafter) carrying an N-terminal Venus tag (BRET acceptor). The expression levels of the Venus-tagged β-arrestins remained unaffected by the nature of the co-expressed M3R construct (Figure 3A, B). As expected, carbachol treatment of cells expressing the M3R-WT construct resulted in a concentration-dependent recruitment of βarr1 (EC₅₀ = 30.8 μM; log EC₅₀ = −4.51 ± 0.21; n=3) and βarr2 (EC₅₀ = 10.3 μM; log EC₅₀ = −4.99 ± 0.14; n=3), respectively. Under the same experimental conditions, carbachol stimulation of M3R-Gq-expressing cells failed to recruit either of the two β-arrestins (Figure 3C, D). In contrast, carbachol treatment of M3R-Arr-expressing cells resulted in a robust recruitment of βarr2 (but not of βarr1; Figure 3C, D). While carbachol stimulated M3R-WT/βarr2 and M3R-Arr/βarr2 (EC₅₀ = 12.4 μM; log EC₅₀ = −4.91 ± 0.26; n=3) interactions within a similar concentration range, the carbachol E_max value for M3R-Arr-mediated βarr2 recruitment was reduced by ~50%, as compared to the M3R-WT/βarr2 interaction (Figure 3D).

Figure 3.

Receptor-mediated recruitment of β-arrestins. Changes in BRET were calculated in COS-7 cells transfected with Venus-tagged βarr1 (A, C) or βarr2 (B, D) constructs and the indicated M3R constructs following treatment with carbachol for 25 min. Expression of the Venus-tagged βarr1 (A) or βarr2 (B) was measured using the fluorescence signal at 535 nm prior to carbachol treatment. Baseline BRET values were similar for cells transfected with the βarr1 plasmid and the different receptor constructs: M3R-WT, 1.42±0.04; M3R-Gq, 1.34±0.03; M3R-arr, 1.36±0.05 (n=3). The same was true for cells transfected with the βarr2 plasmid and the different M3R constructs: M3R-WT, 1.44±0.03; M3-GqR, 1.31±0.01; M3R-arr, 1.41±0.03 (n=3). Data were analyzed using non-linear regression fit in GraphPad Prism 10. Data are presented as means ± SEM from three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 vs. M3R-WT (2-way ANOVA followed by Tukey’s multiple comparison test).

Agonist-induced Receptor Internalization.

In most cell types, recruitment of β-arrestins to activated GPCRs promotes receptor internalization. To investigate the ability of the different M3R constructs to undergo agonist-induced internalization, we added an EYFP tag to the C-terminus of M3R-WT, M3R-Gq, and M3R-Arr. Following the transient expression of the three receptors in HEK293 cells, we monitored carbachol-dependent receptor internalization in live cells via fluorescence microscopy. Prior to carbachol simulation, all three receptors were predominantly expressed on the cell surface (Figure 4, “no drug”). In contrast, carbachol treatment promoted the internalization of M3R-WT in a concentration-dependent fashion (Figure 4). Cells expressing the M3R-Arr construct showed a very similar pattern of receptor internalization (Figure 4). In contrast, carbachol failed to induce the internalization the M3R-Gq construct (Figure 4).

Figure 4.

Receptor internalization assays. HEK293 cells were transiently transfected with plasmids coding for M3R-WT, M3R-Gq, or M3R-Arr. A EYFP (yellow) sequence was fused to the C-terminus of all receptor constructs. Twenty-four h after transfection, live-cell imaging studies was carried out using a Zeiss Axio Observer microscope. Cell nuclei were counter-stained with Hoechst 33324 (blue). To monitor receptor internalization, cells were incubated for 45 min with increasing concentrations of carbachol. Images were acquired before (no drug) and after carbachol exposure. Representative images are shown. At least two independent experiments were carried out, yielding similar results. Scale bar (upper left panel), 10 μm.

Taken together, these data indicate that the M3R-Gq construct represents a mutant M3R that can activate G proteins but is unable to recruit β-arrestins. The M3R-Arr receptor shows the opposite functional profile. This mutant M3R fails to activate G_q/11 but is able to recruit β-arrestins (βarr2).

Structural Considerations.

The point mutation in the M3R-Arr construct involves the replacement of R165 (Ballesteros-Weinstein number 3.50) with leucine. R165 is a highly conserved arginine within the DRY/ERY motif, which stabilizes the inactive conformation of most class A GPCRs. In G_q- and G_s-coupled receptors, R3.50 also engages in π-cation interactions with a tyrosine side chain (Y243 in human Gα_q) in the G protein’s C-terminal helix (Figure 5). Disruption of this interaction by the R165L mutation most likely destabilizes interactions with the G protein while sparing interactions with β-arrestins, accounting for the intrinsic bias of the M3R-Arr construct.

Figure 5.

Critical site of interaction between the M3R and Gα_q. The M3R (orange) interacts with the C-terminus of the Gα_q in part via π-cation interactions between receptor R165 (R3.50 according to the Ballesteros-Weinstein numbering system) and Y243 on Gα_q (PDB ID: 8E9Z). Disruption of this interaction by the R165L point mutation most likely accounts for the intrinsic β-arrestin bias of the M3R-Arr construct. TM3, transmembrane helix 3; TM7, transmembrane helix 7.

The M3R-Gq construct lacks most of the third intracellular loop (A274 to K469), which contains phosphorylation sites important for β-arrestin recruitment^{14, 39}. In total, this deletion leads to the removal of 31 serine and 14 threonine residues. In addition, the M3R-Gq mutant receptor includes the C542A, C546S, C560S, and C562S point mutations. C542 and C546 are located in TM7 and directly contact TM6. As a result, alanine (or serine) substitution of these two cysteine residues may affect local conformations of these helices in a way that interferes with β-arrestin coupling. However, experimental structures of M3R-β-arrestin complexes are required to test this hypothesis. It is likely that the approaches used to create M3R-Arr and M3R-Gq can be generalized to other family A GPCRs, since the G protein interaction with R3.50 and β-arrestin interactions with phosphorylation sites are broadly conserved^40–43.

DISCUSSION

In this study, we constructed two mutant M3Rs that can be used as novel tools to study the cellular and physiological roles of M3R-dependent G protein activation vs. M3R-mediated β-arrestin recruitment and signaling. Specifically, we generated two biased mutant M3Rs referred to as M3R-Gq and M3R-Arr. Following agonist activation, the M3R-Gq receptor activates G proteins of the G_q/11 family but fails to recruit β-arrestins and to undergo internalization, at least in cultured COS-7 and HEK293 cells. Ligand activation of the M3R-Arr receptor promotes the recruitment of βarr2 and stimulates receptor internalization but is severely impaired in its ability to activate G_q/11-type G proteins.

Expression of the two mutant M3Rs in physiologically relevant cell types in vitro should shed light on the signaling pathways and functional outcomes triggered by M3R-mediated G protein activation versus M3R-dependent β-arrestin recruitment. Moreover, generation of knockin mice involving the replacement of the WT M3R sequence with either the M3R-Gq or M3R-Arr sequence should yield mouse models useful for exploring the in vivo roles of M3R-dependent G protein vs. β-arrestin signaling. The novel information gained from the analysis of such mouse strains could guide the development of biased M3R agonists for various pathophysiological disorders including metabolic diseases characterized by deficits in glucose and energy homeostasis^{12, 44, 45}.

Interestingly, Tobin and coworkers generated a knockin mouse strain in which the endogenous M3R was replaced by a phosphorylation-deficient version of the M3R (M3R_phos-neg)^{14, 39}. This mutant M3R was impaired in its ability to recruit β-arrestins but retained the ability to stimulate G_q/11 signaling in a WT M3R-like fashion^{14, 39}. Thus, the M3R_phos-neg receptor represents a G protein-biased mutant GPCR. Interestingly, phenotyping studies showed that M3R_phos-neg knockin mice displayed deficits in hippocampal learning and memory¹⁴. This finding suggests that muscarinic agonists that strongly promote M3R/β-arrestin interactions may prove clinically useful to improve cognition in certain disorders of the CNS.

The M3R_phos-neg knock-in mice also displayed impaired glucose tolerance that correlated with deficits in M3R-mediated stimulation of insulin secretion³⁹. This observation suggests that M3R/β-arrestin interactions contribute to the ability of muscarinic agonists to augment glucose-induced insulin release. Clearly, this finding is highly relevant for the potential clinical use of muscarinic agonists as antidiabetic drugs^{12, 44}.

It should be noted in this context that agonist activation of the M3R_phos-neg receptor strongly reduced but did not abolish the recruitment of β-arrestins^{14, 39}. In contrast, the M3R-Gq receptor described here was completely devoid of receptor/β-arrestin interactions. For this reason, the analysis of M3R-Gq knockin mice may reveal novel phenotypes that were not observed with M3Rp_hos-neg knock-in mice.

Moreover, ligand-activated M3R-Gq receptors were unable to undergo agonist-induced internalization, raising the possibility that the lack of M3R-Gq/β-arrestin interactions is responsible for the defect in M3R-Gq internalization. In a previous study, Lee et al.⁴⁶ used a dominant-negative mutant version of βarr1 to disrupt arrestin-dependent receptor internalization in HEK-tsA201 cells transiently expressing the M3R. Under these experimental conditions, carbachol-induced M3R internalization remained unaffected, suggesting that M3R internalization occurs via an arrestin-independent pathway. In seeming contrast, a related study showed that co-expression of the M3R with dominant negative mutants of βarr1 in HEK-293 tsA201 cells led to pronounced reductions in carbachol-induced M3R sequestration¹⁶, indicative of a major role of β-arrestins in mediating M3R internalization. Most likely, these discrepant findings were caused by differences in experimental conditions including differences in receptor and β-arrestin expression levels and/or agonist concentrations and incubation times. Clearly, more systematic studies are needed to clarify the precise roles of the two β-arrestins in mediating M3R internalization in different cell types.

As reviewed recently⁴⁷, several biased muscarinic agonists that preferentially recruit either G proteins or β-arrestins have been identified during the past decades. Such agents have been found for all mAChR subtypes except for the M5R⁴⁷. For example, Pera et al.⁴⁸ showed that agonist (methacholine) activation of M3Rs expressed by immortalized human airway smooth muscle (ASM) cells leads to an increase in [Ca²⁺]_i, as expected. Interestingly, the magnitude of this response (E_max) was significantly reduced following siRNA-mediated knockdown of βarr1 expression. In agreement with this observation, methacholine-induced bronchoconstriction was impaired in βarr1-deficient mice in vivo⁴⁸. These data indicate that M3R-dependent βarr1 recruitment/signaling contributes to ASM contractions triggered by M3R activation, suggesting that the clinical use of muscarinic agonists unable to recruit βarr1 to the M3R may reduce the risk of bronchoconstriction.

Determining the exact structural basis underlying the intrinsic bias of M3R-Gq and M3R-Arr will require structural studies of complexes of the M3R with G proteins and β-arrestins. However, the nature of the mutations that are present in the two mutant M3R receptors suggests likely explanations, namely attenuation of favorable interactions between receptor and G protein in the M3R-Arr mutant, and ablation of multiple phosphorylation sites in the M3R-Gq receptor, respectively. It is likely that similar approaches can be applied to other class A GPCRs, since the interaction of R3.50 of the receptor with the tyrosine side chain present in the C-terminal helix of Gα (G_q/11 and G_s) is broadly conserved. In agreement with the key role of R3.50 in promoting GPCR/G protein interactions, Donthamsetti et al.⁴⁹ recently reported the development a β-arrestin-biased mutant dopamine D₂ receptor that contained the R3.50L point mutation, similar to M3R-Arr.

Interestingly, the high-resolution crystal structure of the human δ-opioid receptor (δ-OR) revealed the key role of a sodium ion in the allosteric control of δ-OR function⁵⁰. Mutational modification of specific sodium-anchoring residues of the δ-OR resulted in marked changes in the functional selectivity of the δ-OR by augmenting constitutive β-arrestin-mediated signalling. This observation suggests that GPCR coupling bias can be achieved via diverse structural mechanisms.

β-arrestin interactions with phospho-sites are known to be critical for efficient β-arrestin recruitment by most GPCRs, and deletion of phosphorylatable serine and threonine in the receptor usually impairs the recruitment of β-arrestins. G protein bias has been described for several other mutant GPCRs. For example, introduction of the Y219A point mutation (position Y5.58 according to the Ballesteros-Weinstein numbering system) into TM5 of the β₂-adrenergic receptor (β₂-AR) resulted in a mutant receptor that was devoid of β-arrestin recruitment but retained the ability to couple to G proteins⁵¹. Mechanistic studies demonstrated that the Y219A point mutation interfered with β-arrestin recruitment by reducing GRK-mediated phosphorylation of the β₂-AR, indicative of the importance of GRKs in determining GPCR signaling outcomes.

In this context, an example of clinically relevant GPCR mutations that result in G protein bias deserves mentioning. Certain mutations in the C-terminal tail of the C-X-C chemokine receptor type 4 (CXCR4) cause WHIM (Warts, Hypogammaglobulinaemia, Infections and Myelokathexis) syndrome, a rare form of a congenital immunodeficiency^{52, 53}. Most of these disease-causing mutations lead to truncated receptor proteins that show greatly increased signaling strength due to impaired CXCR4 internalization. As discussed in detail recently^{52, 53}, several of these truncated receptors have been shown to be impaired in their ability to recruit β-arrestins, providing a possible explanation for the increased G protein coupling efficiency displayed by these mutant receptors.

CONCLUSIONS

The availability of functionally biased GPCRs should facilitate the development of novel classes of GPCR ligands that may prove clinically superior, as compared to non-biased GPCR agonists. For example, in the case of the M3R, functionally biased ligands may prove beneficial for the treatment of certain disorders of the CNS and deficits in glucose homeostasis caused by insufficient insulin secretion.

ABBREVIATIONS

ACh

acetylcholine

BRET

Bioluminescence Resonance Energy Transfer

M3R

M₃ muscarinic receptor

mAChR

muscarinic acetylcholine receptor

GPCR

G protein-coupled receptor

WT

wild type

βarr1

β-arrestin-1

βarr2

β-arrestin-2

DREADD

designer receptor exclusively activated by a designer drug

FBS

fetal bovine serum

[Ca²⁺]_i

intracellular calcium concentration

ASM

airway smooth muscle