Allosteric modulators restore orthosteric agonist binding to mutated CB₁ receptors

Abstract

Objectives

To determine if diminished orthosteric agonist binding due to mutations in extracellular loops 1 or 2 of the cannabinoid receptor 1 (CB₁) can be overcome by an allosteric modulator and restore agonist binding.

Methods

Binding assays were performed using a range of concentrations of orthosteric compound, in the presence or absence of a set concentration of the allosteric modulator PSNCBAM-1 to determine the EC₅₀ in its absence or presence.

Key findings

Single mutations in extracellular loop 1 or 2 of CB₁ showed weak or no binding of agonist CP55940 to the receptor. Interestingly, upon addition of the allosteric modulator PSNCBAM-1, this binding was restored typically to wild-type CB₁ levels. In a few cases, the allosteric modulator ORG27569 was compared with PSNCBAM-1 for CP55940 binding and it also restored binding. Further, wild-type levels of inverse agonist bound the CB₁ mutants in the absence of modulator, suggesting the mutants were originally folded like the wild type.

Conclusions

Based on our findings, we provide evidence of a therapeutic application for allosteric modulators in situations where a mutation in the receptor may hinder its function. By utilizing allosteric modulators, restoration of orthosteric binding may be possible.

Introduction

The cannabinoid receptor 1 (CB₁) and cannabinoid receptor 2 (CB₂) belong to the group of rhodopsin-like G protein-coupled receptors (GPCRs).^[1] CB₁ is highly expressed in the central nervous system, particularly in the brain, while CB₂ is predominantly expressed in peripheral tissues but can also be found in the brain.^[2,3] Ligands which bind the orthosteric site on CB₁ include the endogenous cannabinoids anandamide and 2-arachidonyl glycerol, the plant-derived compound Δ⁹-tetrahydrocannabinol (Δ9-THC), the synthetic agonist CP55940, and the synthetic inverse agonist SR141716A.^[4–10]

Allosteric modulators bind a site that is topographically distinct from the orthosteric site.^[11,12] Allosteric modulators may affect the affinity and/or the efficacy of an orthosteric agonist, and/or may also elicit signalling on their own.^[13] Allosteric modulators for CB₁ include ORG27569, an indole-2-carboxamide, and PSNCBAM-1, a diarylurea.^[12,14] These compounds are specific for CB₁ and do not bind CB_2.^[14,15] They hold promise as potential scaffolds for the development of treatments for pain and neurodegenerative disorders.^[16] PSNCBAM-1 increased CP55940 binding and reduced binding of the inverse agonist SR141716A, consistent with it promoting an active conformation of the receptor.^[14] ORG27569 functioned similarly.^[12] In addition, ORG27569 treatment of cells expressing CB₁ has been shown to induce ERK phosphorylation via the β-arrestin 1 pathway.^[17] Allosteric modulation of receptors including GPCRs is considered a valuable approach to treat some disorders.^[18] For example, Cinacalcet is a positive allosteric modulator (PAM) of the calcium-sensing receptor and activates the receptor.^[19]

Here, we sought to determine if allosteric modulation of CB₁ could restore orthosteric agonist binding as a first step to addressing issues that may result from mutation of this receptor. Residues of extracellular loop 1 (EC1) of CB₁ seem to be important for agonist binding, but not for inverse agonist binding.^[20] This is based on evidence that includes insertions of amino acids at positions 181 or 188 that caused a complete loss of CP55940 binding and alanine substitutions of single residues in EC1 of CB₁ that caused a decrease in CP55940 binding, but inverse agonists can still bind comparably to wild-type.^[20] Similar results were seen upon mutation of some residues in extracellular loop 2 of CB₁.^[21,22] This further suggests that these mutant CB₁ receptors were largely folded since an inverse agonist could still bind them.

Mutations of a receptor may result in a compromised receptor structure and/or coupling mechanism and/or a loss in ability to respond to ligands. Allosteric modulators may be able to overcome some types of dysfunction by restoring the orthosteric binding site and generating a conformation that will promote ligand binding. In addition, mutated M₄ muscarinic receptors showed reduced responses to acetylcholine when compared to wild type, but the binding and response improved once allosteric modulators were introduced.^[23]

In this study, we have generated single residue mutations at residues in the EC1 and EC2 regions of CB₁. The amino acids used for substitution were chosen either because they were the corresponding residue in CB_2, or an alanine residue since previous studies have shown that alanine mutations at these sites decreased binding of the agonist CP55940.^[20] We observed that these mutations caused a marked decrease in binding of the orthosteric agonist CP55940, but upon addition of the allosteric compound PSNCBAM-1, binding was recovered. This suggests that at least some mutations of CB₁ can be overcome by allosteric modulator treatment.

Materials and Methods

Materials

Cold CP55940, SR141716A, ORG27569 and PSNCBAM-1 were purchased from Tocris (Minneapolis, MN). Methanandamide is from Cayman Chemical (Ann Arbor, MI, USA). [³H] CP55940 (150.2 Ci/mmol) and [³H] SR141716A (56 Ci/mmol) were purchased from Perkin Elmer (Boston, MA, USA).

Mutagenesis

Human CB₁ receptor cDNA in the mammalian expression vector pcDNA3.1 was used for the template for site-directed mutagenesis via the PCR-based QuikChange system. Mutations were verified by DNA sequencing. Single amino acid changes were made in the extracellular loop 1 (EC1) or extracellular loop 2 (EC2) region of CB₁ (Table 1).

Table 1.

CB₁ receptor mutation regions

CB1 receptor	Mutation region	Codon change involved
WT
R182G	EC1	CGC → GGC
K183V	EC1	AAA → GTA
D184A	EC1	GAT → GCT
H270A	EC2	CAC → GCC
I271A	EC2	ATT → GCT

CB₁ receptor mutation regions: The mutation name is given, along with a representation on the GPCR of its location, and the codon change involved in the mutation.

Receptor expression and membrane preparation

Human embryonic kidney 293T (HEK293T) cells were grown and seeded in Dulbecco's modified Eagle's medium with 3.5 mg/ml glucose and 10% foetal bovine serum at 37°C and 5% CO₂. To express the wild-type or mutated CB₁ receptors, HEK293T cells were seeded at 1 000 000 cells/100-mm plate. The following day, cells were transfected via the calcium phosphate method with 10 µg of pcDNA3.1 containing mutant or wild-type human CB₁ receptor.^[24] Membranes of the transfected cells were prepared as previously described, using cells harvested 21 h after transfection.^[25] Briefly, cells were harvested by washing with phosphate buffered saline (PBS: 137 mm NaCl, 2.7 mm KCl, 10 mm Na₂HPO₄, 1.8 mm KH₂PO₄, pH 7). Cells were then resuspended in PBS with protease inhibitor cocktail (4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride, pepstatin A, E-64, bestatin hydrochloride, leupeptin hemisulfate salt and aprotinin; Sigma-Aldrich, St. Louis, MO), lysed using nitrogen cavitation at 750 psi for 5 min, then centrifuged at 4°C for 10 min at 500g. The supernatant was further spun at 4°C for 45 min at 100 000g. The pellet containing the membrane was resuspended in TME buffer (25 mm Tris-HCl, 5 mm MgCl₂ and 1 mm EDTA, pH 7.4) with 7% w/v sucrose.

Ligand binding assays

[³H]CP55940 assays: To perform competition binding assays, approximately 5 μg of membrane preparation expressing the wild-type or mutant CB₁ receptors was incubated with 1.5 nm radioligand tracer and nine concentrations of unlabelled compound (typically ranging from 100 pm and 10 µm) with or without the addition of 1 µm of allosteric modulator PSNCBAM-1. Some receptors were also tested for binding with nine concentrations of either allosteric modulator ORG27569 or PSNCBAM-1 using 0.5 nm [³H]CP55940 as a tracer. The tracer and the cold orthosteric compound were homologous except when the agonist methanandamide is used as the cold compound and then the tracer is [³H]CP55940. Nonspecific binding was established by treating with 10 µm of unlabelled compound.

[³H]SR141716A assays: Approximately, 5 µg of membrane preparation expressing the wild-type or mutant CB₁ receptors was incubated with nine concentrations of unlabelled SR141716A with 2 nm [³H]SR141716A, an inverse agonist of CB₁. Nonspecific binding was established by treating with 1 µm of unlabelled SR141716A.

The membranes were incubated at 30°C for 60 min, and the reaction was terminated with the addition of 300 µl of TME buffer with 5% BSA. The mixture was harvested by filtration through a Brandel cell harvester with Whatman GF/C filter paper. Liquid scintillation counting was used to measure bound radioactivity. All ligand binding assays were carried out in duplicate, and at least three independent experiments were performed.

Data analysis

Data are presented in Table 2 as the mean EC₅₀ values and the corresponding 95% confidence limits, and in the graphs as the mean ± SE (error bars). The data collected were subjected to nonlinear regression, fitted to a log (dose) vs response curve to determine the EC₅₀ using Prism 7.02 (Graphpad Software, LaJolla, CA, USA). Comparison was done using analysis of variance (ANOVA) for agonist binding plus and minus the allosteric modulator for the mutant receptors followed by Bonferroni's post hoc test and P < 0.05 was considered statistically significant.

Table 2.

Addition of PSNCBAM-1 can improve agonist binding

CB1 receptor	EC50 (nm)	EC50 (nm)	EC50 (nm)
	CP55940	CP55940 + 1 μm PSNCBAM-1	SR141716A
WT	2.3 (0.96–8.6)	1.1 (0.81–1.6)	4.8 (2.4–10)
R182G	NB	1.6 (0.82–3.1)	2.3 (1.5–3.6)
K183V	6.6 (2.6–18)	0.9 (0.60–1.5)	3.4 (2.1–5.4)
D184A	NB	1.0 (0.64–1.5)	ND
H270A	NB	2.0 (0.72–4.7)	ND
I271A	NB	NB	ND

NB: No binding detected up to 1 μm of CP55940, ND: not determined.

Addition of PSNCBAM-1 can improve agonist binding. Binding parameters given for mutations tested are given as EC₅₀ values with 95% confidence intervals in parentheses.

Results

The effect of PSNCBAM-1 on agonist binding

Five individual amino acid mutations were generated singly in either EC1 or EC2 of CB₁ (Table 1). These involved an alanine substitution or the amino acid that is in the equivalent position in CB₂. These mutations were thought to impact agonist CP55940 binding, but not inverse agonist SR141716A binding, indicating that the mutation probably did not impact folding of the receptor since the orthosteric site (the main active site of the receptor) seems to bind some orthosteric compounds.^[20,21] Following transfection and membrane preparation, and unlike the wild-type CB₁ (Figure 1a), three of the mutant receptors (R182G, D184A and H270A) did not bind the agonist, CP55940, at a detectable level in the absence of allosteric modulator (Figure 1b, 1d and 1e) and a fourth mutant receptor, K183V, exhibited somewhat weaker binding (EC₅₀ = 6.6 nm) in its absence (Figure 1c). However, these four mutant receptors showed wild-type levels (Figure 1f) of CP55940 (EC₅₀ = 0.93 to 1.6 nm) binding upon treatment with the allosteric modulator, PSNCBAM-1 (Figure 1g, 1h, 1i, 1j). The I271A mutant receptor was devoid of agonist CP55940 binding, regardless of the presence of the allosteric modulator (data not shown). Furthermore, to demonstrate that folding was largely unimpacted, inverse agonist SR141716A binding was not affected in the mutants R182G and K183V, with EC₅₀ values of 2.3 nm and 3.4 nm, respectively, compared to the wild-type CB₁ EC₅₀ of 4.8 nm (Figure 2). The EC₅₀ values for these data are summarized in Table 2.

Figure 1.

Restoration of CP55940 binding to CB₁ receptor mutants with allosteric modulator (1 μm PSNCBAM-1) treatment. (a) wild type (b) R182G (c) K183V (d) D184A (e) H270A CB₁ receptors in homologous competition assays with CP55940 (●). The lack of binding detected is evident with R182G, D184A and H270A. While K183V did bind CP55940, the EC₅₀ = 6.6 nm, indicates it is weaker than in the presence of allosteric modulator. CP55940 homologous competition in the presence of 1 μm PSNCBAM-1 with the (f) wild type (g) R182G (h) K183V (i) D184A (j) H270A CB₁ receptors caused restoration of agonist binding (■). Results were determined from at least three experiments performed in duplicate except when no binding is observed and then a representative assay is given. Data are presented as the mean ± SE (error bars).

Figure 2.

Binding of SR141716A by the mutant receptors was comparable to wild-type CB₁. Homologous competition binding assays of (a) wild-type (b) R182G and (c) K183V CB₁ receptors with SR141716A. Results were determined from at least three experiments performed in duplicate. Data are presented as the mean ± SE (error bars).

The D184A mutant receptor form was tested for binding of methanandamide, with [³H] CP55940 as a tracer. This was to determine if an endogenous agonist, made more stable with the addition of a methyl group, also exhibited the same affect. As expected, the agonist alone did not bind the D184A receptor (Figure 3a). However, upon addition of 1 µm PSNCBAM-1, to membranes containing the D184A mutant receptor, methanandamide binding was restored (Figure 3b).

Figure 3.

Restoration of methandamide binding to the D184 CB₁ receptor mutant with allosteric modulator (1 μm PSNCBAM-1) treatment. (a) Binding of methanandamide to the CB₁ D184A receptor in the absence (●) and the (b) presence (■) of 1 µm PSNCBAM-1. [³H] CP55940 was used as the tracer, in the presence of varying amounts of methanandamide. Results were determined from at least three experiments performed in duplicate, except when no binding is observed and then a representative assay is given. Data are presented as the mean ± SE (error bars).

ORG27569 also restores agonist binding

To ensure that restoration of agonist binding was not allosteric modulator specific, another modulator, ORG27569, was also tested with the R182G mutant and separately with the K183V mutant receptors. Dose–response experiments were performed in the presence of varying concentrations of either PSNCBAM-1 or ORG27569 to determine if these enhanced binding of agonist [³H]CP55940. Indeed, both allosteric modulators enhanced the binding of [³H]CP55940 in a dose-dependent manner (Figure 4).

Figure 4.

Allosteric modulators ORG27569 and PSNCBAM-1 restore agonist binding. Binding of ORG27569 (○) and PSNCBAM-1 (△) to CB₁ receptor variants R182G (a) and K183V (b), with varying concentrations of allosteric compound. Results were determined from at least three experiments performed in duplicate. Data are presented as the mean ± SE (error bars).

Discussion

The allosteric modulator PSNCBAM-1 has the ability to restore agonist CP55940 binding near wild-type CB₁ levels, that is lost due to mutations in extracellular loops 1 and 2 of CB₁. Future work will include determining how PSNCBAM-1 is able to restore the binding. It may act as a chaperone, or alternatively, it may stabilize the conformation of the mutated receptor in the active form which promotes agonist binding.

With the allosteric modulator the affinity for CP55940 is restored like wild-type CB₁. If the mutant CB₁ with the modulator was not activated to the same extent, and coupled, it would have a different value for the EC₅₀ and different from wild type. However, they are the same (see Table 2). The concentration required to bind ligand to 50% of receptor, relative to the wild-type receptor, reflects the activation and coupling levels, as well as the thermal stability of the receptor, and the cellular location where ligand binds.^[26,27]

These mutants impacted the binding of orthosteric agonists such as CP55940 and methanandamide. However, the I271A mutation did not have restored binding of agonist CP55940 upon co-treatment with the allosteric modulator PSNCBAM-1, which may indicate the importance of this residue in agonist CP55940 binding. Indeed, previous studies support this idea; I271 forms part of the ligand binding site for CP55940 and interacts directly with the ligand.^[21] Nonetheless, the binding of the inverse agonist SR141716A is unaffected either shown here or in previous studies.^[20,22] Further, H270A and I271A were tested for their effects on SR141716A binding in Bertalovitz et al.^[22] and were found to be comparable to wild-type CB₁, while D184A was tested for its effects on SR141716A binding in Murphy and Kendall^[20] and retained binding.

Wild-type levels of inverse agonist binding strictly speak only to the folding of the orthosteric site (i.e. active site) which binds agonists and inverse agonists of CB₁; they compete with each other in binding assays. However, it also suggests that folding overall is comparable to wild-type CB₁, and the receptor is not likely to be aberrantly transported in the endoplasmic reticulum. Indeed, the observation that the inverse agonist SR141716A binds mutant receptors with similar potency to wild-type CB₁ also implies that cell expression overall of CB₁ is not affected by these mutations.

These results support the use and development of allosteric modulators also for other drug targets, since they suggest that allosteric modulators may restore the function of a deleterious mutation. Allosteric modulators could be useful in genetic diseases in which a receptor has little or no activity, in order to restore orthosteric binding to that receptor. This restoration of orthosteric binding includes the endogenous ligand, thus reestablishing the function of the receptor.

The concept of the involvement of an allosteric modulator to restore function to a mutated receptor has also been proposed in the context of the α1R271Q/L glycine receptor, which is a mutated receptor that causes hyperekplexia. Upon replacement of the mutated R271Q/L residue on the α1 subunit of the glycine receptor through the generation of a 12 amino acid chimeric receptor with the homologous amino acid sequence from the β subunit of the glycine receptor, the function of the receptor was restored. In that case, it was further suggested that the local conformation was modified, moving the receptor off utilization of an allosteric pathway that involved residue 271. New drug development to target this condition could focus on the allosteric site, which would allow for a conformational shift that could result in a functional receptor.^[28]

Mutational studies on the dopamine receptor suggest that some mutations cause a weaker affinity of dopamine for the receptor, but the allosteric modulator DETQ in addition to dopamine was able to induce an increased maximum response when compared with dopamine alone.^[29] The restoration of the function of a mutated receptor by an allosteric modulator was also shown in M₄ muscarinic receptors which were mutated such that they were rendered functionally inactive. The M₄ muscarinic receptor allosteric modulator LY2033298 was able to rescue the function of the orthosteric ligand, acetylcholine, in ERK1/2 phosphorylation assays, further demonstrating that an allosteric compound can rescue an otherwise inactive receptor.^[23]

Restoration of CB₁ agonist binding by an allosteric modulator could restore function to an otherwise nonfunctional mutant receptor. The importance of a functional CB₁ receptor is demonstrated in CB₁ knockout mice experiments; while the CB₁ knockout mice were fertile with no physiological abnormalities, they had an increased mortality rate when compared with wild-type mice. Also, the animals exhibited abnormal behaviours such as a decrease in pain sensitivity and a decrease in locomotor activity.^[30] For example, allosteric modulators may enhance the potency of the CB₁ splice variant, CB₁A, which is shorter than CB₁ by 61 amino acids and differs from CB₁ in its first 28 amino acids. The prevalence of the CB₁A splice variant is up to 20% of CB₁ in certain tissues.^[31] This splice variant exhibits the same properties as CB₁, but to a slightly lesser extent.^[32] An allosteric modulator may further enhance the activity of these splice variants, thus contributing to the overall activity of the receptors in the body.

By studying mutations in regions which were already known to have an influence on agonist binding, we provide proof-of-principle that an allosteric modulator has the ability to restore binding properties to these mutated receptors. Further studies would apply this concept to naturally occurring mutations in the human population, such as the mutations present in a CB₁ gene sequence from an epilepsy patient, which contains three amino acid differences, two in the third transmembrane domain (F200L and I216V) and one in the fourth transmembrane domain (V246A).^[33]

Conclusions

The capabilities of allosteric modulators appear to extend beyond their ability to fine-tune endogenous orthosteric binding and signalling. Here, we demonstrate that allosteric modulators may also be capable of rehabilitating a receptor with an otherwise deleterious mutation.

Declarations

Conflict of interest

The authors declare they have no conflict of interest to disclose.

Acknowledgements

This work was supported in part by National Institutes of Health Grant DA039942 (to D.A.K.).