GPCRs (G-protein-coupled receptors) exist in a spontaneous equilibrium between active and inactive conformations that are stabilized by agonists and inverse agonists respectively. Because ligand binding of agonists and inverse agonists often occurs in a competitive manner, one can assume an overlap between both binding sites. Only a few studies report mutations in GPCRs that convert receptor blockers into agonists by unknown mechanisms. Taking advantage of a genetically modified yeast strain, we screened libraries of mutant M3Rs {M3 mAChRs [muscarinic ACh (acetylcholine) receptors)]} and identified 13 mutants which could be activated by atropine (EC50 0.3–10 μM), an inverse agonist on wild-type M3R. Many of the mutations sensitizing M3R to atropine activation were located at the junction of intracellular loop 3 and helix 6, a region known to be involved in G-protein coupling. In addition to atropine, the pharmacological switch was found for other M3R blockers such as scopolamine, pirenzepine and oxybutynine. However, atropine functions as an agonist on the mutant M3R only when expressed in yeast, but not in mammalian COS-7 cells, although high-affinity ligand binding was comparable in both expression systems. Interestingly, we found that atropine still blocks carbachol-induced activation of the M3R mutants in the yeast expression system by binding at the high-affinity-binding site (Ki ∼10 nM). Our results indicate that blocker-to-agonist converting mutations enable atropine to function as both agonist and antagonist by interaction with two functionally distinct binding sites.

INTRODUCTION

GPCRs (G-protein-coupled receptors) recognize an extraordinary spectrum of chemically different natural and synthetic compounds. Residues within the seven TMDs (transmembrane domains) and extracellular loops (ECLs) 1–3 form the ligand-binding pocket of most rhodopsin-like GPCRs. The bundle-like structure of the transmembrane core is maintained only by hydrophobic interactions and by a few highly conserved determinants. Multiple combinations of other residues assure evolutionary adaptability towards almost any chemical structure that may serve as a ligand. However, even at the level of an individual GPCR, several chemically distantly-related compounds can bind to and activate the receptor. For example, in addition to ACh (acetylcholine), the piperidine alkaloid arecoline and the imidazole alkaloid pilocarpine are agonists of mAChRs (muscarinic ACh receptors). Antagonists or inverse agonists of mAChRs are even more diverse and include the tropane alkaloids atropine and scopolamine, the pyrido-benzodiazepine derivative pirenzepine, the piperidine derivate 4-DAMP (4-diphenylacetoxy-N-methylpiperidine methobromide) and peptides [1,2]. In view of such structural diversity of ligands, what determines the pharmacological property of a ligand to be an agonist, antagonist or inverse agonist?

It is generally accepted that agonist-induced conformational changes, involving mainly the TMD regions, lead to G-protein activation [3]. However, previous studies have suggested that TMD movements are not only caused by agonist and partial agonist binding [4,5], but also during antagonist and inverse agonist binding [68]. Therefore multi-state models have been invoked to explain the complex behaviour of GPCRs in the presence of agonists, antagonists and other binding partners [9]. To explain the molecular mechanism that determines the agonistic or partial agonistic properties of a ligand, it has been hypothesized for the β2 adrenergic receptor that differences in the interaction of the aromatic ring of catecholamines with a conserved cluster of aromatic residues in TMD6 play an important role in triggering full or partial receptor activation [10]. This already indicates that the direct interplay of the receptor with the ligand may determine the pharmacological properties of a ligand. However, not only changes in the chemical structure of the ligand, but also mutations of the receptor can increase agonistic activity [11] and even convert an antagonist/inverse agonist into an agonist [1214]. This has created the possibility of ‘designing’ GPCRs for special purposes, e.g. as biosensors or to help study distinct GPCR signalling pathways. As a first step towards this goal, so-called RASSLs (receptors activated solely by synthetic ligands) were developed [15]. RASSLs are insensitive to their native agonists; but synthetic ligands exhibit either similar, lower or higher agonist efficacy than demonstrated following binding to wt (wild-type) receptors. To date, most mutations converting the pharmacological properties of a ligand were accidentally identified by site-directed mutagenesis approaches. Although classical site-directed mutagenesis and alanine-scanning approaches have resulted in many important insights into the molecular mechanisms governing GPCR function, these strategies are limited by the use of a relatively small number of mutant proteins and combinations of mutations are rarely studied. The combination of random mutagenesis with a functional screening assay in yeast has proven to be a powerful tool and has provided insights into the structure and function of many GPCRs [1618]. This approach allows for screening of several hundred thousand mutant receptors in parallel [19,20]. Recently, this random mutagenesis approach was used to generate a mAChR that is not activated by the natural ligand, but by a pharmacologically inert substance [21].

In this present study, we screened libraries of randomly mutated M3Rs (M3 mAChRs) in a yeast expression system for mutations that convert the muscarinic antagonist/inverse agonist atropine into an agonist. Screening of more than 10 million clones led to the identification of 13 atropine-activatable M3R mutants, all containing two or more missense mutations. Interestingly, most mutations were located at the ICL3 (intracellular loop 3)/TMD6 transition and far from the putative ligand-binding site. None of the individual missense mutations enabled atropine to act as a full agonist; only combinations of mutations were effective, highlighting the usefulness of a combinatorial mutagenesis approach. Strikingly, antagonist-to-agonist switching was observed not only for atropine, but also for all other M3R ligands tested. We provide clear evidence that atropine-sensitizing mutations in M3Rs generate a new binding site where muscarinic blockers can act as agonists.

EXPERIMENTAL

Materials

CCh (carbachol), atropine sulfate, scopolamine hydrochloride, butylscopolamine (n-butyl scopolamine bromide), oxybutynine chloride, pirenzepine dihydrochloride, 4-DAMP, p-F-HHSiD (para-fluoro-hexahydrosila-difenidol) and 3-AT (3-amino-1,2,4-triazole) were obtained from Sigma. Yeast medium components were purchased from Sigma (amino acids) and from BD Biosciences. Restriction enzymes were purchased from New England Biolabs.

Construction of plasmids and mAChR mutants

All mutations were introduced into rat M3R [22] lacking the central part of ICL3 (amino acids Ala274–Lys469) and containing an N-terminal haemagglutinin tag and a C-terminal FLAG tag. Previous studies have shown that these modifications have no significant effect on receptor function in mammalian expression systems [2224]. Further loop deletion was necessary for efficient expression of M3R in yeast [25]. This modified receptor is referred to as wt M3RΔi3 (M3R lacking ICL3). For yeast expression, wt M3RΔi3 was subcloned into the yeast expression plasmid p416GPD (provided by Dr Mark Pausch, Wyeth Research, Princeton, NJ, U.S.A.). All mutant M3RΔi3 constructs containing a single mutation were generated by PCR-based site-directed mutagenesis and fragment replacement strategies. All PCR-derived sequences were confirmed by restriction enzyme analysis and DNA sequencing.

Generation of yeast libraries containing random mutagenized M3RΔi3

To generate libraries of mutated M3Rs, an error-prone PCR approach was used where wt M3RΔi3 served as the template. Mutagenesis was carried out from ECL1 (901 bp fragment) and TMD5 (558 bp fragment) to the stop codon of the cDNA using the following primers: sense 5′-GGCACTGGGGAACTTAGCC-′3 and 5′-CTATTTTATACTGGAGGATCTA-′3 respectively, and antisense 5′-ACTCGAGGTCGACGGTATC-′3. Error-prone PCR was performed using 7 mM MgCl2, 50 mM KCl, 20 mM Tris/HCl (pH 8.4), 1 mM dCTP, 1 mM dTTP, 0.2 mM dATP, 0.2 mM dGTP, 0.5 mM MnCl2, 1 μM sense primer and antisense primer, 200 ng template and 0.5 unit Taq polymerase in a final volume of 100 μl. Two separate PCRs were carried out and pooled prior to transformation into yeast. This was counted as an independent experiment. Twenty of these independent experiments (ten experiments covering M3RΔi3 ECL1–Stop and ten experiments covering M3RΔi3 TMD5–Stop) were performed and the transformants were cultured on U/H (synthetic dropout medium lacking both uracil and histidine) agar plates containing 100 μM atropine for selection. Three additional independent experiments were performed and transformants were cultured on U (synthetic dropout medium lacking uracil) agar plates to estimate the number of mutants transformed into yeast.

To express random mutagenized M3RΔi3 constructs in yeast we used a gap-repair protocol [26]. In brief, the yeast expression plasmid p416GPD was digested with MscI and SpeI (for 901 bp fragment, see above) or with BstXI (558 bp fragment). Linearized plasmid was mixed with the error-prone PCR product and co-transformed into yeast (see below). The PCR product was inserted into the plasmid by homologous recombination in yeast.

Yeast strains, transformation growth and selection

The haploid Saccharomyces cerevisiae yeast strain MPY578q5 (provided by Dr Mark Pausch) was used for the expression of the mutant M3RΔi3 libraries. Cells were transformed with plasmid DNA using electroporation. Briefly, an overnight culture grown at 30 °C in YPAD (yeast extract/peptone/dextrose medium with adenine) was diluted to an attenuance at 600 nm (D600) of 0.2 in 50 ml YPAD. This culture was incubated at 30 °C until the attenuance at 600 nm (D600) reached 1.2–1.5. Cells were then harvested by centrifugation at 2500 g for 5 min at room temperature (21 °C) and washed with 50 ml and 25 ml of ice-cold water and with 10 ml of 1 M ice-cold sorbitol. The pellet was dissolved in an equal volume of 1 M ice-cold sorbitol. Cells were then mixed with the plasmid and products from error-prone PCR and transformed by electroporation using the Gene Pulser Xcell electroporator (Bio-Rad).

For selection of atropine-sensitive clones, cells were plated on agar plates containing U/H and atropine. After incubation at 30 °C for 4 days, clones were transferred individually into 96-well plates and screened for agonist-dependent or agonist-independent growth.

Growth response curves were recorded in 96-well plates. Cells transformed with wt M3RΔi3 or the respective M3RΔi3 mutants were pre-cultured at 30 °C in U/H with 100 μM CCh to an attenuance (D600) of approx. 1.5. Pre-culturing of cells in U revealed lower receptor expression levels compared with pre-culturing cells in U/H in the presence of 100 μM agonist. To remove CCh, cells were washed three times with water and then diluted to an attenuance (D600) of 0.1. Of this cell suspension, 100 μl was pipetted into each well, and to this 100 μl of a 2× ligand solution was added. Background growth was suppressed by the addition of 20 mM 3-AT. EC50 values were calculated using GraphPad Prism 4 software (GraphPad).

COS-7 cell culture, transfection and IP (inositol phosphate) assays

COS-7 cells were grown in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin and 100 μg/ml streptomycin at 37 °C in a humidified 7% CO2 incubator. Rotifect® (Carl Roth) was used for transient transfection of COS-7 cells with the M3R constructs in the mammalian expression vector pcDps [22]. Thus cells were split into 12-well plates (1.5×105 cells/well) and transfected with a total of 0.5 μg of plasmid DNA/well. To measure IP formation, transfected COS-7 cells were incubated with 2 μCi/ml myo-[3H]inositol (18.6 Ci/mmol, PerkinElmer) for 18 h at 37 °C. Then cells were washed once with serum-free DMEM containing 10 mM LiCl, followed by incubation with 100 μM atropine or 100 μM CCh for 30 min at 37 °C. Intracellular IP levels were determined by anion-exchange chromatography as described previously [27].

Radioligand binding studies

Yeast cells were grown overnight at 30 °C in 100 ml of U/H supplemented with 10 mM CCh for the M3RΔi3 R179K/A489P mutant or 100 μM CCh for wt M3RΔi3 and all other mutant M3RΔi3 constructs. Cells were harvested at an attenuance (D600) of 1.2–1.5, counted, washed three times with 50 ml of water and stored at −80 °C until use. For COS-7 cell binding studies, cells were washed twice with ice-cold PBS 48 h after transfection, harvested and homogenized in 25 mM Tris/HCl (pH 7.4) and 1 mM EDTA. Following removal of nuclei by low-speed centrifugation at 500 g for 10 min at room temperature, COS-7 cell membranes were prepared by ultracentrifugation at 50000 g for 20 min at 4 °C (SS-34 rotor; Sorvall RC28S). [3H]NMS (N-methylscopolamine) binding studies were performed with yeast cells (2×;107 cells/tube) and with COS-7 cell membranes (membranes from 8×104 cells/tube) in 25 mM sodium phosphate buffer (pH 7.4) supplemented with 5 mM MgCl2 in a final volume of 250 μl. To avoid ligand depletion, membrane amounts were adjusted so that less than 10% of the total activity was bound to the membranes. For saturation binding assays, six different concentrations of [3H]NMS (41 pM–10 nM; 82 Ci/mmol, PerkinElmer) were used. Non-specific binding was determined by incubation in the presence of 5 μM atropine. For [3H]NMS displacement binding studies (0.5 nM [3H]NMS), 11 different concentrations of the competing ligand (CCh: 0.17 μM–10 mM and atropine: 0.17 nM–10 μM) were used. All binding assays were incubated at room temperature for 2 h, and bound and non-bound radioligands were separated on a filter harvester (Brandel, Gaithersburg, MD, U.S.A.). Results were analysed using the algorithms implemented in GraphPad Prism 4.

RESULTS

M3RΔi3 expressed in yeast shows the expected pharmacological properties

M3RΔi3 cDNA was introduced into the yeast expression plasmid p416GPD and transformed into the S. cerevisiae strain MPY578q5. This yeast strain was modified genetically such that it required productive receptor/G-protein coupling for growth in histidine-deficient medium [16]. The MPY578q5 strain contains a mutant version of the GPA1 gene encoding a chimeric yeast/mammalian G protein α-subunit in which the last five amino acid residues of Gpa1p have been replaced with the corresponding mammalian Gαq residues [25]. As shown in Figure 1(A), agonist stimulation of wt M3RΔi3 resulted in concentration-dependent yeast growth. Growth for 24 h was found to be sufficient for receptor activity measurements and was not limited by the energy sources in the growth medium. Next, M3RΔi3-expressing yeast cells were grown in medium containing 10 μM CCh and various receptor antagonists (1 μM and 10 μM atropine, scopolamine, ipratropium, oxybutynine, 4-DAMP, butylscopolamine, pirenzepine and p-F-HHSiD). As expected, all antagonists inhibited cell growth: atropine, scopolamine, ipratropium, oxybutynine and 4-DAMP with high potency, and butylscopolamine, pirenzepine and p-F-HHSiD with lower potency (Figure 1B).

Function of the M3R expressed in yeast

Figure 1
Function of the M3R expressed in yeast

(A) Yeast were transformed with the yeast expression plasmid p416GPD containing M3RΔi3 cDNA (see the Experimental section). Yeast growth (OD600nm) measured at various CCh concentrations showed CCh dependency over time. (B) Yeast expressing M3RΔi3 were stimulated with 10 μM CCh in the presence of different M3R antagonists (0, 1 and 10 μM) and cell growth was measured at 600 nm (OD600 nm) after 24 h. All results are means±S.E.M. (n=3), with each experiment performed in triplicate.

Figure 1
Function of the M3R expressed in yeast

(A) Yeast were transformed with the yeast expression plasmid p416GPD containing M3RΔi3 cDNA (see the Experimental section). Yeast growth (OD600nm) measured at various CCh concentrations showed CCh dependency over time. (B) Yeast expressing M3RΔi3 were stimulated with 10 μM CCh in the presence of different M3R antagonists (0, 1 and 10 μM) and cell growth was measured at 600 nm (OD600 nm) after 24 h. All results are means±S.E.M. (n=3), with each experiment performed in triplicate.

Several inhibitors, including atropine, act as inverse agonists on mAChR [28]. To test whether this property is retained in the yeast expression system, a constitutively active M3RΔi3 mutant (L158Q/W503R, Figure 2A) was incubated with increasing concentrations of atropine. As shown in Figure 2(B), the increased basal activity displayed by this mutant receptor was reduced by atropine in a concentration-dependent manner. All these results indicate that agonistic and inverse agonistic properties of M3R ligands for M3RΔi3 and the constitutively active mutant are retained in the yeast expression system.

Atropine is an inverse agonist on a constitutively active M3RΔi3 mutant

Figure 2
Atropine is an inverse agonist on a constitutively active M3RΔi3 mutant

(A) Yeast cells were transformed with wt M3RΔi3 or the constitutively active M3RΔi3 L158Q/W503R (L158Q/W503R) mutant and yeast growth (OD600 nm) was measured for up to 24 h. This mutant was identified by its atropine-independent growth after random mutagenesis (see the Results section). This M3RΔi3 mutant showed high agonist-independent growth when compared with wt M3RΔi3. (B) Agonist-independent growth of M3RΔi3 L158Q/W503R (L158Q/W503R) was inhibited by atropine, whereas CCh acted as an agonist. Results are means±S.E.M. (n=3), with each experiment performed in triplicate.

Figure 2
Atropine is an inverse agonist on a constitutively active M3RΔi3 mutant

(A) Yeast cells were transformed with wt M3RΔi3 or the constitutively active M3RΔi3 L158Q/W503R (L158Q/W503R) mutant and yeast growth (OD600 nm) was measured for up to 24 h. This mutant was identified by its atropine-independent growth after random mutagenesis (see the Results section). This M3RΔi3 mutant showed high agonist-independent growth when compared with wt M3RΔi3. (B) Agonist-independent growth of M3RΔi3 L158Q/W503R (L158Q/W503R) was inhibited by atropine, whereas CCh acted as an agonist. Results are means±S.E.M. (n=3), with each experiment performed in triplicate.

Identification of M3RΔi3 mutants activated by atropine

To test the hypothesis that mutations in M3R can convert the inverse agonist atropine into an agonist, error-prone PCR was performed covering the sequence from ECL1 to the C-terminus and TMD5 to the C-terminus. First, we determined the mutation rate by sequencing 30 clones randomly picked from transformed yeast cultured on non-selective U agar plates (medium without uracil selects only for plasmid-containing yeast clones). The mutation rate was 3.0±1.5 nucleotide changes/kbp.

Next, 20 independent random mutagenesis experiments were performed and approx. 1×107 clones were selected for growth on U/H agar plates containing atropine. The latter number was calculated from the number of recombinant clones growing on U agar plates. Following selection on U/H agar plates containing atropine, each clone was subdivided into one well of two 96-well plates containing either U/H alone or U/H with 100 μM atropine to discriminate between atropine-activated and constitutively active mutants. From 150 clones primarily selected on atropine-containing U/H agar plates, the growth of only 14 clones was atropine-dependent. The M3RΔi3 plasmids of these clones were sequenced and the mutations determined are listed in Table 1. All clones contained two or more missense mutations. Although random mutagenesis experiments were performed independently, mutation of at least one of the following positions (Arg179, Glu485, Gln490 and Ile495) was found in all mutants. Q490L/I495N was the most frequent combination of mutations occurring alone (two clones) or together with other mutations (six clones).

Table 1
M3RΔi3 mutants that can be activated by atropine

Error-prone PCR-based mutagenesis was performed (using rat wt M3RΔi3 as a template) between ECL1 to the C-terminus (ECL1/C-term) and TMD5 to the C-terminus (TMD5/C-term). Mutant libraries were transformed into yeast and atropine-activatable M3RΔi3 mutants were selected on atropine-containing U/H. To differentiate atropine-activatable mutants and constitutively active mutants, cells were further characterized in U/H with and without atropine. Mutants devoid of constitutive growth were maintained and the M3RΔi3 coding sequence was determined after transformation into Escherichia coli and DNA sequencing (see the Experimental section). Fourteen atropine-activatable mutants were isolated from 20 independent random mutagenesis experiments (experiment number is stated in parenthesis). All random mutagenesis experiments yielded constitutively active mutants, but not all mutants were atropine-activatable. Although random mutagenesis experiments were performed independently, R179, E485, Q490 and/or I495 were found in all clones and are indicated in bold.

Mutant numberRandom mutagenesisMutations
TMD5/C-term (4) Y250N, K478N, I495P, L589S 
ECL1/C-term (9) K484R, E485V, L492H 
ECL1/C-term (9) Q490L, I495N, Q576Stop 
ECL1/C-term (9) N152S, L173F, R179S, K476E, E485G, K566R 
TMD5/C-term (10) Q490L, I495N, Q573Stop 
ECL1/C-term (11) P228S, A238V, Q490L, I495N 
ECL1/C-term (13) R179K, E227K, I230T, E485G 
ECL1/C-term (13) S154T, F221S, E485G, K486M, Y572F, new sequence from F581 onwards (SQASAGAGLGL) due to frameshift 
TMD5/C-term (16) Q490L, I495N, K522N 
10 ECL1/C-term (17) K259R, Q490L, I495N 
11* ECL1/C-term (17, 18) Q490L, I495N 
12 ECL1/C-term (17) R179K, A489P 
13 TMD5/C-term (18) K262R, I474F, K487R, Q490L, I495N 
Mutant numberRandom mutagenesisMutations
TMD5/C-term (4) Y250N, K478N, I495P, L589S 
ECL1/C-term (9) K484R, E485V, L492H 
ECL1/C-term (9) Q490L, I495N, Q576Stop 
ECL1/C-term (9) N152S, L173F, R179S, K476E, E485G, K566R 
TMD5/C-term (10) Q490L, I495N, Q573Stop 
ECL1/C-term (11) P228S, A238V, Q490L, I495N 
ECL1/C-term (13) R179K, E227K, I230T, E485G 
ECL1/C-term (13) S154T, F221S, E485G, K486M, Y572F, new sequence from F581 onwards (SQASAGAGLGL) due to frameshift 
TMD5/C-term (16) Q490L, I495N, K522N 
10 ECL1/C-term (17) K259R, Q490L, I495N 
11* ECL1/C-term (17, 18) Q490L, I495N 
12 ECL1/C-term (17) R179K, A489P 
13 TMD5/C-term (18) K262R, I474F, K487R, Q490L, I495N 
*

This mutant was found in two independent random mutagenesis experiments.

Several clones which grew independently of atropine on U/H agar plates were maintained and the M3RΔi3 cDNA of these clones was sequenced. The constitutively active mutant L158Q/W503R was used for control purposes (see above and Figure 2A).

Conversion of the pharmacological property of M3R blockers is non-selective

To characterize further the functional properties of the recovered M3RΔi3 mutants, concentration–growth curves were obtained. Thus CCh, atropine and other ligands of mAChR (scopolamine, butylscopolamine, ipratropium, pirenzepine, oxybutynine, 4-DAMP and p-F-HHSiD) were tested. On the basis of differences between ligand potencies (Table 2), the mutants were categorized into three groups. Group 1: atropine=CCh (mutants 1, 3–5, 10, 11 and 13, Figure 3A); group 2: atropine>CCh (mutants 2, 6, 8 and 9); and group 3: atropine>>>CCh (mutants 7 and 12, Figure 3B). In general, potencies and efficacies of scopolamine and oxybutynine were very similar to atropine for all mutants. Ipratropium, butylscopolamine and pirenzepine were approx. 2.8-fold, 145-fold and 450-fold less potent respectively, but the EC50 ratios to atropine were not significantly different (P>0.01). 4-DAMP and p-F-HHSiD were only tested qualitatively, indicating that most mutants responded to both structurally different compounds. To exclude the possibility that the agonistic effect was the result of some non-receptor-specific effects, several other GPCR ligands and substances were tested on yeast cells expressing wt M3RΔi3 and the mutant M3RΔi3 constructs. For example, the local-anaesthetic drug lidocaine and the β-adrenergic blocker propranolol did not stimulate growth of transformed yeast cells (see Supplementary Figures S1A–S1C at http://www.BiochemJ.org/bj/412/bj4120103add.htm).

Table 2
Pharmacological properties of wt and mutant M3RΔi3 expressed in yeast

To evaluate the functional properties of wt and mutant M3RΔi3, basal receptor activity [measured as growth without ligand at an attenuance at 600 nm (D600)], ligand efficacy (Emax, measured as fold change compared with wt basal) and potency (pEC50) were determined from ligand concentration–growth curves (10 nM to 1 mM ligand or 10 mM ligand for pirenzepine, for details see Experimental section) after 24 h of growth. The basal attenuance (D600) of M3R wt was 0.25±0.05 (n=20). Results are means±S.E.M., with the number of experiments indicated in parenthesis and each experiment carried out in triplicate. n.d., not determinable; n.g., no growth.

CChAtropineScopolamineButylscopolamine
MutantBasal activity (fold of wt basal)EmaxpEC50EmaxpEC50EmaxpEC50EmaxpEC50
wt M3RΔi3 4.20±0.12 (17) 5.56±0.06 n.g. n.d. n.g. n.d. n.g. n.d. 
0.88±0.08 (6) 4.44±0.08 (5) 5.52±0.22 4.00±0.14 (7) 5.42±0.07 4.15±0.21 (4) 5.47±0.20 3.29±0.33 (4) 3.56±0.17 
1.10±0.13 (6) 4.25±0.23 (5) 3.79±0.11 3.90±0.10 (7) 5.64±0.10 3.84±0.18 (4) 5.87±0.25 3.21±0.28 (4) 4.30±0.33 
1.49±0.27 (6) 4.26±0.15 (5) 6.67±0.44 4.17±0.08 (7) 6.57±0.26 4.25±0.09 (4) 6.12±0.37 4.45±0.12 (4) 3.85±0.29 
0.78±0.24 (6) 4.29±0.12 (5) 5.44±0.38 3.76±0.15 (7) 5.87±0.25 3.68±0.17 (4) 5.50±0.26 3.52±0.18 (4) 3.59±0.19 
1.12±0.12 (6) 4.26±0.18 (5) 6.82±0.46 4.14±0.08 (7) 6.47±0.29 4.12±0.14 (4) 6.43±0.32 4.19±0.22 (4) 4.15±0.56 
0.93±0.15 (7) 3.50±0.23 (6) 3.78±0.24 4.15±0.06 (8) 6.10±0.11 4.26±0.10 (4) 6.23±0.18 4.11±0.19 (4) 4.29±0.09 
1.04±0.16 (7) 1.90±0.67 (4) n.d. 4.13±0.13 (6) 5.10±0.11 3.92±0.27 (3) 4.97±0.15 2.09±0.78 (3) 3.41±0.36 
1.15±0.22 (8) 3.20±0.39 (5) 3.34±0.18 3.96±0.13 (7) 5.47±0.09 3.98±0.22 (4) 5.28±0.08 3.37±0.42 (4) 4.02±0.29 
1.13±0.21 (6) 4.41±0.14 (5) 4.97±0.07 4.21±0.10 (7) 5.53±0.10 4.15±0.20 (4) 5.22±0.09 4.15±0.20 (4) 3.34±0.17 
10 0.99±0.10 (7) 4.17±0.08 (3) 5.05±0.24 3.85±0.01 (3) 5.22±0.24 3.52±0.25 (3) 4.96±0.03 2.19±0.21 (3) <2.70 
11 1.14±0.12 (13) 3.86±0.24 (11) 5.47±0.18 3.61±0.25 (13) 5.57±0.18 4.14±0.21 (3) 5.64±0.03 3.59±0.13 (3) 3.56±0.08 
12 0.83±0.09 (7) n.g. (4) n.d. 3.53±0.12 (3) 4.95±0.12 3.48±0.31 (3) 5.07±0.07 2.01±0.16 (3) <3.00 
13 0.91±0.10 (7) 4.01±0.11 (3) 4.72±0.20 3.43±0.01 (3) 4.92±0.19 3.01±0.25 (3) 4.58±0.02 2.32±0.32 (3) < 2.90 
cont.
 IpratropiumOxybutyninePirenzepine4-DAMPp-F-HHSid 
MutantEmaxpEC50EmaxpEC50EmaxpEC50Emax (at 10 μM)Emax (at 10 μM)
wt M3RΔi3 n.g. n.d. n.g. n.d. n.g. n.d. n.g. n.g.  
2.89±0.20 (4) 4.66±0.23 2.40±0.25 (3) n.d. 1.83±0.62 (3) n.d. 3.08±0.04 2.28±0.28  
3.76±0.24 (4) 5.16±0.13 3.25±0.27 (3) 6.27±0.26 3.15±0.43 (3) <2.80 2.76±0.20 1.36±0.04  
4.13±0.17 (4) 5.83±0.20 3.38±0.15 (3) 5.73±0.08 3.99±0.20 (3) 3.86±0.03 2.76±0.20 1.24±0.04  
3.45±0.29 (4) 5.08±0.20 3.41±0.14 (3) 5.92±0.15 3.73±0.31 (3) 3.62±0.15 2.68±0.04 1.12±0.04  
4.14±0.19 (4) 5.66±0.34 3.15±0.17 (3) 5.67±0.18 3.84±0.35 (3) 3.74±0.11 2.24±0.08 0.76±0.04  
4.27±0.06 (4) 5.85±0.14 4.06±0.14 (3) 4.87±0.25 3.89±0.23 (3) 4.22±0.07 3.80±0.04 3.04±0.20  
4.01±0.28 (3) 5.07±0.22 4.24±0.18 (3) 5.44±0.13 3.17±0.58 (3) <2.90 3.76±0.36 3.20±0.16  
3.97±0.23 (4) 5.08±0.12 3.96±0.15 (3) 5.65±0.12 3.86±0.39 (3) 3.16±0.20 3.08±0.32 1.16±0.40  
4.24±0.16 (4) 5.06±0.02 3.73±0.18 (3) 5.26±0.18 3.35±0.34 (3) <2.80 2.40±0.28 1.60±0.04  
10 3.87±0.15 (3) 4.74±0.07 2.56±0.23 (3) 4.75±0.05 3.09±0.14 (3) <2.40 1.72±0.04 0.68±0.04  
11 4.18±0.31 (3) 5.24±0.12 2.92±0.09 (3) 5.39±0.17 3.85±0.55 (3) <2.90 2.28±0.08 1.36±0.12  
12 3.55±0.08 (4) 4.99±0.05 3.66±0.26 (3) 5.37±0.01 2.33±0.22 (3) <1.90 3.28±0.04 0.68±0.04  
13 3.48±0.19 (3) 4.57±0.07 1.94±0.21 (3) 4.65±0.02 2.35±0.09 (3) <2.30 1.04±0.04 0.56±0.04  
CChAtropineScopolamineButylscopolamine
MutantBasal activity (fold of wt basal)EmaxpEC50EmaxpEC50EmaxpEC50EmaxpEC50
wt M3RΔi3 4.20±0.12 (17) 5.56±0.06 n.g. n.d. n.g. n.d. n.g. n.d. 
0.88±0.08 (6) 4.44±0.08 (5) 5.52±0.22 4.00±0.14 (7) 5.42±0.07 4.15±0.21 (4) 5.47±0.20 3.29±0.33 (4) 3.56±0.17 
1.10±0.13 (6) 4.25±0.23 (5) 3.79±0.11 3.90±0.10 (7) 5.64±0.10 3.84±0.18 (4) 5.87±0.25 3.21±0.28 (4) 4.30±0.33 
1.49±0.27 (6) 4.26±0.15 (5) 6.67±0.44 4.17±0.08 (7) 6.57±0.26 4.25±0.09 (4) 6.12±0.37 4.45±0.12 (4) 3.85±0.29 
0.78±0.24 (6) 4.29±0.12 (5) 5.44±0.38 3.76±0.15 (7) 5.87±0.25 3.68±0.17 (4) 5.50±0.26 3.52±0.18 (4) 3.59±0.19 
1.12±0.12 (6) 4.26±0.18 (5) 6.82±0.46 4.14±0.08 (7) 6.47±0.29 4.12±0.14 (4) 6.43±0.32 4.19±0.22 (4) 4.15±0.56 
0.93±0.15 (7) 3.50±0.23 (6) 3.78±0.24 4.15±0.06 (8) 6.10±0.11 4.26±0.10 (4) 6.23±0.18 4.11±0.19 (4) 4.29±0.09 
1.04±0.16 (7) 1.90±0.67 (4) n.d. 4.13±0.13 (6) 5.10±0.11 3.92±0.27 (3) 4.97±0.15 2.09±0.78 (3) 3.41±0.36 
1.15±0.22 (8) 3.20±0.39 (5) 3.34±0.18 3.96±0.13 (7) 5.47±0.09 3.98±0.22 (4) 5.28±0.08 3.37±0.42 (4) 4.02±0.29 
1.13±0.21 (6) 4.41±0.14 (5) 4.97±0.07 4.21±0.10 (7) 5.53±0.10 4.15±0.20 (4) 5.22±0.09 4.15±0.20 (4) 3.34±0.17 
10 0.99±0.10 (7) 4.17±0.08 (3) 5.05±0.24 3.85±0.01 (3) 5.22±0.24 3.52±0.25 (3) 4.96±0.03 2.19±0.21 (3) <2.70 
11 1.14±0.12 (13) 3.86±0.24 (11) 5.47±0.18 3.61±0.25 (13) 5.57±0.18 4.14±0.21 (3) 5.64±0.03 3.59±0.13 (3) 3.56±0.08 
12 0.83±0.09 (7) n.g. (4) n.d. 3.53±0.12 (3) 4.95±0.12 3.48±0.31 (3) 5.07±0.07 2.01±0.16 (3) <3.00 
13 0.91±0.10 (7) 4.01±0.11 (3) 4.72±0.20 3.43±0.01 (3) 4.92±0.19 3.01±0.25 (3) 4.58±0.02 2.32±0.32 (3) < 2.90 
cont.
 IpratropiumOxybutyninePirenzepine4-DAMPp-F-HHSid 
MutantEmaxpEC50EmaxpEC50EmaxpEC50Emax (at 10 μM)Emax (at 10 μM)
wt M3RΔi3 n.g. n.d. n.g. n.d. n.g. n.d. n.g. n.g.  
2.89±0.20 (4) 4.66±0.23 2.40±0.25 (3) n.d. 1.83±0.62 (3) n.d. 3.08±0.04 2.28±0.28  
3.76±0.24 (4) 5.16±0.13 3.25±0.27 (3) 6.27±0.26 3.15±0.43 (3) <2.80 2.76±0.20 1.36±0.04  
4.13±0.17 (4) 5.83±0.20 3.38±0.15 (3) 5.73±0.08 3.99±0.20 (3) 3.86±0.03 2.76±0.20 1.24±0.04  
3.45±0.29 (4) 5.08±0.20 3.41±0.14 (3) 5.92±0.15 3.73±0.31 (3) 3.62±0.15 2.68±0.04 1.12±0.04  
4.14±0.19 (4) 5.66±0.34 3.15±0.17 (3) 5.67±0.18 3.84±0.35 (3) 3.74±0.11 2.24±0.08 0.76±0.04  
4.27±0.06 (4) 5.85±0.14 4.06±0.14 (3) 4.87±0.25 3.89±0.23 (3) 4.22±0.07 3.80±0.04 3.04±0.20  
4.01±0.28 (3) 5.07±0.22 4.24±0.18 (3) 5.44±0.13 3.17±0.58 (3) <2.90 3.76±0.36 3.20±0.16  
3.97±0.23 (4) 5.08±0.12 3.96±0.15 (3) 5.65±0.12 3.86±0.39 (3) 3.16±0.20 3.08±0.32 1.16±0.40  
4.24±0.16 (4) 5.06±0.02 3.73±0.18 (3) 5.26±0.18 3.35±0.34 (3) <2.80 2.40±0.28 1.60±0.04  
10 3.87±0.15 (3) 4.74±0.07 2.56±0.23 (3) 4.75±0.05 3.09±0.14 (3) <2.40 1.72±0.04 0.68±0.04  
11 4.18±0.31 (3) 5.24±0.12 2.92±0.09 (3) 5.39±0.17 3.85±0.55 (3) <2.90 2.28±0.08 1.36±0.12  
12 3.55±0.08 (4) 4.99±0.05 3.66±0.26 (3) 5.37±0.01 2.33±0.22 (3) <1.90 3.28±0.04 0.68±0.04  
13 3.48±0.19 (3) 4.57±0.07 1.94±0.21 (3) 4.65±0.02 2.35±0.09 (3) <2.30 1.04±0.04 0.56±0.04  

Functional characterization of M3RΔi3 mutants activated by atropine

Figure 3
Functional characterization of M3RΔi3 mutants activated by atropine

(A) The M3RΔi3 Q490L/I495N (Q490L/I495N) mutant expressed in yeast responds by yeast growth (measured as OD600nm) in the presence of various concentrations of atropine and CCh, whereas atropine has no effect on wt M3RΔi3. (B) The M3RΔi3 R179K/A489P (R179K/A489P) mutant is atropine-activatable (measured by yeast growth at OD600nm), but the potency of CCh is significantly reduced. All results are means±S.E.M. (n=3), with each experiment performed in triplicate.

Figure 3
Functional characterization of M3RΔi3 mutants activated by atropine

(A) The M3RΔi3 Q490L/I495N (Q490L/I495N) mutant expressed in yeast responds by yeast growth (measured as OD600nm) in the presence of various concentrations of atropine and CCh, whereas atropine has no effect on wt M3RΔi3. (B) The M3RΔi3 R179K/A489P (R179K/A489P) mutant is atropine-activatable (measured by yeast growth at OD600nm), but the potency of CCh is significantly reduced. All results are means±S.E.M. (n=3), with each experiment performed in triplicate.

M3RΔi3 activation by atropine requires a combination of mutations

The positions Arg179, Glu485, Gln490 and Ile495 were found mutated frequently in the atropine-sensitive clones. As shown in Supplementary Figure S2 (at http://www.BiochemJ.org/bj/412/bj4120103add.htm), these positions are highly conserved in M3R orthologues and other mAChRs. To analyse the contribution of the individual mutations to the observed switch in ligand properties, several single mutant constructs were generated. Surprisingly, all single mutations at positions Glu485, Gln490 and Ile495 displayed high basal activity but none showed atropine-induced increases in cell growth (Figures 4A–4C). Dissection of M3RΔi3 R179K/A489P displayed a slightly different picture. M3RΔi3 R179K showed almost wt-like pharmacological properties, whereas M3RΔi3 A489P revealed an increase in basal activity and a small but significant agonistic activity of atropine and CCh (Figure 4D). In summary, our results indicate that the functional properties of the atropine-sensitive M3RΔi3 clones can only be achieved by a combination of mutations.

Contribution of individual mutations to M3RΔi3 activation by atropine

Figure 4
Contribution of individual mutations to M3RΔi3 activation by atropine

All mutants found in the yeast genetic screen contained two or more missense mutations. To analyse the contribution of the individual mutations, the most frequent mutations, (A) E485G (E485G) and E485V (E485V), (B) Q490L (Q490L), (C) I495P (I495P) and I495N (I495N) and (D) A489P (A489P) and R179K (R179K), were individually introduced into M3RΔi3 and expressed in yeast. Yeast growth in U/H (measured as OD 600nm) was determined in the presence of various concentrations of CCh and atropine after 24 h. For comparison purposes, the growth curve of wt M3RΔi3 is shown in Figure 3(A). Results are means±S.E.M. (n=3), with each experiment performed in triplicate. The increase in cell density of A489P-expressing yeast cells between CCh and atropine concentrations of 10 nM and 1 mM is statistically significant as determined by the Student's t test (P<0.01).

Figure 4
Contribution of individual mutations to M3RΔi3 activation by atropine

All mutants found in the yeast genetic screen contained two or more missense mutations. To analyse the contribution of the individual mutations, the most frequent mutations, (A) E485G (E485G) and E485V (E485V), (B) Q490L (Q490L), (C) I495P (I495P) and I495N (I495N) and (D) A489P (A489P) and R179K (R179K), were individually introduced into M3RΔi3 and expressed in yeast. Yeast growth in U/H (measured as OD 600nm) was determined in the presence of various concentrations of CCh and atropine after 24 h. For comparison purposes, the growth curve of wt M3RΔi3 is shown in Figure 3(A). Results are means±S.E.M. (n=3), with each experiment performed in triplicate. The increase in cell density of A489P-expressing yeast cells between CCh and atropine concentrations of 10 nM and 1 mM is statistically significant as determined by the Student's t test (P<0.01).

Atropine-activatable mutants behave differently in a mammalian expression system

To examine whether the recovered M3RΔi3 mutants retain atropine-induced receptor activation in a mammalian expression system, COS-7 cells were transfected with all of the mutant constructs. In contrast to the findings in the yeast expression system, none of the mutants showed atropine-induced activity in IP accumulation assays, as shown for M3RΔi3 R179K/A489P and M3RΔi3 Q490L/I495N in Figure 5. CCh-induced IP formation was shown for M3RΔi3 R179K/A489P, whereas M3RΔi3 Q490L/I495N showed a 2-fold elevation compared with basal activity, and CCh-induced IP formation was reduced when compared with wt M3RΔi3 (Figure 5). To exclude the possibility that atropine is a partial agonist in the COS-7 cell expression system and that constitutive activity of the mutants may mask the partial agonistic effect, we measured IP accumulation after 10 min and 1 h in addition to the 30 min timepoint. However, IP levels in the presence of 100 μM atropine were similar or even lower at any incubation time examined when compared with the non-stimulated mutants (results not shown). Also, the introduction of the Q490L/I495N mutation into full-length wt M3R resulted in an increase in basal activity, but no stimulation by atropine was observed when expressed in COS-7 cells (see Supplementary Figure S3 at http://www.BiochemJ.org/bj/412/bj4120103add.htm).

Characterization of selected M3RΔi3 mutants in COS-7 cells

Figure 5
Characterization of selected M3RΔi3 mutants in COS-7 cells

COS-7 cells were transiently transfected with wt and mutant M3RΔi3 Q490L/I495N (Q490L/I495N), R179K/A489P (R179K/A489P) or E485G (E485G). IP accumulation assays were performed 72 h after transfection as described in the Experimental section. Results are expressed as fold increase compared with basal IP accumulation for GFP (green fluorescent protein)-transfected COS-7 cells (fold over GFP, 873±274 c.p.m./well). Results are means±S.E.M. (n=4), with each experiment performed in triplicate.

Figure 5
Characterization of selected M3RΔi3 mutants in COS-7 cells

COS-7 cells were transiently transfected with wt and mutant M3RΔi3 Q490L/I495N (Q490L/I495N), R179K/A489P (R179K/A489P) or E485G (E485G). IP accumulation assays were performed 72 h after transfection as described in the Experimental section. Results are expressed as fold increase compared with basal IP accumulation for GFP (green fluorescent protein)-transfected COS-7 cells (fold over GFP, 873±274 c.p.m./well). Results are means±S.E.M. (n=4), with each experiment performed in triplicate.

In the yeast expression system receptor activation is measured indirectly, and several components are involved to mediate cell division on receptor stimulation. In contrast, measurement of IP levels, although still indirect, is more upstream in the receptor signalling cascade. Therefore GTP[S] (guanosine 5′-[γ-thio]triphosphate) binding and GTPase activity (for assays see [29,30]) measurements were performed using yeast cell membranes. Because of the high level of non-specific GTP binding and GTPase activity, these assays were unsuitable to measure even wt receptor activation (results not shown).

To analyse whether the lack of atropine-induced receptor activation is the result of a loss of atropine binding when mutants were expressed in the mammalian COS-7 cell expression system, radioligand binding assays were performed. As shown in Table 3, wt and mutant M3RΔi3 constructs showed high affinity for the antagonist NMS when expressed in COS-7 cells. In general, Kd values of the mutant M3RΔi3 constructs were 6–10-fold lower in the mammalian expression system, but the ratio between Ki values of atropine and CCh showed no significant differences between yeast and COS-7 cells (P>0.01). Next, we addressed the question of whether the shifted concentration–response curve of M3RΔi3 R179K/A489P towards higher CCh concentrations (Figure 3B and Table 2) is caused by changes in the ability to bind CCh. However, Ki values of CCh were not significantly different between wt M3RΔi3 and M3RΔi3 R179K/A489P (Table 3).

Table 3
Ligand-binding properties of mutant M3RΔi3 expressed in yeast or in COS-7 cells

Wt M3RΔi3 and the M3RΔi3 mutants indicated were expressed in the MPY578q5 yeast strain and in COS-7 cells. Kd and Bmax (maximal binding) values were calculated from saturation binding studies using different concentrations of [3H]NMS (for details see the Experimental section). In [3H]NMS displacement binding studies, pIC50 values of CCh and atropine were determined and converted into pKi values using the Cheng–Prusoff equation [48]. Results are means±S.E.M. (n=2–4), with each experiment performed in duplicate.

[3H]NMS saturation binding studies[3H]NMS displacement studies
ConstructBmax (receptors/cell)Kd (nM)pKi atropinepKi CCh
Yeast cell binding     
 Wt M3RΔi3 291±38 0.86±0.14 8.03±0.03 3.40±0.11 
 Q490L/I495N (11) 589±167 1.39±0.15 8.11±0.06 5.13±0.16 
 R179K/A489P (12) 233±41 1.20±0.17 7.93±0.01 3.59±0.10 
 E485G 494±76 2.71±0.22 7.60±0.06 3.82±0.08 
COS-7 cell binding     
 Wt M3RΔi3 (3.99±1.67)×105 0.14±0.05 8.49±0.04 3.83±0.11 
 Q490L/I495N (11) (1.20±0.44)×105 0.16±0.01 8.81±0.01 5.85±0.04 
 R179K/A489P (12) (1.33±0.45)×105 0.18±0.03 8.51±0.04 4.57±0.01 
 E485G (2.29±0.95)×105 0.27±0.02 8.34±0.05 4.46±0.13 
[3H]NMS saturation binding studies[3H]NMS displacement studies
ConstructBmax (receptors/cell)Kd (nM)pKi atropinepKi CCh
Yeast cell binding     
 Wt M3RΔi3 291±38 0.86±0.14 8.03±0.03 3.40±0.11 
 Q490L/I495N (11) 589±167 1.39±0.15 8.11±0.06 5.13±0.16 
 R179K/A489P (12) 233±41 1.20±0.17 7.93±0.01 3.59±0.10 
 E485G 494±76 2.71±0.22 7.60±0.06 3.82±0.08 
COS-7 cell binding     
 Wt M3RΔi3 (3.99±1.67)×105 0.14±0.05 8.49±0.04 3.83±0.11 
 Q490L/I495N (11) (1.20±0.44)×105 0.16±0.01 8.81±0.01 5.85±0.04 
 R179K/A489P (12) (1.33±0.45)×105 0.18±0.03 8.51±0.04 4.57±0.01 
 E485G (2.29±0.95)×105 0.27±0.02 8.34±0.05 4.46±0.13 

Different post-translational modifications in both expression systems may account for the functional differences observed for the mutant receptors in the yeast and mammalian expression systems. Therefore the five potential asparagine glycosylation sites in the N-terminus of M3RΔi3 (Asn6/Asn15/Asn41/Asn48/Asn52) were mutated to glutamine residues. Wt and mutant M3RΔi3 constructs with no glycosylation sites were tested in the yeast and the COS-7 cell expression system. As shown in Supplementary Figure S4 (http://www.BiochemJ.org/bj/412/bj4120103add.htm), the lack of potential N-glycosylation sites neither changed wt nor mutant M3RΔi3 receptor function.

Atropine-sensitizing mutations create a new agonistic binding site

It should be noted that the Ki values of atropine and agonistic potencies of atropine differ by approx. 1000-fold (see Tables 2 and 3). A potential mechanism which would account for this phenomenon is that atropine has two binding sites at the receptor: one for the measured high-affinity antagonistic binding site and one lower-affinity binding site for receptor activation. There are only a few examples in the literature (CGP 12177 binding to the β1 adrenergic receptor and zinc binding to the melanocortin type 1 receptor) where a ligand can act as antagonist and agonist at one receptor depending on the concentration of the ligand (summarized in [31,32]). On the basis of our results, we would expect an affinity difference between the two binding sites of three orders of magnitude, which almost excludes detection of the low-affinity binding site by [3H]NMS binding experiments. Therefore we used a similar approach to the one used for the β1 adrenergic receptor, where the agonist-stimulated receptor is incubated with increasing concentrations of the tested ligand [32]. Thus yeast cells expressing wt M3RΔi3 and M3RΔi3 Q490L/I495N were grown under CCh stimulation and increasing concentrations of atropine were applied. As shown in Figure 6(A), growth of wt M3RΔi3-transformed yeast cells was inhibited in a concentration-dependent manner. Similarly, the initial cell density of M3RΔi3 Q490L/I495N decreased significantly in the presence of increasing concentrations of atropine but, in contrast to the wt receptor, increased again at higher atropine concentrations (Figure 6C). This striking pharmacological property of an M3R blocker was also found for scopolamine on wt M3RΔi3 and M3RΔi3 Q490L/I495N (Figures 6B and 6D). These results strongly support the existence of an additional binding site in the M3RΔi3 Q490L/I495N mutant where M3R blockers act as agonists.

Effect of atropine and scopolamine on CCh-stimulated M3RΔi3 and mutant M3RΔi3 Q490L/I495N

Figure 6
Effect of atropine and scopolamine on CCh-stimulated M3RΔi3 and mutant M3RΔi3 Q490L/I495N

Yeast cells were transformed with wt M3RΔi3 (A, B) and M3RΔi3 Q490L/I495N (C, D) and incubated with fixed concentrations of CCh and increasing concentrations of atropine (A, C) or scopolamine (B, D), and the yeast growth (OD 600nm) was measured. Results are means±S.E.M. (n=3–4), with each experiment performed in triplicate. *, P<0.01 compared with the negative control (without atropine or scopolamine) using the Student's t test.

Figure 6
Effect of atropine and scopolamine on CCh-stimulated M3RΔi3 and mutant M3RΔi3 Q490L/I495N

Yeast cells were transformed with wt M3RΔi3 (A, B) and M3RΔi3 Q490L/I495N (C, D) and incubated with fixed concentrations of CCh and increasing concentrations of atropine (A, C) or scopolamine (B, D), and the yeast growth (OD 600nm) was measured. Results are means±S.E.M. (n=3–4), with each experiment performed in triplicate. *, P<0.01 compared with the negative control (without atropine or scopolamine) using the Student's t test.

DISCUSSION

Mutational switching of ligand properties can provide valuable insights into the molecular basis of GPCR function. However, only a few mutations have been identified in GPCR which convert an antagonist or inverse agonist into a full agonist [1214], and the molecular basis of how this is achieved is mainly unsolved. We chose the rat M3R as a model system to address this question, because structure/function relationships of this receptor have been intensively studied [33] and mAChR can be functionally expressed in a genetically modified yeast strain for high-throughput mutagenesis approaches [25,34]. The pharmacology of several ligands on M3RΔi3 when expressed in yeast is identical to that found in mammalian expression systems (see Figure 1). For example, the EC50 value of CCh (2.8 μM) determined in yeast expressing wt M3RΔi3 is almost identical with the potency (3.9 μM) found for full-length M3R expressed in mammalian (COS-7) cells [22]. Furthermore, atropine shows inverse agonistic activity when tested on constitutively active receptor mutants (see Figure 2). This property of atropine and other mAChR blockers is well studied in mammalian cell systems [28,35]. These results, together with studies of other GPCRs expressed in yeast [11,36,37], indicate that ligand pharmacology is comparable between yeast and mammalian expression systems.

To generate ligand-property-switching M3RΔi3 mutants, we applied error-prone PCR and screened approx. 1×107 mutant M3RΔi3 clones for activation by atropine. We identified 13 different clones harbouring mutations which enable M3RΔi3 to be activated by atropine. All clones contained two or more missense mutations. To our knowledge, these are the first ligand-property-switching mutants discovered so far in an mAChR. In a previous report, a missense mutation in TMD6 (F451P) of M5R was identified which switched atropine to a partial agonist [38]. However, we were unable to reproduce these results for the M5R F451P mutant (see Supplementary Figure S5A at http://www.BiochemJ.org/bj/412/bj4120103add.htm). Introduction of this mutation at the corresponding position of M3R (F499P) resulted in a constitutively active receptor and atropine still acted as an inverse agonist on this mutant (Supplementary Figure S5A). Furthermore, expression of M3RΔi3 F499P in yeast displayed similar properties and no activation by atropine (Supplementary Figure S5B).

Seven clones were found to contain Q490L and I495N mutations. This double mutation is sufficient to enable atropine to act as an agonist (see Figure 3A). Interestingly, the individual mutations displayed no further activation by atropine but were constitutively active, indicating that the Q490L and I495N mutations silence each other. Silencing of constitutively active mutants by ‘second site’ mutations has been described previously [34]. Similarly, substitutions of Glu485 were found in several atropine-activatable mutants (mutants 2, 4, 7 and 8) and mutation of Glu485 to valine or glycine residues alone induced constitutive activity (Figure 4A). These mutants (mutants 2, 4, 7 and 8) contained additional substitutions and some positions (Ile230 and Lys484) have previously been found to silence the constitutive active Q490L mutation [34]. This may implicate a specific conformational rearrangement, necessary to allow for atropine-induced receptor activation, which is achieved by ‘second site’ mutations. However, we cannot exclude the possibility that the individual mutants still contain a second low-affinity atropine binding site. In this scenario, high constitutive activity may mask the agonistic effect of atropine because the receptor is already in an almost completely active conformational state. In contrast with other constitutively active mutants (e.g. L158Q/W503R, see Figure 1), atropine is not an inverse agonist at these single mutant positions (E485G, E485V, Q490L, I495N and I495P, see Figure 4). One can also speculate that the agonistic effect overlaps with the inverse agonistic effect of atropine, resulting in no obvious change in the concentration–response curves.

To test whether the pharmacological switch is restricted to atropine, a number of structurally different muscarinic ligands was tested. Strikingly, all blockers of wt M3R function displayed agonistic activity at all recovered M3RΔi3 mutants. We found distinct differences in agonist potencies between the different mutants. The most obvious differences were found between atropine and CCh (see Table 2). For example, mutant 6 showed a high potency for atropine, but a ∼100-fold increase in the EC50 value of CCh. One can speculate that the additional mutations (P228S and A238V) somehow reduce CCh potency, because the M3RΔi3 Q490L/I495N double mutant alone (mutant 11) displayed unchanged CCh potency. Indeed, mutagenesis studies with the M1R suggest a participation of Ala196 (corresponding to Ala238 in M3R) in ACh binding [39]. A greater than 1000-fold reduced potency of CCh was found for the M3RΔi3 R179K/A489P mutant (mutant 12). Here, ligand-binding data indicated no disturbance of the CCh binding site (see Table 3), but rather suggested a reduction in the potency of CCh to activate R179K/A489P.

It should be noted that activation by atropine disappeared when the M3RΔi3 mutants were were expressed in COS-7 cells. Unfortunately, all our attempts to identify the reasons for this obvious difference between mammalian and yeast expression systems were unsuccessful. We excluded differences in receptor glycosylation (see Supplementary Figure S3) and approaches (GTP[S] binding and GTPase assays) to directly measure receptor activation in yeast failed because of high non-specific GTP binding/GTPase activity. Because we had measured yeast growth, which is a very downstream readout of receptor activation, we tested whether atropine-induced receptor activation is seen in the MAPK (mitogen-activated protein kinase) pathway in transfected COS-7 cells. Although full-length M3RΔi3 and M3RΔi3 R179K/A489P showed CCh-induced increases in ERK1/2 (extracellular-signal-regulated kinase 1/2) phosphorylation, atropine had no stimulatory effect (see Supplementary Figure S6 at http://www.BiochemJ.org/bj/412/bj4120103add.htm).

Furthermore, when wt and mutant M3RΔi3 constructs (e.g. M3RΔi3 Q490L/I495N) were expressed in COS-7 and yeast cells, we found no significant differences in Ki values (see Table 3). If found, these might have indicated differences in receptor folding, e.g. due to differences in lipid composition of the plasma membrane. However, ligand-binding data of the mutant M3RΔi3 constructs verified only the existence of the high-affinity orthosteric binding site in yeast and mammalian expression systems. We cannot exclude the possibility that the low-affinity allosteric binding site is not formed in the mammalian expression system. But differences in transferring functional findings from the yeast expression system appear to have a more general cause, because this phenomenon has been observed for other functionally different mutations [40]. Both systems greatly differ in their G proteins and downstream signalling components. In yeast, the endogenous G protein is modified for efficient coupling to mammalian GPCRs by exchanging the C-terminal five amino acids of the yeast G protein (Gpa1p) with the corresponding mammalian Gq sequence. One can speculate that conformational stages or the dynamics of conformational changes which are responsible for functional G protein/receptor interaction differ between mammalian and yeast cells. This assumption is supported by the fact that functional properties of mutants cannot always be transferred directly from yeast to mammalian expression systems. For example, M3RΔi3 E485G expression in yeast stimulated cell growth to a maximum which could not be further increased by CCh (Figure 4A). In the mammalian expression system, M3RΔi3 E485G is still constitutively active, but to a lesser extent, and CCh can still increase receptor activation (see Figure 5). Therefore it is more likely that atropine-induced conformational changes within the atropine-activatable mutant M3RΔi3 constructs are recognized by the chimaeric yeast G protein as an agonistic signal more efficiently when compared with endogenous mammalian Gq. This would support the hypothesis that the pharmacological property of a ligand, to be an agonist or antagonist, is not only determined by certain ligand/receptor interactions, but also by certain properties of the receptor/G protein interplay.

Provoked by the discrepancy between Ki values of atropine in the nanomolar range but EC50 values in the micromolar range, we postulated the presence of a second binding site which mediates the agonistic action of atropine. Strikingly, we showed that blockers like atropine and scopolamine functioned as antagonists of CCh-triggered receptor activation, but at higher concentrations activated the mutant receptor (Figure 6). It has been shown that the adrenergic ligand CGP 12177 can act as high-affinity antagonist and low-affinity agonist at the wt β1 adrenergic receptor [41]. Similarly, the apparent concentration–response curve of atropine on CCh-activated M3RΔi3 Q490L/I495N (Figure 6C) is the summation of the ligand inhibition curve (Figure 6A) and the ligand activation curve (Figure 3A). To explain the experimental findings, the existence of two agonistic sites on the β1 adrenoceptor or at least more than one conformation of the agonistic catecholamine site was discussed [32]. In this present study, we show for the first time that this unusual pharmacological property can be induced by mutations.

Many key mutations that switch atropine to an agonist are located at the ICL3/TMD6 junction, away from the binding pocket. Similarly, agonist/antagonist switching mutations in the β1 adrenergic receptor were found outside of the proposed binding pocket [42]. The ICL3/TMD6 transition has been mainly implicated in the mechanics of receptor/G-protein interactions [5,4346]. Therefore one must assume that the structural changes leading to the second agonistic binding site are indirect or that the mutations now enable a pre-formed second binding site to induce G-protein activation when occupied by atropine. Nevertheless, our functional results obtained in yeast suggest that the mutations within the ICL3/TMD6 transition permit translation of structural movements into G-protein activation on atropine binding to a second binding site.

In summary, previous studies have already demonstrated that GPCRs with new ligand-binding and signal-transduction abilities created by sophisticated site-directed mutagenesis [15,47] and in vitro selection [21] can lead to powerful tools for biomedical research. Our study extends these opportunities by showing that GPCR architecture still provides room for the generation of additional ligand-binding sites. As shown in this present study with the mAChR, ligand binding to such new allosteric sites can activate the receptor even when the receptor is occupied by an inverse agonist at the orthosteric binding site. Receptors with a designed additional agonistic binding site may be of value as biosensors or ligand-concentration-dependent off/on bioswitches.

We thank Professor Jürgen Wess [NIDDK (National Institute of Diabetes and Digestive and Kidney Disease), NIH (National Institutes of Health), Bethesda, MD, U.S.A.)], Professor Roland Seifert (Department of Pharmacology and Toxicology, University of Regensburg, Germany), Dr Holger Römpler (Institute of Biochemistry, Medical Faculty, University of Leipzig, Germany) and the anonymous reviewers for critical reading of the manuscript and helpful suggestions. We thank Dr Erich Schneider (Department of Pharmacology and Toxicology, University of Regensburg, Germany) for helping with the GTPase assay, and Dr Mark Pausch (Wyeth Research, Princeton, NJ, U.S.A.) for the p416GPD yeast vector and for the S. cerevisiae MPY578q5 yeast strain. This work was supported by the Deutsche Forschungsgemeinschaft (Sfb 610) and the IZKF (Interdisziplinäres Zentrum für Klinische Forschung) Leipzig.

Abbreviations

     
  • ACh

    acetylcholine

  •  
  • 3-AT

    3-amino-1,2,4-triazole

  •  
  • butylscopolamine

    n-butyl scopolamine bromide

  •  
  • CCh

    carbachol

  •  
  • 4-DAMP

    4-diphenylacetoxy-N-methylpiperidine methobromide

  •  
  • DMEM

    Dulbecco's modified Eagle's medium

  •  
  • ECL

    extracellular loop

  •  
  • GPCR

    G-protein-coupled receptor

  •  
  • GTP[S]

    guanosine 5′-[γ-thio]triphosphate

  •  
  • ICL

    intracellular loop

  •  
  • IP

    inositol phosphate

  •  
  • mAChR

    muscarinic ACh receptor

  •  
  • M3R etc.

    M3 mAChR etc.

  •  
  • M3RΔi3

    M3R lacking ICL3

  •  
  • NMS

    N-methylscopolamine

  •  
  • p-F-HHSiD

    para-fluoro-hexahydrosila-difenidol

  •  
  • RASSL

    receptor activated solely by synthetic ligands

  •  
  • TMD

    transmembrane domain

  •  
  • U

    synthetic dropout medium lacking uracil

  •  
  • U/H

    synthetic dropout medium lacking both uracil and histidine

  •  
  • wt

    wild-type

  •  
  • YPAD

    yeast extract/peptone/dextrose medium with adenine

References

References
1
Wess
 
J.
Blin
 
N.
Mutschler
 
E.
Blüml
 
K.
 
Muscarinic acetylcholine receptors: structural basis of ligand binding and G protein coupling
Life Sci.
1995
, vol. 
56
 (pg. 
915
-
922
)
2
Onali
 
P.
Adem
 
A.
Karlsson
 
E.
Olianas
 
M. C.
 
The pharmacological action of MT-7
Life Sci.
2005
, vol. 
76
 (pg. 
1547
-
1552
)
3
Kristiansen
 
K.
 
Molecular mechanisms of ligand binding, signaling, and regulation within the superfamily of G-protein-coupled receptors: molecular modeling and mutagenesis approaches to receptor structure and function
Pharmacol. Ther.
2004
, vol. 
103
 (pg. 
21
-
80
)
4
Ghanouni
 
P.
Gryczynski
 
Z.
Steenhuis
 
J. J.
Lee
 
T. W.
Farrens
 
D. L.
Lakowicz
 
J. R.
Kobilka
 
B. K.
 
Functionally different agonists induce distinct conformations in the G protein coupling domain of the β2 adrenergic receptor
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
24433
-
24436
)
5
Ghanouni
 
P.
Steenhuis
 
J. J.
Farrens
 
D. L.
Kobilka
 
B. K.
 
Agonist-induced conformational changes in the G-protein-coupling domain of the β2 adrenergic receptor
Proc. Natl. Acad. Sci. U.S.A.
2001
, vol. 
98
 (pg. 
5997
-
6002
)
6
Takezako
 
T.
Gogonea
 
C.
Saad
 
Y.
Noda
 
K.
Karnik
 
S. S.
 
“Network leaning” as a mechanism of insurmountable antagonism of the angiotensin II type 1 receptor by non-peptide antagonists
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
15248
-
15257
)
7
Vilardaga
 
J. P.
Steinmeyer
 
R.
Harms
 
G. S.
Lohse
 
M. J.
 
Molecular basis of inverse agonism in a G protein-coupled receptor
Nat. Chem. Biol.
2005
, vol. 
1
 (pg. 
25
-
28
)
8
Li
 
J. H.
Han
 
S. J.
Hamdan
 
F. F.
Kim
 
S. K.
Jacobson
 
K. A.
Bloodworth
 
L. M.
Zhang
 
X.
Wess
 
J.
 
Distinct structural changes in a G protein-coupled receptor caused by different classes of agonist ligands
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
26284
-
26293
)
9
Perez
 
D. M.
Karnik
 
S. S.
 
Multiple signaling states of G-protein-coupled receptors
Pharmacol. Rev.
2005
, vol. 
57
 (pg. 
147
-
161
)
10
Swaminath
 
G.
Deupi
 
X.
Lee
 
T. W.
Zhu
 
W.
Thian
 
F. S.
Kobilka
 
T. S.
Kobilka
 
B.
 
Probing the β2 adrenoceptor binding site with catechol reveals differences in binding and activation by agonists and partial agonists
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
22165
-
22171
)
11
Ault
 
A. D.
Broach
 
J. R.
 
Creation of GPCR-based chemical sensors by directed evolution in yeast
Protein Eng. Des. Sel.
2006
, vol. 
19
 (pg. 
1
-
8
)
12
Gerber
 
B. O.
Meng
 
E. C.
Dotsch
 
V.
Baranski
 
T. J.
Bourne
 
H. R.
 
An activation switch in the ligand binding pocket of the C5a receptor
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
3394
-
3400
)
13
Blaker
 
M.
Ren
 
Y.
Gordon
 
M. C.
Hsu
 
J. E.
Beinborn
 
M.
Kopin
 
A. S.
 
Mutations within the cholecystokinin-B/gastrin receptor ligand ‘pocket’ interconvert the functions of nonpeptide agonists and antagonists
Mol. Pharmacol.
1998
, vol. 
54
 (pg. 
857
-
863
)
14
Claeysen
 
S.
Joubert
 
L.
Sebben
 
M.
Bockaert
 
J.
Dumuis
 
A.
 
A single mutation in the 5-HT4 receptor (5-HT4-R D100(3.32)A) generates a Gs-coupled receptor activated exclusively by synthetic ligands (RASSL)
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
699
-
702
)
15
Coward
 
P.
Wada
 
H. G.
Falk
 
M. S.
Chan
 
S. D.
Meng
 
F.
Akil
 
H.
Conklin
 
B. R.
 
Controlling signaling with a specifically designed Gi-coupled receptor
Proc. Natl. Acad. Sci. U.S.A.
1998
, vol. 
95
 (pg. 
352
-
357
)
16
Pausch
 
M. H.
 
G-protein-coupled receptors in Saccharomyces cerevisiae: high-throughput screening assays for drug discovery
Trends Biotechnol.
1997
, vol. 
15
 (pg. 
487
-
494
)
17
Celic
 
A.
Connelly
 
S. M.
Martin
 
N. P.
Dumont
 
M. E.
 
Intensive mutational analysis of G protein-coupled receptors in yeast
Methods Mol. Biol.
2004
, vol. 
237
 (pg. 
105
-
120
)
18
Beukers
 
M. W.
Ijzerman
 
A. P.
 
Techniques: how to boost GPCR mutagenesis studies using yeast
Trends Pharmacol. Sci.
2005
, vol. 
26
 (pg. 
533
-
539
)
19
Geva
 
A.
Lassere
 
T. B.
Lichtarge
 
O.
Pollitt
 
S. K.
Baranski
 
T. J.
 
Genetic mapping of the human C5a receptor. Identification of transmembrane amino acids critical for receptor function
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
35393
-
35401
)
20
Li
 
B.
Nowak
 
N. M.
Kim
 
S. K.
Jacobson
 
K. A.
Bagheri
 
A.
Schmidt
 
C.
Wess
 
J.
 
Random mutagenesis of the M3 muscarinic acetylcholine receptor expressed in yeast: identification of second-site mutations that restore function to a coupling-deficient mutant M3 receptor
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
5664
-
5675
)
21
Armbruster
 
B. N.
Li
 
X.
Pausch
 
M. H.
Herlitze
 
S.
Roth
 
B. L.
 
Evolving the lock to fit the key to create a family of G protein-coupled receptors potently activated by an inert ligand
Proc. Natl. Acad. Sci. U.S.A.
2007
, vol. 
104
 (pg. 
5163
-
5168
)
22
Schöneberg
 
T.
Liu
 
J.
Wess
 
J.
 
Plasma membrane localization and functional rescue of truncated forms of a G protein-coupled receptor
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
18000
-
18006
)
23
Zeng
 
F. Y.
Söldner
 
A.
Schöneberg
 
T.
Wess
 
J.
 
Conserved extracellular cysteine pair in the M3 muscarinic acetylcholine receptor is essential for proper receptor cell surface localization but not for G protein coupling
J. Neurochem.
1999
, vol. 
72
 (pg. 
2404
-
2414
)
24
Maggio
 
R.
Barbier
 
P.
Fornai
 
F.
Corsini
 
G. U.
 
Functional role of the third cytoplasmic loop in muscarinic receptor dimerization
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
31055
-
31060
)
25
Erlenbach
 
I.
Kostenis
 
E.
Schmidt
 
C.
Hamdan
 
F. F.
Pausch
 
M. H.
Wess
 
J.
 
Functional expression of M1, M3 and M5 muscarinic acetylcholine receptors in yeast
J. Neurochem.
2001
, vol. 
77
 (pg. 
1327
-
1337
)
26
Oldenburg
 
K. R.
Vo
 
K. T.
Michaelis
 
S.
Paddon
 
C.
 
Recombination-mediated PCR-directed plasmid construction in vivo in yeast
Nucleic Acids Res.
1997
, vol. 
25
 (pg. 
451
-
452
)
27
Berridge
 
M. J.
 
Rapid accumulation of inositol trisphosphate reveals that agonists hydrolyse polyphosphoinositides instead of phosphatidylinositol
Biochem. J.
1983
, vol. 
212
 (pg. 
849
-
858
)
28
Nelson
 
C. P.
Nahorski
 
S. R.
Challiss
 
R. A.
 
Constitutive activity and inverse agonism at the M2 muscarinic acetylcholine receptor
J. Pharmacol. Exp. Ther.
2006
, vol. 
316
 (pg. 
279
-
288
)
29
Ward
 
S. D.
Hamdan
 
F. F.
Bloodworth
 
L. M.
Wess
 
J.
 
Conformational changes that occur during M3 muscarinic acetylcholine receptor activation probed by the use of an in situ disulfide cross-linking strategy
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
2247
-
2257
)
30
Preuss
 
H.
Ghorai
 
P.
Kraus
 
A.
Dove
 
S.
Buschauer
 
A.
Seifert
 
R.
 
Mutations of Cys-17 and Ala-271 in the human histamine H2 receptor determine the species selectivity of guanidine-type agonists and increase constitutive activity
J. Pharmacol. Exp. Ther.
2007
, vol. 
321
 (pg. 
975
-
982
)
31
Schwartz
 
T. W.
Holst
 
B.
 
Allosteric enhancers, allosteric agonists and agoallosteric modulators: where do they bind and how do they act?
Trends Pharmacol. Sci.
2007
, vol. 
28
 (pg. 
366
-
373
)
32
Baker
 
J. G.
Hill
 
S. J.
 
Multiple GPCR conformations and signalling pathways: implications for antagonist affinity estimates
Trends Pharmacol. Sci.
2007
, vol. 
28
 (pg. 
374
-
381
)
33
Wess
 
J.
Liu
 
J.
Blin
 
N.
Yun
 
J.
Lerche
 
C.
Kostenis
 
E.
 
Structural basis of receptor/G protein coupling selectivity studied with muscarinic receptors as model systems
Life Sci.
1997
, vol. 
60
 (pg. 
1007
-
1014
)
34
Schmidt
 
C.
Li
 
B.
Bloodworth
 
L.
Erlenbach
 
I.
Zeng
 
F. Y.
Wess
 
J.
 
Random mutagenesis of the M3 muscarinic acetylcholine receptor expressed in yeast. Identification of point mutations that “silence” a constitutively active mutant M3 receptor and greatly impair receptor/G protein coupling
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
30248
-
30260
)
35
Burstein
 
E. S.
Spalding
 
T. A.
Brann
 
M. R.
 
Pharmacology of muscarinic receptor subtypes constitutively activated by G proteins
Mol. Pharmacol.
1997
, vol. 
51
 (pg. 
312
-
319
)
36
Pausch
 
M. H.
Lai
 
M.
Tseng
 
E.
Paulsen
 
J.
Bates
 
B.
Kwak
 
S.
 
Functional expression of human and mouse P2Y12 receptors in Saccharomyces cerevisiae
Biochem. Biophys. Res. Commun.
2004
, vol. 
324
 (pg. 
171
-
177
)
37
Niebauer
 
R. T.
Gao
 
Z. G.
Li
 
B.
Wess
 
J.
Jacobson
 
K. A.
 
Signaling of the human P2Y(1) receptor measured by a yeast growth assay with comparisons to assays of phospholipase C and calcium mobilization in 1321N1 human astrocytoma cells
Purinergic Signal.
2005
, vol. 
1
 (pg. 
241
-
247
)
38
Spalding
 
T. A.
Burstein
 
E. S.
Henderson
 
S. C.
Ducote
 
K. R.
Brann
 
M. R.
 
Identification of a ligand-dependent switch within a muscarinic receptor
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
21563
-
21568
)
39
Allman
 
K.
Page
 
K. M.
Curtis
 
C. A.
Hulme
 
E. C.
 
Scanning mutagenesis identifies amino acid side chains in transmembrane domain 5 of the M1 muscarinic receptor that participate in binding the acetyl methyl group of acetylcholine
Mol. Pharmacol.
2000
, vol. 
58
 (pg. 
175
-
184
)
40
Li
 
B.
Scarselli
 
M.
Knudsen
 
C. D.
Kim
 
S. K.
Jacobson
 
K. A.
McMillin
 
S. M.
Wess
 
J.
 
Rapid identification of functionally critical amino acids in a G protein-coupled receptor
Nat. Methods.
2007
, vol. 
4
 (pg. 
169
-
174
)
41
Baker
 
J. G.
Hall
 
I. P.
Hill
 
S. J.
 
Agonist actions of “beta-blockers” provide evidence for two agonist activation sites or conformations of the human β1-adrenoceptor
Mol. Pharmacol.
2003
, vol. 
63
 (pg. 
1312
-
1321
)
42
Behr
 
B.
Hoffmann
 
C.
Ottolina
 
G.
Klotz
 
K. N.
 
Novel mutants of the human β1-adrenergic receptor reveal amino acids relevant for receptor activation
J. Biol. Chem.
2006
, vol. 
281
 (pg. 
18120
-
18125
)
43
Ward
 
S. D.
Hamdan
 
F. F.
Bloodworth
 
L. M.
Siddiqui
 
N. A.
Li
 
J. H.
Wess
 
J.
 
Use of an in situ disulfide cross-linking strategy to study the dynamic properties of the cytoplasmic end of transmembrane domain VI of the M3 muscarinic acetylcholine receptor
Biochemistry.
2006
, vol. 
45
 (pg. 
676
-
685
)
44
Altenbach
 
C.
Yang
 
K.
Farrens
 
D. L.
Farahbakhsh
 
Z. T.
Khorana
 
H. G.
Hubbell
 
W. L.
 
Structural features and light-dependent changes in the cytoplasmic interhelical E-F loop region of rhodopsin: a site-directed spin-labeling study
Biochemistry.
1996
, vol. 
35
 (pg. 
12470
-
12478
)
45
Jensen
 
A. D.
Guarnieri
 
F.
Rasmussen
 
S. G.
Asmar
 
F.
Ballesteros
 
J. A.
Gether
 
U.
 
Agonist-induced conformational changes at the cytoplasmic side of transmembrane segment 6 in the β2 adrenergic receptor mapped by site-selective fluorescent labeling
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
9279
-
9290
)
46
Rasmussen
 
S. G.
Jensen
 
A. D.
Liapakis
 
G.
Ghanouni
 
P.
Javitch
 
J. A.
Gether
 
U.
 
Mutation of a highly conserved aspartic acid in the β2 adrenergic receptor: constitutive activation, structural instability, and conformational rearrangement of transmembrane segment 6
Mol. Pharmacol.
1999
, vol. 
56
 (pg. 
175
-
184
)
47
Scearce-Levie
 
K.
Coward
 
P.
Redfern
 
C. H.
Conklin
 
B. R.
 
Engineering receptors activated solely by synthetic ligands (RASSLs)
Trends Pharmacol. Sci.
2001
, vol. 
22
 (pg. 
414
-
420
)
48
Cheng
 
Y.
Prusoff
 
W. H.
 
Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction
Biochem. Pharmacol.
1973
, vol. 
22
 (pg. 
3099
-
3108
)

Supplementary data