An Antibody Biosensor Establishes the Activation of the M1 Muscarinic Acetylcholine Receptor during Learning and Memory*♦

Establishing the in vivo activation status of G protein-coupled receptors would not only indicate physiological roles of G protein-coupled receptors but would also aid drug discovery by establishing drug/receptor engagement. Here, we develop a phospho-specific antibody-based biosensor to detect activation of the M1 muscarinic acetylcholine receptor (M1 mAChR) in vitro and in vivo. Mass spectrometry phosphoproteomics identified 14 sites of phosphorylation on the M1 mAChR. Phospho-specific antibodies to four of these sites established that serine at position 228 (Ser228) on the M1 mAChR showed extremely low levels of basal phosphorylation that were significantly up-regulated by orthosteric agonist stimulation. In addition, the M1 mAChR-positive allosteric modulator, 1-(4-methoxybenzyl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid, enhanced acetylcholine-mediated phosphorylation at Ser228. These data supported the hypothesis that phosphorylation at Ser228 was an indicator of M1 mAChR activation. This was further supported in vivo by the identification of phosphorylated Ser228 on the M1 mAChR in the hippocampus of mice following administration of the muscarinic ligands xanomeline and 1-(4-methoxybenzyl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid. Finally, Ser228 phosphorylation was seen to increase in the CA1 region of the hippocampus following memory acquisition, a response that correlated closely with up-regulation of CA1 neuronal activity. Thus, determining the phosphorylation status of the M1 mAChR at Ser228 not only provides a means of establishing receptor activation following drug treatment both in vitro and in vivo but also allows for the mapping of the activation status of the M1 mAChR in the hippocampus following memory acquisition thereby establishing a link between M1 mAChR activation and hippocampus-based memory and learning.

Establishing the in vivo activation status of G protein-coupled receptors would not only indicate physiological roles of G protein-coupled receptors but would also aid drug discovery by establishing drug/receptor engagement. Here, we develop a phospho-specific antibody-based biosensor to detect activation of the M 1 muscarinic acetylcholine receptor (M 1 mAChR) in vitro and in vivo. Mass spectrometry phosphoproteomics identified 14 sites of phosphorylation on the M 1 mAChR. Phosphospecific antibodies to four of these sites established that serine at position 228 (Ser 228 ) on the M 1 mAChR showed extremely low levels of basal phosphorylation that were significantly up-regulated by orthosteric agonist stimulation. In addition, the M 1 mAChR-positive allosteric modulator, 1-(4-methoxybenzyl)-4oxo-1,4-dihydroquinoline-3-carboxylic acid, enhanced acetylcholine-mediated phosphorylation at Ser 228 . These data supported the hypothesis that phosphorylation at Ser 228 was an indicator of M 1 mAChR activation. This was further supported in vivo by the identification of phosphorylated Ser 228 on the M 1 mAChR in the hippocampus of mice following administration of the muscarinic ligands xanomeline and 1-(4-methoxybenzyl)-4oxo-1,4-dihydroquinoline-3-carboxylic acid. Finally, Ser 228 phosphorylation was seen to increase in the CA1 region of the hippocampus following memory acquisition, a response that correlated closely with up-regulation of CA1 neuronal activity. Thus, determining the phosphorylation status of the M 1 mAChR at Ser 228 not only provides a means of establishing receptor activation following drug treatment both in vitro and in vivo but also allows for the mapping of the activation status of the M 1 mAChR in the hippocampus following memory acquisition thereby establishing a link between M 1 mAChR activation and hippocampus-based memory and learning.
G protein-coupled receptors (GPCRs) 3 respond to the binding of their cognate ligands by transitioning from an inactive to an active conformation capable of engaging with intracellular signaling cascades (1)(2)(3). Whereas this process has been described in exquisite pharmacological detail (4,5), and biophysically in recent crystal structures (2,6), correlating the activation state of a GPCR subtype in vivo to a physiological response or drug treatment is extremely challenging and presents a considerable barrier to establishing the physiological role of GPCRs and the on-target action of GPCR ligands.
Progress in this area has been made in transfected systems where it has been possible to monitor receptor conformational changes in response to ligand occupation using fluorescent resonance energy transfer (FRET). In these studies, changes in the energy transfer between FRET-acceptor and FRET-donor moieties, engineered within the receptor sequence, provide a readout of changes in receptor conformation on agonist binding (7)(8)(9)(10). Alternatively, a green fluorescent protein (GFP) biosensor based on a conformation-sensitive antibody (nanobody-80, Nb80) that preferentially recognizes the active state of the ␤ 2 -adrenoreceptor has recently been employed to determine the active conformation of the ␤ 2 -adrenoreceptor at the plasma membrane and within intracellular compartments (11). These approaches, however, require transfection of either mutated receptors (7)(8)(9)(10) or a GFP biosensor (11), and therefore, although able to monitor receptor conformational changes quantitatively in living cells in real time, are restricted to heterologous systems.
An alternative approach considered here is to monitor the phosphorylation status of GPCRs as a read-out of receptor activation. This is based on the "classical" principle that the conformation adopted by a receptor upon agonist occupation reveals phosphorylation sites, often within the third intracellular loop and C-terminal tail, that otherwise are not accessible in the inactive receptor conformation (12). In this scenario, the phosphorylation status of a particular GPCR may serve as a read-out of the proportion of receptors that have adopted an active conformation. Therefore, phospho-specific antibodies to agonist-dependent receptor phosphorylation events could potentially be used as a probe for the activated receptor in not only recombinant systems, but also in physiologically relevant tissues. If this were the case, then it might be possible to correlate the activation status of GPCRs with physiological responses and, importantly for drug discovery, could be used to assess receptor engagement with synthetic ligands.
We test this notion here by focusing on the M 1 -muscarinic acetylcholine receptor (M 1 mAChR), which is one of five muscarinic receptor subtypes (M 1 -M 5 ) that respond to the natural ligand acetylcholine and is a subtype that has been implicated in a number of neurological processes (13,14) most notably learning and memory (15)(16)(17). We have shown previously that this receptor subtype is rapidly phosphorylated by agonist addition likely via members of the G protein-coupled receptor kinase family (18), although the involvement of other receptor kinases has not been ruled out. Here, we used mass spectrometry-based phosphoproteomics to determine the sites of receptor phosphorylation from which we developed a series of phospho-specific antibodies. This included an antibody to phosphoserine 228 (Ser(P) 228 ) in the third intracellular loop, which we show is a phosphorylation event highly sensitive to agonist stimulation. This antibody was used here to probe the phosphorylation status of the M 1 mAChR following engagement with orthosteric and allosteric muscarinic ligands, both in vitro and in vivo. Furthermore, this antibody was used to determine that the M 1 mAChR was activated in specific regions of the hippocampus following fear conditioning in a manner that maps to the regions of the hippocampus showing increased synaptic activity during memory acquisition. In this way we not only link the activation status of the M 1 mAChR to memory acquisition but also establish the principle that phosphorylation sites can be used to probe the activation status of GPCRs during physiological responses and on drug treatment.

Experimental Procedures
Materials-Unless otherwise stated, all chemicals were purchased from Sigma. Xanomeline was obtained from Lilly (Erl Wood Manor, Windlesham, Surrey, UK).
Generation of Cell Lines-Chinese hamster ovary (CHO) cells that stably and constitutively expressed the C-terminal HA epitope-tagged mouse M 1 mAChR (CHO-M1 cells) were generated using the Flp-In TM system. CHO Flp-In cells were cotransfected with pcDNA5FRT containing M 1 mAChR and pOG44, and transfected cells were selected with hygromycin B, and expression of M 1 mAChR was confirmed by immunoblotting with anti-HA antibodies and by N-[ 3 H]methylscopolamine (NMS) radioligand binding.
[ 32 P]Orthophosphate Labeling and Immunoprecipitation-Cells were plated in 6-well plates at 200,000 cells/well 24 h  before experimentation. For phosphorylation experiments,  cells were washed three times with Krebs/HEPES buffer without phosphate (118 mM NaCl, 1.3 mM CaCl 2 , 4.3 mM KCl,1.17 MgSO 4 , 4.17 mM NaHCO 3 , 11.7 mM glucose, 10 mM HEPES (pH 7.4)) and incubated in this buffer containing 100 Ci/ml [ 32 P]orthophosphate for 1 h at 37°C. Cells were stimulated for 5 min with test compounds and immediately lysed by addition of buffer containing 20 mM Tris (pH 7.4), 150 mM NaCl, 3 mM EDTA, 1% Nonidet P-40, 0.5% sodium deoxycholate. The M 1 mAChR was immunoprecipitated from the cleared lysates using anti-HA affinity matrix (Roche Applied Science). The washed immunoprecipitates were separated by SDS-PAGE on 8% gels that were then dried, and radioactive bands were revealed using autoradiography film. The films were scanned and bands were quantified using AlphaImager software (Alpha Innotech, San Leandro, CA). M 1 mAChR Purification and Mass Spectrometry-Stably transfected CHO-M1 cells expressing mouse HA-tagged M 1 mAChR were grown until confluent in expanded surface rolling bottles at 0.25 rpm in a humidified CO 2 incubator. For receptor purification, cells from four rolling bottles were harvested, resuspended in 40 ml of Krebs/HEPES buffer, and stimulated with 100 M acetylcholine for 5 min. Membranes were then prepared and solubilized by addition of 5 ml of PBS containing 1% Nonidet P-40 plus a mixture of protease and phosphatase inhibitors (Complete, Roche Applied Science). After centrifugation at 20,000 ϫ g, the resulting supernatant was diluted 1:1 with PBS, and the receptor was then purified on anti-HA affinity matrix. After extensive washing with solubilization buffer containing 0.5% Nonidet P-40, the resin was resuspended in 2ϫ SDS-PAGE sample buffer. The sample was resolved by SDS-PAGE on 10% gels and stained with colloidal Coomassie Blue. The band associated with the M 1 mAChRs was excised from the polyacrylamide and washed three times for 15 min with 100 mM triethylammonium bicarbonate (TEAB). Reduction and alkylation of cysteines were performed by addition of 10 mM dithiothreitol in 50 mM TEAB at 60°C for 30 min followed by addition of 100 mM iodoacetamide in 50 mM TEAB for 30 min in the dark. Gel slices were washed three times for 5 min with 50 mM TEAB containing 50% acetonitrile, finally resuspended in TEAB containing 10% acetonitrile, and incubated overnight at 37°C with 1 g of sequencing grade trypsin (Promega, Southampton, UK). The resulting tryptic peptides were dried and resuspended in 1 ml of buffer containing 250 mM acetic acid and 30% acetonitrile, and phosphorylated peptides were enriched by addition of 20 l of PHOS-Select TM iron affinity resin and incubation at room temperature for 2 h with mixing. After washing the resin twice with loading buffer and once with water, tryptic phosphopeptides were eluted by addition of 200 l of buffer containing 400 mM ammonium hydroxide and 30% acetonitrile. LC-MS/MS was carried out on each sample using an LTQ Orbitrap mass spectrometer (Thermo Fisher Scientific, Rockford, IL). Peptides resulting from in-gel digestion were loaded at a high flow rate onto a reverse-phase trapping column (0.3-mm inner diameter ϫ 1 mm containing 5 m C 18 300-Å Acclaim PepMap medium (Dionex, UK) and eluted through a reverse phase capillary column (75-m inner diameter ϫ 150 mm) containing Symmetry C 18 100-Å medium (Waters, Elsetree, UK) that was self-packed using a high pressure packing device (Proxeon Biosystems, Odense, Denmark). The resulting spectra were searched against the UniProtKB/Swiss-Prot database using Mascot (Matrix Science Ltd.) software with peptide tolerance set to 5 ppm and the MS/MS tolerance set to 0.6 Da. Fixed modifications were set as carbamidomethylcysteine with variable modifications of phosphoserine, phosphothreonine, phosphotyrosine, and oxidized methionine. The enzyme was set to trypsin/Pro, and up to two missed cleavages were allowed. Peptides with a Mascot score greater than 20 and for which the probability (p) that the observed match was a random event was Ͻ0.05 were included in the analysis. The spectra of peptides reported as being phosphorylated were interrogated manually to confirm the precise sites of phosphorylation. . Nonspecific binding and filtration were carried out as above. Membrane-bound ligand was separated from free ligand by rapid filtration onto GF/B glass microfiber filters followed by three rapid washes with ice-cold 0.9% NaCl. Membrane bound radioactivity was determined by liquid scintillation (Ultima Gold, PerkinElmer Life Sciences) counting. Nonspecific binding was determined by the inclusion of atropine (1 M) during the incubation with [ 3 H]NMS. M 1 mAChR Immunoprecipitation from Tissue Samples-To establish the presence and phosphorylation status of M 1 mAChRs in preparations from the hippocampus, tissues from adult wild-type or adult M 1 mAChR knock-out mice on a C57Bl6/NTAC background were dissected into ice-cold Hanks' balanced salt solution buffer containing protease inhibitors and phosphatase inhibitors (Complete, Roche Diagnostics). Membranes were prepared and solubilized in buffer composed of 20 mM Tris (pH 7.4), 150 mM NaCl, 3 mM EDTA, and 1% Nonidet P-40. The receptor was immunoprecipitated with the anti M 1 mAChR polyclonal antibody. Immune complexes were washed three times in solubilization buffer and resuspended in 2ϫ SDS-PAGE sample buffer. Receptors were separated by SDS-PAGE on 8% gels, transferred PVDF membranes, and immuno-blotted with anti-phosphoserine 228 antibodies or antibodies against the M 1 mAChR protein.
Immunocytochemistry-CHO-M1 cells expressing mouse HA-tagged M 1 AChR were seeded onto 20-mm glass coverslips for 24 h prior to experimentation. Cells were washed and incubated for 1 h in Krebs/HEPES buffer prior to treatment. Cells were fixed in PBS containing 4% paraformaldehyde and 0.1% glutaraldehyde for 30 min at room temperature. Anti-HA antibody was used at 50 ng/ml followed by Alexa Fluor TM 546 goat anti rat secondary antibody at 1:1000. Anti-phospho-specific serine 228 antibodies were used at 0.25 g/ml followed by goat anti-rabbit Alexa Fluor TM 488 secondary antibody at 1:1000. Data were acquired using an Axiovert 200 M confocal laser scanning microscope (Zeiss).
Total [ 3 H]Inositol Phosphate Accumulation Assay-Cells seeded at 100,000 cells/well in 24-well plates were labeled with 2.5 Ci/ml myo-[ 3 H]inositol (PerkinElmer Life Sciences) for 24 h at 37°C. Cells were washed twice in Krebs/HEPES buffer (pH 7.4) and incubated with 10 mM LiCl for 20 min at 37°C. Appropriate concentrations of agonist were added for 20 min to stimulate [ 3 H]inositol phosphate production. Where the allosteric interactions were studied, cells were pre-incubated with the allosteric modulator for 2 min prior to the addition of agonist. Incubations were terminated by aspiration of buffer and rapid addition of 500 l of ice-cold trichloroacetic acid (TCA) (0.5 M). After extraction on ice for 20 -30 min, samples were transferred to tubes containing 100 l of EDTA (10 mM (pH 7.0)) and 500 l of a 1:1 mixture of tri-n-octylamine, and 1,1,2trichlorofluoroethane was added for 15 min at room temperature. After centrifugation at 20,000 ϫ g for 2 min, 400 l of the upper aqueous phase was transferred to fresh tubes containing 60 mM NaHCO 3 . [ 3 H]inositol mono-, bis-, and trisphosphate ([ 3 H]InsPx) fraction was recovered by anion-exchange chromatography on Dowex AG1-X8 formate columns. Columns were regenerated with 10 ml of ammonium formate (2 M)/formic acid (0.1 M) and washed thoroughly with distilled water. Samples were applied to the columns and the columns washed with 10 ml of distilled water. Columns were then washed with ammonium formate (60 mM)/sodium tetraborate (10 mM) solution. Total [ 3 H]InsPx was eluted in 10 ml of ammonium formate (0.75 M)/formic acid (0.1 M) and collected in large scintillation vials. A 5-ml aliquot from the eluate was mixed with 10 ml of SafeFluor scintillation mixture, and radioactivity was detected by liquid scintillation counting.
Fear Conditioning Training-Male C57Bl6/NTAC mice (8 -15 weeks old) were placed in the conditioning chamber (Stoelting ANY-maze fear conditioning system), and after a 2-min adaptation period, the mice received three tone/foot shock pairings, where a tone (conditioned stimulus, 2.8 kH, 85 db, 30 s) always co-terminated with the foot shock (unconditioned stimulus, 2 s, 0.4 mA). The conditioned stimulus-unconditioned stimulus pairings were separated by 1-min intervals. After completion of training, the mice remained in the conditioning chamber for 1 min and were returned to their home cages. For the immediate foot shock control, mice were put in the fear conditioning chamber, and a foot shock was immediately delivered (2 s, 0.4 mA); 30 s latter, they were returned to their home cages. Mice remained in their home cages for 30 min before being anesthetized with 3% isoflurane (2 liters/min O 2 ) and transcardially perfused with 4% PFA. Following fixation, brains were immediately removed and further fixed overnight in 4% PFA.
Immunohistochemistry of Mouse Brain-Brains were processed in paraffin wax and sliced at 5 m using a microtome. Following antigen retrieval, sections were washed in TBS containing 0.1% Triton X-100 and blocked for 2 h at room temperature in TBS, 0.1% Triton X-100, 10% goat serum, and 5% BSA. Sections were incubated with antibodies to c-Fos (Santa Cruz Biotechnology), ARC (Santa Cruz Biotechnology), and M 1 AChR phospho-specific serine 228 at 2.5 g/ml in blocking buffer (overnight at 4°C). Sections were washed three times, and incubated with Alexa Fluor TM 488 fluorescent secondary antibodies for 1 h at room temperature in blocking buffer. Following three washes, slices were mounted in Vectashield Hard-Set TM mounting medium with DAPI. All images were taken using a Zeiss confocal microscope with Zen software (Zeiss).
Assay for ERK Activity-Extracellular signal-regulated protein kinase 1/2 phosphorylation assay in CHO-M1 cells were seeded into transparent 96-well plates at 35,000 cells/well and grown overnight at 37°C. Cells were washed with PBS and incubated in serum-free ␣-minimal essential medium at 37°C for at least 4 h. Cells were incubated with varying concentrations of agonist in the presence and absence of increasing concentrations of an allosteric compound for 5 min (compounds were co-added). The reaction was terminated by the removal of compounds and addition of 50 l of SureFire lysis buffer. The lysates were agitated for 30 min at room temperature, and 4 l of each lysate was transferred into a 384-well opaque Optiplate. SureFire detection mix (7 l/well) consisting of detection reagent, activation reagent, donor beads, and acceptor beads (660:110:11:11 v/v) were added to the plate. Plates were incubated in the dark at room temperature for 2 h with gentle agitation before the fluorescence signal was measured using a PheraStar TM plate reader.
Generation of M1 DREADD Mice-Transgenic C57BL6/J mice, which express a humanized, mutated form of the M 1 mAChR, were generated by genOway (Lyon, France). Briefly, humanizing mutations V5A, S254T, K320R, G337A, and V413I were introduced into the mouse coding sequence to make the corresponding amino acid sequence identical to the human M 1 mAChR. In addition, two point mutations, Y106A and A195G, were introduced that render the M 1 mAChR insensitive to acetylcholine and promote sensitivity to the synthetic compound clozapine-N-oxide (CNO). This receptor mutant was termed the M1 DREADD receptor. Finally, an HA tag sequence YPY-DVPDYA was appended to the C terminus followed by a stop codon. The gene-targeting construct consisted of long and short homology arms. The long arm consisted of a 5268-bp fragment containing the 3Ј part of intron 2; this was generated by PCR from genomic DNA template using primers 85426GAlm 5Ј GGGGACAACTTTGTATAGAAAAGTTGTTGGGAGG-ATGCAGCCTCTCCA and 85427GAlm 5Ј GGGGACTGC-TTTTTTGTACAAACTTGGGCGCGCCACGTACGCCT-CTTCAAGGCTTAAGTGGAATGAAGGGGCAGCCCC. The short arm consisted of a 3188-bp fragment containing the 3ЈUTR sequence of exon 3 that was amplified from genomic DNA using primers 85424GAsm 5Ј ACTGTCAACCCCAT-GTGCTACGCAC and 85425GAsm 5Ј GGGCTACAAGGGA-GCATGAACAAGC. A diphtheria toxin negative selection cassette was included to allow for selection of ES cells that had undergone homologous recombination, and a positive selection neomycin gene and a STOP of transcription cassette flanked by loxP sites were inserted upstream of the CHRM1 coding sequence. This construct was transfected into C57BL6 ES cells that were selected using neomycin. Resistant colonies were screened to verify correct integration of the 5Ј and 3Ј homology arms by PCR using primers 85488 5Ј GCAGGTCGA-GGGACCTAATAACTTCG and 85489sa 5Ј AAGCTCTCAA-GGTCCAACAGTTTCTGG for short arm integration and primers 85487la 5Ј AGTCTTACCAGGGAAGAAAGCCTG-ATCC and 0070-Neo-16219sa 5Ј CCTGCTCTTTACTGAAG-GCTCTTTACTATTGC for long arm integration.
Positive clones were expanded and injected into blastocysts derived from C57BL6/J female mice to generate chimeric animals that were then bred with C57BL6 wild-type C57BL6 mice to generate heterozygous F1 animals. The neomycin/STOP cassette was excised by breeding heterozygous animals with Cre deleter mice to generate M1 DREADD constitutive knock-in mice. Genotyping was performed by PCR on genomic DNA obtained from tail biopsies using primers mM1 451 forward 5Ј TTG GTT TCC TTC GTT CTC TGG GC and mM1 1227 reverse 5Ј GAC GTA GCA AAG CCA GTA GCC CAG C. The resulting PCR products were subjected to restriction digestion with SacI, the products of which discriminated between WT C57BL6 and heterozygous and homozygous M1 DREADD animals.

Results
Determination of the Phosphorylation Sites on the M 1 mAChR-Our initial evaluation of the phosphorylation status of the M 1 mAChR was conducted on Chinese hamster ovary cells expressing the mouse M 1 mAChR (CHO-M1 cells) stimulated with the natural ligand acetylcholine or the M 1 /M 4 -preferring agonist xanomeline (Fig. 1A) (19,20). These two ligands stimulated inositol phosphate production in a concentration-dependent manner with acetylcholine and xanomeline both acting as full agonists (Fig. 1A).
In phosphorylation studies, where total receptor phosphorylation was determined using metabolically labeled CHO-M1 cells, both muscarinic agonists increased the phosphorylation status of M 1 mAChRs (Fig. 1B). To establish the precise sites of phosphorylation, a mass spectrometry-based phosphoproteomic study was conducted that identified nine serine and three threonine phosphorylation sites in the third intracellular loop of the receptor, as well as two serine phosphorylation sites in the C-terminal tail (Fig. 1, C-F). Four of these sites at serine residues 228, 273, 322, and 451 (Ser 228 , Ser 273 , Ser 322 , and Ser 451 ) were selected for the generation of phospho-specific antibodies based on the antigenicity of the sites.
The phospho-specific nature of the antibodies generated was tested by immunoprecipitation of M 1 mAChRs from CHO-M1 cells (using antibodies that recognized the HA epitope engineered at the C terminus of the receptor) followed by treatment of the immunoprecipitate with calf intestinal alkaline phosphatase (CIAP) to remove the phosphate groups ( Fig. 2A). Control immunoprecipitates were treated with vehicle. The immunoprecipitates were then probed in Western blots with the phos-pho-specific antibodies. In these experiments, treatment with CIAP removed the immunoreactivity for the M 1 mAChR of all four phospho-specific antibodies (Fig. 2B) confirming that the antibodies were all phospho-specific.
Agonist-dependent M 1 mAChR Phosphorylation-We next used the phospho-specific antibodies to determine whether phosphorylation on Ser 228 , Ser 273 , Ser 322 , and Ser 451 was regulated by agonist stimulation. Probing lysates derived from CHO-M1 cells stimulated with vehicle, acetylcholine, or xanomeline in Western blots using the phospho-specific antibodies revealed that there were low levels of basal phosphorylation on residues Ser 228 and Ser 273 but that agonist treatment significantly increased phosphorylation at these residues (Fig.  3). This is in contrast to phosphorylation at Ser 451 that showed notable basal phosphorylation that was only increased to a small extent in response to agonist treatment (Fig. 3). This was compared with phosphorylation at Ser 322 that demonstrated high levels of basal phosphorylation and showed no substantive change in response to agonist (Fig. 3). In this respect, our panel of phospho-specific antibodies revealed that different phosphorylation sites on the M 1 mAChR had different sensitivities to agonist stimulation.
Because the antibody used to detect phosphorylation at Ser 228 gave a strong signal and phosphorylation at this site was exquisitely agonist-sensitive, this antibody was considered an excellent candidate for a biosensor that would identify the activated M 1 mAChR receptor and as such was characterized further.
Determination of Changes in M 1 mAChR Phosphorylation in Response to an Allosteric Ligand-The above data demonstrated that the activated state of the M 1 mAChR in response to two orthosteric agonists (acetylcholine and xanomeline) could be established by monitoring the phosphorylation status of the receptor at Ser 228 . It is now well understood that the active  A, schematic representation of the experimental procedure used to characterize M 1 mAChR phosphorylation-specific antibodies where CIAP was used to dephosphorylate immunoprecipitated M 1 mAChR receptor derived from cells that had been treated with acetylcholine (ACh) before being probed in Western blots with the phospho-specific antibodies. B, results of Western blots from the experiments that are illustrated in A. Blots were probed with phospho-specific antibodies directed toward Ser(P) 228 , Ser(P) 273 , Ser(P) 322 , and Ser(P) 451 . Also shown is the loading control from this particular experiment using an anti-HA antibody to detect the epitope-tagged M 1 mAChR. The * indicates slight cross-reactivity of the phospho-specific antibodies with CIAP.
conformation of GPCRs can also be promoted by ligands that bind at allosteric sites (Fig. 4A) (21,22). Previous studies had established that the compound, 1-(4-methoxybenzyl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid (benzyl quinolone carboxylic acid; BQCA), is a selective positive allosteric modulator (PAM) to the M 1 mAChR (23)(24)(25). We confirm these studies by establishing in our system that BQCA can promote acetylcholine binding at the M 1 mAChR with a cooperativity factor of Ͼ100 (␣ factor ϭ 128; Fig. 4B). The enhanced binding of acetylcholine observed in the presence of BQCA was correlated with an augmentation of acetylcholine signaling as demonstrated by a leftward shift in the InsPx concentration-response curve to acetylcholine when conducted in the presence of BQCA (cooperativity ␣␤ factor ϭ 85; Fig. 4C). Hence, BQCA was seen to promote the active conformation of the M 1 mAChR, which was reflected in the ability of BQCA to augment agonist binding and agonist-mediated signaling.
Based on these observations, and given that phosphorylation at Ser 228 likely reflects the proportion of M 1 mAChRs in an agonist-occupied (R*) conformation, we reasoned that BQCA would promote acetylcholine-mediated phosphorylation at Ser 228 . This was indeed the case, with BQCA mediating an ϳ100-fold shift in the potency of acetylcholine-mediated phosphorylation at Ser 228 (pEC 50 (Ϫlog M) for acetylcholine ϭ 7.5 Ϯ 0.1 and pEC 50 for acetylcholine in the presence of BQCA (3 M) ϭ 8.9 Ϯ 0.3; Fig. 4E). Thus, phosphorylation at Ser 228 not only reflected the action of orthosteric ligands at the M 1 mAChR (Fig. 3), but also reported the action of an allosteric modulator at this receptor.
These phosphorylation experiments were conducted with BQCA at a concentration that in itself gave no response (i.e. 3 M), but rather it acted to enhance acetylcholine responses (i.e. a concentration where BQCA acted as a PAM). However, at higher concentrations, BQCA could be seen to act as an agonist in its own right (23,25). The agonist properties of BQCA can be readily determined in pERK1/2 assays that, due to the high coupling efficiency between the M 1 mAChR and pERK1/2 signaling in this system, were sufficiently sensitive to detect the agonist action of BQCA (Fig. 4D). Thus, high concentrations of BQCA in a pERK1/2 assay showed a direct agonist action (Fig.  4D). This agonist activity was reflected in complex concentration-response curves, where increasing concentrations of BQCA in the presence of very low concentrations of acetylcholine revealed the agonist activity of BQCA (green arrow, Fig.  4D), whereas higher concentrations of acetylcholine revealed mixed agonist and PAM activities of BQCA (black arrow, Fig.  4D). Importantly, when examined in phosphorylation studies, it was seen that at a high concentration (100 M) BQCA mediated phosphorylation at Ser 228 to a similar extent as that seen with a maximal concentration of acetylcholine (500 M) (Fig. 4F). In contrast, a low concentration of BQCA (3 M) alone did not stimulate any change in receptor phosphorylation (Fig. 4F).
Phosphorylation at Ser 228 was further characterized in immunocytochemistry studies. In these studies, the Ser 228 phospho-specific antibody detected very little immunoreactivity in non-transfected CHO cells and in CHO-M1 cells under basal conditions (Fig. 4G). However, following agonist treatment (acetylcholine, 5 min) or treatment with a high concentration of BQCA (100 M), punctate cytoplasmic staining was detected (Fig. 4G). Interestingly, the punctate localization of the phosphorylated M 1 mAChR represented only a small proportion of the total receptor population (identified using the antibodies against the HA epitope tag) that, after a 5-min stimulation, appeared concentrated at or very close to the plasma membrane (Fig. 4G).
By combining all these data, it can be concluded that muscarinic ligands that promote the transition of the receptor from an inactive R-state to an active R*-state promote phosphorylation of the M 1 mAChR on Ser 228 . This includes orthosteric ligands, such as acetylcholine and xanomeline, as well as allosteric ligands (i.e. BQCA). In this sense, Ser 228 phosphorylation is an ideal sensor of receptor activation, because this phosphorylation event reflects activation of the receptor regardless of the mechanism of activation.
Determination of the Active Conformation of the M 1 mAChR during Memory Acquisition and Drug Treatment-The data presented above support the notion that phosphorylation of Ser 228 on the M 1 mAChR can act as a biosensor for receptor activation in response to either orthosteric or allosteric ligands. To test whether this could be used to monitor the activation status of the M 1 mAChR in vivo, we first employed a novel mouse model where the wild-type M 1 mAChR gene was humanized so that the translated amino acid sequence was identical to the human M1 mAChR and where an HA epitope tag was inserted at the C terminus. Two further point mutations were introduced into the orthosteric binding site (Y106A and A195G) (Fig. 5A), which resulted in an HA-tagged humanized M 1 mAChR mutant that was unable to be activated by the natural ligand acetylcholine but rather could be activated by the synthetic drug CNO (26,27). This receptor mutant has been previously described and characterized in vitro and termed as a M1 DREADD (26). The design of this mouse means that expression of the M1 DREADD will be under the control of the native M 1 mAChR promoter ensuring that the M1 DREADD would be expressed in the same cell types and at the same expression levels as the native M 1 mAChR.
Analysis of the ability of acetylcholine and CNO to induce phosphorylation of the M1 DREADD was tested in transfected cells. It was found that acetylcholine had no significant effect on M1 DREADD phosphorylation at Ser 228 (Fig. 5B). In contrast, the M1 DREADD was phosphorylated by CNO in a concentration-dependent manner (Fig. 5B). Immunohistochemical labeling of the HA epitope tag identified expression of the M1 DREADD mutant in the molecular layer of the CA1 region of the hippocampus (Fig. 5C). Using the Ser 228 phospho-specific antibody in M1 DREADD identified light immunohistochemical staining in the CA1 region of the hippocampus that was up-regulated within 30 min of intraperitoneal administration of CNO (Fig. 5D).
Co-staining with anti-HA, to reveal the expression of the M1-DREADD HA-tagged receptor, and the Ser 228 phosphospecific antibody revealed neurons where the M1-DREADD receptor existed in a phosphorylated state (indicated by arrows in Fig. 5E). Interestingly, not all of the M1-DREADD-expressing neurons stained with the phospho-specific Ser 228 antibody suggesting that in some neurons the M1 receptor remained non-phosphorylated (at least at this residue) (Fig. 5E). It should be noted that M1-KO controls also revealed some weak nonspecific nuclear staining with the Ser 228 phospho-specific antibody (Fig. 5, D and E).
The data in the M1 DREADD mice supported the notion that the Ser 228 phospho-specific antibody can be used to monitor M 1 mAChR activation following stimulation by a synthetic ligand. We therefore tested this further by the administration of xanomeline (5 mg/kg i.p.) into wild-type mice followed by immunohistochemical analysis of phosphorylation at Ser 228 . Pharmacokinetic analysis determined that xanomeline levels in the brain peaked 30 min after injection. At this time the animal was sacrificed by perfusion fixation, and the brain was removed and sectioned for immunohistochemical staining. This procedure revealed phosphorylation of the M 1 mAChR at Ser 228 in the neuronal cell bodies in the pyramidal layer of the CA1 region of the hippocampus (Fig. 6A). In contrast, low levels of staining were observed in vehicle-treated mice (Fig. 6A).
To test whether phosphorylation at Ser 228 could also act as a biosensor for M 1 mAChR activation in response to the administration of a PAM, BQCA was administered at 15 mg/kg, which resulted in a free concentration of BQCA in the hippocampus of 51.8 Ϯ 3.6 nM; a concentration where BQCA would show PAM activity at the M 1 mAChR. In these experiments, Western blots of hippocampal lysates prepared from BQCA-treated animals showed an increase in the phosphorylation status of the M 1 mAChR at Ser 228 in response to BQCA (Fig. 6, B and C).
M1 mAChR Is Activated during Memory Acquisition-Having demonstrated that the action of an orthosteric agonist and a PAM at the M 1 mAChR in the hippocampus could be monitored using the phosphorylation status of the M 1 mAChR at Ser 228 , we next asked whether the phosphorylation of Ser 228 could also be used as a biosensor for M 1 mAChR activation during a physiological process. Previous studies had implicated a role for the M 1 mAChR in learning and memory (15)(16)(17). Because the M 1 mAChR is highly expressed in the hippocampus (28), an area associated with learning and memory, we asked whether changes in the phosphorylation status of Ser 228 on M 1 mAChRs could be detected in the hippocampus following fear conditioning training, a classical protocol used to induce a learning response (29). The mice were therefore subjected to fear conditioning training and 30 min were later sacrificed, tissues fixed by transcardial perfusion, and brains sectioned and probed with phospho-specific antibodies to Ser 228 phosphorylation. In these experiments, an increase in the phosphorylation status of Ser 228 was observed in the stratum pyramidale of the CA1 region of the hippocampus following fear conditioning training when compared with an immediate unpaired foot shock control (Fig. 6D). Importantly, this increase in Ser 228 phosphorylation correlated with an increase in c-Fos immunoreactivity, a marker of neuronal activity (Fig.  6D).

Discussion
Here, we show that phosphorylation at serine 228 in the third intracellular loop of the M 1 mAChR can be used as a read-out for receptor activation. Using this antibody as a biosensor for the activated receptor, we show that the M 1 mAChR is activated in the hippocampus following fear conditioning training. This correlates with an increase in neuronal activity in the hippocampus thereby strongly supporting a link between hippocampal M 1 mAChR activity and hippocampus-based learning and memory (15)(16)(17). Our study also illustrates how analysis of GPCR phosphorylation status can be used to assess the activation of receptors following drug treatment. Establishing target-drug engagement is highly valued in drug discovery, but up to now, no method has been described to monitor this for a GPCR target. Cell lysates were prepared, resolved by SDS-PAGE, and Western blots probed with the phospho-specific serine 228 antibody. The IgG band was used as a loading control in these experiments. Also shown is the mean data Ϯ S.E. of Western blots from three independent experiments. G, CHO-M1 cells stimulated with acetylcholine or BQCA were fixed and processed for immunocytochemistry using an anti-HA antibody to show total M 1 mAChR or with the phospho-specific serine 228. Non-transfected CHO cells were similarly treated. All the images and gels shown were typical of at least three independent experiments. All the graphical data represent the mean Ϯ S.E. of at least three independent experiments. Statistical analysis uses Student's paired t test.
To establish the phospho-specific antibody to Ser 228 as a sensor for M 1 mAChR activation, we first conducted a mass spectrometry-based phosphoproteomic analysis of the M 1 mAChR expressed in a recombinant system. This revealed that the M 1 mAChR, like many other GPCRs (30,31), is multiply phosphorylated at sites within the third intracellular loop and C-terminal tail. The multisite nature of M 1 mAChR phosphorylation is consistent with the suggestion made by us, and others, that the complex pattern of receptor phosphorylation is cell type-specific. This idea led to the notion that a receptor expressed in different cell types might show a different phosphorylation pattern, and this pattern would contribute to cell type-specific receptor signaling and regulation (30,32,33). This notion has been coined the receptor phosphorylation bar code (30, 32) and has been further extended to suggest that different agonists might drive different patterns of phosphorylation in a manner that encodes for different signaling outcomes (30,32,33). In this way, the phosphorylation bar code is thought to contribute to stimulus bias where a ligand can direct signaling down one pathway in preference to another (34,35), possibly by mediating a specific pattern of receptor phosphorylation.
Given the notion that phosphorylation of GPCRs might be a dynamic process regulated, at least in part, by the pharmacological properties of the agonist used to stimulate the receptor, FIGURE 6. M 1 mAChR is activated in the hippocampus following drug treatment and memory acquisition. A, C57/BL6/NTAC mice were injected (i.p.) with vehicle or xanomeline (5 mg/kg). After 30 min, tissues were fixed by transcardial perfusion, and sections were obtained and stained with the phospho-specific serine 228 antibody and with DAPI stain to reveal the nuclei. Shown are representative sections through the CA1 region of the hippocampus. B, C57/BL6/NTAC mice (Wild-Type) or M 1 mAChR-knock-out mice (M1-KO) were injected (intraperitoneally) with BQCA (15 mg/kg) or vehicle, and after 30 min hippocampal membranes were prepared from which the M 1 mAChR was immunoprecipitated. The sample was then processed in Western blots, which were probed with phospho-specific serine 228 antibody or an M 1 mAChR-specific antibody to detect total M 1 mAChR. C, quantification of Western blots from B. The data are presented as means Ϯ S.E. (n ϭ 3). Statistical analysis uses Student's paired t test. D, C57/BL6/NTAC mice were subjected to a fear conditioning training protocol or to an unpaired immediate foot shock as a control; 30 min later tissue was fixed by transcardial perfusion, and sections obtained and stained with phosphorylated Ser 228 -specific antibody (upper panel) or anti-c-FOS antibody (lower panel). All the data shown are typical of at least three independent experiments.
it was important in this study to carefully characterize the agonist dependence of the phosphorylation events revealed by the mass spectrometry studies. Hence, we generated a panel of phospho-specific antibodies to the M 1 mAChRs that could be used to further probe at least 4 of the 14 sites of phosphorylation. First, all four of the antibodies generated were phosphospecific recognizing phosphorylation at Ser 228 , Ser 273 , Ser 322 , and Ser 451 . These antibodies established that there were high levels of constitutive phosphorylation at Ser 322 and Ser 451 . In contrast, phosphorylation at Ser 228 and Ser 273 was seen to be highly sensitive to agonist stimulation and showed little phosphorylation in the basal state. Thus, phosphorylation at Ser 228 and Ser 273 appeared to fit the traditionally held view that only the agonist-occupied receptor undergoes phosphorylation (36,37), and thus these sites were ideal candidates as sensors of the activated receptor.
Because of the high signal-to-noise ratio obtained with the phospho-specific antibody to phosphorylated Ser 228 , we used this in further studies and established that orthosteric agonists, acetylcholine and xanomeline, mediated phosphorylation of Ser 228 . Furthermore, the positive allosteric modulator, BQCA, was also able to drive Ser 228 phosphorylation. Hence, it would appear that at least in the case of Ser 228 phosphorylation, all ligands that promoted the transition of the receptor from an inactive R-conformation to an active R*-conformation resulted in phosphorylation of the receptor on Ser 228 despite having different pharmacological properties. In this sense, an antibody that specifically recognized Ser 228 phosphorylation could be considered as a biosensor of the active R*-conformation of M 1 mAChR.
It is important to point out that not all phosphorylation sites on GPCRs will show the characteristics of Ser 228 phosphorylation described here. Careful evaluation of the agonist sensitivity of the phosphorylation event with a range of pharmacological ligands would need to be conducted because other sites of phosphorylation might be less sensitive to agonist stimulation (such as Ser 322 and Ser 451 ) or might show different levels of phosphorylation in response to ligands. Such phosphorylation sites would not be suitable as indicators of receptor activation.
To determine whether the phospho-specific Ser 228 antibody could be used to detect phosphorylation (and thereby activation) of the M 1 mAChR in vivo, we employed a mutant mouse line that expressed a M1 DREADD receptor in place of the wild-type M1 mAChR. This receptor could be selectively activated by CNO in vivo (26). Importantly, our in vitro studies had established that CNO stimulated phosphorylation at Ser 228 on the M1 DREADD. We show here that the M1 DREADD expressed in the CA1 region of the hippocampus was phosphorylated on Ser 228 following administration of CNO. These data confirmed the hypothesis that the phospho-specific Ser 228 antibody could be used to identify activated receptors in the hippocampus. This was further supported by receptor stimulation using xanomeline and BQCA in wild-type animals, where receptor activation could similarly be identified by an up-regulation of phosphorylation at Ser 228 .
Given that phosphorylation at Ser 228 fit the criteria for use as a biosensor for receptor activation, we investigated whether the M 1 mAChR was activated in the hippocampus during memory acquisition. Previous studies had established a role for muscarinic receptor signaling in learning and memory (14,38,39). The involvement of the M 1 mAChR in this process is suspected not only because this receptor subtype is highly expressed in the hippocampus (28) but also because gene knock-out and pharmacological disruption of M 1 mAChRs resulted in defective learning and memory (14). Thus, there is a great deal of interest in targeting the M 1 mAChR as a mechanism to treat cognitive deficits in neurodegenerative disease, such as Alzheimer disease (14,40,41). Despite this, it is still unclear whether M 1 mAChRs acting directly at the level of the hippocampus mediate memory processing or whether these receptors are more subtly involved by promoting interactions between the prefrontal cortex and hippocampus (16,(42)(43)(44). We were therefore interested to test the possibility that M 1 mAChRs were activated in the hippocampus during memory acquisition. This was found to be the case with a significant up-regulation of M 1 mAChR phosphorylation at Ser 228 following fear conditioning training in the CA1 region of the hippocampus. Importantly, this region was also seen to show high levels of neuronal activity, as indicated by an increase in c-FOS expression, following fear conditioning training. These data support the hypothesis that hippocampal M 1 mAChRs are activated during memory acquisition and that this may contribute to the process of learning and memory.
In conclusion, by determining the phosphorylation sites within the M 1 mAChR and generating and characterizing the phospho-specific antibodies, we have established that at least one of the phosphorylation events (Ser 228 ) is exquisitely sensitive to agonist stimulation. Using this phosphorylation event as a biosensor for receptor activation, we have been able to establish the activation of the M 1 mAChR in vivo in the CA1 region of the hippocampus following drug treatment and memory acquisition.