Molecular Mechanisms of Action of M5 Muscarinic Acetylcholine Receptor Allosteric Modulators S

Home , Muscarinic agonist

Supplemental material to this article can be found at: http://molpharm.aspetjournals.org/content/suppl/2016/07/26/mol.116.104182.DC1

1521-0111/90/4/427–436$25.00 http://dx.doi.org/10.1124/mol.116.104182 MOLECULAR PHARMACOLOGY Mol Pharmacol 90:427–436, October 2016 Copyright ª 2016 by The American Society for Pharmacology and Experimental Therapeutics

Molecular Mechanisms of Action of M5 Muscarinic Acetylcholine Receptor Allosteric Modulators s

Alice E. Berizzi, Patrick R. Gentry, Patricia Rueda, Sandra Den Hoedt, Patrick M. Sexton, Christopher J. Langmead, and Arthur Christopoulos Drug Discovery Biology, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville, Victoria, Australia Received March 3, 2016; accepted July 18, 2016

ABSTRACT Downloaded from Recently, the first subtype-selective allosteric modulators of pilocarpine response. Radioligand-binding affinity coopera- the M5 muscarinic acetylcholine receptor (mAChR) have been tivity estimates were consistent with values derived from described, but their molecular mechanisms of action remain functional assays in some instances but not others, suggest- unknown. Using radioligand-binding and functional assays of ing additional allosteric effects on orthosteric ligand efficacy. inositol phosphate (IP) accumulation and Ca21 mobilization in For ML375 this was confirmed in IP assays performed after a recombinant cell line stably expressing the human M5 reduction of receptor reserve by the alkylating agent, phenox- mAChR, we investigated the effects of the positive allosteric ybenzamine, as it reduced the maximal ACh response. In molpharm.aspetjournals.org modulator (PAM), ML380, and negative allosteric modulator, contrast, ML380 enhanced only ACh potency after receptor ML375. In functional assays, ML380 caused robust enhance- alkylation, with no effect on maximal response, consistent ments in the potency of the full agonists, acetylcholine (ACh), with studies of the M1 mAChR with the prototypical PAM, carbachol, and oxotremorine-M, while significantly increasing BQZ12. Interaction studies between ML380 and ML375 also the maximal response to the partial agonist, pilocarpine. indicated that they most likely used an overlapping allosteric ML380 also demonstrated direct allosteric agonist activity. site. Our findings indicate that novel small-molecule modula- In contrast, ML375 displayed negative cooperativity with tors of the M5 mAChR display mixed mechanisms of action each of the agonists in a manner that varied with the path- compared with previously characterized modulators of other way investigated and progressively reduced the maximal mAChRs. at ASPET Journals on October 2, 2021

Introduction To date, the majority of studies focusing on mAChRs in CNS disorders have targeted the M1 and M4 subtypes (Bymaster et al., The muscarinic acetylcholine receptors (mAChRs) are pro- 1997; Shannon et al., 2000; Brady et al., 2008; Chan et al., 2008; totypical family A guanine nucleotide-binding protein-coupled Ma et al., 2009; Shirey et al., 2009). However, despite only receptors (GPCRs); the M1,M3, and M5 mAChR subtypes representing 2% of the total mAChR population in the brain primarily couple to Gq/11 proteins to activate the inositol (Yamada et al., 2003), the M5 mAChR is expressed in highly phosphate (IP) pathway, whereas M2 and M4 mAChR sub- discrete regions of therapeutic interest. For example, it is the only types primarily couple to Gi/o proteins to mediate the in- mAChR subtype for which mRNA transcript can be identified in hibition of adenylyl cyclase (Porter et al., 2002; Matsui et al., dopamine-containing neurons of the substantia nigra pars 2004; Langmead et al., 2008). The mAChRs are widely compacta and the ventral tegmental area (Vilaró et al., 1990; expressed in the central nervous system (CNS) and periphery, Weiner et al., 1990; Yasuda et al., 1993), leading to the hypothesis where they control a number of physiologic processes. In the that the M5 mAChR may regulate midbrain dopamine trans- CNS, they have been implicated as important therapeutic mission and reward mechanisms and thus represent a target for targets for a range of disorders, including schizophrenia, treating drug addiction (Vilaró et al., 1990; Weiner et al., 1990). ’ ’ Alzheimer s disease, addiction, and Parkinson sdisease The phenotype of M5 mAChR knockout (KO) mice supports (Wess et al., 2007; Langmead et al., 2008). the validity of the receptor as an addiction target; KO mice self-administer less cocaine and morphine, and are unable to develop conditioned place preference to the same drugs, with This work was supported by the National Health and Medical Research blunted withdrawal symptoms (Basile et al., 2002; Fink- Council of Australia [Grant APP1055134]. A.C. is a Senior Principal Research Fellow, and P.M.S. is a Principal Research Fellow of the National Health and Jensen et al., 2003; Thomsen et al., 2005). However, until Medical Research Council of Australia. recently, there has been a lack of sufficiently potent and dx.doi.org/10.1124/mol.116.104182. s This article has supplemental material available at molpharm. selective tool compounds with which to probe receptor function aspetjournals.org. and augment findings from constitutive M5 mAChR KO mice.

ABBREVIATIONS: ACh, acetylcholine; CCh, carbachol; CHO, Chinese hamster ovary; CNS, central nervous system; GPCR, guanine nucleotide- binding protein-coupled receptor; IP, inositol phosphate; KO, knockout; mAChR, muscarinic ACh receptor; NAM, negative allosteric modulator; NMS, N-methylscopolamine; oxo-M, oxotremorine-M; PAM, positive allosteric modulator; PBZ, phenoxybenzamine.

427 428 Berizzi et al.

A major breakthrough in this regard has been the discovery of were generous gifts of C. Lindsley (Vanderbilt University, Nashville, small-molecule allosteric modulators for these mAChRs. By TN). BQZ12 was synthesized in-house, as described previously targeting topographically distinct binding sites from those used (Abdul-Ridha et al., 2014a,b). All other chemicals were purchased V by orthosteric ligands, allosteric modulators can selectively from Sigma-Aldrich. For all procedures, purified water (18.2 M cm) target receptor subtypes to enhance or inhibit the function of from a Milli-Q PF Plus system was used. Cell Culture. The wild-type human (hM ) mAChR construct was acetylcholine (ACh) while maintaining the spatial and temporal 5 isogenically integrated into Flp-In CHO cells (Invitrogen), and cells nature of endogenous neurotransmission. Allosteric modula- were selected in the presence of 600 mg/mL hygromycin B at 37°C, 5% tors also offer the potential of reduced propensity for on-target CO2, as previously described for the hM1 mAChR (Abdul-Ridha et al., mediated toxicity and engendering signal-pathway bias 2014a). CHO-hM1 and CHO-hM5 cells were subcultured in a 1:3 ratio (Gregory et al., 2007, 2010; Leach et al., 2007; Conn et al., every 2 days after reaching 80% confluence with 6–10 mL Versene to 2009; Valant et al., 2012; Kruse et al., 2014). Highly selective detach cells and incubated in a humidified atmosphere at 37°C, 5% – positive allosteric modulators (PAMs) have been identified for CO2, for 2 5 minutes. Cells were counted by a hemocytometer to be seeded at 25,000 cells/well in 96-well cell culture plates 24 hours prior both the M1 and M4 mAChR subtypes (Chan et al., 2008; to assay, unless otherwise stated. Shirey et al., 2008, 2009; Ma et al., 2009), enabling their role in 3 animal models of cognition and schizophrenia to be elucidated. [ H]-NMS Equilibrium Binding. CHO-hM5 cells were seeded into 96-well Isoplates (PerkinElmer Life Sciences) in 100 mL culture Most recently, screening and medicinal chemistry studies media and incubated in a humidified atmosphere at 37°C, 5% CO2, for identified the first M5 mAChR-selective allosteric ligands, at least 6 hours. After 6 hours, media was removed and cells were Downloaded from including the exemplar PAM, ML380 [1-((1H-indazol-5-yl) washed with 100 mL phosphate-buffered saline, before physiologic sulfonyl)-N-ethyl-N-(2-(trifluoromethyl)benzyl)piperidine-4- HEPES buffer (100 mL; 110 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2, carboxamide], and the negative allosteric modulator (NAM), 1 mM MgSO4, 25 mM glucose, 50 mM HEPES, and 58 mM sucrose, pH ML375 [(S)-9b-(4-chlorophenyl)-1-(3,4-difluorobenzoyl)-2,3- 7.4) was added. Binding was determined in the presence of 0.3 nM dihydro-1H-imidazo[2,1-a]isoindol-5(9bH)-one] (Gentry et al., [3H]-NMS. Nonspecific binding of the radioligand was assessed in the 2013, 2014) (Fig. 1). However, the molecular mechanisms presence of 10 mM atropine and total binding by radioligand in the molpharm.aspetjournals.org underlying their activity remain largely unexplored. Given presence of HEPES buffer alone. Binding assays were terminated that GPCR allosteric modulators can display a range of after a 2-hour incubation with radioligand at 37°C by removal of any unbound radioligand and the repeated washing of cells (3 times) with activities, and that the type of activity can determine the 100 mL 0.9% NaCl at 4°C, followed by addition of 100 mL Microscint degree of in vivo efficacy in a context-sensitive manner scintillation liquid. Bound radioactivity was assessed by liquid (Christopoulos, 2014), the aim of the current study was to scintillation counting by a MicroBeta2 Plate Counter (PerkinElmer characterize the molecular mechanisms of action of ML380 Life Sciences, Glen Waverley, Australia). For inhibition-binding and ML375 in combination with different orthosteric probes at experiments, varying concentrations of individual (or combinations the human M5 mAChR. Our studies confirm that ML380 and of) agonists, allosteric modulators, and antagonist in HEPES buffer ML375 exhibit PAM agonist and NAM activity, respectively, were added to the cells in the presence of [3H]-NMS in a final assay of with the potential to modulate orthosteric ligand affinity and/ volume of 100 mL. at ASPET Journals on October 2, 2021 or intrinsic efficacy, a mechanism not observed with allosteric IP Accumulation. The IP-One assay kit (Cisbio) was used for the direct quantitative measurement of myo-inositol 1-phosphate in CHO- modulators of the related M1 mAChR subtype. Moreover, functional interaction studies between the PAM and NAM hM5 or CHO-hM1 cells, as described previously (Abdul-Ridha et al., 2014a), with the exception that cells were seeded at 25,000 per well suggest that they exert their opposing effects on receptor onto a transparent 96-well cell culture plate, and the following day function via a common, albeit currently uncharacterized, cells were stimulated with various concentrations of orthosteric and/or allosteric binding site. allosteric ligands, and were then incubated for 40 minutes at 37°C, 5%

CO2. Responses were normalized to that to a single, maximal concentration of ACh (10 mM).

Materials and Methods Calcium Mobilization. CHO-hM5 cells were cultured overnight in 96-well cell culture plates at 50,000 cells/well. The following day, Materials. Flp-In-Chinese hamster ovary (CHO) cells were media was removed and replaced with 50 mL1mM Fluo-4-AM Ca21 ’ obtained from Life Technologies (Mulgrave, Australia). Dulbecco s sensitive dye in assay buffer (Hank’s balanced salt solution supple- modified Eagle medium was purchased from Invitrogen (Carlsbad, mented with 1.26 mM CaCl2,20mMHEPES,and2.5mMpro- CA). Fetal bovine serum was purchased from ThermoTrace (Melbourne, benecid, pH 7.4), and the cells were incubated 45 minutes at 37°C, Australia). [3H]-N-methylscopolamine ([3H]-NMS; specific activity, 5% CO2. Dye was then removed and replaced with 50 mL fresh assay 84 Ci/mmol) and MicroScint scintillation liquid were purchased from buffer. Test compounds were added to the cells in a FLEXStation II Perkin-Elmer Life Sciences. The IP-One assay kit was purchased from (Molecular Devices, Sunnyvale, CA). For interaction studies, ligands Cisbio (Codolet, France). ACh, carbachol (CCh), pilocarpine, oxotre- were added simultaneously to cells. Ca21/Fluo-4 fluorescence was morine-M (oxo-M), atropine, and phenoxybenzamine (PBZ) were measured at an emission wavelength of 525 nm for 90 seconds purchased from Sigma-Aldrich (St. Louis, MO). ML380 and ML375 (19 seconds prior to and 71 seconds following the addition of ligand) at room temperature. Receptor Alkylation Studies. To determine functional affinity and efficacy estimates for the agonists, additional IP accumulation assays were performed, as described above, but on cells pretreated for 30 minutes at 37°C with the irreversible orthosteric-site alkylating agent, PBZ (or vehicle buffer control), followed by three washes with 10% phosphate-buffered saline. PBZ reduces functional receptor availability by irreversible steric hindrance for orthosteric agonists (ACh, CCh, oxo-M) or due to the irreversible formation of an inactive state for which the positive allosteric modulator, ML380, has minimal Fig. 1. Structures of (A) ML380 and (B) ML375. affinity due to high negative cooperativity. Mechanisms of M5 Muscarinic Receptor Allosteric Modulation 429

Data Analysis. GraphPad Prism version 6.04 (San Diego, CA) was All interaction radioligand-binding studies were analyzed accord- used for all statistical analysis and curve fitting. For the empirical ing to the following adapted form of an allosteric ternary complex analysis of functional concentration–response data, agonist model that accounts for the interaction of two orthosteric ligands and concentration–response curves were analyzed using a standard one allosteric ligand on a receptor (Christopoulos and Mitchelson, logistic function to determine measures of potency (as negative 1997): logarithms; pEC50) and maximal response (Emax). For direct de- Bmax×½A termination of functional agonist dissociation constants (KA) and Y 5 1 × = a 1 × 1 = 1 = 1 a = × operational efficacies (t) from the alkylation experiments, agonist IP ½A ðKA KBÞ ð A½B KBÞ fg1 ð½I KIÞ ð½B KBÞ ½ð I ½I½BÞ ðKI KBÞ accumulation concentration–response curves (in the presence or (4) absence of PBZ pretreatment) were globally fitted to the operational model of agonism (Black and Leff, 1983): where [A], [B], and [I] represent the concentrations of the radio- ( ) ligand ([3H]-NMS), allosteric ligand, and orthosteric inhibitor, re- . n spectively, and K ,K , and K represent their respective equilibrium Log½A Log½A A B I Y 5 Basal 1 ðEM 2 BasalÞ 1 1 KA 1 10 t 1 10 dissociation constants, and Bmax is as defined above. The value KA was fixed to 0.3 nM. aA and aI represent the affinity cooperativity values (1) between the allosteric ligand and the radioligand or orthosteric inhibitor, respectively; values greater than 1 indicate positive cooper- where [A] represents the concentration of the agonist, KA repre- ativity; values ,1 (but .0) indicate negative cooperativity; and

sents the agonist equilibrium dissociation constant, Basal is the Downloaded from values of unity indicate neutral cooperativity. All potency, affinity, response in the absence of ligand, EM is the maximal system response, t represents the agonist operational efficacy, which subsumes both and cooperativity parameters were estimated as logarithms receptor density and stimulus–response coupling efficiency, and n is (Christopoulos, 1998). Where appropriate, fitted parameters were the slope of the transducer function linking occupancy to response. A compared by extra sum-of-squares F-test (Motulsky and Christopoulos, modified form of eq. 1 was also applied to the data for the partial 2004). agonist, pilocarpine, using a four-parameter logistic fit of the ACh

control (full agonist) concentration–response curve to provide esti- molpharm.aspetjournals.org mates of EM, transducer slope (n), and Basal for fitting the pilocarpine Results data. The effects of ML380 and ML375 were studied in canonical For functional interaction studies between full orthosteric agonists 21 and allosteric ligands in the IP accumulation or calcium assays, the Gq/11-linked IP accumulation and Ca mobilization func- 5 following simplified operational model of allosterism was applied tional assays. ACh stimulates both IP accumulation (pEC50 21 (Leach et al., 2007): 6.68 6 0.05; nH 5 1.0 6 0.1; n 5 3) and Ca mobilization (pEC 5 7.63 6 0.06; n 5 1.3 6 0.2; n 5 4) in CHO-hM cells 50 H 5 n ðEM 2 BasalÞ× ½AðKB 1 ab½BÞ 1 tB½B×EC50 (Fig. 2). The PAM, ML380, also robustly stimulated IP Y 5 Basal 1 21 n n n accumulation and Ca mobilization (pEC values 5.33 6 ½AðKB 1 ab½BÞ 1 tB½B×EC50 1 EC50 ×ðKB½BÞ 50

5 6 5 6 5 6 at ASPET Journals on October 2, 2021 (2) 0.06; nH 1.4 0.2; n 5 and 5.71 0.07; nH 1.4 0.3; n 5 3, respectively; Fig. 2), appearing as a high efficacy agonist where Basal is the response in the absence of ligand, EC50 is the in its own right. The NAM, ML375, was inactive on its own in – midpoint of the full agonist concentration response curve, KB is the either signaling assay. equilibrium dissociation constant of the allosteric ligand (fixed to To determine the degree of relative intrinsic efficacy of the t the KA value from alkylation studies for ML380 for the IP assay), B allosteric agonist, ML380, relative to that of the orthosteric denotes the capacity of the allosteric ligand to exhibit agonism agonists, concentration–response curves to ACh, CCh, oxo-M, (constrained to 0 for ML375), and ab is the net affinity/efficacy cooperativity parameter describing the combined effect of the alloste- or ML380 in IP accumulation assays were re-established after pretreatment (followed by washout) of the CHO-hM5 cells with ric modulator on agonist function (both affinity and efficacy). EM and n are as described above. As this model is only suitable for full agonists PBZ to irreversibly reduce receptor accessibility through that remain full agonists in the presence of allosteric modulator direct steric hindrance (orthosteric agonists) or high negative (Aurelio et al., 2009; Leach et al., 2010), the pilocarpine datasets were cooperativity (ML380). As shown in Fig. 3, this resulted in a analyzed according to the complete model of allostery and agonism reduction in both potencies and maximal responses for all (Leach et al., 2007): agonists. Application of the operational model of agonism to the resulting families of curves for each agonist (Table 1) E 2 Basal × t A K 1 ab B 1 t K n 5 1 ð M Þ A½ ð B ½ Þ B B Y Basal n n ½AKB 1 KAKB 1 KA½B 1 a½A½B 1 tA½A KB 1 ab½B 1 tB½B×KB (3)

where [A] and [B] represent the concentrations of the orthosteric agonist and allosteric ligand, respectively, and KA and KB represent their respective equilibrium dissociation constants (constrained as above for ML380). tA and tB represent the relative efficacies of the orthosteric and allosteric ligands, receptively, and n represents the slope of the transducer function that links occupancy to response. a represents the affinity cooperativity between the orthosteric agonist and allosteric modulator, and b represents a scaling factor that Fig. 2. Effect of ACh, ML380, and ML375 on (A) IP accumulation and (B) describes the magnitude and direction of the effect of the allosteric 2+ Ca mobilization in CHO-hM5 cells. Data are expressed as a percentage of modulator on orthosteric agonist efficacy. Where required to facilitate maximal ACh response and represent the mean 6 S.E.M. of at least four convergence of global model fitting, the value of n was constrained to independent experiments performed in triplicate; fitted curves are from a unity and indicated as such. four-parameter logistic equation. 430 Berizzi et al.

Fig. 3. Effect of PBZ (0.1–100 mM) on IP accumulation stimulated by (A) ACh, (B) CCh, (C) oxo-M, and (D) ML380 in CHO-hM5 cells. Data are expressed as a percentage of maximal control ACh response and represent the mean 6 S.E.M. of at least five independent experiments performed in triplicate. Fitted curves are from global analysis of the datasets according to eq. 1 (in Materials and Methods), with parameter estimates shown in Table 1. Downloaded from

revealed that PBZ reduces the freely accessible receptor 1.78 6 0.15 (ab 5 60), which was significantly greater than population by 48 6 13% at 1 mM, 88 6 4.4% at 10 mM, and that observed against the full agonists and not consistent 99 6 0.4% at 100 mM. All of the orthosteric and allosteric with the predictions of a simple two-state model of allostery molpharm.aspetjournals.org ligands had similar functional affinities for their respective as exemplified by prior studies of the related M1 mAChR sites on the receptor (pKA 5 4.8–5.3), but ML380 displayed the with PAMs such as BQCA and BQZ12 (Canals et al., 2012; lowest operational efficacy (t 5 4) compared with the orthos- Abdul-Ridha et al., 2014a,b). teric agonists (t 5 20–33; Table 1). The NAM, ML375, produced parallel, rightward shifts in As well as exhibiting intrinsic agonist activity, ML380 concentration–response curves to ACh in both IP and Ca21 potentiated the activity of ACh in both IP and Ca21 mobilization assays (Fig. 4, C and D). Analysis of the latter mobilization assays (Fig. 4, A and B), thus acting as a yielded estimates of allosteric site affinity (pKB 5 6.62 6 0.06; PAM agonist (Christopoulos, 2014) with net cooperativity n 5 4) and saturable negative cooperativity [Log ab 521.34 6 (Log ab)valuesof0.966 0.26 (ab 5 9.1) and 1.69 6 0.26 (ab 5 0.04 (ab 5 0.05); n 5 4]. In the IP assay, however, the 49), respectively (Table 2). The stronger positive cooperativity reduction in agonist potency did not appear to reach a ceiling at ASPET Journals on October 2, 2021 for ML380 in the calcium assay is consistent with its higher level over the entire concentration range tested, indicative of degree of stimulus–response coupling, also reflected by the either a competitive interaction or, more likely, a very high greater intrinsic agonism for ML380 at this endpoint versus degree of negative cooperativity. As such, the data could the IP assay (Log tB 5 1.18 6 0.18; tB 5 15, versus Log tB 5 be fitted to an allosteric model in which the negative 0.69 6 0.06; tB 5 4.9). To evaluate potential probe de- cooperativity could be constrained to a very small value (Log pendence of the PAM activity, interaction studies were also ab 523.0) under which the allosteric ternary complex model conducted with CCh, oxo-M, and the partial agonist, pilo- and a competitive interaction model become indistinguish- carpine, in the IP assay. The positive modulation of CCh- able (Keov et al., 2014). This yielded an estimate of the ML375 and oxo-M–mediated responses by ML380 was similar to pKB 5 6.22 6 0.13 (n 5 6), which was in excellent agreement that for ACh (Fig. 5; Table 2), indicating little probe with the allosteric ternary complex analysis of the Ca21 dependence with full agonists. Interestingly, ML380 mark- mobilization data. As was observed for the ML380 experi- edly potentiated the potency and maximal response to ments, there was little difference between the observed the partial agonist, pilocarpine, with a Log ab value of modulation of ACh-, CCh-, and oxo-M–stimulated IP accumulation, with ML375 causing parallel rightward shifts with high negative cooperativity and similar estimates of affinity irrespective of the agonist tested (Fig. 5; Table 2). Again, as TABLE 1 observed with the PAM with ML380, the modulation of Operational model parameters for agonist-stimulated IP accumulation pilocarpine activity by the NAM yielded a different profile to with and without PBZ pretreatment in CHO-hM cells 5 the full agonists, with ML375 virtually abolishing the maxi- a b c Agonist pKA Log t (t) n mal agonist response (Fig. 5; Table 2). ACh 5.28 6 0.11 1.29 6 0.11 (20) 0.93 6 0.06 To further understand the mechanisms underlying the CCh 4.78 6 0.18 1.42 6 0.18 (26) 0.64 6 0.07 positive and negative cooperativity exhibited by ML380 and Oxo-M 4.65 6 0.22 1.52 6 0.23 (33) 0.53 6 0.06 ML375, respectively, whole-cell [3H]-NMS–binding studies 6 6 6 ML380 4.79 0.16 0.61 0.11 (4) 1.39 0.18 were subsequently performed to directly determine the co- Estimated parameters represent the mean 6 S.E.M. of at least four experiments operative effects on agonist binding alone (a). [3H]-NMS performed in triplicate. Functional IP accumulation responses were analyzed according to eq. 1 (Materials and Methods). bound to whole CHO-hM5 cells in a monophasic and saturable a Negative logarithm of the agonist equilibrium dissociation constant. manner with an estimated pKD 5 9.64 6 0.12 (KD 5 0.26 6 bNegative logarithm of the operational efficacy parameter (antilog shown in 5 6 parentheses). 0.07 nM; n 4) and maximal binding capacity (Bmax)of7.3 5 cSlope of the transducer function linking receptor occupancy to response. 1.2 fmol/10 cells. Mechanisms of M5 Muscarinic Receptor Allosteric Modulation 431

Fig. 4. Effect of ML380 (A, B) or ML375 (C, D) on ACh- stimulated IP accumulation (A, C) and Ca2+ mobilization (B, D) in CHO-hM5 cells. Data are expressed as a percentage of maximal ACh response and represent the mean 6 S.E.M. of at least five independent experiments performed in triplicate. Fitted curves are from global analysis of datasets according to eq. 2 (in Materials and Methods) with parameter estimates shown in Table 2 and Results. Downloaded from

ML380 increased the affinity of ACh, CCh, and oxo-M to agonist function not only by modulating orthosteric agonist 3 inhibit the binding of [ H]-NMS from CHO-hM5 cells with affinity, but also intrinsic efficacy. As these effects were not estimates of positive cooperativity (Log a) similar to the Log apparent in the functional assays with the full orthosteric molpharm.aspetjournals.org ab values determined in the IP accumulation assays (Fig. 6; agonists [because their higher efficacies (Table 1) allowed Table 3). In addition, ML380 exerted weak negative coopera- them to achieve the maximum cellular effect in both the tivity with respect to the inverse agonist radioligand, [3H]-NMS, absence and presence of modulator under control conditions], and neutral affinity cooperativity with respect to the weak we exploited the ability of PBZ to irreversibly alkylate the partial agonist, pilocarpine (Log a 5 0.00 6 0.18; Fig. 6; mAChRs and effectively reduce the number of spare receptors, Table 3). thus enforcing conditions in which even high efficacy agonists Similar [3H]-NMS–binding studies revealed that ML375 like ACh would require maximal receptor occupancy to caused a concentration-dependent decrease in the affinities achieve maximal observed effect. Accordingly, pretreatment of ACh, CCh, and, to a lesser extent, oxo-M and pilocarpine with PBZ (3 mM), followed by extensive washout, yielded a at ASPET Journals on October 2, 2021 (Fig. 7; Table 3). Conversely, ML375 exhibited neutral-to- depression in the ACh maximal response and potency to 3 weak positive cooperativity with respect to [ H]-NMS itself. stimulate IP accumulation in CHO-hM5 cells (control pEC50 The affinity (pKB) estimates for ML375 were broadly in line 5 7.00 6 0.13, Emax 5 98.3 6 3.2%; PBZ-treated pEC50 5 6.03 with those generated from the functional assays (Table 2). 6 0.10, Emax 5 47.2 6 1.9%; Fig. 8A). In the presence of ML380 However, the degree of negative cooperativity with respect (10 mM), the potency (pEC50 5 7.09 6 0.15; P , 0.0001, F-test), to agonist binding was not sufficient to account for the but not the maximal response to ACh (49.7 6 1.8%; P 5 0.35, higher negative cooperativity noted in the functional assays F-test), was significantly increased compared with control. (Table 2). There was a similar lack of effect on the maximal ACh The discrepancies between cooperativity values across response in the presence of a higher concentration of ML380 binding and functional studies, and the direct effects observed (30 mM; data not shown) and also a small elevation in the basal with either PAM or NAM on the maximal effect of the partial response, reflecting residual ML380 agonist activity (Fig. 8A). agonist, pilocarpine, thus led to the hypothesis that these An almost identical profile was seen when the experimental allosteric modulators may be exerting their overall effects on paradigm was repeated using oxo-M as the orthosteric agonist

TABLE 2 Operational model estimates for the interactions between ML380 or ML375 and the indicated orthosteric agonists in IP accumulation assays in CHO-hM5 cells

ML380 ML375

a b c d b d Agonist pKB Log ab (ab) Log tB n pKB Log ab (ab) n ACh 4.79 0.96 6 0.26 (9.1) 0.69 6 0.06 =1.00 6.22 6 0.13 23.0 (0.001) 0.92 6 0.11 CCh 4.79 1.1 6 0.32 (12.3) 0.68 6 0.07 =1.00 6.22 6 0.13 23.0 (0.001) 1.08 6 0.15 Oxo-M 4.79 0.81 6 0.19 (6.5) 0.51 6 0.04 =1.00 6.35 6 0.11 23.0 (0.001) 1.29 6 0.21 Pilocarpine 4.79 1.8 6 0.15 (60.2) 0.45 6 0.05 =1.00 6.62 6 0.42 23.0 (0.001) 0.56 6 0.18

Estimated parameters represent the mean 6 S.E. of at least five experiments performed in triplicate. Functional IP accumulation responses were analyzed according to eq. 2 (or eq. 3 for pilocarpine; Materials and Methods). aNegative logarithm of the allosteric modulator equilibrium dissociation constant (constrained to the value obtained from the receptor alkylation studies; Table 1). bLogarithm of functional cooperativity between agonist and allosteric modulator (antilog values shown in parentheses). cLogarithm of operational efficacy parameter. dSlope of the transducer function linking receptor occupancy to response (constrained to unity for ML380 datasets). 432 Berizzi et al.

BQZ12, a high-affinity congener of the M1 mAChR PAM, BQCA, that has been proposed to act according to a simple two-state model of allostery (Canals et al., 2012; Abdul-Ridha

et al., 2014a). As with the M5 mAChR experiments, PBZ pretreatment reduced ACh potency and maximal response to

stimulate IP accumulation in CHO-hM1 cells (control pEC50 5 6.66 6 0.06, Emax 5 88.5 6 1.5%; PBZ-treated pEC50 5 5.18 6 0.11, Emax 5 49.8 6 2.2%; Fig. 8C). In the presence of BQZ12 (10–30 nM), the ACh potency was significantly enhanced by approximately 50-fold (P , 0.0001, F-test), but there was no effect on the maximal response (P 5 0.65, F-test; Fig. 8C),

suggesting that ML380 may exert its effects at the M5 mAChR in a similar manner to BQZ12 at the M1 mAChR, that is, primarily by modulation of full agonist affinity. Interestingly, when profiled under identical PBZ-treated conditions, ML375 (10–30 mM) significantly decreased both

potency (control pEC50 5 6.85 6 0.05, Emax 5 98.8 6 1.1%; Downloaded from PBZ-treated 5 5.45 6 0.09, Emax 5 48.2 6 1.4%; pEC50 5 3.98 6 0.12 at 10 mM ML375; 3.75 6 0.17 at 30 mM ML375; P , 0.01, F-test) and maximal response (27.1 6 1.6% at 10 mM; 22.1 6 2.0% at 30 mM; P , 0.05, F-test) of ACh to stimulate IP accumulation compared with control (Fig. 8D), indicating that

it is a NAM of both ACh affinity and signaling efficacy. molpharm.aspetjournals.org Finally, we investigated whether ML380 and ML375 mediate their PAM and NAM effects, respectively, from a common

site on the M5 receptor. Using the IP accumulation assay, we determined the ability of ML375 or the orthosteric antagonist, Fig. 5. Effect of ML380 (A, C, E) or ML375 (B, D, F) on CCh, oxo-M, or atropine, to inhibit the agonist response to a single, fixed pilocarpine-stimulated IP accumulation in CHO-hM5 cells. Data are m expressed as a percentage of maximal ACh response and represent the concentration of ML380 (10 M). Atropine and ML375 both mean 6 S.E.M. of at least three independent experiments performed in fully inhibited the response to ML380 (pIC50 values 8.13 6 triplicate. Fitted curves are from global analysis of datasets according to 0.10 and 5.79 6 0.21, respectively; Fig. 9). As ML380 is clearly eq. 2 (or eq. 3 for pilocarpine; in Materials and Methods) with parameter 3 estimates shown in Table 2 and Results. allosteric with respect to ACh and [ H]-NMS (vide supra), the at ASPET Journals on October 2, 2021 inhibition mediated by atropine is almost certainly allosteric, but characterized by a high degree of negative cooperativity. (Fig. 8B), indicating that the ML380-mediated modulation of Moreover, three-way binding studies monitoring the competi- pilocarpine efficacy is not seen in combination with higher tion between atropine, a [3H]-NMS, in the absence or presence efficacy agonists. of ML380 (Supplemental Fig. 1) revealed that atropine has To compare this mechanism with that of other well-studied neutral affinity cooperativity with respect to modulator (Log a allosteric modulators of mAChRs, we performed similar 5 0.00 6 0.12), suggesting that the inhibition seen in the alkylation and IP accumulation studies in CHO cells stably functional assay must be mediated by high negative efficacy expressing the Gq/11-coupled M1 mAChR subtype using cooperativity.

Fig. 6. Three-way radioligand-binding studies to demonstrate the effect of ML380 (0.3–30 mM) on (A) ACh, (B) CCh, (C) oxo-M, and (D) pilocarpine-mediated inhibition of [3H]- NMS binding in whole CHO-hM5 cells. Cells were incubated with radioligand for 2 h at 37°C; data are expressed as a percentage of control specific binding and represent the mean 6 S.E.M. of at least three independent experiments performed in triplicate. Fitted curves are from global analysis of datasets according to eq. 4 (in Materials and Methods)withparameter estimates shown in Table 3. Mechanisms of M5 Muscarinic Receptor Allosteric Modulation 433

TABLE 3 Allosteric ternary complex model parameters for the interaction between ML380 or ML375 with the orthosteric antagonist/inverse agonist, [3H]-NMS, and the indicated agonists

ML380 ML375 a b Agonist pKI pKB c d b c d b Log a (a) Log a9 (a9) pKB Log a (a) Log a9 (a9) pKB ACh 4.64 6 0.07 4.79 0.87 6 0.12 (7.4) 20.17 6 0.06 (0.67) 4.94 6 0.09 21.37 6 0.12 (0.04) 0.08 6 0.03 (1.2) 6.87 6 0.14 CCh 4.32 6 0.06 4.79 0.74 6 0.11 (5.5) 20.36 6 0.06 (0.44) 4.43 6 0.13 21.52 6 0.18 (0.03) 0.06 6 0.03 (1.1) 6.82 6 0.20 Oxo-M 4.86 6 0.06 4.79 1.06 6 0.11 (12) 20.40 6 0.06 (0.40) 4.94 6 0.09 20.85 6 0.13 (0.14) 0.16 6 0.03 (1.4) 6.52 6 0.15 Pilocarpine 4.83 6 0.05 4.79 0.00 6 0.18 (1.0) 20.37 6 0.05 (0.43) 4.88 6 0.09 20.57 6 0.11 (0.27) 20.01 6 0.02 (0.98) 6.72 6 0.30

Estimated parameters represent the mean 6 S.E. of three experiments performed in duplicate. Three-way binding interactions were analyzed according to eq. 4 (Materials and Methods). aNegative logarithm of the orthosteric agonist equilibrium dissociation constant. bNegative logarithm of the allosteric modulator equilibrium dissociation constant. cLogarithm of affinity cooperativity between orthosteric agonist and allosteric modulator (antilog values shown in parentheses). dLogarithm of affinity cooperativity between [3H]-NMS and allosteric modulator (antilog values shown in parentheses).

Unfortunately, the data for ML375 provide less resolution— ACh-stimulated calcium mobilization at both human and rat Downloaded from the full inhibition of the ML380 response could indicate a muscarinic M5 mAChRs, which was shown to be allosteric by competitive interaction for a common allosteric site, or a high its ability to slow [3H]-NMS dissociation from the receptor degree of negative cooperativity with respect to a second (Gentry et al., 2013). Despite these data, relatively little is allosteric binding site. Assuming the former, analysis of the known about the mechanisms of modulation by these alloste- ML375 antagonist inhibition curve using a modified compet- ric ligands. The aim of the current studies was to fully itive Cheng-Prusoff equation (Leff et al., 1993) yielded an characterize the pharmacology of the PAM, ML380, and estimated pKB of 6.14 6 0.10, which is similar to the affinity the NAM, ML375, at the muscarinic M5 mAChR using molpharm.aspetjournals.org estimates of the modulator obtained in our other assays based radioligand-binding and canonical Gq/11-linked signaling as- on allosteric model analyses (vide supra). The most parsimo- says (IP accumulation and Ca21 mobilization). By doing so, we nious explanation, thus, is that both modulators interact with reveal that these allosteric ligands have minimal effects on a common site. antagonists, but have the potential to modulate both orthosteric agonist affinity and intrinsic efficacy of agonists in a Discussion probe-dependent manner. Our studies confirm that ML380 and ML375 display PAM Recent screening and medicinal chemistry efforts have agonist and NAM behavior, respectively, when interacting

resulted in the first subtype-selective small-molecule positive with a range of orthosteric agonists in assays of IP accumu- at ASPET Journals on October 2, 2021 21 and negative allosteric modulators of muscarinic M5 mAChRs lation and Ca mobilization, with ML375 possessing sub- (Bridges et al., 2010; Gentry et al., 2013, 2014). ML380 has micromolar affinity for the M5 mAChR and ML380 been described as a PAM of ACh-stimulated calcium mobili- displaying similar affinity as CCh and oxo-M. Three-way zation at both human and rat muscarinic M5 mAChRs, with radioligand-binding studies with the inverse agonist radio- 3 moderate selectivity versus the M1 and M3 mAChR subtypes. ligand, [ H]-NMS and ACh, CCh, or oxo-M suggest that Radioligand-binding studies also suggested that ML380 could ML380 and ML375 exert their effects in a manner broadly increase the affinity of ACh for the M5 mAChR (Gentry consistent with a two-state mechanism (Fig. 2; Table 2), as et al., 2014). ML375 is a highly subtype-selective NAM of previously described for PAMs of the muscarinic M1 receptor,

Fig. 7. Three-way radioligand-binding studies to demonstrate the effect of ML375 (0.3–30 mM) on (A) ACh-, (B) CCh-, (C) oxo- M–, and (D) pilocarpine-mediated inhibition of [3H]-NMS binding in whole CHO-hM5 cells. Cells were incubated with radioligand for 2 h at 37°C; data are expressed as a percentage of control specific binding and represent the mean 6 S.E.M. of at least three independent experiments performed in triplicate. Fitted curves are from global analysis of datasets according to eq. 4 (in Materials and Methods) with parameter estimates shown in Table 3. 434 Berizzi et al.

Fig. 8. Effect of ML380 (10 mM) on (A) ACh- and (B) oxo-M– stimulated IP accumulation in CHO-hM5 cells; (C) BQZ12 (10–30 nM) on ACh-stimulated IP accumulation in CHO- hM1 cells; and (D) ML375 (10–30 mM) on ACh-stimulated IP accumulation in CHO-hM5 cells, all after 30-minute pretreatment with 3 mM PBZ, followed by three phosphate- buffered saline washes. Dashed line indicates the control (untreated) response to ACh. Data points represent the mean 6 S.E.M. of at least three independent experiments performed in duplicate or triplicate. Fitted curves to ACh (6 PBZ) are from a global fit to eq. 1; curves of ACh in the presence of allosteric modulator are fitted according to a three-parameter logistic equation. Downloaded from

BQCA and BQZ12 (Canals et al., 2012; Abdul-Ridha et al., and maximal response of ACh to stimulate IP accumulation in 2014a). However, the modulation of the partial agonist, CHO-hM cells and opened a window to visualize both

5 molpharm.aspetjournals.org pilocarpine, by both ML380 and ML375 indicated that negative and positive effects on the maximal response to the additional mechanisms might contribute to their pharmacol- agonist. Under these conditions, ML380 clearly increased ogy. ML380 displayed neutral affinity cooperativity with the potency of both ACh and oxo-M, but did not increase the respect to pilocarpine, yet potentiated its function in IP maximal response (Fig. 8), suggesting that it modulates only accumulation assays by approximately 60-fold (Tables the affinity of full agonists. Similar studies with the M1 2 and 3). In addition, the PAM agonist displayed a steeper mAChR verified that this PAM effect was conserved across concentration–response relationship when mediating IP ac- mAChRs; under identical assay conditions, the M1 mAChR cumulation in both the absence and presence of orthosteric PAM, BQZ12, caused only a leftward shift in the agonists (Figs. 3 and 5), suggesting that it may bind co- concentration–response curve to ACh, with no change in operatively with respect to itself or that it has the potential to maximal agonist response (Fig. 8). However, the NAM, at ASPET Journals on October 2, 2021 change the sensitivity of the coupling of the receptor to the ML375, clearly decreased the maximal response to ACh, Gq/11 pathway. Finally, ML375 displayed only weak negative respectively (Fig. 8), suggesting that it engenders changes in affinity cooperativity with pilocarpine, yet ablates the func- the ability of the M5 mAChR to couple to G protein(s) as well as tional response to this agonist (Tables 2 and 3). These effects modulating the affinity of agonist binding. The effect of ML380 cannot be accommodated within a simple model allowing only on pilocarpine signaling efficacy is also a clear indicator that for affinity modulation and indicate that both ML380 and the PAM can exert probe-dependent effects on receptor– ML375 might additionally modulate the signaling efficacy of effector coupling as well. the M5 mAChR. Although the location of allosteric sites on the M5 mAChR As effects of allosteric modulators on relative efficacy are has not been extensively studied, mutagenesis studies, as well often difficult to detect in functional assays (due to the use of as computational and structural insights, support the notion full agonists that reach the maximal system response and/or that there is at least one common allosteric binding site in the assays with high degrees of stimulus–response coupling), we extracellular vestibule of mAChRs (Huang et al., 2005; Dror used the receptor-alkylating agent, PBZ, to reduce levels of et al., 2013; Kruse et al., 2013). Thus, despite topographical receptor reserve. Pretreatment with PBZ reduced the potency similarities, these data raise the potential of differential molecular modes of allosteric modulation within the same receptor family. Across the wider GPCR superfamily, different modes of allosteric modulation have already been observed: at the family C mGlu1 and mGlu5 receptors, NAMs such as CPCCOEt and MPEP do not alter the ability of orthosteric agonists to bind to the receptor, but exclusively negatively modulate agonist efficacy (Litschig et al., 1999; Bradley et al., 2011). At the other end of the spectrum are the prototypical mAChR NAMs, gallamine and C7/3-phth, and the M1 mAChR- selective PAMs, BQCA and BQZ12, which appear to modulate only ACh affinity (Canals et al., 2012; Abdul-Ridha et al., Fig. 9. Effect of atropine or ML375 on IP accumulation in CHO-hM5 cells 2014a). Many allosteric modulators of GPCRs exert a mixed m stimulated by a single, fixed concentration of ML380 (10 M). Data points mechanism with respect to agonist affinity and efficacy, represent the mean 6 S.E.M. of at least three independent experiments performed in triplicate. Fitted curves are from a four-parameter logistic including muscarinic M2 and M4 receptor PAMs (Valant equation. et al., 2012; Croy et al., 2014) and mGlu4 and mGlu5 receptor Mechanisms of M5 Muscarinic Receptor Allosteric Modulation 435

PAMs (Mathiesen et al., 2003; Bradley et al., 2011). The Chan WY, McKinzie DL, Bose S, Mitchell SN, Witkin JM, Thompson RC, Christo- poulos A, Lazareno S, Birdsall NJ, Bymaster FP, et al. (2008) Allosteric modulation studies described in this work indicate that positive modula- of the muscarinic M4 receptor as an approach to treating schizophrenia. Proc Natl – tion of the M5 mAChR can occur via regulation of either Acad Sci USA 105:10978 10983. Christopoulos A (1998) Assessing the distribution of parameters in models of ligand- affinity (for ACh and oxo-M) or efficacy (as for pilocarpine) and receptor interaction: to log or not to log. Trends Pharmacol Sci 19:351–357. that negative modulation is mediated by changes in both Christopoulos A (2014) Advances in G protein-coupled receptor allostery: from function to structure. Mol Pharmacol 86:463–478. properties. Christopoulos A and Mitchelson F (1997) Application of an allosteric ternary complex If, as our findings suggest, this is achieved via interaction model to the technique of pharmacological resultant analysis. J Pharm Pharmacol with a common allosteric site, it would be interesting for 49:781–786. Conn PJ, Christopoulos A, and Lindsley CW (2009) Allosteric modulators of GPCRs: future studies to determine whether this site represents the a novel approach for the treatment of CNS disorders. Nat Rev Drug Discov 8:41–54. prototypical allosteric site proposed for other mAChRs in the Croy CH, Schober DA, Xiao H, Quets A, Christopoulos A, and Felder CC (2014) Characterization of the novel positive allosteric modulator, LY2119620, at the extracellular vestibule (Kruse et al., 2014), or whether the M5 muscarinic M(2) and M(4) receptors. Mol Pharmacol 86:106–115. mAChR modulators interact with a hitherto unidentified site Dror RO, Green HF, Valant C, Borhani DW, Valcourt JR, Pan AC, Arlow DH, Canals M, Lane JR, Rahmani R, et al. (2013) Structural basis for modulation of a on this receptor. More importantly, this may have implica- G-protein-coupled receptor by allosteric drugs. Nature 503:295–299. tions to the ultimate use of such compounds and how their Fink-Jensen A, Fedorova I, Wörtwein G, Woldbye DP, Rasmussen T, Thomsen M, Bolwig TG, Knitowski KM, McKinzie DL, Yamada M, et al. (2003) Role for M5 effects translate into native tissue or in vivo studies. For muscarinic acetylcholine receptors in cocaine addiction. J Neurosci Res 74:91–96. instance, a recent investigation by Foster et al. (2014) using an Foster DJ, Gentry PR, Lizardi-Ortiz JE, Bridges TM, Wood MR, Niswender CM, Sulzer D, Lindsley CW, Xiang Z, and Conn PJ. (2014) J Neurosci 34:3253–3262. earlier generation M5 mAChR PAM (ML129) (Bridges et al., Gentry PR, Kokubo M, Bridges TM, Kett NR, Harp JM, Cho HP, Smith E, Chase P, Downloaded from 2009) found that somatodendritic M5 mAChRs increase sub- Hodder PS, Niswender CM, et al. (2013) Discovery of the first M5-selective and CNS penetrant negative allosteric modulator (NAM) of a muscarinic acetylcholine stantia nigra pars compacta dopamine neuron excitability, receptor: (S)-9b-(4-chlorophenyl)-1-(3,4-difluorobenzoyl)-2,3-dihydro-1H-imidazo[2,1-a] – whereas terminal M5 mAChRs inhibit dopamine release. It is isoindol-5(9bH)-one (ML375). J Med Chem 56:9351 9355. Gentry PR, Kokubo M, Bridges TM, Noetzel MJ, Cho HP, and Lamsal A, Smith E, envisaged that the availability of newer generation PAMs and Chase P, Hodder PS, Niswender CM et al. (2014). Development of a highly potent, NAMs with different modes of allosteric modulator effects can novel M5 positive allosteric modulator (PAM) demonstrating CNS exposure: also be used to shed further insights into the physiology and 1-((1H-indazol-5-yl)sulfoneyl)-N-ethyl-N-(2-(trifluoromethyl)benzyl)piperidine- 4-carboxamide (ML380). J Med Chem 57:7804–7810. molpharm.aspetjournals.org potential therapeutic targeting of the M5 mAChR. Gregory KJ, Hall NE, Tobin AB, Sexton PM, and Christopoulos A (2010) Identifi- cation of orthosteric and allosteric site mutations in M2 muscarinic acetylcholine Authorship Contributions receptors that contribute to ligand-selective signaling bias. J Biol Chem 285: 7459–7474. Participated in research design: Berizzi, Gentry, Langmead, Gregory KJ, Sexton PM, and Christopoulos A (2007) Allosteric modulation of mus- – Christopoulos. carinic acetylcholine receptors. Curr Neuropharmacol 5:157 167. Huang XP, Prilla S, Mohr K, and Ellis J (2005) Critical amino acid residues of the Conducted experiments: Berizzi, Gentry, Rueda, Den Hoedt. common allosteric site on the M2 muscarinic acetylcholine receptor: more simi- Performed data analysis: Berizzi, Gentry, Rueda, Den Hoedt, larities than differences between the structurally divergent agents gallamine and Langmead, Christopoulos. bis(ammonio)alkane-type hexamethylene-bis-[dimethyl-(3-phthalimidopropyl) ammonium]dibromide. Mol Pharmacol 68:769–778. Wrote or contributed to the writing of the manuscript: Berizzi, Keov P, López L, Devine SM, Valant C, Lane JR, Scammells PJ, Sexton PM, and Christopoulos A (2014) Molecular mechanisms of bitopic ligand engagement

Gentry, Rueda, Sexton, Langmead, Christopoulos. at ASPET Journals on October 2, 2021 with the M1 muscarinic acetylcholine receptor. J Biol Chem 289:23817–23837. References Kruse AC, Kobilka BK, Gautam D, Sexton PM, Christopoulos A, and Wess J (2014) Muscarinic acetylcholine receptors: novel opportunities for drug development. Nat Abdul-Ridha A, Lane JR, Mistry SN, López L, Sexton PM, Scammells PJ, Rev Drug Discov 13:549–560. Christopoulos A, and Canals M (2014a) Mechanistic insights into allosteric Kruse AC, Ring AM, Manglik A, Hu J, Hu K, Eitel K, Hübner H, Pardon E, Valant C, structure-function relationships at the M1 muscarinic acetylcholine receptor. J Sexton PM, et al. (2013) Activation and allosteric modulation of a muscarinic Biol Chem 289:33701–33711. acetylcholine receptor. Nature 504:101–106. Abdul-Ridha A, López L, Keov P, Thal DM, Mistry SN, Sexton PM, Lane JR, Canals Langmead CJ, Watson J, and Reavill C (2008) Muscarinic acetylcholine receptors as M, and Christopoulos A (2014b) Molecular determinants of allosteric modulation at CNS drug targets. Pharmacol Ther 117:232–243. the M1 muscarinic acetylcholine receptor. J Biol Chem 289:6067–6079. Aurelio L, Valant C, Figler H, Flynn BL, Linden J, Sexton PM, Christopoulos A, Leach K, Loiacono RE, Felder CC, McKinzie DL, Mogg A, Shaw DB, Sexton PM, and Scammells PJ (2009) 3- and 6-Substituted 2-amino-4,5,6,7-tetrahydrothieno[2,3- and Christopoulos A (2010) Molecular mechanisms of action and in vivo validation of an M4 muscarinic acetylcholine receptor allosteric modulator with potential c]pyridines as A1 adenosine receptor allosteric modulators and antagonists. Bioorg – Med Chem 17:7353–7361. antipsychotic properties. Neuropsychopharmacology 35:855 869. Basile AS, Fedorova I, Zapata A, Liu X, Shippenberg T, Duttaroy A, Yamada M, Leach K, Sexton PM, and Christopoulos A (2007) Allosteric GPCR modulators: taking and Wess J (2002) Deletion of the M5 muscarinic acetylcholine receptor attenuates advantage of permissive receptor pharmacology. Trends Pharmacol Sci 28: – morphine reinforcement and withdrawal but not morphine analgesia. Proc Natl 382 389. Acad Sci USA 99:11452–11457. Leff P, Dougall IG, and Harper D (1993) Estimation of partial agonist affinity by Black JW and Leff P (1983) Operational models of pharmacological agonism. Proc R interaction with a full agonist: a direct operational model-fitting approach. Br J – Soc Lond B Biol Sci 220:141-162. Pharmacol 110:239 244. Bradley SJ, Langmead CJ, Watson JM, and Challiss RA (2011) Quantitative analysis Litschig S, Gasparini F, Rueegg D, Stoehr N, Flor PJ, Vranesic I, Prézeau L, Pin JP, reveals multiple mechanisms of allosteric modulation of the mGlu5 receptor in rat Thomsen C, and Kuhn R (1999) CPCCOEt, a noncompetitive metabotropic gluta- astroglia. Mol Pharmacol 79:874–885. mate receptor 1 antagonist, inhibits receptor signaling without affecting glutamate – Brady AE, Jones CK, Bridges TM, Kennedy JP, Thompson AD, Heiman JU, binding. Mol Pharmacol 55:453 461. Breininger ML, Gentry PR, Yin H, Jadhav SB, et al. (2008) Centrally active Ma L, Seager MA, Wittmann M, Jacobson M, Bickel D, Burno M, Jones K, Graufelds allosteric potentiators of the M4 muscarinic acetylcholine receptor reverse VK, Xu G, Pearson M, et al. (2009) Selective activation of the M1 muscarinic amphetamine-induced hyperlocomotor activity in rats. J Pharmacol Exp Ther acetylcholine receptor achieved by allosteric potentiation. Proc Natl Acad Sci USA 327:941–953. 106:15950–15955. Bridges TM, Kennedy JP, Cho HP, Breininger ML, Gentry PR, Hopkins CR, Conn PJ, Mathiesen JM, Svendsen N, Bräuner-Osborne H, Thomsen C, and Ramirez MT and Lindsley CW (2010) Chemical lead optimization of a pan G(q) mAChR M(1), M (2003) Positive allosteric modulation of the human metabotropic glutamate re- (3), M(5) positive allosteric modulator (PAM) lead. Part I: Development of the first ceptor 4 (hmGluR4) by SIB-1893 and MPEP. Br J Pharmacol 138:1026–1030. highly selective M(5) PAM. Bioorg Med Chem Lett 20:558–562. Matsui M, Yamada S, Oki T, Manabe T, Taketo MM, and Ehlert FJ (2004) Functional Bridges TM, Marlo JE, Niswender CM, Jones CK, Jadhav SB, Gentry PR, Plumley analysis of muscarinic acetylcholine receptors using knockout mice. Life Sci 75: HC, Weaver CD, Conn PJ, and Lindsley CW (2009) Discovery of the first highly 2971–2981. M5-preferring muscarinic acetylcholine receptor ligand, an M5 positive allosteric Motulsky H and Christopoulos A (2004) Fitting models to biological data using linear modulator derived from a series of 5-trifluoromethoxy N-benzyl isatins. JMed and nonlinear regression: a practical guide to curve fitting, Oxford University Chem 52:3445–3448. Press, Oxford. Bymaster FP, Whitesitt CA, Shannon HE, DeLapp N, Ward JS, Calligaro DO, Porter AC, Bymaster FP, DeLapp NW, Yamada M, Wess J, Hamilton SE, Nathanson Shipley LA, Buelke-Sam JL, Bodick NC, Farde L, et al. (1997) Xanomeline: a NM, and Felder CC (2002) M1 muscarinic receptor signaling in mouse hippo- selective muscarinic agonist for the treatment of Alzheimer’s disease. Drug Dev Res campus and cortex. Brain Res 944:82–89. 40:158–170. Shannon HE, Rasmussen K, Bymaster FP, Hart JC, Peters SC, Swedberg MD, Canals M, Lane JR, Wen A, Scammells PJ, Sexton PM, and Christopoulos A (2012) A Jeppesen L, Sheardown MJ, Sauerberg P, and Fink-Jensen A (2000) Xanomeline, Monod-Wyman-Changeux mechanism can explain G protein-coupled receptor an M(1)/M(4) preferring muscarinic cholinergic receptor agonist, produces (GPCR) allosteric modulation. J Biol Chem 287:650–659. antipsychotic-like activity in rats and mice. Schizophr Res 42:249–259. 436 Berizzi et al.

Shirey JK, Brady AE, Jones PJ, Davis AA, Bridges TM, Kennedy JP, Jadhav SB, Weiner DM, Levey AI, and Brann MR (1990) Expression of muscarinic acetylcholine and Menon UN, Xiang Z, Watson ML, et al. (2009) A selective allosteric potentiator of dopamine receptor mRNAs in rat basal ganglia. Proc Natl Acad Sci USA 87:7050–7054. the M1 muscarinic acetylcholine receptor increases activity of medial prefrontal Wess J, Eglen RM, and Gautam D (2007) Muscarinic acetylcholine receptors: mutant cortical neurons and restores impairments in reversal learning. J Neurosci 29: mice provide new insights for drug development. Nat Rev Drug Discov 6:721–733. 14271–14286. Yamada M, Basile AS, Fedorova I, Zhang W, Duttaroy A, Cui Y, Lamping KG, Faraci Shirey JK, Xiang Z, Orton D, Brady AE, Johnson KA, Williams R, Ayala JE, FM, Deng CX, and Wess J (2003) Novel insights into M5 muscarinic acetylcholine Rodriguez AL, Wess J, Weaver D, et al. (2008) An allosteric potentiator of M4 receptor function by the use of gene targeting technology. Life Sci 74:345–353. mAChR modulates hippocampal synaptic transmission. Nat Chem Biol 4:42–50. Yasuda RP, Ciesla W, Flores LR, Wall SJ, Li M, Satkus SA, Weisstein JS, Spagnola Thomsen M, Woldbye DP, Wörtwein G, Fink-Jensen A, Wess J, and Caine SB (2005) BV, and Wolfe BB (1993) Development of antisera selective for m4 and m5 mus- Reduced cocaine self-administration in muscarinic M5 acetylcholine receptor- carinic cholinergic receptors: distribution of m4 and m5 receptors in rat brain. Mol deficient mice. J Neurosci 25:8141–8149. Pharmacol 43:149–157. Valant C, Felder CC, Sexton PM, and Christopoulos A (2012) Probe dependence in the allosteric modulation of a G protein-coupled receptor: implications for detection and validation of allosteric ligand effects. Mol Pharmacol 81: Address correspondence to: Dr. Christopher J. Langmead or Prof. Arthur 41–52. Christopoulos, Monash Institute of Pharmaceutical Sciences, Monash Univer- Vilaró MT, Palacios JM, and Mengod G (1990) Localization of m5 muscarinic receptor sity, 381 Royal Parade, Parkville, VI 3052, Australia. E-mail: chris.langmead@ mRNA in rat brain examined by in situ hybridization histochemistry. Neurosci Lett monash.edu or [email protected] 114:154–159. Downloaded from molpharm.aspetjournals.org at ASPET Journals on October 2, 2021 Molecular Mechanisms of Action of M5 Muscarinic Acetylcholine Receptor Allosteric Modulators Alice Berizzi, Patrick R. Gentry, Patricia Rueda, Sandra den Hoedt, Patrick M. Sexton, Christopher J Langmead & Arthur Christopoulos, Molecular Pharmacology

[ML380] (µM)

100 0 3

10 50 30

% specific H]NMS binding

[ 0

-12 -10 -8 -6 -4

Log [Atropine] (M) FIGURE S1 Three-way radioligand binding studies to demonstrate the effect of ML380 (0.3 - 30 µM) on 3 atropine-mediated inhibition of [ H]-NMS binding in whole CHO-hM5 cells. Cells were incubated with radioligand for 2h at 37oC; data are expressed as a % of control specific binding and represent the mean ± SEM three independent experiments performed in triplicate. Fitted curves are from global analysis of datasets according to Equation 4 (in Methods) with parameter estimates shown in the Results.