Muscarinic Acetylcholine Receptors: Novel Opportunities for Drug Development Andrew C

REVIEWS

Muscarinic acetylcholine receptors: novel opportunities for drug development Andrew C. Kruse1,2, Brian K. Kobilka1, Dinesh Gautam3, Patrick M. Sexton4, Arthur Christopoulos4 and Jürgen Wess3 Abstract | The muscarinic acetylcholine receptors are a subfamily of G protein-coupled receptors that regulate numerous fundamental functions of the central and peripheral nervous system. The past few years have witnessed unprecedented new insights into muscarinic receptor physiology, pharmacology and structure. These advances include the first structural views of muscarinic receptors in both inactive and active conformations, as well as a better understanding of the molecular underpinnings of muscarinic receptor regulation by allosteric modulators. These recent findings should facilitate the development of new muscarinic receptor subtype-selective ligands that could prove to be useful for the treatment of many severe pathophysiological conditions.

The muscarinic acetylcholine receptors (mAChRs) com- Recent studies with novel mAChR mouse models prise a family of five related G protein-coupled receptors have provided additional insights into the physiolog (GPCRs) belonging to the α-branch of class A GPCRs1. ical roles of the different mAChR subtypes4. Moreover, The mAChR family consists of five distinct subtypes, during the past few years, several laboratories have suc-

denoted M1 to M5 (and encoded by the genes CHRM1 ceeded in developing ligands that show high selectivity 1 5,6 Department of Molecular to CHRM5). Three of these receptor subtypes (M1, M3 for specific mAChR subtypes . In contrast to conven- and Cellular Physiology, and M5) have been shown to couple to G proteins of the tional muscarinic receptor ligands, which bind to the Stanford University School G family, whereas the remaining two subtypes (M orthosteric receptor site, most of these new ligands bind of Medicine, Stanford, q/11 2 California 94305, USA. and M4) preferentially signal through the Gi/o family of to distinct allosteric sites. Such allosteric ligands can 2 2Present address: G proteins . The mAChRs have a central role in human influence the potency and efficacy of orthosteric ligands, Department of Biological physiology, regulating heart rate, smooth muscle con- and they may possess agonistic or inverse agonist activ- Chemistry and Molecular traction, glandular secretion and many fundamental ity in their own right. Interestingly, bitopic muscarinic Pharmacology, Harvard 3 Medical School, Boston, functions of the central nervous system (CNS) . receptor ligands with preference for certain subtypes 7–9 Massachusetts 02115, USA. Currently, drugs targeting muscarinic receptors are have also been developed recently . Such agents, which 3Molecular Signaling Section, used for the treatment of several pathophysiological con- can interact with both allosteric and orthosteric receptor Laboratory of Bioorganic ditions, including chronic obstructive pulmonary disease, sites simultaneously, offer new opportunities to target Chemistry, National Institute overactive bladder and Sjögren’s syndrome3. Despite the specific mAChR subtypes for therapeutic purposes. of Diabetes and Digestive and Kidney Diseases, powerful and diverse pharmacological actions of mus- The recent determination of the first mAChR struc- US National Institutes of carinic receptor agonists and antagonists, the development tures (M2 and M3 subtypes) represents a milestone in the Health, Bethesda, Maryland of these drugs for other clinical applications has prob- mAChR field10–12. These studies have provided the first 20892–0810, USA. ably been held back, at least in part, by the lack of small- molecular views of mAChRs in both their inactive10,11 and 4 Monash Institute of 12 Pharmaceutical Sciences molecule ligands that can inhibit or activate specific active conformations , revealing the molecular nature and Department of mAChRs with high selectivity. As a result, the precise of the binding sites for orthosteric muscarinic receptor Pharmacology, Monash physiological and pathophysiological roles of the individ- ligands. Moreover, structural and computational stud- University, Parkville, ual mAChR subtypes have, until recently, remained poorly ies have identified the mechanisms by which drug-like Victoria 3052, Australia. defined. However, during the past 15 years the generation allosteric modulators bind to the M subtype12,13. Correspondence to J.W. 2 e‑mail: [email protected] and phenotypic analysis of Chrm1- to Chrm5‑knockout In this Review, we summarize and discuss a series doi:10.1038/nrd4295 mice has been instrumental in improving our under of recent studies that have advanced our knowledge of Published online 6 June 2014 standing of the biology of the individual mAChRs3. mAChR biology, structure and pharmacology. In particular,

NATURE REVIEWS | DRUG DISCOVERY VOLUME 13 | JULY 2014 | 549

we emphasize the potential therapeutic implications of Chrm1‑knockout mice exhibited an age-dependent these novel findings. As mAChRs are prototypic class A cognitive decline in tasks that they performed normally GPCRs, the topics covered herein have implications for at a younger age17. Most importantly, several recent stud-

other members of this receptor superfamily. ies have shown that compounds that act as PAMs at M1 receptors have cognition-enhancing activity in rodents Novel mAChR mouse models and are able to improve impaired cognition in mouse Since the late 1990s, studies with Chrm1- to Chrm5- models of Alzheimer’s disease18,19. Taken together, these knockout mice have elucidated many important physio new findings provide a rational basis for the develop- 3 logical functions of the individual mAChR subtypes . ment of M1-selective drugs (PAMs or agents that can

The outcome of this work suggested that modulating activate M1 receptors directly) for the treatment of the activity of specific mAChR subtypes by selective Alzheimer’s disease and related disorders. ligands might prove to be beneficial for the treatment of many CNS disorders and other clinical conditions3. Schizophrenia. Schizophrenia, which is a chronic dis Until recently, essentially all phenotyping studies were abling brain disorder that affects ~1% of the general carried out with conventional, constitutive mAChR- population20, is characterized by enhanced central dopa- knockout mice, which lack one or more of the genes minergic signalling. Studies with conventional mAChR- encoding mAChRs throughout development in all cells knockout mice have indicated that the lack of central

of the body. The proper interpretation of the phenotypes M1 and/or M4 receptors leads to a ‘dopamine hypersen- observed with conventional mAChR-knockout mice is sitivity phenotype’, which suggests that agents that can often complicated because individual mAChR subtypes enhance signalling through these receptor subtypes may are expressed in many different tissues and cell types, be endowed with antipsychotic activity3. In agreement and compensatory molecular or physiological changes with this concept, two clinical trials demonstrated that

can occur during development. xanomeline, an M1- and M4-preferring muscarinic recep- To circumvent these difficulties, novel conditional tor agonist21, was effective in ameliorating psychosis-like knockout mice have been created that lack specific symptoms22,23. A recent study with cell type-specific mAChR subtypes only in certain tissues or cell types. Chrm4‑knockout mice strongly suggests that a distinct

For these studies, mice containing ‘floxed’ versions of subpopulation of neuronal M4 receptors expressed by

specific genes encoding mAChRs were crossed with D1 dopamine receptor-expressing neurons may be of transgenic mice expressing Cre recombinase in a cell particular importance in mediating the antipsychotic (TABLE 1) 24 type- or tissue-selective fashion . mAChR actions triggered by M4 receptor activation . In addition, knock-in mice have also been reported, in which the Dencker et al.25 reported that the antipsychotic-like effects

native genes encoding M3 or M4 receptors — CHRM3 of xanomeline were almost completely abolished in two

or CHRM4, respectively— were replaced with mutant mouse models of psychosis in which mice lacked M4 (TABLE 1) versions . Here, we summarize several recent receptors selectively in D1 receptor-expressing neurons. studies that have used novel mAChR-mutant mice as well In agreement with these findings, several recent preclini-

as conventional mAChR-knockout mice. These studies cal studies suggest that PAMs of M1 and/or M4 receptors are discussed in the context of selected human diseases in may prove to be clinically useful for the treatment of which the development of novel classes of drugs targeting schizophrenia (reviewed in REFS 18,26–28). muscarinic receptors might be beneficial. Drug addiction. Acetylcholine has a key role in modulat- Targeting mAChRs in disease ing the neurochemical and behavioural CNS responses Alzheimer’s disease. Alzheimer’s disease, which is the to drugs of abuse29. Phenotypic analysis of Chrm5‑

most common form of dementia, is recognized as a major knockout mice suggested that centrally active M5 public health crisis14. At present, there are no drugs avail- receptor blockers might prove to be useful for the treat- able that are highly effective in preventing Alzheimer’s ment of drug abuse, including cocaine or opioid addic- disease or slowing its progress. Based on behavioural, tion3. More recent studies with other mAChR-knockout

pharmacological, anatomical and neurochemical evi- lines indicate that signalling through central M1 and

dence, M1 receptor-selective agonists seem to have the M4 receptors may also modulate drug-seeking behav- potential to ameliorate the symptoms of Alzheimer’s iour. For example, recent work demonstrated that 15 disease and related cognitive disorders . In fact, several allosteric M1 receptor agonists and xanomeline can

recent studies support the concept that M1 receptor- attenuate the reinforcing and discriminative stimulus selective agonists or positive allosteric modulators effects of cocaine30,31. These effects were greatly reduced (PAMs) may prove to be useful as cognition-enhancing or abolished in Chrm1‑knockout mice or in mice that 16 –/– drugs. Davis et al. demonstrated that M1 receptor defi- were deficient in both M1 and M4 receptors (Chrm1 ciency increases amyloidogenic processing of amyloid Chrm4–/– mice). Moreover, Schmidt et al.32 reported precursor protein (APP) — a key feature of Alzheimer’s that the reinforcing effects of chronic cocaine self- disease — in a mouse model of Alzheimer’s disease. In a administration are significantly increased in Chrm4‑ related study, Medeiros et al.17 reported that the lack of knockout mice. These findings indicate that centrally

M1 receptors further exacerbated cognitive processes active drugs that can selectively stimulate M1 and/or M4 and Alzheimer’s disease‑related pathological features receptors may prove to be useful in the treatment of drug in a mouse model of Alzheimer’s disease. In addition, addiction.

550 | JULY 2014 | VOLUME 13 www.nature.com/reviews/drugdisc

Table 1 | Novel mAChR-mutant mouse models Mouse strain Major phenotypes Refs Chrm1‑mutant mice selectively Lack of mGluR-mediated LTD in the hippocampus, accompanied 116

lacking M1 receptors in the forebrain or by changes in presynaptic neurotransmitter release hippocampal CA3 pyramidal neurons

(use of floxed M1 receptor mice) Mice carrying a floxed Chrm2 gene N/A *

(floxed M2 receptor mice) Chrm3‑mutant mice selectively lacking Pancreatic β‑cells: deficits in glucose tolerance, accompanied by 34 M3 receptors in specific cell types impaired insulin release in vivo (use of floxed M receptor mice) 3 Brain (neurons and glial cells): reduced body size; hypoplasia of 117 the anterior pituitary gland; reduced levels of pituitary and serum growth hormone, as well as prolactin Brain (neurons and glial cells): low bone mass, probably owing to 118 enhanced central sympathetic outflow Hepatocytes: none (no changes in glucose homeostasis and 119 hepatic glucose fluxes) Osteoblasts: none (no changes in bone formation and bone mass) 118 Smooth muscle cells (airways): restoration of normal lung 120 function in obese mice

Chrm3‑knock-in mice containing Greatly reduced M3 receptor phosphorylation and impaired 121 15 Ser→Ala point mutations within arrestin recruitment following M3 receptor activation; deficits in the i3 loop of the M3 receptor fear conditioning, learning and memory Impaired glucose tolerance, accompanied by reduced 40

M3 receptor‑mediated insulin release from pancreatic islets Chrm4‑mutant mice selectively lacking Increased dopamine efflux in the nucleus accumbens; modulation 24

M4 receptors in D1 dopamine receptor- of dopamine-dependent behaviours expressing neurons (use of floxed M4 receptor mice) Strong reduction of the antipsychotic-like effects of xanomeline 25 in two mouse models of psychosis Chrm5‑knock-in mice containing an Reduction in evoked dopamine release from striatal slices 122

18‑amino-acid deletion within the M5 receptor‑i3 loop (residues 369–386)

LTD, long-term depression; mAChR, muscarinic acetylcholine receptor (five distinct subtypes, denoted M1 to M5, which are encoded by the genes Chrm1 to Chrm5 in mice); mGluR, metabotropic glutamate receptor. *J.W. & J. Jeon, unpublished observations.

Type 2 diabetes. Type 2 diabetes has emerged as a major arrestin-dependent β‑cell pathways may further enhance threat to public health worldwide33. One of the key patho- insulin secretion40,41. Currently, ligands that are able to

physiological features of type 2 diabetes is that pancreatic selectively promote signalling through M3 receptors are β‑cells are unable to secrete sufficient insulin to overcome not available, and the development of such agents would peripheral insulin resistance. Studies with β‑cell-selective greatly facilitate studies aimed at testing the potential use- 34,35 Chrm3‑mutant mice showed that strategies aimed at fulness of targeting β‑cell M3 receptors for the treatment

enhancing the activity of β‑cell M3 receptors may prove to of type 2 diabetes and related metabolic disorders. be useful for the treatment of type 2 diabetes. To further test this hypothesis, two recent studies36,37 analysed trans- Cancer. Accumulating evidence suggests that mAChR-

genic mice that expressed an M3 receptor-based designer dependent signalling pathways can promote cell prolifer- 42,43 receptor (Gq DREADD; where DREADD stands for ation and cancer progression . Interestingly, Raufman ‘designer receptor exclusively activated by a designer drug’) et al.44 demonstrated that Chrm3‑knockout mice dis-

selectively in pancreatic β‑cells. This modified M3 receptor played reduced epithelial cell proliferation, tumour is unable to bind acetylcholine, the endogenous ligand, but number and size in a mouse model of colon neoplasia can be selectivity activated by clozapine N‑oxide (CNO), (azoxymethane-induced colon cancer). In a related 36,38 45 a drug that is otherwise pharmacologically inert . study , M3 receptor deficiency was associated with a Systematic metabolic analysis of this mouse line con- pronounced reduction in tumour number and volume firmed that CNO-dependent activation of this designer in a genetic model of intestinal neoplasia (ApcMin/+ mice). 46 M3 receptor has numerous beneficial effects on β‑cell Magnon et al. reported that M1 receptor deficiency function and is able to prevent experimentally induced inhibited mAChR-mediated prostate cancer invasion diabetes36,37. Although there is evidence that the ability of and metastasis in two mouse models of prostate cancer.

β‑cell M3 receptors to promote insulin release depends on These findings support the concept that subtype-selective 39 the presence of G proteins of the Gq family , two recent mAChR antagonists may prove to be clinically useful for

studies suggest that M3 receptor-mediated activation of the treatment of certain forms of cancer.

NATURE REVIEWS | DRUG DISCOVERY VOLUME 13 | JULY 2014 | 551

Novel mAChR pharmacology and by the potential for direct activation of receptor A major shift in drug discovery has been the recognition signalling via the allosteric site52. PAMs enhance orthos- that most — if not all — GPCRs possess spatially distinct teric activity, negative allosteric modulators (NAMs) allosteric sites that can be exploited by small molecules inhibit it, and agents that occupy an allosteric site but do to modulate the activity of orthosteric ligands6,47,48. The not change the activity of orthosteric ligands are referred mAChRs are arguably the pre-eminent models for con- to as neutral allosteric ligands (NALs). Allosteric com- tributing to our understanding of GPCR allostery5,49–51 pounds that directly activate GPCRs are called allosteric (see BOX 1 for a representative list of muscarinic allosteric agonists. However, these designations are contextual, as modulators). they are dependent on the receptor, orthosteric ligand and 52–54 the assay system used . For example, the M1 receptor- Pharmacological characteristics of allosteric modulators. selective modulator BQCA (REF. 55) acts solely as a PAM Allosteric ligands promote conformational changes in of acetylcholine activity when assayed in a cell line with 56 the receptor that manifest as an alteration in the proper- low M1 receptor expression and/or at signalling path 57 ties of a ligand bound to the classical, orthosteric site, ways that are weakly coupled to the M1 receptor , but it acts as both a full allosteric agonist and as a PAM in a system 57 with a high M1 receptor reserve . Box 1 | Selected allosteric modulators of mAChRs Characteristically, the effects of allosteric modula- The allosteric modulators listed here have been used tors are saturable: that is, no further activity is observed to provide important insights into muscarinic above a certain limit, irrespective of modulator dose. acetylcholine receptor (mAChR) allostery. Note that This phenomenon is driven by the cooperativity between this list is by no means exhaustive. Some of the orthosteric and allosteric sites, and can range from modulators bind to all five mAChR subtypes but are subtle fine-tuning to very large modulation. For instance, listed under the mAChR subtype for which they have N‑chloromethyl brucine potentiates acetylcholine affin- been most extensively studied. 58 ity by threefold at the M3 receptor , whereas BQCA M1 receptor potentiates acetylcholine affinity at the M1 receptor by • Tacrine ~100‑fold57. A second property is probe dependence59, • MT7 whereby the magnitude and direction of the allosteric • Brucine effect mediated by the same modulator acting at the • Staurosporine same receptor varies depending on the orthosteric ligand • BQCA used to probe receptor function. A striking example is • VU0119498 LY2033298, which is a PAM of oxotremorine‑M signalling at the M receptor, but a NAM of xanomeline at the • VU0029627 2 same receptor60. • ML169 The third characteristic of GPCR allostery is biased M2 receptor agonism, which refers to the ability of different ligands • Gallamine to stabilize a subset of functionally relevant GPCR con- • Alcuronium formations such that different signalling outputs are 54,61,62 • C7/3‑phth achieved at the exclusion of others . Biased agonism • W84 can be imposed on the signalling of orthosteric agonists • DUO3 by co‑bound modulators. One example is the compound • Tacrine VU0029767, which potentiates acetylcholine-mediated phospholipase C activity via the M receptor, but does • LY2033298 1 not affect acetylcholine-induced phospholipase D acti- • LY2119620 vation63. Finally, a crucial property of allosteric modu-

M3 receptor lators is potential mAChR-subtype selectivity. This • N‑chloromethyl brucine can be attained either by targeting a less conserved site • WIN62577 on the receptor or via selective cooperativity with the • VU0119498 orthosteric ligand, even if the allosteric site is shared between subtypes64. Indeed, selective cooperativity is M4 receptor • MT3 the major mechanism by which compounds such as thiochrome and LY2033298 show high selectivity for • Thiochrome the M receptor64–66. • LY2033298 4 • LY2119620 Bitopic muscarinic ligands. In recent years, research • VU0010010 has turned to the rational design of hybrid molecules, • VU0152100 which are designed to simultaneously bridge orthosteric and allosteric sites within a single receptor; such M5 receptor • VU0119498 compounds have been termed ‘bitopic’, ‘dualsteric’ or 7,67,68 FIG. 1 • VU0238429 ‘multivalent’ . As summarized in , each of the different modes of targeting mAChRs has its pros and • ML375 cons. The orthosteric site can provide high affinity and is

552 | JULY 2014 | VOLUME 13 www.nature.com/reviews/drugdisc

supported by decades of studies of structure–activity study of the M1 receptor-selective agonist TBPB also relationships (SARs), but mAChR-subtype selectivity identified orthosteric and allosteric fragments of this remains largely unattainable. Allosteric sites provide molecule53. This latter study is of particular relevance

the opportunity to achieve greater receptor selectivity because TBPB, together with other novel M1 receptor- and promote more physiological response patterns. selective agonists such as AC‑42 and 77‑LH‑28‑1, was However, high-affinity modulators have not been originally classified as an allosteric agonist77–80. However, reported to date, and certain modulator scaffolds — like McN‑A-343, these agents display mixed modes of like many orthosteric drugs — show minimal improve- orthosteric or allosteric pharmacology depending on the ments in activity despite numerous modifications69. experimental assay conditions48,53,81. It is thus likely that The bitopic ligand approach attempts to target the allos- such compounds adopt novel poses within the mAChRs teric site to achieve selectivity and the orthosteric site that bridge both orthosteric and allosteric sites, hence to provide high affinity. accounting for their unique pharmacology81. Despite some challenges7,68,70, the pursuit of bitopic ligands has yielded interesting pharmacological findings. Towards the molecular mechanisms of allostery. The Disingrini et al.71 were the first to combine a nonselec- simplest mechanism to account for allosteric behaviour tive, high-affinity orthosteric agonist (iperoxo) with an is the Monod–Wyman–Changeux (MWC) model82,83,

M2 receptor-selective allosteric modulator to generate which postulates that receptors exist in an equilibrium

an M2 receptor-selective agonist. This concept was between different states, and that orthosteric and allos- extended by Mohr and colleagues72–74. Interestingly, teric ligands select one or more of these states over although bitopic agonists with mAChR-subtype selec- others84. Because this occurs via topographically distinct tivity and signalling bias have been described, sub- sites, the state that is stabilized by one class of ligand can stantial gains in affinity were generally not observed7. present a modified binding surface for the other class, By contrast, the bitopic muscarinic receptor antagonist which is manifested as the allosteric interaction. If this THRX‑160209 demonstrated remarkable gains in both mechanism were restricted to two states, it would be 75 affinity and selectivity for the M2 receptor , which high- expected that any ligand favouring the active state would lights that the appropriate combination of an orthosteric display some degree of agonism, whereas compounds antagonist, NAM and linker can yield ligands with the stabilizing the inactive state would show some degree desired pharmacological properties. of inverse agonism. Moreover, allosteric ligands that These findings raise the possibility that previously favoured the active state would be PAMs for agonists but described selective and/or biased agonists may repre- NAMs for inverse agonists; this is the simplest molecu- sent unappreciated bitopic ligands. Consistent with this lar explanation for probe dependence. Interestingly,

concept, a study of the biased M2 receptor partial agonist these predictions were confirmed in a recent study of 57 McN‑A-343 unmasked a pure orthosteric agonist and BQCA activity at the M1 receptor , which indicates that pure NAM of agonist efficacy following reverse engi- some mAChR modulators can mediate their effects pre- neering of McN‑A-343 into fragments76. A more recent dominantly through a simple conformational selection mechanism. However, this simple model does not explain the probe dependence observed with compounds such as

Orthosteric Allosteric Bitopic LY2033298 at the M2 receptor, which can show PAM or NAM activity with different agonists60. It also does not explain biased allosteric modulation, or why compounds such as gallamine are NAMs for agonists and antagonists85. Thus, additional mechanisms seem to be involved, most obviously involving multiple active receptor states86. Prior studies have also shown that compounds such as staurosporine and WIN62577 interact with a second allosteric site on mAChRs that is distinct from the common site87,88. The location of this site Well-deﬁned SAR ✓ ✗ ✗ remains unidentified, but it may be intracellular89. If so, High aﬃnity ✓ ✗ ✓ one can envisage more complex scenarios that involve Subtype selectivity ✗ ✓ ✓ an allosteric mechanism with respect to the orthosteric Spatial and temporal ✗ ✓ ✗ site, but potentially a steric mechanism with respect to selectivity of receptor–transducer interactions. Similarly, it has been endogenous signalling postulated that gallamine may utilize a two-step binding Figure 1 | Modes of targeting mAChRs (GPCRs) by different classes of process, involving a peripheral domain that is distinct ligands. Orthosteric ligands (green) bind to the site recognizedNature Reviews by the endogenous | Drug Discovery from the common allosteric site, in its interaction with agonist (acetylcholine) for the receptor. Allosteric ligands (yellow) bind to a the receptor85. Finally, the simplest allosteric models do topographically distinct site. Bitopic ligands (blue) concomitantly interact with both not explicitly incorporate receptor oligomers as the key orthosteric and allosteric sites. The key properties generally associated with each functional units; a tetrameric arrangement has been mode of receptor targeting are also indicated. GPCR, G protein-coupled receptor; proposed, at least for the M2 receptor, to account for mAChR, muscarinic acetylcholine receptor; SAR, structure–activity relationship. the observed cooperative effects of both orthosteric and

NATURE REVIEWS | DRUG DISCOVERY VOLUME 13 | JULY 2014 | 553

90,91 allosteric ligands . Thus, despite the identification of Extracellular a vast repertoire of behaviours for muscarinic allosteric ligands, the molecular mechanisms underlying these behaviours are only starting to be elucidated. It is par- ticularly exciting, therefore, that recent structural and computational biology breakthroughs have shed new light on this matter.

Structural insights into ligand binding mode. Although mAChRs are among the most extensively studied GPCRs in terms of their pharmacology and biological function, the molecular structures of these receptors have, until recently, remained poorly understood. In 2012 the first Intracellular mAChR structures were published10,11, revealing the molecular organization of the M and M receptor sub- 2 3 Figure 2 | Overall structure of the M and M receptors. types in inactive, antagonist (inverse agonist)-bound 2 3 The structures of the inactive, antagonist-bound M2 (blue) conformations. This work was made possible owing to and M (yellow) receptors areNature highly Reviews similar | toDrug each Discovery other advances in GPCR biochemistry and crystallography, 3 in overall fold. For the sake of clarity, only the M3 receptor principal among them the T4 lysozyme fusion method92, ligand (tiotropium) is shown. The overall architecture 93 new detergents , the lipidic cubic phase crystallization of the M2 and M3 muscarinic acetylcholine receptors method94,95 and micro-focus X‑ray diffraction data (mAChRs) is very similar to that of other biogenic collection96. amine G protein-coupled receptors (GPCRs), with similar orthosteric ligand binding sites (orange spheres). The structures of the M2 and M3 receptors revealed that, like other biogenic amine receptors, members of the mAChR family share the seven-pass transmembrane topology and overall fold of other GPCRs, with a few fea- (FIG. 2) tures extending outside of the membrane plane . Strikingly, in the structures of the M2 and M3 receptors, The ligand-binding pocket is deeply buried within the the orthosteric ligand is almost completely occluded membrane (FIGS 2,3a), and is situated similarly to the from solvent, with a tyrosine ‘lid’ located directly above orthosteric binding sites of the related histamine97, dopa- — that is, extracellular to — the ligand (FIG. 3). This lid mine98, adrenaline99–101 and serotonin102,103 receptors. The divides a large, solvent-accessible cavity into two distinct

ligands bound to the M2 and M3 receptors in the two regions, only one of which interacts with the bound crystal structures — quinuclidinyl benzilate (QNB) and antagonists (FIG. 3a). The upper portion of this cavity, tiotropium, respectively — are both antagonists (inverse termed the extracellular vestibule, is lined with residues agonists) with similar chemical structures. In each case, that have been implicated in the binding of allosteric the ligand occupies a similar pose (FIG. 3b), engaging modulators107. As discussed below, LY2119620, which is in extensive hydrophobic contact with the receptor. A a PAM of mAChRs, binds to this outer cavity12. Although smaller number of polar contacts are evident: a pair of the backbone fold of the structures surrounding this hydrogen bonds between Asn6.52 and the ligand hydroxyl allosteric site is remarkably well conserved between the

and ketone, and a charge–charge interaction between the M2 and M3 receptors, the amino acid side chains lining cationic amine of the ligand and the conserved Asp3.32 it show more divergence10. Hence, the development of (superscripts denote Ballesteros–Weinstein104 GPCR ligands that interact with non-conserved amino acids numbering) (FIG. 3b). The latter interaction is seen in all in the allosteric site provides a possible means by which biogenic amine receptor structures solved to date and subtype-selective therapeutics might be developed.

has been shown to make a major energetic contribution Molecular dynamics studies of the M3 receptor indi- to binding105. The paired hydrogen bonding between the cate that the allosteric site may have a role in facilitating ligand and Asn6.52 seems to be a unique feature of the the binding of orthosteric ligands. Simulations of tio-

mAChR family and has been proposed to be an impor- tropium binding to the M3 receptor suggested that this tant factor in slow ligand dissociation from mAChRs106. ligand may rapidly interact with the extracellular ves- Comparison of the orthosteric ligand-binding pockets tibule of the receptor, but in these simulations tiotro- 10 of the M2 and M3 receptors reveals a probable under pium failed to enter the orthosteric site . This may be lying factor for the limited success in the development due to limitations in the timescale accessible to molecu- of mAChR-subtype-selective ligands. The residues lar dynamics simulations, as well as the slow binding lining the orthosteric binding site are conserved abso- kinetics of tiotropium10. Notably, it has been shown lutely in sequence and, moreover, are positioned almost experimentally that orthosteric ligands can bind to an (FIG. 3b) identically in space in both receptors . Although allosteric site on the M2 receptor, albeit with low affin- there are a few subtle differences between the struc- ity90. Interactions with the extracellular vestibule may

tures of the M2 and M3 receptors, attempts to exploit initiate desolvation of the ligand, which is required for these minor structural variations to obtain ligands with the occupation of the orthosteric pocket. If binding to increased mAChR-subtype selectivity have thus far been the allosteric site is indeed a metastable state for orthos- unsuccessful105. teric muscarinic ligands, deliberate exploitation of this

554 | JULY 2014 | VOLUME 13 www.nature.com/reviews/drugdisc

a b conformational coupling between orthosteric and Extracellular surface allosteric sites. Specifically, molecular dynamics simulations indicated that bulky modulators (for example, Extracellular TM3 alcuronium) are likely to stabilize an open extracellular vestibule Y1043.33 vestibule conformation reminiscent of that seen in the 11 D1033.32 crystal structure of the M2 receptor bound to QNB . By contrast, smaller modulators (for example, C7/3‑phth) may preferentially stabilize a closed conformation of the 13 7.43 extracellular vestibule , similar to that seen in molecular Y430 10 TM5 dynamics simulations of the unliganded M3 receptor . Y4267.39 Thus, in the inactive state the smaller modulators gen- N4046.52 erally exhibit negative cooperativity with antagonists Y4036.51 (that is, they destabilize antagonist binding), whereas the bulkier modulators typically enhance antagonist bind- TM6 TM7 ing. Consistent with this model, the addition of bulky substituents to C /3‑phth improved the binding of this Intracellular surface View from extracellular side 7 NAM to the M2 receptors occupied by radiolabelled 3 13 Tyrosine lid M2 receptor M3 receptor H-N-methylscopolamine . Taken together, these data support the concept that orthosteric antagonist bind- Figure 3 | Structure of the orthosteric mAChR ligand binding site. a | A cross-section ing can be enhanced or diminished by a modulator that through the M receptor structure reveals a large, solvent-accessibleNature Reviews cavity | Drug extending Discovery 2 preferentially stabilizes a particular conformation of the through the receptor. The QNB ligand (orange spheres) is bound within the receptor extracellular vestibule. transmembrane core (black, receptor protein; blue, receptor surface). b | The orthosteric ligand (antagonist)-binding sites for the M2 and M3 receptors are almost identical in both structure and sequence. Polar contacts (red dotted lines) between the receptor and Activation and positive allosteric modulation of the bound antagonist (M , QNB; M , tiotropium) are identical for the two receptors. Residues M2 receptor. The initial structures of the M2 and M3 2 3 receptors were obtained in complex with high-affinity are numbered according to the human M2 receptor sequence. mAChR, muscarinic acetylcholine receptor. antagonists (inverse agonists) and consequently repre- sent inactive receptor conformations. Obtaining crystals of active GPCRs has thus far proved to be extremely chal- lenging; this is probably due in large part to the confor- feature could potentially facilitate the development of mational heterogeneity induced by agonist binding108. To muscarinic ligands with increased mAChR-subtype date, these problems have been surmounted in the case of 109–111 112,113 selectivity and/or improved kinetic properties. rhodopsin , the β2-adrenergic receptor (β2-AR) 12 and, more recently, the M2 receptor . In the last two

Molecular basis for NAM function at the M2 receptor. cases, the first active-state structures have been obtained Recently, the molecular mechanism of allosteric modu- with the aid of conformationally selective antibody

lator binding to the inactive state of the M2 receptor fragments, which mimic G proteins and stabilize the was investigated by computational molecular dynam- active conformation of the receptors to which they bind. ics and site-directed mutagenesis studies13. Using the This approach should prove to be useful for obtaining

inactive crystal structure of the M2 receptor as a starting active-state structures of other GPCRs in the future.

model, the binding of a series of structurally diverse The most striking feature of the active-state M2 recep-

cationic small-molecule modulators was simulated using tor structure, like that of active β2-AR and rhodopsin, is the long-timescale molecular dynamics. These simulations outward rotation of transmembrane domain 6 to create suggested that modulator binding to the extracellular a G protein-binding cavity on the intracellular surface vestibule may be driven largely by interactions between of the receptor (FIG. 4a,b). The active-state structure of

cationic amines on the modulators and aromatic residues the M2 receptor shows substantial similarity to the active

on the receptor. The allosteric pocket appears to con- states of both β2-AR and rhodopsin, with a slightly closer tain two cation‑π centres: one including Trp4227.35 and resemblance to the latter structure. Tyr177ECL2 (extracellular loop 2), and the other bounded One of the most intriguing structural features of active 2.61 2.64 by Tyr80 and Tyr83 . The importance of these regions β2-AR and rhodopsin is that both receptors show far more in modulator binding was confirmed by site-directed subtle rearrangements in the ligand-binding pocket and mutagenesis experiments, which showed that affinity extracellular surface than on the intracellular side. This for the NAM heptane‑1,7‑bis(dimethyl‑3’-phthalimido observation suggests that these regions are relatively fixed

propyl)ammonium bromide (C7/3‑phth) could be conformationally, with few changes associated with ago- enhanced by the addition of more aromatic residues or nist binding or activation. By contrast, the active struc- 13 acidic residues to the extracellular vestibule . ture of the M2 receptor revealed large changes throughout Dror et al.13 also examined the interaction between the receptor, including an unexpected contraction of the the modulators and orthosteric antagonists to obtain extracellular vestibule, in addition to changes on the sur- insights into cooperativity. Two key mechanisms were face of the intracellular receptor that closely parallel those (FIG. 4c,d) proposed. The first was electrostatic repulsion between seen in β2-AR and rhodopsin . This extracellular the modulator and the antagonist. The second was rearrangement results in a narrower allosteric site and

NATURE REVIEWS | DRUG DISCOVERY VOLUME 13 | JULY 2014 | 555

a b TM6 Helix 8

90° TM5 TM5 TM7 TM1 TM3 TM1 TM6

TM2 ICL2 TM4 Helix 8

c d Extracellular surface TM3 D1033.32 Y1043.33

TM2

Y4267.39 TM5

TM7 6.52 N404 6.51 TM6 Y403

Inactive M2 receptor Active M2 receptor

Figure 4 | Activation and allosteric modulation of the M2 receptor. As shown in part a and part b, the intracellular tip | of transmembrane domain 6 (TM6) rotates outwards in the active M2 receptor structureNature (orange) Reviews relative Drug to the Discovery inactive

state (blue). As shown in part c, the orthosteric binding site contracts upon M2 receptor activation, enclosing the agonist iperoxo (yellow) in a smaller binding site, as compared to the antagonist (QNB; green) binding cavity. Residues

are numbered according to the human M2 receptor sequence. As shown in part d, LY2119620 (magenta), a muscarinic

positive allosteric modulator, binds to the extracellular vestibule of the M2 receptor directly above the orthosteric agonist iperoxo (yellow). ICL2, intracellular loop 2.

a smaller orthosteric site that is occluded entirely from In addition to the substantial contraction of the solvent (FIG. 4d). The much smaller size of the orthosteric orthosteric site upon activation, the extracellular vestibule site is reflected by the lower molecular weight of mus- is similarly much smaller in the active conformation of

carinic agonists compared to antagonists and inverse the M2 receptor, even in the absence of a bound allosteric agonists. modulator. Isomorphous crystals were grown in the pres- 12 In the active state of the M2 receptor, the small orthos ence of LY2119620, a PAM of mAChRs , and inspection teric agonist iperoxo engages Asp1033.32 with its cationic of electron density maps revealed binding of this ligand to head group, whereas the polar isoxazoline tail contacts the extracellular vestibule (FIG. 4d). Strikingly, structures of 6.52 Asn404 with a single hydrogen bond, paralleling the the active M2 receptor with and without LY2119620 were interactions of the inactive receptor with the antagonist highly similar, even in the extracellular vestibule12. Indeed, QNB (FIG. 4c). These two interactions are bridged by an the only substantial change in this pocket was a reorienta- acetylene moiety in iperoxo, which runs perpendicular tion of Trp4227.35, which directly interacts with the bound to transmembrane domain 3. Within the binding site, LY2119620. In addition, Phe181ECL2 was poorly resolved iperoxo exhibits a high degree of shape complementa- in the allosteric complex, which suggests that this residue rity to the active conformation, filling the pocket almost may become more mobile upon binding of the receptor entirely (FIG. 4d). By contrast, in the inactive conforma- to this modulator. As Phe181ECL2 is a leucine residue in all

tion the M2 receptor presents a large orthosteric site that is other mAChR subtypes, interaction with this residue may

better suited to the binding of larger antagonists and allow modulators to have different effects on M2 receptors inverse agonists. The exceptional complementarity of compared to other mAChR subtypes. iperoxo with the active — but not inactive — orthosteric The remarkable similarity of the extracellular vesti-

site of the receptor may account for its agonistic activity. bule conformation in the active M2 receptor with and

556 | JULY 2014 | VOLUME 13 www.nature.com/reviews/drugdisc

DRUG DISCOVERY to further favour active receptor conformations. LY2119620 to the active Figure 5 | allosteric site but of also thereceptor as awhole equilibrium towards active conformations not only of the In doing so, PAMs may receptor shiftthe conformational is, active) conformation of extracellular the vestibule. preferentially binding to and stabilizing (that closed the modulation.allosteric PAMs may by achieve effects their without bound LY2119620 suggests amechanism for equilibrium in favour of active receptor conformations. a ligand, the receptor adopts inactive conformations, which are relatively more stable. receptor with the orthosteric and allosteric binding sites highlighted in green and red, respectively. In the absence of Comparison of active- the and inactive-state structures binding and pocket Gprotein-binding the site extracellular the vestibule, orthosteric the structurally: sites that have pharmacologically defined been and ensemble of receptor the infavour of inactive states. inactive structures, thereby skewing conformationalthe conformation of extracellular the vestibule inthe seen Analogously, NAMs may preferentially bind open the a c b With agonistandPAM No ligands With agonist In total, mAChRs the contain at least three binding Inactive M

Hypothetical mechanismfor allosteric modulationoftheM 2 receptor © 2014 Macmillan Publishers Limited. Allrights reserved -state receptor enhances the affinity of the receptor for the agonist and shifts the equilibrium G protein bindingsite Orthosteric site Allosteric site Active M ( ( FIG. 5 FIG. 5 PAM Agonist c 2 receptor | Binding of a positive allosteric modulator (PAM) such as ) ) . .

separable conformational changes at each of three the state. Activation of mAChRs the might similarly involve the intracellular surface remains in a closed, inactive orthosteric site are indicative of receptor activation, but bound to agonists is suggested by structures of the A independentlyoccur of one another. Such apossibility but it remains changes whether seen to be ineach site can conformationalcerted change throughout three sites, all picture. tor conformation and thus provide, at an best, incomplete structures presentcrystal only static of views asingle recep activation. However, it is important to recognize that the substantial conformational rearrangements upon receptor of M the The structures solved to date are suggestive of acon 2 receptor shows that three sites all can undergo 2 receptor byaPAM.

114 Energy Energy Energy , 115 . Conformational changes inthe b Nature Reviews | Agonist binding shifts the

VOLUME 13 Reaction coordinate Reaction coordinate Reaction coordinate R R R

a | Scheme of the M 2A REVIEWS adenosine receptor

| |

JULY 2014 Drug Discovery R* R* R* 2

557 - -

REVIEWS

sites rather than a single concerted activation through- active conformations, as well as in complex with different out the entire receptor. Spectroscopic and other bio- classes of ligands (orthosteric, allosteric or bitopic physical assays will probably be required to definitively ligands). Such studies may reveal a broad spectrum establish the mechanism of receptor activation and the of mAChR conformations depending on the receptor means through which ligands stabilize active receptor subtype and activation state, as well as receptor-bound conformations. ligands. Identifying the structural basis that underlies the ability of mAChRs, as well as other GPCRs, to inter- Conclusions and future challenges act with G proteins, arrestins, GPCR kinases (GRKs) The past few years have witnessed unprecedented pro- and other GPCR-associated signalling molecules should gress in our understanding of the biology, pharmacology also be of great general interest. Taken together, the and structure of mAChRs. The use of novel mutant mouse outcome of these studies should considerably facilitate models, including mAChR knock-in mice with altered the development of novel muscarinic receptor agents signalling properties and mutant mice in which distinct and tools that show a high degree of mAChR sub- mAChR subtypes can be deleted in a temporally and type selectivity and hopefully cause fewer side effects. spatially controlled fashion, should provide more detailed Finally, lessons learned from work in the mAChR insights into mAChR biology. field may be applicable to GPCRs in general, offer- From a structural point of view, the most obvious ing new insights into how to design better therapeu- future goal is to obtain high-resolution structures for tic agents targeting this important class of cell surface all five mAChR subtypes, both in their inactive and receptors.

1. Fredriksson, R., Lagerstrom, M. C., Lundin, L. G. & This study represents a computational biology 28. Dencker, D. et al. Muscarinic acetylcholine receptor Schioth, H. B. The G‑protein-coupled receptors in the breakthrough in delineating the molecular subtypes as potential drug targets for the treatment of human genome form five main families. Phylogenetic mechanisms governing the allosteric modulation schizophrenia, drug abuse and Parkinson’s disease.

analysis, paralogon groups, and fingerprints. Mol. of the M2 receptor. ACS Chem. Neurosci. 3, 80–89 (2012). Pharmacol. 63, 1256–1272 (2003). 14. Ballard, C. et al. Alzheimer’s disease. Lancet 377, 29. Sofuoglu, M. & Mooney, M. Cholinergic functioning in 2. Hulme, E. C., Birdsall, N. J. & Buckley, N. J. Muscarinic 1019–1031 (2011). stimulant addiction: implications for medications receptor subtypes. Annu. Rev. Pharmacol. Toxicol. 30, 15. Langmead, C. J., Watson, J. & Reavill, C. Muscarinic development. CNS Drugs 23, 939–952 (2009). 633–673 (1990). acetylcholine receptors as CNS drug targets. 30. Thomsen, M. et al. Attenuation of cocaine’s reinforcing

3. Wess, J., Eglen, R. M. & Gautam, D. Muscarinic Pharmacol. Ther. 117, 232–243 (2008). and discriminative stimulus effects via muscarinic M1 acetylcholine receptors: mutant mice provide new 16. Davis, A. A., Fritz, J. J., Wess, J., Lah, J. J. & acetylcholine receptor stimulation. J. Pharmacol. Exp.

insights for drug development. Nature Rev. Drug Levey, A. I. Deletion of M1 muscarinic acetylcholine Ther. 332, 959–969 (2010).

Discov. 6, 721–733 (2007). receptors increases amyloid pathology in vitro and 31. Thomsen, M. et al. Contribution of both M1 and M4 4. Wess, J. Novel muscarinic receptor mutant mouse in vivo. J. Neurosci. 30, 4190–4196 (2010). receptors to muscarinic agonist-mediated attenuation

models. Handb. Exp. Pharmacol. 208, 95–117 17. Medeiros, R. et al. Loss of muscarinic M1 receptor of the cocaine discriminative stimulus in mice. (2012). exacerbates Alzheimer’s disease-like pathology and Psychopharmacol. 220, 673–685 (2012). 5. Conn, P. J., Jones, C. K. & Lindsley, C. W. Subtype- cognitive decline. Am. J. Pathol. 179, 980–991 32. Schmidt, L. S. et al. Increased cocaine self-

selective allosteric modulators of muscarinic receptors (2011). administration in M4 muscarinic acetylcholine receptor for the treatment of CNS disorders. Trends Pharmacol. 18. Melancon, B. J., Tarr, J. C., Panarese, J. D., knockout mice. Psychopharmacol. 216, 367–378 Sci. 30, 148–155 (2009). Wood, M. R. & Lindsley, C. W. Allosteric modulation of (2011). 6. Wootten, D., Christopoulos, A. & Sexton, P. M. the M1 muscarinic acetylcholine receptor: improving 33. Lam, D. W. & LeRoith, D. The worldwide diabetes Emerging paradigms in GPCR allostery: implications cognition and a potential treatment for schizophrenia epidemic. Curr. Opin. Endocrinol. Diabetes Obes. 19, for drug discovery. Nature Rev. Drug Discov. 12, and Alzheimer’s disease. Drug Discov. Today 18, 93–96 (2012).

630–644 (2013). 1185–1199 (2013). 34. Gautam, D. et al. A critical role for β cell M3 muscarinic 7. Lane, J. R., Sexton, P. M. & Christopoulos, A. 19. Davie, B. J., Christopoulos, A. & Scammells, P. J. acetylcholine receptors in regulating insulin release

Bridging the gap: bitopic ligands of G‑protein- Development of M1 mAChR allosteric and bitopic and blood glucose homeostasis in vivo. Cell. Metab. 3, coupled receptors. Trends Pharmacol. Sci. 34, ligands: prospective therapeutics for the treatment of 449–461 (2006). 59–66 (2013). cognitive deficits. ACS Chem. Neurosci. 4, 1026–1048 35. Gautam, D. et al. Beneficial metabolic effects caused

8. Bock, A. & Mohr, K. Dualsteric GPCR targeting and (2013). by persistent activation of β-cell M3 muscarinic

functional selectivity: the paradigmatic M2 muscarinic 20. van Os, J. & Kapur, S. Schizophrenia. Lancet 374, acetylcholine receptors in transgenic mice. acetylcholine receptor. Drug Discov. Today Technol. 635–645 (2009). Endocrinology 151, 5185–5194 (2010). 10, e245–e252 (2013). 21. McKinzie, D. L. & Bymaster, F. P. Muscarinic 36. Guettier, J. M. et al. A chemical-genetic approach to 9. De Amici, M., Dallanoce, C., Holzgrabe, U., Tränkle, C. mechanisms in psychotic disorders. Handb. Exp. study G protein regulation of β cell function in vivo. & Mohr, K. Allosteric ligands for G protein-coupled Pharmacol. 213, 233–265 (2012). Proc. Natl Acad. Sci. USA 106, 19197–19202 receptors: a novel strategy with attractive therapeutic 22. Bodick, N. C. et al. Effects of xanomeline, a selective (2009). opportunities. Med. Res. Rev. 30, 463–549 (2010). muscarinic receptor agonist, on cognitive function 37. Jain, S. et al. Chronic activation of a designer

10. Kruse, A. C. et al. Structure and dynamics of the M3 and behavioral symptoms in Alzheimer disease. Gq-coupled receptor improves β cell function. muscarinic acetylcholine receptor. Nature 482, Arch. Neurol. 54, 465–473 (1997). J. Clin. Invest. 123, 1750–1762 (2013). 552–556 (2012). 23. Shekhar, A. et al. Selective muscarinic receptor This study shows that chronic, exogenous

This study reports the first high-resolution agonist xanomeline as a novel treatment approach for ligand-induced activation of an M3

structure of the M3 receptor in complex with schizophrenia. Am. J. Psychiatry 165, 1033–1039 receptor-derived designer receptor expressed tiotropium, a clinically used muscarinic antagonist (2008). by pancreatic β‑cells prevents diabetes in

and inverse agonist. 24. Jeon, J. et al. A subpopulation of neuronal M4 different mouse models. 11. Haga, K. et al. Structure of the human M2 muscarinic muscarinic acetylcholine receptors plays a critical 38. Armbruster, B. N., Li, X., Pausch, M. H., Herlitze, S. acetylcholine receptor bound to an antagonist. Nature role in modulating dopamine-dependent behaviors. & Roth, B. L. Evolving the lock to fit the key to create 482, 547–551 (2012). J. Neurosci. 30, 2396–2405 (2010). a family of G protein-coupled receptors potently In this study, the authors present the first 25. Dencker, D. et al. Involvement of a subpopulation of activated by an inert ligand. Proc. Natl Acad. Sci. USA

high-resolution structure of the M2 receptor in neuronal M4 muscarinic acetylcholine receptors in the 104, 5163–5168 (2007).

complex with an orthosteric muscarinic antagonist antipsychotic-like effects of the M1/M4 preferring 39. Sassmann, A. et al. The Gq/G11‑mediated signaling and inverse agonist — QNB. muscarinic receptor agonist xanomeline. J. Neurosci. pathway is critical for autocrine potentiation of insulin 12. Kruse, A. C. et al. Activation and allosteric modulation 31, 5905–5908 (2011). secretion in mice. J. Clin. Invest. 120, 2184–2193 of a muscarinic acetylcholine receptor. Nature 504, 26. Foster, D. J., Jones, C. K. & Conn, P. J. Emerging (2010).

101–106 (2013). approaches for treatment of schizophrenia: 40. Kong, K. C. et al. M3‑muscarinic receptor promotes This study provides the first high-resolution modulation of cholinergic signaling. Discov. Med. insulin release via receptor phosphorylation/arrestin- structural information of an agonist-activated 14, 413–420 (2012). dependent activation of protein kinase D1. Proc. Natl

mAChR (the M2 subtype) and reveals how a PAM 27. Jones, C. K., Byun, N. & Bubser, M. Muscarinic Acad. Sci. USA 107, 21181–21186 (2010).

interacts with the M2 receptor. and nicotinic acetylcholine receptor agonists and This analysis of phosphorylation-deficient M3 13. Dror, R. O. et al. Structural basis for modulation of a allosteric modulators for the treatment of receptor knock‑in mice strongly suggests that G‑protein-coupled receptor by allosteric drugs. Nature schizophrenia. Neuropsychopharmacology 37, arrestin-dependent signalling pathways contribute

503, 295–299 (2013). 16–42 (2012). to M3 receptor-stimulated insulin release.

558 | JULY 2014 | VOLUME 13 www.nature.com/reviews/drugdisc

41. Nakajima, K. & Wess, J. Design and functional 62. Stallaert, W., Christopoulos, A. & Bouvier, M. Ligand 83. Changeux, J. P. Allosteric receptors: from electric characterization of a novel, arrestin-biased designer functional selectivity and quantitative pharmacology organ to cognition. Annu. Rev. Pharmacol. Toxicol. 50, G protein-coupled receptor. Mol. Pharmacol. 82, at G protein-coupled receptors. Expert Opin. Drug 1–38 (2010). 575–582 (2012). Discov. 6, 811–825 (2011). 84. Canals, M., Sexton, P. M. & Christopoulos, A. Allostery 42. Shah, N., Khurana, S., Cheng, K. & Raufman, J. P. 63. Marlo, J. E. et al. Discovery and characterization of in GPCRs: ‘MWC’ revisited. Trends Biochem. Sci. 36,

Muscarinic receptors and ligands in cancer. novel allosteric potentiators of M1 muscarinic 663–672 (2011). Am. J. Physiol. Cell Physiol. 296, C221–C232 receptors reveals multiple modes of activity. Mol. 85. Ehlert, F. J. & Griffin, M. T. Two-state models and the

(2009). Pharmacol. 75, 577–588 (2009). analysis of the allosteric effect of gallamine at the M2 43. Spindel, E. R. Muscarinic receptor agonists and 64. Lazareno, S., Dolezal, V., Popham, A. & Birdsall, N. J. muscarinic receptor. J. Pharmacol. Exp. Ther. 325, antagonists: effects on cancer. Handb. Exp. Thiochrome enhances acetylcholine affinity at 1039–1060 (2008).

Pharmacol. 208, 451–468 (2012). muscarinic M4 receptors: receptor subtype selectivity 86. Abdul-Ridha, A., Lane, J. R., Sexton, P. M., Canals, M. 44. Raufman, J. P. et al. Genetic ablation of M3 via cooperativity rather than affinity. Mol. Pharmacol. & Christopoulos, A. Allosteric modulation of a muscarinic receptors attenuates murine colon 65, 257–266 (2004). chemogenetically modified G protein-coupled epithelial cell proliferation and neoplasia. Cancer Res. 65. Chan, W. Y. et al. Allosteric modulation of the receptor. Mol. Pharmacol. 83, 521–530 (2013).

68, 3573–3578 (2008). muscarinic M4 receptor as an approach to treating 87. Lazareno, S., Popham, A. & Birdsall, N. J. Allosteric 45. Raufman, J. P. et al. Muscarinic receptor subtype‑3 schizophrenia. Proc. Natl Acad. Sci. USA 105, interactions of staurosporine and other gene ablation and scopolamine butylbromide 10978–10983 (2008). indolocarbazoles with N-[methyl-3H]scopolamine and treatment attenuate small intestinal neoplasia in 66. Suratman, S. et al. Impact of species variability and acetylcholine at muscarinic receptor subtypes: Apcmin/+ mice. Carcinogenesis 32, 1396–1402 (2011). ‘probe-dependence’ on the detection and in vivo identification of a second allosteric site. Mol.

46. Magnon, C. et al. Autonomic nerve development validation of allosteric modulation at the M4 Pharmacol. 58, 194–207 (2000). contributes to prostate cancer progression. Science muscarinic acetylcholine receptor. Br. J. Pharmacol. 88. Lazareno, S., Popham, A. & Birdsall, N. J. 341, 1236361 (2013). 162, 1659–1670 (2011). Analogs of WIN 62,577 define a second allosteric

This study reports that M1 receptor deficiency 67. Valant, C., Sexton, P. M. & Christopoulos, A. site on muscarinic receptors. Mol. Pharmacol. 62, inhibits mAChR-mediated prostate cancer invasion Orthosteric/allosteric bitopic ligands: going hybrid at 1492–1505 (2002). and metastasis in two mouse models of prostate GPCRs. Mol. Interv. 9, 125–135 (2009). 89. Espinoza-Fonseca, L. M. & Trujillo-Ferrara, J. G.

cancer. 68. Mohr, K. et al. Rational design of dualsteric GPCR The existence of a second allosteric site on the M1 47. Christopoulos, A. Allosteric binding sites on cell- ligands: quests and promise. Br. J. Pharmacol. 159, muscarinic acetylcholine receptor and its implications surface receptors: novel targets for drug discovery. 997–1008 (2010). for drug design. Bioorg. Med. Chem. Lett. 16, Nature Rev. Drug Discov. 1, 198–210 (2002). 69. Melancon, B. J. et al. Allosteric modulation of seven 1217–1220 (2006). 48. May, L. T., Leach, K., Sexton, P. M. & Christopoulos, A. transmembrane spanning receptors: theory, practice, 90. Redka, D. S., Pisterzi, L. F. & Wells, J. W. Binding of

Allosteric modulation of G protein-coupled receptors. and opportunities for central nervous system drug orthosteric ligands to the allosteric site of the M2 Annu. Rev. Pharmacol. Toxicol. 47, 1–51 (2007). discovery. J. Med. Chem. 55, 1445–1464 (2012). muscarinic cholinergic receptor. Mol. Pharmacol. 74, 49. Christopoulos, A., Lanzafame, A. & Mitchelson, F. 70. Valant, C., Robert Lane, J., Sexton, P. M. & 834–843 (2008). Allosteric interactions at muscarinic cholinoceptors. Christopoulos, A. The best of both worlds? Bitopic 91. Shivnaraine, R. V., Huang, X. P., Seidenberg, M., Ellis, J. Clin. Exp. Pharmacol. Physiol. 25, 185–194 (1998). orthosteric/allosteric ligands of G protein-coupled & Wells, J. W. Heterotropic cooperativity within and

50. Birdsall, N. J. & Lazareno, S. Allosterism at muscarinic receptors. Annu. Rev. Pharmacol. Toxicol. 52, between protomers of an oligomeric M2 muscarinic receptors: ligands and mechanisms. Mini Rev. Med. 153–178 (2012). receptor. Biochemistry 51, 4518–4540 (2012). Chem. 5, 523–543 (2005). 71. Disingrini, T. et al. Design, synthesis, and action of 92. Rosenbaum, D. M. et al. GPCR engineering yields

51. Conn, P. J., Christopoulos, A. & Lindsley, C. W. oxotremorine-related hybrid-type allosteric high-resolution structural insights into β2‑adrenergic Allosteric modulators of GPCRs: a novel approach for modulators of muscarinic acetylcholine receptors. receptor function. Science 318, 1266–1273 (2007). the treatment of CNS disorders. Nature Rev. Drug J. Med. Chem. 49, 366–372 (2006). 93. Chae, P. S. et al. Maltose-neopentyl glycol (MNG) Discov. 8, 41–54 (2009). 72. Antony, J. et al. Dualsteric GPCR targeting: a novel amphiphiles for solubilization, stabilization and 52. Keov, P., Sexton, P. M. & Christopoulos, A. Allosteric route to binding and signaling pathway selectivity. crystallization of membrane proteins. Nature Methods modulation of G protein-coupled receptors: a FASEB J. 23, 442–450 (2009). 7, 1003–1008 (2010). pharmacological perspective. Neuropharmacology 73. Kebig, A., Kostenis, E., Mohr, K. & Mohr-Andra, M. 94. Landau, E. M. & Rosenbusch, J. P. Lipidic cubic 60, 24–35 (2011). An optical dynamic mass redistribution assay reveals phases: a novel concept for the crystallization of 53. Keov, P. et al. Reverse engineering of the selective biased signaling of dualsteric GPCR activators. membrane proteins. Proc. Natl Acad. Sci. USA 93, agonist TBPB unveils both orthosteric and allosteric J. Recept. Signal Transduct. Res. 29, 140–145 14532–14535 (1996).

modes of action at the M1 muscarinic acetylcholine (2009). 95. Caffrey, M. & Cherezov, V. Crystallizing membrane receptor. Mol. Pharmacol. 84, 425–437 (2013). 74. Bock, A. et al. The allosteric vestibule of a seven proteins using lipidic mesophases. Nature Protoc. 4, 54. Kenakin, T. & Christopoulos, A. Signalling bias in new transmembrane helical receptor controls G‑protein 706–731 (2009). drug discovery: detection, quantification and coupling. Nature Commun. 3, 1044 (2012). 96. Smith, J. L., Fischetti, R. F. & Yamamoto, M. Micro- therapeutic impact. Nature Rev. Drug Discov. 12, 75. Steinfeld, T., Mammen, M., Smith, J. A., Wilson, R. D. crystallography comes of age. Curr. Opin. Struct. Biol. 205–216 (2013). & Jasper, J. R. A novel multivalent ligand that bridges 22, 602–612 (2012).

55. Ma, L. et al. Selective activation of the M1 muscarinic the allosteric and orthosteric binding sites of the M2 97. Shimamura, T. et al. Structure of the human histamine

acetylcholine receptor achieved by allosteric muscarinic receptor. Mol. Pharmacol. 72, 291–302 H1 receptor complex with doxepin. Nature 475, potentiation. Proc. Natl Acad. Sci. USA 106, (2007). 65–70 (2011). 15950–15955 (2009). 76. Valant, C. et al. A novel mechanism of G protein- 98. Chien, E. Y. et al. Structure of the human dopamine 56. Shirey, J. K. et al. A selective allosteric potentiator of coupled receptor functional selectivity. Muscarinic D3 receptor in complex with a D2/D3 selective

the M1 muscarinic acetylcholine receptor increases partial agonist McN‑A-343 as a bitopic orthosteric/ antagonist. Science 330, 1091–1095 (2010). activity of medial prefrontal cortical neurons and allosteric ligand. J. Biol. Chem. 283, 29312–29321 99. Warne, T. et al. Structure of a β1‑adrenergic G‑protein- restores impairments in reversal learning. J. Neurosci. (2008). coupled receptor. Nature 454, 486–491 (2008). 29, 14271–14286 (2009). This is the first study to show that functionally 100. Rasmussen, S. G. et al. Crystal structure of the human

57. Canals, M. et al. A Monod–Wyman–Changeux selective ligands may mediate their behaviour β2 adrenergic G‑protein-coupled receptor. Nature 450, mechanism can explain G protein-coupled receptor via a bitopic mechanism. 383–387 (2007). (GPCR) allosteric modulation. J. Biol. Chem. 287, 77. Spalding, T. A. et al. Discovery of an ectopic activation 101. Cherezov, V. et al. High-resolution crystal structure of

650–659 (2012). site on the M1 muscarinic receptor. Mol. Pharmacol. an engineered human β2‑adrenergic G protein-coupled This study presents a chemical biology framework 61, 1297–1302 (2002). receptor. Science 318, 1258–1265 (2007). with which to study and classify the simplest 78. Langmead, C. J. et al. Probing the molecular 102. Wacker, D. et al. Structural features for functional allosteric ligand behaviours. mechanism of interaction between 4‑n‑butyl‑1-[4- selectivity at serotonin receptors. Science 340, 58. Lazareno, S. & Birdsall, N. J. Detection, quantitation, (2‑methylphenyl)-4‑oxo‑1‑butyl]-piperidine (AC‑42) 615–619 (2013).

and verification of allosteric interactions of agents and the muscarinic M1 receptor: direct 103. Wang, C. et al. Structural basis for molecular with labeled and unlabeled ligands at G protein- pharmacological evidence that AC‑42 is an allosteric recognition at serotonin receptors. Science 340, coupled receptors: interactions of strychnine agonist. Mol. Pharmacol. 69, 236–246 (2006). 610–614 (2013). and acetylcholine at muscarinic receptors. 79. Jones, C. K. et al. Novel selective allosteric activator of 104. Ballesteros, J. & Weinstein, H. Integrated methods for

Mol. Pharmacol. 48, 362–378 (1995). the M1 muscarinic acetylcholine receptor regulates the construction of three-dimensional models and 59. Kenakin, T. New concepts in drug discovery: amyloid processing and produces antipsychotic-like computational probing of structure-function relations collateral efficacy and permissive antagonism. activity in rats. J. Neurosci. 28, 10422–10433 in G protein-coupled receptors. Methods Neurosci. Nature Rev. Drug Discov. 4, 919–927 (2005). (2008). 25, 366–428 (1995). 60. Valant, C., Felder, C. C., Sexton, P. M. & 80. Gregory, K. J., Hall, N. E., Tobin, A. B., Sexton, P. M. & 105. Kruse, A. C. et al. Muscarinic receptors as model Christopoulos, A. Probe dependence in the allosteric Christopoulos, A. Identification of orthosteric and targets and antitargets for structure-based ligand

modulation of a G protein-coupled receptor: allosteric site mutations in M2 muscarinic acetylcholine discovery. Mol. Pharmacol. 84, 528–540 (2013). implications for detection and validation of allosteric receptors that contribute to ligand-selective signaling 106. Tautermann, C. S. et al. Molecular basis for the long ligand effects. Mol. Pharmacol. 81, 41–52 (2012). bias. J. Biol. Chem. 285, 7459–7474 (2010). duration of action and kinetic selectivity of tiotropium This study highlights the importance of probe 81. Avlani, V. A. et al. Orthosteric and allosteric modes of for the muscarinic M3 receptor. J. Med. Chem. 56,

dependence in the study of the effects of allosteric interaction of novel selective agonists of the M1 8746–8756 (2013). modulators. muscarinic acetylcholine receptor. Mol. Pharmacol. 107. Gregory, K. J., Sexton, P. M. & Christopoulos, A. 61. Rajagopal, S., Rajagopal, K. & Lefkowitz, R. J. 78, 94–104 (2010). Allosteric modulation of muscarinic acetylcholine Teaching old receptors new tricks: biasing seven- 82. Monod, J., Wyman, J. & Changeux, J. P. On the receptors. Curr. Neuropharmacol. 5, 157–167 (2007).

transmembrane receptors. Nature Rev. Drug Discov. nature of allosteric transitions: a plausible model. 108. Nygaard, R. et al. The dynamic process of β2-adrenergic 9, 373–386 (2010). J. Mol. Biol. 12, 88–118 (1965). receptor activation. Cell 152, 532–542 (2013).

NATURE REVIEWS | DRUG DISCOVERY VOLUME 13 | JULY 2014 | 559

109. Scheerer, P. et al. Crystal structure of opsin in its 116. Kamsler, A., McHugh, J., Gerber, D., Huang, S. Y. & 121. Poulin, B. et al. The M3-muscarinic receptor regulates G‑protein-interacting conformation. Nature 455, Tonegawa, S. Presynaptic M1 muscarinic receptors learning and memory in a receptor phosphorylation/ 497–502 (2008). are necessary for mGluR long-term depression in arrestin-dependent manner. Proc. Natl Acad. Sci. USA 110. Choe, H. W. et al. Crystal structure of metarhodopsin II. the hippocampus. Proc. Natl Acad. Sci. USA 107, 107, 9440–9445 (2010). Nature 471, 651–655 (2011). 1618–1623 (2010). 122. Bendor, J. et al. AGAP1/AP‑3‑dependent endocytic

111. Standfuss, J. et al. The structural basis of agonist- 117. Gautam, D. et al. Neuronal M3 muscarinic recycling of M5 muscarinic receptors promotes induced activation in constitutively active rhodopsin. acetylcholine receptors are essential for dopamine release. EMBO J. 29, 2813–2826 (2010). Nature 471, 656–660 (2011). somatotroph proliferation and normal somatic 112. Rasmussen, S. G. et al. Structure of a nanobody- growth. Proc. Natl Acad. Sci. USA 106, 6398–6403 Acknowledgements We apologize to all investigators whose important contribu‑ stabilized active state of the β2 adrenoceptor. (2009). Nature 469, 175–180 (2011). 118. Shi, Y. et al. Signaling through the M muscarinic tions could not be acknowledged owing to space limitations. 3 The work of A.C.K. and B.K.K. was supported by a US National 113. Rasmussen, S. G. et al. Crystal structure of the β2 receptor favors bone mass accrual by decreasing adrenergic receptor‑Gs protein complex. Nature sympathetic activity. Cell. Metab. 11, 231–238 Science Foundation Graduate Research Fellowship (A.C.K.) and 477, 549–555 (2011). (2010). by the National Science Foundation grant CHE-1223785 and US National Institutes of Health (NIH) grant U19GM106990 This crystal structure represents the first high- 119. Li, J. H. et al. Hepatic muscarinic acetylcholine (B.K.K.). A.C. and P.M.S. received funds from Program Grant resolution view of the active-state ternary receptors are not critically involved in maintaining No. APP1055134 of the National Health and Medical complex composed of an agonist-occupied GPCR glucose homeostasis in mice. Diabetes 58, Research Council (NHMRC) of Australia. A.C. and P.M.S. are ( -AR) and a G protein (nucleotide-free G 2776–2787 (2009). β2 s NHMRC Principal Research Fellows. The research of D.G. and et al. heterotrimer). 120. Arteaga-Solis, E. Inhibition of leptin regulation of J.W. was supported by the Intramural Research Program of the 114. Lebon, G. et al. Agonist-bound adenosine A parasympathetic signaling as a cause of extreme body 2A National Institute of Diabetes and Digestive and Kidney receptor structures reveal common features weight-associated asthma. Cell. Metab. 17, 35–48 Diseases (NIDDK) at the NIH. We thank all our co-workers and of GPCR activation. Nature 474, 521–525 (2013). collaborators for their invaluable contributions to the work (2011). This study reports that leptin signalling in the summarized in this Review. 115. Xu, F. et al. Structure of an agonist-bound human A2A brain promotes bronchodilation by inhibiting adenosine receptor. Science 332, 322–327 parasympathetic signalling through airway smooth Competing interests statement

(2011). muscle M3 receptors. The authors declare no competing interests.

560 | JULY 2014 | VOLUME 13 www.nature.com/reviews/drugdisc