Molecular properties of muscarinic acetylcholine receptors.
1.1. Distributions of cholinergic neurons and muscarinic receptors. The concept of muscarinic receptors originated from a report by Dale in 1914, (1) although he did not use the term receptor. He reported that the actions of several kinds of choline esters or other derivatives in various tissues could be divided into muscarinic and nicotinic ones, which are mimicked typically by muscarine and nicotine, respectively. Muscarine is a natural product of certain mushrooms including Amanita muscaria, and mimics the actions of parasympathetic neurons, which include vasodilation of vessels, slowing of the heart rate, constriction of bronchioles, elicitation of saliva secretion, constriction of the eye pupils, and so on. The action of muscarine is antagonized by atropine, which is a product of the plant called Atropa bella-donna (beautiful lady): the plant shrub has been used for visual appeal, as atropine causes dilation of the pupils (mydriasis).
The endogenous ligand for muscarinic receptors as well as for nicotinic receptors was confirmed to be acetylcholine (ACh) through a series of experiments by Loewi, Dale and others in 1921-1934 (see review ref. (2). ACh is now known to be a neurotransmitter in both the peripheral (PNS) and central (CNS) nervous systems: neurons containing ACh as a neurotransmitter are called cholinergic neurons. Cholinergic neurons in the PNS comprise motor neurons, postganglionic parasympathetic neurons, and preganglionic sympathetic and parasympathetic neurons. Muscarinic receptors are receptors for ACh in postganglionic parasympathetic neurons and are present in the target organs of parasympathetic neurons, which cover almost all organs, whereas nicotinic receptors are receptors for ACh in motor neurons and preganglionic neurons, and are present in skeletal muscles and in post-ganglionic neurons. Muscarinic receptors are also present in preganglionic neurons, where they are involved in regulation of the actions of nicotinic receptors, in sweat glands which are regulated by sympathetic but cholinergic neurons, and in some lymphoma cells.
Cholinergic neurons in the CNS reside in the forebrain including the medial septal nuclei, diagonal band nuclei, Meynert nuclei, and so on: these neurons project to various areas of the cerebrum. (3) Cholinergic neurons are also present in the caudate nucleus as interneurons, which are targeted to GABA-containing neurons competing with dopaminergic neurons. Muscarinic receptors are present in various areas of the brain, which are innervated by cholinergic neurons, both on neurons and glias, and both on postsynaptic and presynaptic membranes. Muscarinic receptors in the CNS are thought to be involved in learning-memory, sleep-waking, attention focusing, motor control, and so on.
1.2. Muscarinic receptors as binding sites for [[sup.3]H]QNB or [[sup.3]H]NMS. The term "receptor", including the muscarinic receptor, used to be a functional concept, leaving the molecular entity unknown. Muscarinic receptors should have at least two functions (1) recognition of ACh and (2) initiation of a physiological response. Attempts to define and characterize muscarinic receptors as molecular entities were facilitated through the introduction of radioactive, high-affinity ligands (antagonists) such as tritium-labeled quinuclidinyl benzylate ([[sup.3]H]QNB) by Yamamura and Snyder (4) (1974), and N-methylscopolamine ([[sup.3]H]NMS) by Birdsall and Hulme (5) (1976). The binding components of [[sup.3]H]QNB or [[sup.3]H]NMS retain the first function of muscarinic receptors at least, and used to be regarded as muscarinic receptors themselves, leaving the second function unknown. [[sup.3]H]QNB or [[sup.3]H]NMS binds to muscarinic receptors in membrane preparations with high affinity, the equilibrium dissociation constants being estimated to be 15-80 pM and 50-700 pM, respectively. (6) These binding activities can be easily measured in membrane preparations from brain and other tissues. The affinites of various compounds for muscarinic receptors can be determined by measuring their ability to compete with [[sup.3]H]QNB or [[sup.3]H]NMS.
The density of muscarinic receptors in the CNS of rhesus monkey, as assessed as [[sup.3]H]QNB binding activity, was reported to range from 30 to 1200 pmol/g protein, and to be high in the caudate nucleus, putamen, and cerebral cortex, in accord with expectation during previous physiological experiments. (7) The distribution of muscarinic receptors was also found to be roughly in parallel with that of high affinity choline uptake activity, which was assumed to be a prerequisite step for ACh synthesis and to be present in cholinergic nerve terminals specifically. (8,9) These results support the idea that a major portion of acetylcholine receptors in the brain are muscarinic. (7)
These binding experiments indicated that muscarinic receptors in membrane preparations derived from different tissues exhibit different affinities for specific antagonists depending on the origins of the tissues, although they show similar affinities for atropine, [[sup.3]H]QNB and [[sup.3]H]NMS. Two kinds of muscarinic receptors, termed the [M.sub.1] and [M.sub.2] subtypes, were suggested to be distinguishable based on their different affinities for an antagonist, pirenzepine, and to be dominant in cerebral and atrial tissues, respectively. (10) In addition, a third subtype, [M.sub.3], was suggested to be present in smooth muscles, and characterized by low affinity for pirenzepine and high affinity for 4-DAMP (4-diphenylacetoxy-N-methylpiperidine methiodide), for which the [M.sub.2] subtype exhibits low affinity. (11)
These receptors may be assumed to represent three distinct entities as the simplest interpretation, but the possibility remained that they might represent three different states of a single entity. The situation was further complicated by the finding that an agonist shows three different affinities for muscarinic receptors in cerebral membranes, although [[sup.3]H]NMS and atropine each shows a single affinity for them. (12) This result may also be interpreted by means of two different assumptions: (1) a single kind of receptor in three states with the same affinity for an antagonist but different affinities for an agonist, and (2) multiple receptors with the same affinity for an antagonist but different affinities for an agonist.
1.3. Solubilization of muscarinic receptors.
Muscarinic receptors, as radioligand binding components, were demonstrated to be solubilizable from membrane preparations of various tissues. Beld and Ariens (13) (1974) succeeded in solubilizing a stereo-specific binding activity with [[sup.3]H](+)-benzetimide (another radiolabeled ligand for muscarinic receptors) from bovine smooth muscle and caudate nucleus using digitonin, which had been used for the extraction of visual pigments from suspensions of rod cells. This is the first demonstration that muscarinic receptors may be macromolecules extractable from cell membranes. It is interesting to note that the ligand binding activity is lost on treatment with most common detergents like Triton X-100 or sodium cholate, but not digitonin, and that [beta] adrenergic receptors as well as rhodopsin and muscarinic receptors were also found to be solubilized in their active forms only with digitonin (14) when these three proteins were not thought to belong to the same family.
A problem of using digitonin as a solubilizing agent is that it exists as large micelles, which makes it difficult to characterize solubilized muscarinic receptors in terms of molecular mass: the ligand binding activity in a digitonin extract was found in the fraction corresponding to a mass of around a million. (15) On the other hand, the [[sup.3]H]QNB-receptor complex, which was solubilized with Lubrol PX from cerebral membranes pretreated with [[sup.3]H]QNB, was estimated to have a mass of approximately 80 kDa on gel permeation chromatography and sucrose density gradient centrifugation in [H.sub.2]O and [D.sub.2]O (Haga, T., (16) 1980). The mass of 80 kDa was close to the value estimated on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of muscarinic receptors covalently labeled with muscarinic ligand [[sup.3]H]PrBCM (propylbenzilylcholine mustard), (17) suggesting that muscarinic receptors exist as monomers in contrast with nicotinic receptors, which had been indicated to be composed of five subunits.
The bands labeled with either [[sup.3]H]QNB or [[sup.3]H]PrBCM were much broader than that expected for a single protein. This suggests one or all of the following possibilities, that (1) muscarinic receptors may be bound covalently with a sugar moiety, (2) muscarinic receptors may consist of multiple subtypes of different sizes, (3) muscarinic receptors may exist in a equilibrium of monomers and dimers, and/or (4) muscarinic receptors may bind with other components in extracts with Lubrol PX or SDS. These possibilities remained to be clarified by further works.
2. Molecular entities of muscarinic receptors
2.1. Development of an affinity chromatog raphy system. In 1981 we started to attempt to purify muscarinic receptors at Burgen's laboratory in the National Institute for Medical Research in London. In order to purify muscarinic receptors it appeared essential to develop an affinity chromatography system with high specificity and high recovery, because the expression level of muscarinic receptors was known to be very low, an order of pmol/mg protein at most, which requires approximately [10.sup.4]-fold purification. Haga, K., who had been engaged in organic synthesis in a laboratory of Ajinomoto Co., started the project with the help of Peck by synthesizing a muscarinic ligand with an amino group: the amino group was expected to bind the ligand with an agarose gel through a linker. The first amino compound (APCPP; No. 11 in Table 1) synthesized there exhibited high affinity, with a dissociation constant of nM order for solubilized muscarinic receptors, but was not useful when bound to an agarose gel for unknown reasons. Then, systematic studies for preparation of an affinity gel were restarted in Ichiyama's laboratory in Hamamatsu University School of Medicine in 1982. At first we consulted with Sakakibara in Nagoya City University and obtained advice as to synthetic methods. Then we decided to synthesize various derivatives of a model compound (2-benzhydryloxyN-ethyl piperazine, BP; No. 1 in Table 1) with an amino group at different positions (No. 2 to 8 in Table 1) in order to determine where the linker between the ligand and the agarose gel should be positioned. The addition of an amino group decreased the affinity for muscarinic receptors by five- to twenty-fold, but the effects of differences in the position of the amino group were relatively small. On the other hand, the effects of the positions became greater when a heptanoic acid moiety (a linker) was coupled to the amino group: the decrease in the affinity for muscarinic receptors was 120-fold on the introduction of the linker to the piperazine group of compound No. 2 in Table 1, whereas the decrease was only 9-fold on introduction of the linker to the phenyl ring of compound No. 8 in Table 1. This indicated that the linker should be placed on the phenyl group. Then, compounds with higher affinity for muscarinic receptors were synthesized by introducing a tropine or quinuclidinyl group instead of a piperazine group (Nos. 9 and 10 in Table 1). Finally ABT (aminobenztropine) was chosen as a ligand for the affinity gel. ABT exhibits a high affinity for solubilized muscarinic receptors, with an equilibrium dissociation constant of 7 nM.
We then attempted to optimize the efficiency of the affinity gel, that is, to increase the amount of bound muscarinic receptors and to decrease nonspecific binding of other proteins, according to advice from Nakata and Yamauchi in Asahikawa University School of Medicine and others. Several factors were found to be involved in determination of the efficiency, which included (a) chemical species of the linker, (b) species of the agarose gel, (c) density of bound ligands, (d) concentration of salt in the incubation medium, (e) ratio of gel to applied proteins, and so on. CH-Sepharose gel with a hydrophobic linker such as [(C[H.sub.2]).sup.5] showed higher capacity for muscarinic receptors but lower specificity as compared to Epoxy-Sepharose gel (see the structure of ABT agarose in Fig. 1). In addition, it was important to add an appropriate concentration of salt, e.g. 0.1-0.2 M NaCl, to the incubation medium, as greater non-specific binding was observed at lower salt concentrations, and less binding of muscarinic receptors at higher salt concentrations. (18) Finally, we prepared an affinity gel comprising agarose gel bound with a muscarinic ligand, amino-benztropine (ABT agarose) (Haga and Haga, (19) 1983). Approximately 1,000-fold purification was attained with a single step of affinity chromatography with ABT agarose.
2.2. Purification of muscarinic receptors. Muscarinic receptors were purified to apparent homogeneity from a digitonin/cholate-extract of porcine brain by ABT agarose and hydroxylapatite chromatography. (20) SDS-PAGE of the purified preparation gave an apparently single band corresponding to approximately 70kDa. The specific[[sup.3]H]NMS binding activity of the purified preparation was estimated to be 13 nmol/mg of protein, which is close to the theoretical value of 14 nmol/mg of protein based on the assumption of a single binding site per 70 kDa. Muscarinic receptors were also purified as a [[sup.3]H]QNB binding component from porcine atria using the ABT agarose and several other chromatography sytems. (21) SDS-PAGE of the purified preparation gave two bands corresponding to apparent molecular weights of 79 kDa and 15 kDa, the 79 kDa component being shown to bind with [[sup.3]H]PrBCM. These results provide direct evidence that muscarinic receptors, which are defined as [[sup.3]H]NMS or [[sup.3]H]QNB binding components, are protein molecules of 70-80 kDa.
Muscarinic receptors purified from porcine brain and atria had been expected to correspond to the [M.sub.1] and [M.sub.2] subtypes, which can be discriminated according to the difference in the affinity for pirenzepine: pirenzepine was reported to have a higher affinity for cerebral membrane preparations by a factor of approximately 30 than for atrial ones. The situation, however, was complicated because pirenzepine was indicated to show similar affinity for the two receptor preparations solubilized with digitonin from either rat brain or heart. (22) Thus, purification of muscarinic receptors from both cerebral and atrial membranes was not sufficient to prove that the [M.sub.1] and [M.sub.2] receptors are distinct molecules. This problem could be solved by cloning cDNA for each receptor, as described later.
As for the affinity of pirenzepine, Nishiyama et al. (23) confirmed that muscarinic receptors purified from either porcine cerebrum or atrium exhibit the same affinity for pirenzepine. This indicates that the affinity for pirenzepine of muscarinic receptors in membrane preparations is not determined solely by their protein moiety but may be affected by other factors such as bound phospholipids. Berstein et al. (24) demonstrated that pirenzepine had a higher affinity for muscarinic receptors purified from the cerebrum than those purified from the atrium when they were reconstituted into cell membranes pretreated with PrBCM, and that the result was not affected whether cell membranes were prepared from the cerebrum or the atriam. Furthermore, the membrane preparations could be replaced by particular sets of lipid preparations containing cholesteryl hemisuccinate and phosphatidylcholine. Cerebral and atrial muscarinic receptors in membranes appear to be differentially regulated by the same set of cholesterol and phosphatidylcholine to show different affinities for pirenzepine. It is possible that the pirenzepine binding site may involve the lipid moiety or may be affected by interaction with lipids, whereas the binding site for atropine, QNB and NMS does not involve and is not affected by the lipid moiety.
Muscarinic receptors purified from porcine atrium (21,23) as well as ones purified from porcine cerebrum (20) showed heterogeneous affinities for a given agonist, although a single kind of muscarinic receptor had been expected to be present in atrium. This indicates that muscarinic receptors may exist in different states with different affinities for agonists. A possible candidate for the different states is the redox state of the cysteine residues: full reduction of muscarinic receptors purified from either cerebra or atria with high concentrations of dithiothreitol (DTT) caused decreases in the affinities for ligands including agonists, and modification by SH reagent dithiobis(2-nitro benzoic acid)(DTNB) caused increases in the affinities for agonists by approximately 30-fold. (23,25) Treatment of purified muscarinic receptors with these SH reagents did not affect the size of the muscarinic receptors, as judged from sucrose density gradient centrifugation and SDSPAGE, indicating that intra-molecular, but not inter-molecular, S-S bond(s) is (are) involved in determination of the affinities for ligands.
2.3. Cloning of complementary DNAs for muscarinic receptors. Purification of muscarinic receptors enabled us to determine their partial amino acid sequences, which could be used to clone complementary DNAs (cDNAs) for muscarinic receptors. We collaborated with Numa's and Matsuo's laboratories in Kyoto University and Miyazaki University School of Medicine, respectively. Muscarinic receptors were purified from porcine cerebra in Hamamatsu University School of Medicine and sent to Miyazaki University School of Medicine, where Kangawa et al. partially hydrolyzed them and determined amino acid sequences of five peptides. By utilizing the amino acid sequences of these peptides, Kubo et al. (26) (1986) in Kyoto University cloned a cDNA from a library constructed from the mRNAs of porcine cerebrum and confirmed that Xenopus oocytes expressing the cDNA respond to ACh with a specificity expected for muscarinic receptors. (26) Furthermore, another cDNA was cloned by using a different amino acid sequence from the library constructed from mRNA of porcine atrium. (27) Difference in the tissue distributions of mRNAs indicated that two cDNAs obtained from libraries of porcine cerebrum and atrium encode the muscarinic receptor [M.sub.1] and [M.sub.2] subtypes, respectively. A cDNA corresponding to the [M.sub.2] subtype was independently cloned by using partial amino acid sequences obtained for muscarinic receptors purified from porcine atria. (28) These results provide definite evidence that muscarinic receptors are protein molecules, and that the [M.sub.1] and [M.sub.2] subtypes are two distinct protein molecules with different amino acid sequences. In this paper, hereafter, they are termed [M.sub.1] and [M.sub.2] receptors respectively.
The amino acid sequences of [M.sub.1] and [M.sub.2] receptors were deduced from the nucleotide sequences of cDNAs. Three of the five sequences determined for muscarinic receptors purified from porcine brain were found in the sequence of [M.sub.1] receptors, and two in that of [M.sub.2] receptors. [M.sub.1] and [M.sub.2] receptors consist of 460 and 466 amino acid residues with calculated molecular weights of 51 and 52kDa, respectively. The difference between these values and the value of 70 kDa estimated from SDSPAGE of cerebral muscarinic receptors was supposed to be accounted for by the contribution of the carbohydrate moiety, and was confirmed by endoglycosidase treatment of cerebral and atrial receptors. (29,30) In fact, the sequences of [M.sub.1] and [M.sub.2] receptors contain two and three potential N-glycosylation sites with the consensus sequence of NXS/ TX in the amino terminal portion.
The amino acid sequences of [M.sub.1] and [M.sub.2] receptors revealed that they are similar to each other but not to nicotinic receptors, and that they contain seven hydrophobic regions, with a length of 20-24 amino acid residues, which were supposed to constitute transmembrane segments. The aminoterminal portion with glycosylation sites should be in the extracellular phase and the carboxyterminal tail in the intracellular phase. In accord with previous observations on the effects of sulfhydryl reagents, cysteine residues were found in both [M.sub.1] and [M.sub.2] receptors. Uchiyama et al. (31) suggested that there is an S--S bond between Cys98 in the first extracellular loop (E1: a loop between transmembrane segment 2 and 3 ([TM.sub.2] and [TM.sub.3])) and Cys 178 in the second extracellular loop (E2 loop): here cerebral receptors labeled with [[sup.3]H]PrBCM were analyzed by SDS PAGE with or without partial hydrolysis, reduction by DTT, endoglycosidase F treatment, and reaction with antibodies against peptides in the specific regions. In the same series of experiments, the binding site of [[sup.3]H]PrBCM was located between the aminoterminus and the E2 loop. The presence of the S--S bond and localization of [[sup.3]H]PrBCM binding site to an Asp residue in T[M.sub.3] were directly demonstrated by peptide map analysis by a group of Hulme. (32)
Screening for cDNAs or genes exhibiting homology to cDNAs of [M.sub.1] and [M.sub.2] receptors in a cDNA library or genomic library resulted in the identification of other muscarinic receptors, that were termed as [M.sub.3],[M.sub.4] and [M.sub.5] receptors. (33-35) The coding regions of these five receptors were found to be in a single exon and to contain no intron, and thus genes with full length coding region could be easily cloned from the genomic library. The presence of [M.sub.1] and [M.sub.2] receptors, and possibly [M.sub.3] receptors, was expected from pharmacological studies, but [M.sub.4] and [M.sub.5] receptors had not been anticipated before their cDNAs or genes were cloned. The pharmacological properties of these five muscarinic receptors were determined by expressing them initially on Xenopus oocytes, and later on cultured cells such as COS, CHO or HEK293 cells, and are summarized in the refernece. (6)
The presence of five different muscarinic receptors appears to provide an opportunity to develop subtype-specific ligands that may affect only a certain subtype and thus a specific physiological function. In addition, multiple subtypes of muscarinic receptors appear to contribute to subtle regulation by ACh of many physiological functions. For example, Kajimura et al. (36) demonstrated that acid secretion by parietal cells is regulated in a dual way by muscarinic agonists, i.e. stimulation and inhibition with their low and high concentrations, respectively, most probably through different muscarinic receptor subtypes.
2.4. A protein family with seven transmembrane segments. The presence of seven transmembrane segments in rhodopsin (37) and O adrenergic receptors (38) was suggested by their amino acid sequences, which were determined in 1980 by protein analysis and in 1986 by cDNA screening, respectively. Rhodopsin, [beta] adrenergic receptors, and muscarinic receptors are typical receptors for external stimuli, hormones, and neurotransmitters, respectively. The similarity among these three proteins had been suggested by experiments on their interaction with G proteins, as described in the following section, and was confirmed by cloning of cDNAs for [beta] adrenergic receptors, and [M.sub.1] and [M.sub.2] muscarinic receptors in 1986, and by identifying the seven putative transmembrane segments. It is interesting to note that the amino acid sequences of [M.sub.1] and [M.sub.2] receptors are very homologous except the third intracellular loop (I3 loop), but homology among [M.sub.1] receptors, rhodopsin and [beta] adrenergic receptors is rather limited in spite of the common property of seven transmembrane segments (Fig. 2). Subsequently, however, the cDNAs for a lot of odorant-, hormone-, and neurotransmitter-receptors have been cloned on the basis of homology to rhodopsin, [beta] adrenergic, and muscarinic receptors. They are called G protein-coupled receptors (GPCRs) or seven-transmembrane (7-TM) receptors.
In 2000, a human genome sequence was reported, and thus the systematic search for GPCRs became possible. We thought that we might find GPCR genes in the human genome because we knew that all GPCRs, which had been identified so far, have the seven hydrophobic region and many of them have no intron in the coding region, and so decided to collaborate with the laboratory of Mitaku in the Tokyo University of Agriculture and Technology, who is a theoretical biophysicist. During the collaboration, Takeda et al. (39) (2002) succeeded in identifying 581 GPCR candidates without an intron in the coding region, 301 of which had not been identified until then. The total number of GPCRs were estimated to be 948, on the assumption that the ratio of intron-less GPCRs is not different between already-registered and not yet-registered GPCRs. The estimated numbers of GPCRs (the proportion of intron-less GPCRs) are as follows: odorant receptors 481 (100%), taste receptors 28 (77%), receptors for endogenous ligands 330 (66%), and ones without homology to already-registered GPCRs 109 (assumed to be 66%).
GPCRs constitute one of the largest super-families in the human genome, and function as cell sensors for external stimuli, hormones, and neurotransmitters. GPCRs are believed to have the common properties that they have 7-TM segments and activate GTP-binding regulatory proteins (G proteins). Strictly speaking, however, direct evidence for 7-TM segments has been obtained for limited number of members, including muscarinic receptors, rhodopsin and [beta]2 adrenergic receptors and several others. Direct evidence for coupling with G proteins is also limited to only several members. On the other hand, GPCRs appear to have other functions besides G protein activation. (40) [M.sub.2] receptors, together with G protein [beta][gamma] subunits, has been shown to activate G protein-coupled receptor kinase 2 in an agonist-dependent manner, as described later.
All five muscarinic receptors are distinct from most GPCRs in that they have a long third intracellular loop (I3 loop, loop between T[M.sub.5] and TM6) with 160-240 amino acid residues. These long I3 loops are found in [alpha]2 adrenergic receptors and some 5HT receptors. A long I3 loop is not involved in G protein activation by a receptor but appears to be involved in its regulation.
3. Functions of muscarinic receptors
3.1. [beta] Adrenergic receptor--G protein (Gs)--adenylate cyclase system. Studies on functions of muscarinic receptors have been built on the prior work on [beta] adrenergic receptor--G protein (Gs)--adenylate cyclase system, and therefore the historical background of the latter system is introduced here.
Biochemical studies on intracellular events initiated by extracellular hormones originated from the discovery of cyclic AMP (cAMP) by Rall and Sutherland in the late 1950s. (41) They showed that a certain group of hormones, including adrenaline and glucagon, induces increases in cAMP concentration in the cell. Rodbell et al. studied glucagon-stimulated cAMP formation in membrane preparations from rat liver, and demonstrated in the early 1970s that guanine nucleotides (GTP or GDP) decrease the affinity for glucagon and that GppNHp (hydrolysis-resistant analogue of GTP) activates the cAMP forming reaction in a pseudo-irreversible manner. (42) Guanine nucleotides have also been shown to decrease the affinity of agonists, but not of antagonists, for [beta] adrenergic receptors in membrane preparations of cultured cells. (43) These results suggested that this signal transduction system requires at least three functional components. It was not known, however, if the hormone receptor, adenylate cyclase, and GTP-recognizing component represent three independent proteins or three functions of a single or two protein(s). In this context, it is interesting to note that the same question was raised for the hormone-dependent increase in cyclic GMP (cGMP), and it turned out later that ANP (atrial natriuretic peptide)-dependent cGMP formation is accomplished by a single protein of ANP receptor/guannyl cyclase, which possesses an extracellular ligand-binding domain, a single transmembrane segment, and an intracellular guanylate cyclase catalytic domain, (44) whereas ACh-dependent cGMP formation needs a series of components, as described later.
Initial biochemical studies indicated that the [beta] adrenergic receptor, which was assessed as the binding component of [[sup.125]I]IHYP (iodohydroxybenzylindolol), and the component with GppNHp-activated adenylate cyclase activity could be physically separated as distinct entities. (45) The molecular sizes of the [beta] adrenergic receptor and adenylate cyclase were estimated to be 75,000 and 220,000, respectively.
Molecular identification of a component related with the effect of guanine nucleotides was performed by several groups, particularly Pfeuffer in Germany, Cassel and Selinger in Israel, and Ross and Gilman in USA in the 1970s. Pfeuffer (46) separated the GTP-binding component using a GTP-agarose gel from an extract of pigeon erythrocytes, and demonstrated that the component may confer the adenylate cyclase activity to another component. Cassel and Selinger (47) and Cassel and Pfeuffer (48) showed that O adrenergic agonists enhance the GTPase activity of turkey erythrocyte membranes and proposed the hypothesis that the agonist-bound receptor stimulates the release of GDP from and binding of GTP to the GTP-binding component with GTPase activity, and the component bound with GTP stimulates the adenylate cyclase and is converted to an inactive form bound with GDP by its own GTPase activity. Ross and Gilman used S49 lymphoma cells and their variants ([cyc.sup.-] (49) and unc (50)), and demonstrated through an elegant series of experiments that the cAMP-forming activity in membrane preparations derived from GppNHp-treated S49 lynphoma cells is due to two components: one is labile to high temperature treatment (e.g. 50[degrees]C for 30min) and is present in the [cyc.sup.-] mutant, and the other is stable to high temperature treatment and is absent in the [cyc.sup.-] mutant. (51) The former component was shown to be the adenylate cyclase, and the latter was shown to be the component which confers GppNHp-stimulating activity to the former. The latter was named N or G/F and is now named Gs (stimulatory G protein). Subsequently Gs was purified from a Lubrol PX extract of rabbit liver as a component that confers the adenylate cyclase activity on [cyc.sup.-] membranes (see review ref. (52). Gs was found to be composed of three subunits, [alpha][beta][gamma] and the [alpha] subunit is the GTP-binding component and the substrate of ADP-ribosylation by cholera toxin.
[beta] Adrenergic receptors were purified from a digitonin extract of turkey erythrocytes by an affinity chromatography with a [beta] adrenergic antagonist as a ligand by Strosberg group (53) in France. Interaction between [beta] adrenergic receptors and G protein Gs was studied in details by the groups of Gilman, Lefkowitz, and others, and their direct interaction was confirmed by reconstitution of purified [beta] adrenergic receptors and Gs. (54)
3.2. Receptors--G protein (Gi)--adenylate cyclase inhibition. It appears that muscarinic receptors were generally believed to be linked to cGMP formation or phosphatidylinositol (PI) turnover rather than cAMP in the mid 1970s when the signal transduction system of [beta] receptor-Gs-adenylate cyclase was revealed, although the activation of muscarinic receptors in the heart had been reported to result in inhibition of cAMP formation in the early 1960s. (55) The relation of muscarinic receptors with cGMP formation or PI turnover was studied in detail in the mid 1990s, as described later.
A series of experiments by Rodbell's group in the 1970s showed that a hormone-dependent decrease in cAMP was also mediated through a mechanism including the action of guanine nucleotides. The responsible protein was indicated to have properties different from Gs, (56) and was termed as Ni and now Gi, inhibitory G protein (see review ref. (57). Katada and Ui (58) demonstrated that the hormone-dependent decrease in cAMP, including the ACh-induced decrease in cAMP, (59) is inhibited by treatment with pertussis toxin, which was originally named IAP (Islet-Activating Protein), and that a protein, which is now known to be the [alpha] subunit of Gi, is ADP-ribosylated through the action of the pertussis toxin. Gi was purified from rabbit liver as a substrate of ADP ribosylation by pertussis toxin. (60) Other G protein, which was named as Go, was also purified together with Gi from the brain. (61) Gi and Go are more abundantly present in the brain than Gs, and are estimated to account for 1.5% of membrane protein. Furthermore, light-dependent activation of cGMP phosphodiesterase in the retina was also demonstrated to be mediated by the G protein called transducin and now more often called Gt (see review ref. 62).
All these G proteins, i.e. Gs, Gi, Go and Gt, have been demonstrated to be composed of three subunits as an [[alpha].sub.GDP][beta][gamma] trimer in an inactive form, and as [[alpha].sub.GTP] and [beta][gamma] in an active form. Agonist-bound receptors catalyze the exchange of GDP and GTP, and the a subunit has GTPase activity to convert GTP to GDP. The dissociated [[alpha].sub.GTP] and [beta][gamma] subunit interact with and activate different kinds of effectors. The G proteins are generally classified as to [alpha] subunits, i.e. whether it activates (Gs) or inhibits (Gi, Go) adenylate cyclase, or activates cGMP-phosphodiesterase (Gt). The [alpha] subunit of Gs (Gs-[alpha]) is a substrate of ADP-ribosylation by cholera toxin, and the ADP-ribosylation causes inhibition of its GTPase activity and therefore keeps it in an active form (GTP-bound form). The a subunits of Gi and Go are substrates of ADP-ribosylation by pertussis toxin, and the ADP-ribosylation causes uncoupling between receptors and G proteins. Gt-[alpha] is the substrate of both cholera toxin and pertussis toxin. The O. subunits were also demonstrated to be involved in activation or inhibition of adenylate cyclase (see review ref. 63), and also in other functions such as activation of inward-rectified [K.sup.+] channels. (64)
In the early 1980s, the technique to prepare monoclonal antibodies became common. Nukada, who joined in our group in Hamamatsu University School of Medicine in 1981, succeeded in preparing monoclonal antibodies for Gt-[alpha] and cloned its cDNA by using the antibody in collaboration with Tanabe et al. in Numa's laboratory in Kyoto University. (65) This was the first G protein [alpha] subunit, cDNA for which was cloned, and this was followed by cloning of the Gi-[alpha] (66) and Gs-[alpha] (67) subunits. The amino acid sequences of all these a subunits indicated that they contain a ras domain, which is homologous to the product of an oncogene ras, and another domain, which was later termed the helical domain. The sequences of Gt-[alpha], Gi-[alpha] and Go-[alpha], but not Gs-[alpha], contain a cysteine residue as the fourth amino acid from the carboxy-terminus, which is ADP-ribosylated by pertussis toxin. Thereafter, cloning of homologous proteins revealed the presence of three kinds of Gi-[alpha] genes, seven kinds of [beta] subunits, and 11 kinds of [gamma] subunits (see review ref. 68). In addition, nine kinds of adenylate cyclase genes were identified, which are differentially regulated by various kinds of G-[alpha] and G-[beta]. subunits as well as by [Ca.sup.2+]-calmodulin (see reviews refs. 69, 70).
3.3. Interaction of muscarinic receptors with Gi-type G proteins. Indirect evidence for the interaction between muscarinic receptors and G proteins is based on the observation that agonist but not antagonist binding to muscarinic receptors in membranes of rat heart is affected by GTP and GDP, (71) which is similar to the original effect of guanine nucleotides on the agonist binding to [beta] adrenergic receptors. (43) These results indicated that muscarinic receptors as well as other receptors such as [alpha]2 adrenergic receptors interact with guanine nucleotide-related protein(s). The guanine nucleotide-sensitive high affinity agonist binding of muscarinic receptors in membrane preparations derived from porcine caudate nucleus was found to be abolished by treatment with a sulfhydryl reagent, N-ethylmaleimide (NEM), (72) although guanine nucleotide-sensitive agonist binding to [beta] adrenergic receptors was hardly affected by NEM treatment. These results suggest that the putative G protein involved in agonist binding of muscarinic receptors is different from Gs-type G proteins.
To obtain direct evidence for the interaction between muscarinic receptors and G proteins, we brought muscarinic receptors purified from porcine brain to Ui's laboratory in Hokkaido University, where Gi had been purified. We, together with Katada and Kurose, examined their interaction as muscarinic agonist-enhanced GTPase activity in the reconstituted system. Thus, muscarinic receptors were shown to interact with Gi in a pertussis toxin-sensitive manner. (73) This is the first direct demonstration that at least one of the functions of muscarinic receptors is to interact with and activate Gi-type G proteins. Furthermore, quantitative reconstitution of muscarinic receptors and G proteins Gi and Go indicated that approximately 50% of muscarinic receptors reconstituted with one of either Gi or Go showed GppNHp-sensitive high affinity for ACh (Fig. 3) or other agonists. The proportion of high affinity sites did not increase further on the addition of both Gi and Go. GDP or GTP as well as GppNHp abolished the high-affinity agonist binding in the same concentration ranges, respectively, and no conversion between GTP and GDP was detected under the experimental conditions used (74) (Fig. 3). The simplest interpretation of these results is that (1) 50% of muscarinic receptors interact with either Gi or Go in the absence of GTP and GDP, (2) the muscarinic receptor-Gi/Go complex shows high affinity for agonists, and (3) the complex bound with GTP or GDP dissociates into Gi/Go and muscarinic receptors with low affinity for agonists. These interpretations infer that the muscarinic receptor-guanine nucleotide-free Gi/Go complex exhibits high affinity for agonists and may exist in a stable form at least in vitro, although the muscarinic receptor Gi/Go-GTP (GDP) complex does not exhibit high affinity for agonists or such a complex is not formed in a stable form.
Muscarinic receptors purified from porcine brain were found to interact also with Gn, which was purified from porcine brain as a novel [[sup.35]S]GTP [gamma]S binding component. Gn is a substrate of ADP-ribosylation by pertussis toxin but different from Gi or Go. The gene species of a subunits for Gi and Gn are thought to correspond to [Gi.sub.1][alpha] and [Gi.sub.2][alpha], respectively, but the gene species of [beta] and [gamma] subunits for Gi and Gn have not been determined yet. In the reconstitution system, muscarinic agonists were found to stimulate the release of [[sup.3]H]GDP bound to Gi, Go or Gn, and the amount of released [[sup.3]H]GDP was 5-10 times the amount of muscarinic receptors. These results indicate that a single muscarinic receptor molecule interacts with multiple G proteins and catalytically stimulates the release of [[sup.3]H]GDP bound to them. (75,76)
When the above reconstitution experiments were carried out, it was not known which subtypes of muscarinic receptors are present in cerebral tissues and thus it was not clear which subtypes interacted with Gi, Go or Gn. Based on the present knowledge that all five subtypes are present in the brain, and that [M.sub.2] and [M.sub.4] receptors interact with Gi-type G proteins, it is reasonable to suppose that [M.sub.2] and [M.sub.4] receptors interact with Gi, Go and Gn. More specifically, Ikegaya et al. (77) showed that muscarinic receptors purified from porcine atria ([M.sub.2] receptors) interact with three kinds of Gi-type G proteins, Gi, Go and Gn.
Shiozaki and Haga (78) (1992) examined the interaction between atrial muscarinic receptors and Gi in more detail in the reconstitution system and demonstrated that (1) [M.sub.2] receptors show three different affinities for agonists depending on the presence or absence of Gi and [Mg.sup.2+] ions, (2) agonists cause a decrease in the affinity for GDP, (3) the presence of [Mg.sup.2+] ions is necessary for the agonist-dependent decrease in affinity for GDP, and (4) the effect of an agonist was more prominent on the affinity for GDP rather than on that for GTP. Studies using [M.sub.2]-Gi1[alpha] fusion proteins confirmed these observations, and also indicated that muscarinic partial agonists such as pilocarpine and McN343 caused a partial decrease in the affinity for GDP that was intermediate between those in the presence and absence of a full agonist (79) (Fig. 4). These results indicate that the [M.sub.2]-Gi fusion protein may be used as a ligand-screening system, and also suggest that the [M.sub.4]-Gi fusion protein could be used for the screening of [M.sub.4]-specific agonists or positive allosteric modulators, which may be useful for the treatment of schizophrenia. GPCR-Ga fusion proteins could be used to screen endogenous ligands for orphan receptors (80) and would be useful for searching for agonists for Gi/Go-coupled receptors (see a review ref. 81).
Based on these results, a simplified model was proposed for the interaction between [M.sub.2] and Gi. (76) In the model, the ACh-[M.sub.2]-Gi complex is assumed to be formed as the transition state for the GDP/GTP exchange reaction, which contributes to reduction of the barrier for transition from Gi-GDP to Gi-GTP through Gi. This action of ACh-[M.sub.2] on Gi was noticed to be similar to the action of actin on myosin, where actin facilitates the exchange of ATP/ADP in myosin by forming an actin-ATP/ADP-free myosin complex. In the simplest form of the model, approximately 100-fold difference in the affinity for ACh with or without Gi was thought to correspond to a reduction of the barrier by 3 kcal/mol and to an increase in GDP dissociation rate by 100-fold. Whether the model is accepted or not, it would be safe to conclude that the function of ACh-bound [M.sub.2] is to accelerate the dissociation of GDP from Gi.
3.4. Muscarinic receptors--G protein (Gq)--phospholipase C. Hokin and Hokin (82) showed in the 1950s that stimulation by ACh of pancreas and brain cortex slices causes incorporation of [[sup.32]P]phosphate into phospholipids, particularly phosphatidyl inositol. After a long series of studies by various groups for almost thirty years, a mechanism was proposed by which the stimulation of muscarinic and other receptors leads to activation of phospholipase C, which catalyzes the formation of diacylglycerol and inositol trisphosphate (IP3) from phosphatidylinositol 4,5 bisphosphate (PIP2), and that IP3 acts on intracellular organelles causing the release of [Ca.sup.2+] ions into the cytoplasm (see review ref. 83). This signal transduction system, which is called the PI response, was indicated to involve a lot of receptors and to be ubiquitous in various tissues. [M.sub.1], [M.sub.3] and [M.sub.5] receptors were shown to be coupled with PI turnover, whereas [M.sub.2] and [M.sub.4] are only weakly coupled with PI turnover (see reviews refs. 34, 35).
Receptors and phospholipase C were suggested to be linked through G proteins, and the relevant G proteins were putatively termed Gp, but it took a time to identify the molecular entity of Gp, which is now called Gq-type G proteins. Retrospectively, the difficulty in identifying Gq proteins appears to be due to that (1) a good assay system like [cyc.sup.-] cells for the identification of Gs was not available in the case of Gq, (2) Gq-type G proteins do not bind [[sup.35]S]GTP [gamma]S without activation by agonist-bound receptors, in contrast with Gi and Go proteins, which could be purified as [[sup.35]S]GTP[gamma]S binding components, (3) toxins like cholera and pertussis toxins were not available for Gq-type G proteins, and (4) there was some confusion as to the sensitivity to pertussis toxin as there are two kinds of phospholipase C[beta], which are activated by Gq- and Gi-type G proteins in pertussis toxin-resistant and -sensitive manners, respectively.
We started characterizing the PI turnover initiated by activation of [M.sub.1] receptors with Nukada at Tokyo University in 1989. Nakamura et al. (84) (1991) identified two new G protein a subunits, which were termed GL1a and GL2a, and are not substrates of pertussis toxin, by cloning of the corresponding cDNAs with the use of hybridization with cDNA for [Gi.sub.1][alpha]. [G.sub.L1][alpha] and [G.sub.L2][alpha] were found to be the same as the a subunits of Gq-type G proteins G14a and G11a respectively, which were independently cloned by Strathmann and Simon. (85) We collaborated with Takenawa in Tokyo University and Kikkawa in Kobe University, and constructed a reconstitution system for [M.sub.1] receptors, Gq-type G proteins, and phospholipase C-[beta]1, and could confirm that the function of [M.sub.1] receptors is to interact with and activate Gq-type G proteins, which activate phospholipase C [beta]1: we expressed [M.sub.1] receptor and Gq-type G proteins (G[[beta].sub.1][[gamma].sub.2] + one of [G.sub.L1][alpha] ([G.sub.14][alpha]), [G.sub.L2][alpha] ([G.sub.11][alpha]) or Gq[alpha]) in Sf9 cells using baculovirus, purified them and reconstituted them with purified phospholipase C-[beta] in phospholipid vesicles, and demonstrated that [M.sub.1] receptors can activate all three Gq-type G proteins, but not Go, and that these Gq-type G proteins activate phospholipase C-[beta]1. (86) In these studies, the baculovirus-Sf9 system was very useful for preparing large amounts of proteins, particularly membrane proteins such as muscarinic receptors, which are expressed only in a limited amount in mammalian cultured cells or tissues. The baculovirus-Sf9 system was established in our laboratory by Kameyama, who had learned it at the laboratory of Ross in Texas University. It is interesting to note that guanine nucleotide-sensitive high affinity agonist binding was not detected in the [M.sub.1]-Gq system so clearly as compared with in the [M.sub.2]-Gi system, suggesting that the mechanism of activation by receptors of G proteins might be different between the two systems.
The relation of muscarinic receptors with the formation of cGMP appears to involve many steps of intra- and inter-cellular signal transduction, in contrast with the relation of muscarinic receptors with the inhibition of cAMP formation. In the case of relaxation of arterial smooth muscle, (1) muscarinic [M.sub.3] receptors in endothelial cells respond to ACh and initiate the signal transduction leading to formation of NO (nitric oxide) through many steps including Gq, phospholipase C[beta]1, IP3 receptor, [Ca.sup.2+] /calmodulin, and NO synthase, (2) NO diffuses into smooth muscle and there activates NO receptor with guanylate cyclase activity, and (3) cGMP activates protein kinase G which leads to relaxation of the muscle and then vasodilation (see a review ref. 87).
Recently muscarinic receptors, particularly [M.sub.1] and [M.sub.3] receptors, were shown to be linked to activation of [G.sub.12/13]-type G proteins, which leads to activation of small GTP-binding proteins Rho through activation of RhoGEF (RhoGTPase nucleotide exchange factor) (see a review ref. 88).
4. Regulation of muscarinic receptors
4.1. Synergistic activation of G protein-coupled receptor kinase 2 (GRK2) by agonist-bound muscarinic receptors and G protein [beta][gamma] subunits. In the early 1990s we started studies on the regulatory systems of muscarinic receptors. At first, we examined the phosphorylation by protein kinase C of muscarinic receptors. Muscarinic recep tors purified from porcine brain were found to be phosphorylated at their carboxyterminal tail, but the phosphorylation did not affect their ability to interact with muscarinic agonists and Gi/Go. (89,90) In addition, the reconstituted system of [M.sub.1] receptors, Gq-type G proteins, and phospholipase C-[beta] was also shown to be not affected by treatment with protein kinase C. (86) The phosphorylation by protein kinase C of muscarinic receptors was not dependent on the presence of a muscarinic agonist, in contrast with the phosphorylation by GRK2 described hereafter.
When we started studies on the phosphorylation of muscarinic receptors, rhodopsin and [beta] adrenergic receptors were known to be phosphorylated in a light- or agonist-dependent manner by rhodopsin kinase (GRK1 in the present terminology (91)) and [beta] adrenergic receptor kinase (OARK, GRK2 in the present terminology (92)), respectively, and these phosphorylations were reported to affect their function (see review ref. 93). Kwatra and Hosey (94) (1986) reported that muscarinic receptors in chick heart may be phosphorylated in an agonist-dependent manner by an endogenous, unidentified kinase. We then started to search for this kinase that specifically phosphorylates muscarinic receptors in an agonist-dependent manner. We found that a preparation partially purified from a brain extract may phosphorylate both cerebral and atrial muscarinic receptors in an agonist-dependent manner. (95) Unexpectedly, the activity was found to be affected by G protein Go in a dual manner that is stimulation at lower concentrations of it and inhibition at higher oncentrations of it, and the stimulatory and inhibitory effects were reproduced by the Go-[beta][gamma] subunits and Go-a[beta][gamma] trimer, respectively. (96) The inhibition by the Go-a[beta][gamma] trimer was abolished on the addition of either GTP or GDP, (95) which is consistent with the model that [M.sub.2] receptors form a complex with guanine nucleotide-free Go, but not with GTP or GDP-bound Go. Further experiments indicated that the kinase preparation may also phosphorylate rhodopsin in light- and G[beta][gamma]-dependent manners, whereas the phosphorylation of rhodopsin by rhodopsin kinase was dependent on light but not affected by the presence of [beta][gamma] subunits. 97) Furthermore, any [beta][gamma] subunits derived from Gs, Gi, or Go stimulated the agonist- or light-dependent phosphorylation of muscarinic receptors and rhodopsin. These results suggested that the stimulatory effect of [beta][gamma] subunits would be due to the specific property of the kinase preparation, and not due to the species of substrates or [beta][gamma] subunits.
The partially purified kinase was supposed to be similar to or the same as GRK2, because of the similar purification procedure, and similar effects of salts and heparin, but the stimulatory effect by [beta][gamma] subunits on phosphorylation of rhodopsin (98) and muscarinic receptors (99) by GRK2 had not been reported. Kameyama et al. (100) examined the effect of [beta][gamma] subunits on GRK2, the cDNA of which was supplied by Lefkowitz in Duke University, and showed that GRK2 may phosphorylate muscarinic receptors as well as [beta] adrenergic receptors in agonist-and G[beta][gamma]-dependent manners. In addition, rhodopsin was also phosphorylated by GRK2 in light- and G[beta][gamma]-dependent manners. GRK2 is homologous to rhodopsin kinase except that GRK2 has an extra C-terminal domain of approximately 130 amino acid residues. We assumed that the C-terminal tail may be involved in the interaction with G[beta][gamma]. In accord with this assumption, a mutant with a deletion in the C-terminal tail was shown to phosphorylate muscarinic receptors in an agonist-dependent but G[beta][gamma]-independent manner. (100) In 2003, this assumption was directly confirmed by the crystal structure of the GRK2-G[beta][gamma] complex, where the C-terminal tail of GRK2 is bound with the G[beta][gamma] subunit. (101)
Phosphorylation sites for rhodopsin kinase or GRK2 in rhodopsin or [beta] adrenergic receptors were shown to be in the C-terminal tail, but those in muscarinic receptors had not been determined yet. Nakata et al. (102) located the phosphorylation sites in the central part of the third intracellular loop (I3) but not in the C-terminal tail. Furthermore, the amino acid sequences of putative phosphorylation sites in rhodopsin, [beta] adrenergic receptors, and [M.sub.2] receptors were shown not to be similar to each other. This raises the question why GRK2 specifically phosphorylates only agonist-bound GPCRs. One of the simplest explanations is to assume that the phosphorylation sites are hidden inside GPCRs in their inactive state and are exposed in their active state induced by agonist binding. To test this assumption, we compared the phosphorylation of [M.sub.2] receptors, and a fusion protein of glutathione S-transferase (GST) and a peptide (268-324) containing the phosphorylation sites (I3-GST). If the assumption is the case, I3-GST would be as good a substrate as a gonist-bound [M.sub.2] receptors.
Against the assumption, I3-GST turned out to be a poor substrate for the kinase. The phosphorylation of I3-GST, however, was found to be synergistically enhanced by G protein [beta][gamma] subunits and mastoparan, which is known to mimic agonist-bound receptors. (103) The stimulatory effect of mastoparan could be reproduced by synthetic peptides corresponding to the sequences of intracellular segments adjacent to the transmembrane helices of muscarinic receptors in I2, I3 and the carboxyterminal tail. Furthermore, I3-GST was found to be phosphorylated by GRK2 in an agonist-dependent manner in the presence of G protein [beta][gamma] subunits and I3-deleted [M.sub.2] receptors, which lack the central part of the I3 loop including phosphorylation sites (104) (Fig. 5 (a)). These results indicate that [M.sub.2] receptors interact with GRK2 at two different sites: (1) the phosphorylation sites in the central part of I3 that serves as the substrate and (2) intracellular segments that may serve as a GRK2 activator. It will be reasonable to assume that agonist binding induces a conformational change of [M.sub.2] receptors at sites adjacent to transmembrane segments, which interact with and activate GRK2 together with G protein [beta][gamma] subunits (see review ref. 93). Furthermore, G[alpha][beta][gamma] trimer might compete with GRK2 on these sites, which would explain the finding (95) that the agonist-dependent phosphorylation of muscarinic receptors is inhibited by a guanine nucleotide-free G[alpha][beta][gamma] trimer.
4.2. Phosphorylation-dependent internalization of muscarinic [M.sub.2] receptors. Muscarinic receptors as well as [beta] adrenergic receptors and other GPCRs are subject to agonist-induced desensitization, and agonist-dependent phosphorylation of muscarinic receptors is supposed to be involved in the desensitization. Particularly we have examined the agonist-induced internalization of [M.sub.2] and [M.sub.4] receptors with Kameyama in Tokyo University in the 1990s, and with Yoshida, Ichiyama, and Hashimoto in Gakushuin University in the 2000s. The internalization of muscarinic receptors could be easily determined as the decrease in [[sup.3]H]NMS binding activity of cultured cells as [[sup.3]H]NMS is a quaternary ion and does not penetrate the cell membrane. On the other hand, down regulation, that represents agonist-induced loss of muscarinic receptors from the cell, was determined as the agonist-induced decrease in the binding activity of [[sup.3]H]QNB, which is a tertiary amine, can penetrate the cell membrane, and binds with [M.sub.2] receptors both on the cell surface and in intracellular organelles. Actually, cultured cells expressing [M.sub.2] receptors were treated with an agonist for 0 to 16 hours at 37[degrees]C, washed with cold buffer, and incubated with [[sup.3]H]NMS or [[sup.3]H]QNB for 4hours at 4[degrees]C, and then bound [[sup.3]H]NMS or [[sup.3]H]QNB was determined: here it is presumed that at 4[degrees]C [M.sub.2] receptors may bind with [[sup.3]H]NMS or [[sup.3]H]QNB but may not be internalized from the cell surface nor be recycled back there.
Tsuga et al. (105) showed that coexpression of GRK2 together with [M.sub.2] receptors in COS7 cells causes dramatic increases in both the phosphorylation and internalization of [M.sub.2] receptors in an agonist-dependent manner. Furthermore, both internalization and down regulation of [M.sub.2] receptors were shown to be dependent on the presence of the I3 loop: when CHO cells expressing [M.sub.2] receptors or I3-deleted [M.sub.2] receptors are exposed to an agonist continuously, approximately 90% of the [M.sub.2] receptors were internalized with a half life of 10 min as was shown by decline of [[sup.3]H]NMS binding and 60% of them were down-regulated with a half life of 2.3 hours as was shown by decline of [[sup.3]H]QNB binding, whereas 60% of I3-deleted [M.sub.2] receptors were internalized and down-regulated with a half life of 8-10 hours (106) (Fig. 5(b)). These results indicate that internalization and down regulation of muscarinic receptors are governed by the I3 loop, and that the agonist-induced phosphorylation of the I3 loop by GRK2 is the initial and essential step (Fig. 6).
Agonist-dependent internalization is also observed for muscarinic receptors other than [M.sub.2] receptors. The extent and rate of internalization of muscarinic [M.sub.1]-[M.sub.5] receptors expressed in COS-7 and BHK-21 cells differed from one subtype to another. (107) In addition, the effect of co-expression of GRK2, dominant negative GRK2, GRK4, GRK5, GRK6, dynamin and dominant-negative dynamin was also different depending on the subtype species. These subtype-specific properties are assumed to be due to the different I3 loop sequences among the five subtypes. To test this assumption, we examined the internalization and recycling of the [M.sub.2]-[M.sub.4] chimera with replacement of the I3 loop by another subtype such as [M.sub.2]-[M.sub.4](I3)-[M.sub.2] and [M.sub.4]-[M.sub.2](I3)-[M.sub.4], which were expressed in HEK293 cells. The agonist-dependent internalization of the [M.sub.2]-[M.sub.4](I3)-[M.sub.2] chimera was shown to be greatly reduced on coexpression of the dominant negative dynamin as was the case for [M.sub.4] receptors, whereas the agonist-dependent internalization of the [M.sub.4]-[M.sub.2](I3)-[M.sub.4] chimera was hardly affected by coexpression of the dominant negative dynamin as was the case for [M.sub.2] receptors. In addition, [M.sub.2]-[M.sub.4](I3)-[M.sub.2] chimera and [M.sub.4] receptors were shown to be recycled back to the cell surface after removal of the agonist, whereas no recycling was observed for the [M.sub.4]-[M.sub.2](I3)-[M.sub.4] chimera and [M.sub.2] receptors. * These results demonstrate the major role of the I3 loop in the agonist-dependent internalization/recycling of muscarinic receptors. In contrast with these observations, the possibility remains that other parts of receptors might also be involved in the agonist-dependent internalization of [M.sub.4] receptors, because the internalization of I3-deleted [M.sub.4] receptors lacking major phosphorylation sites was found to be enhanced by coexpression of GRK2. ** It remains to be determined if GRK2 phosphorylates other sites of [M.sub.4] receptors besides the I3 loop, phosphorylates other proteins, or exerts its effect independently of the kinase activity.
We assumed that all five I3 loops have different conformations, each of which recognizes a different protein. To test this assumption, Ichiyama et al. (108) (2006) have expressed a fusion protein of [M.sub.2]-I3 (208-388) with glutathione S-transferase (GST-I3) in E. coli and purified it. The I3 part of GST-I3 was shown to have no secondary structure by circular dichroism (CD) and nuclear magnetic resonance (NMR) measurements. Furthermore, differential CD between wild type [M.sub.2] and I3-deleted [M.sub.2] receptors indicated that the I3 loop does not have any secondary structure even when it is in [M.sub.2] receptors. This finding is rather unexpected but is consistent with the finding that purified [M.sub.2] receptors are very susceptible to proteolysis at the I3 loop during storage at 4[degrees]C. This finding raises the question of how the I3 loop may exert its function.
Hashimoto et al. (109) (2008) attempted to localize the part that is involved in agonist-dependent internalization and recycling of [M.sub.4] receptors. Experiments involving the deletion of various portions of the [M.sub.4]-I3 loop indicated that a segment of 21 amino acid residues is necessary for the internalization and recycling of [M.sub.4] receptors. Furthermore, I3-deleted [M.sub.2] receptors, which could hardly be internalized in response to an agonist, were found to be internalized and then be recycled back to the cell surface in an agonist-dependent manner through insertion of the 21 segment of [M.sub.4] receptors. These results suggest that this segment is recognized by other proteins, and that these hypothetical proteins function to initiate internalization and recycling of receptors. A search for such proteins that specifically interact with the different I3 loops of the five muscarinic subtypes should be fruitful for molecular studies on regulation of muscarinic receptors.
4.3. Other aspects of muscarinic receptor regulations. In a series of experiments on the phosphorylation of muscarinic receptors by GRK2, GRK2 was found to be inhibited by calmodulin in a [Ca.sup.2+] ion-dependent manner. (110) In addition, GRK2 was found to bind with and phosphorylate tubulin, (111) although GRKs had been known to phosphorylate only activated forms of GPCRs. Yoshida et al. (112) (2003) identified the phosphorylation sites as Ser and Thr residues in the C-terminal domain of O-tubulin, which is located on the outermost surface of microtubules. Microtubule formation from tubulin in vitro was shown to be affected by approximately [10.sup.-5] M [Ca.sup.2+] ions, (113) which was later shown to be mediated through calmodulin. These preliminary results regarding the interactions among GRK2, [Ca.sup.2+]-calmodulin, and tubulin might indicate the possible involvement of GRK2 in intracellular signaling or cytoskeleton formation. It is interesting to note that a diverse function of GRK2 and other GRKs, besides their role in receptor desensitization, has been indicated recently. (114)
Muscarinic receptors may be modified after their translation. Muscarinic receptors were supposed to be glycosylated based on the observation that they are bound to lectin gels, (21) affected by endoglycosidase treatment, (29) and so on. Ohara et al. (30) (1990) showed that endoglycosydase F treatment of muscarinic receptors purified from porcine cerebrum caused a decrease in their molecular size from 70 to 50 kDa, as assessed as the mobility of [[sup.3]H]PrBCM-labeled band on SDS-PAGE. The endoglycosidase treatment, however, did not affect their ligand binding activity nor their ability to interact with G proteins, indicating that the glycosylation moiety is not involved in the primary functions of muscarinic receptors, although it may be involved in the intracellular translocation of muscarinic receptors.
Hayashi and Haga (115) expressed muscarinic [M.sub.2] receptors in Sf9 cells using baculovirus and have shown that they are labeled with [[sup.3]H]palmitic acid added to the incubation medium. The labeling rate was greatly accelerated by the presence of an agonist. Such labeling was not observed for a Cys458Ala mutant, indicating that Cys 458 in the C-terminal tail of [M.sub.2] receptors is palmitoylated. The Cys458Ala mutant, which was expressed in Sf9 cells and purified from them, may interact with and activate Gi2- and Go-type G proteins, although the rate of the agonist-dependent increase in [[sup.35]S]GTP [gamma]S binding was slower as compared to the corresponding rate for wild-type [M.sub.2] receptors. These results indicate that the palmitoylation of Cys458 in [M.sub.2] receptors is not necessary for their function but facilitates their interaction with G proteins.
Mieda, (116,117) together with Saffen, studied the regulation of [M.sub.4] gene expression and identified regulatory functions of the 5'-flanking region in the [M.sub.4] receptor gene: the proximal 435-bp sequence contains a constitutive promoter and the 635-bp sequence a cell-type-specific silencer element. Furthermore, the neuron-restrictive silencer/repressor element (NRSE/RE1) upstream of the [M.sub.4] gene was shown to be necessary and sufficient to repress [M.sub.4] promoter activity in non-neuronal cells. The activity of neuronal-restrictive silencer factor/RE1-silencing transcription factor (NRSF/REST) was suggested to depend upon the species of promoter to which it is linked and upon the protein species that bind to this promoter.
5. Structure of muscarinic receptors
5.1. A large scale preparation of purified QNB-[M.sub.2] complex. In the late 1990s, we started to attempt to crystalize [M.sub.2] receptors and to determine their tertiary structures. We chose a [M.sub.2] mutant with a deletion in the central part of I3 because the I3 loop had been shown to be very susceptible to proteolysis and to be broken down during the purification procedure. In addition, asparagine residues in three N-glycosylation consensus sequences in the N-terminal portion were mutated to aspartic acid residues in order to avoid glycosylation, which might cause heterogeneity of [M.sub.2] receptors. We expressed the [M.sub.2] mutant with N2, 3, 6, 9D and a deletion of 235-380 in the I3 loop in cultured insect cells (Sf9) by using the baculovirus system, (118) and purified it by using ABT-agarose, and demonstrated that it has the ability to interact with and activate Gi in agonist-dependent manners. For approximately 10 years since the late 1990s, we have continued large-scale expression of the [M.sub.2] mutant, which will be called here as [M.sub.2] or the [M.sub.2] receptor in this text. The [M.sub.2] receptor was expressed in Sf9 cells, with a scale of 20-40 l of cultured solution per month, which express 1-2mg of [M.sub.2] receptors per l liter of medium, in Wakenyaku Co. and Toyobo Co. initially, and then in Gakushuin University, where Okada, Hayashi, Matsuyama, Watanabe and others contributed. Thus, we expressed 20-80 mg of [M.sub.2] receptors per month and several grams of them in total. Almost all of these [M.sub.2] receptors were used for attempts at crystallization.
In attempt to crystalize [M.sub.2] receptors, one of the most critical problems was to find an appropriate detergent. As described earlier, muscarinic receptors had been known to be solubilized in their active form by digitonin, but the purified [M.sub.2] receptors were found to be bound with more than 10 times as much digitonin as [M.sub.2] receptor proteins. [M.sub.2] receptors lose their ligand-binding activity in most detergents other than digitonin, although some detergents including lysophosphatidyl choline, (119) and dodecyl maltoside (120) were reported to be used for the solubilization of muscarinic receptors with their ligand binding activity. Rinken et al. (121) showed that atrial muscarinic receptors could be solubilized by sucrose monolaurate and partially purified in the detergent. We also tried to solubilize [M.sub.2] receptors in dodecyl maltoside and other alkyl sugars, and to purify them in these detergents, but found that the ligand binding activity was gradually lost during the purification procedure, probably because of the instability of [M.sub.2] receptors in these detergents.
We also noticed that muscarinic receptors could be solubilized in their active forms when porcine cerebral membranes were pre-treated with a muscarinic ligand such as carbamylcholine or atropine, and then solubilized with sodium cholate in the presence of this ligand. (122) [[sup.3]H]QNB, which was pre-bound to muscarinic receptors in membrane preparations, appeared to remain bound to receptors after solubilization in different kinds of detergents. Based on these preliminary results, we adopted the following strategy to prepare muscarinic receptors that could be used for their crystallization. (1) [M.sub.2] receptors are solubilized in the absence of a muscarinic ligand with digitonin, (2) solubilized [M.sub.2] receptors are bound to an ABT-agarose column in digitonin and eluted from the column with atropine, (3) eluted [M.sub.2] receptors are bound to a small column of hydroxyapatite (HA), (4) the HA column are washed and incubated with QNB in digitonin, and washed with a Na cholate solution and then with any chosen detergent, (5) the QNB-[M.sub.2] complex is eluted with a high salt solution in a given detergent, concentrated, and then dialyzed against any chosen medium, and (6) the concentrated QNB-[M.sub.2] complex is subjected to crystallization. Here it was essential to have [M.sub.2] receptors bound with muscarinic ligands such as ABT, atropine, and QNB in digitonin, and then to let digitonin dissociate from [M.sub.2] receptors by washing the HA column with Na cholate before the detergent changes to an alkyl sugar. Digitonin is bound with receptors so tightly that it can not be replaced by detergents like alkyl sugars.
In a typical large scale purification, [M.sub.2] receptors were solubilized with 1% digitonin/0.2% Na cholate from Sf9 cell pellets derived from 20 l of cultured solution expressing 800 nmol (30 mg) of [M.sub.2] receptors, the extract was applied to two columns of ABT-agarose (500 ml each), the eluate with atropine from the ABT columns was led to a HA column (30 ml), and then the QNB-[M.sub.2] receptors were eluted with decyl maltoside from the HA column. The recovery of purified QNB-[M.sub.2] receptors was estimated to be approximately 50% of solubilized receptors by using [[sup.3]H]QNB instead of non-labeled QNB. We also observed that [[sup.3]H]QNB was very tightly bound to [M.sub.2] receptors and dissociation of [[sup.3]H]QNB was hardly observed at 4[degrees]C, indicating that the QNB-[M.sub.2] complex may be stable at 4[degrees]C.
5.2. Crystallization of QNB-[M.sub.2] complex. In 1999, we established a system for large scale purification of [M.sub.2] receptors and started attempt to crystalize them. We consulted X ray crstallography specialists such as Nakasako in Keio University, Yoshikawa and Shinzawa-Itoh in Hyogo Prefectural University, Tsukihara in Osaka University, Harada in Tokyo University and others. We adopted a crystal screen kit (Hampton Research) and attempted to crystalize several kinds of [M.sub.2] preparations purified in different detergents. In 2000, the crystal structure of rhodopsin was reported, (123) which was the first structure determined for a member of the GPCRs. Rhodopsin was shown to have seven-transmembrane segments, as was expected from amino acid sequences, although no direct evidence had been obtained until then. We consulted with and obtained advice from Okada a major contributor to determination of the rhodopsin structure, and tried to crystalize [M.sub.2] receptors under the same or similar conditions as those used for rhodopsin crystallization, that involves ammonium sulfate as a precipitant. However, we could not detect any sign of crystallization of [M.sub.2] receptors under such conditions.
Then we returned to the Hampton kits and examined 500 conditions of the kits for several preparations (more than 2000 conditions in total) by means of the vapor diffusion method. Most attempts were carried out in the cold room around 4[degrees]C. In 2001, we could obtain the first needle-shaped crystal with a few diffraction points, although the resolution was approximately 32 [Angstrom]. The resolution of diffraction points was improved to 18 [Angstrom] in a few months. One year later, we could obtain a rod-shaped crystal with an orthorhombic-type unit cell (111 x 69 x 109 [Angstrom]) by using decylmaltoside as detergent and low molecular weight polyethyleneglycol as the precipitant. The resolution of the diffraction points was 9 [Angstrom] (Fig. 7). The improvement in one year had been so great that we expected to obtain crystals good enough to be used for structure determination in a few years. However, no improvement in the quality of crystals was attained in the next five years, although persistent endeavors were made using a lot of [M.sub.2] receptor preparations and various kinds of crystallization conditions, which included the use of different kinds of alkyl sugars, polyethyleneglycols, the addition or no addition of lipids, various kinds and concentrations of salts and buffers, and various incubation temperatures.
In 2007, Kobilka et al. reported the structure of [[beta].sub.2] adrenergic receptors. (124) They used two unique techniques to facilitate the crystallization of [[beta].sub.2] adrenergic receptors: (1) introduction of lysozyme of bacteriophage T4 (T4L) into the I3 loop of [[beta].sub.2] adrenergic receptors and (2) adoption of the lipidic cubic phase method. T4L had been known to be a well-folded protein and was introduced to increase the hydrophilic regions, which was expected to facilitate inter-molecular interaction. The lipidic cubic phase method was developed by Rosenbusch and Landau, (125) and was used to crystalize bacteriorhodopsin. With the lipidic cubic phase method, membrane proteins are designed to be crystallized in the lipidic phase consisting of oleic acid instead of the aqueous phase. (126) We had also considered adopting the lipidic cubic phase method, and visited and consulted with Rosenbusch in Basel in the late 1990s. However, we did not adopt the method because we could obtain crystals in 2001-2002 with the usual vapor diffusion method and decided to improve the quality of the crystals by means of the same vapor diffusion method.
We had started collaboration with the group of Iwata in Kyoto University, when the structure of O2 adrenergic receptors was reported. We decided to introduce T4L in the I3 loop of [M.sub.2] receptors following the success for [[beta].sub.2] adrenergic receptors. Kobayashi in Iwata's laboratory constructed the cDNA of [M.sub.2]-T4L, and we expressed [M.sub.2]-T4L using the Sf9-baculovirus system and tried to crystalize it at Gakushuin University. Soon we could obtain crystals of QNB-[M.sub.2]-T4L with the vapor diffusion method, but the resolution was not good enough to be used for structure determination. We continued to attempt to improve the quality of the crystals for a few more years but were not successful. Then, we reconsidered to adopting the lipidic cubic phase method, discussed this with Kobayashi and Kobilka by e-mail, and decided to send the purified QNB-[M.sub.2] T4l protein to the laboratory of Kobilka in Stanford University for trial crystallization. We had expected a long series of experiments after the initial survey in Kobilka's laboratory when we sent the sample of purified QNB-[M.sub.2]-T4L in January 2011. To our surprise, a month later we were informed that Kruse succeeded in crystalizing [M.sub.2]-T4L in the first preliminary trial and that the quality of the crystals was so good that the structure would be determined soon. Actually in February 2011, we could see the 3 [Angstrom] structure of the QNB-[M.sub.2]-T4L complex. (127) The combination of T4L and the lipidic cubic phase has been shown to be useful for determination of the structures of many GPCRs including muscarinic [M.sub.3] receptors, (128) opioid receptors, (129) and others.
5.3. The overall structure of [M.sub.2] receptor. As was expected from the amino acid sequence, the structure of [M.sub.2] receptors was demonstrated to include seven transmembrane segments (127) (Fig. 8). Another a helix (Helix 8) was also detected in the carboxyterminal tail. The overall transmembrane structure of [M.sub.2] receptors including Helix 8 is very similar to those of rhodopsin (123) and [[beta].sub.2] adrenergic receptors 124) (Fig. 8). As was expected from biochemical studies, (31,32) a conserved S--S bond between Cys 96 and Cys 173 was detected between the N terminus of T[M.sub.3] and extracellular loop 2 (E2). In addition, another S--S bond between Cys413 and Cys416 was detected in E3 loop. Palmitoylation in the C-terminal tail had been expected to be present from biochemical studies (115) but the palmitoyl residue was not detected in the present structure probably because of the low resolution. In contrast with transmembrane segments, the extracellular and intracellular loops appear to have a specific structure distinct from those of rhodopsin and [[beta].sub.2] adrenergic receptors. QNB bound to [M.sub.2] receptors was found to reside approximately one third-inside from the extracellular phase, to be surrounded by transmem brane segments 2 to 7 (T[M.sub.2] to TM7), and not to be exposed to either the extracellular or intracellular phase. The position of QNB is very similar to that of retinal in rhosopsin and of carazolol ([beta] adrenergic antagonist) in [[beta].sub.2] adrenergic receptors (Fig. 8, see atomic models of ligands colored white).
5.4. Orthosteric binding site of [M.sub.2] receptors. QNB is surrounded by 18 amino acid residues, at least one atom of which is within 4 [Angstrom] from one atom in QNB (Fig. 9). Among them, QNB interacts with Asp103 in T[M.sub.3] electrostatically, with Asp405 in TM6 through hydrogen bonds, and with many Tyr, Trp and Phe residues through van-der-Waals interactions. These multiple interactions explain the extremely high affinity for QNB of [M.sub.2] receptors. Among these 18 residues surrounding QNB, 17 are common to all [M.sub.1] to [M.sub.5] receptors, which is consistent with the previous finding that five muscarinic receptors have similar affinities for QNB. QNB is supposed to bind to the orthosteric site of muscarinic receptors, that is defined as the ACh binding site. Thus the present finding is consistent with and explains the previous results that most orthosteric ligands show similar affinities for the five muscarinic receptors (6) and that it is difficult to develop subtype-specific orthosteric ligands.
Most amino acid residues surrounding QNB were suggested to be involved in the ligand binding by site-directed mutagenesis studies, which had been carried out on [M.sub.1], (130,131) [M.sub.2]132) and [M.sub.3] (133) receptors. This consistence appears to give credit to that the determined structure would represent the active structure of [M.sub.2] receptors, at least as for antagonist binding, and that mutagenesis studies would provide useful information on the ligand binding structure. One of merit of structure determination is that it provides direct information on the role of each residue. For example, the mutation of Asp103 in [M.sub.2] receptors into Ala caused a decrease in the affinity for QNB of more than 100-fold, whereas the mutation of Asp103 into Glu hardly affected the affinity for QNB. 132) This finding apparently indicates the involvement of an electrostatic interaction between Asp103 and QNB. Actually, two oxygen atoms on the carboxyl group of Asp103 were found to be located 3.1 and 3.9 [Angstrom] from the nitrogen atom of the quaternary amine of QNB in the crystal structure of the QNB-[M.sub.2] complex. (127) Furthermore, the mutation of Asn405 to Ala in [M.sub.2] receptors was shown to decrease the affinity for QNB more than 100-fold, (132) but it remained unknown whether this effect is due to the loss of direct interaction between Asn405 and QNB or to a conformational change of [M.sub.2] receptors induced by the mutation. In the crystal structure, Asn405 was shown to form hydrogen bonds with QNB, which is supposed to contribute to the high affinity of QNB. (127)
5.5. Allosteric binding sites of [M.sub.2] receptors. Some muscarinic ligands have been reported to interact with muscarinic receptors but not to compete with an orthosteric ligand such as [[sup.3]H]NMS or ACh. Gallamine (134) and alcuronium (135) are original ligands with such properties: gallamine suppresses the [[sup.3]H]NMS binding of cardiac muscarinic receptors ([M.sub.2] receptors) without competing with [[sup.3]H]NMS, whereas alcuronium enhances the [[sup.3]H]NMS binding. They are presumed to interact with sites distinct from the orthosteric site and are called allosteric ligands, and their binding sites are called allosteric sites. Allosteric ligands, which enhance or suppress the binding of orthosteric ligands positive-or negative-cooperatively, are now called as positive allosteric modulators or negative allosteric modulators respectively. Guo et al. (136) found that lapachone as well as alcuronium acts on [M.sub.2] receptors subtype-specifically and enhances the [[sup.3]H]NMS binding. Another type of allosteric ligand is the allosteric activator, which activates the receptor without being affected by orthosteric antagonist. An allosteric activator TBPB (1-(1'-2-methylbenzyl)-1,4'-bipi peridin-4-yl)-1H-benzo[d]imidazole-2(3H)-one) is reported to be [M.sub.1]-specific and not to affect [M.sub.2]-[M.sub.5] receptors. (137) It is reasonable to assume that allosteric modulators and allosteric activators are subtype-specific because allosteric sites are not the sites of physiological ligands and thus are not conserved, in contrast with the orthosteric site.
Mutagenesis studies of [M.sub.2] receptors indicated that the allosteric sites include amino acid residues in the E2 loop and the aminoterminal end of TM7. (138-140) The allosteric sites were located between the QNB binding site and the extracellular phase in the structure of [M.sub.2] receptors. (127) The amino acid sequences of the allosteric sites are relatively different among the five muscarinic receptors as compared with their orthosteric sites. It is unexpected that the putative allosteric sites of [M.sub.3] receptors have a similar overall structure as those of [M.sub.2] receptors in spite of sequence differences. (128) The same structure of the E2 loop might be required for maintenance of the transmembrane structure. In any cases, the structures of allosteric sites may provide information important for developing ligands acting on the allosteric sites.
6. Future problems
6.1. Interaction of ACh and muscarinic receptors. ACh is known to take on two conformations with a trans or gauche C[alpha]-C[beta] bond. Dose-response analysis involving diasteromers of conformationally rigid ACh analogues has indicated that the trans form of ACh is the active form for muscarinic receptors. (141) Then, we assumed that ACh bound to [M.sub.2] receptors would take on the transform and attempted to confirm this when sufficient amounts of purified receptors became available. Furukawa et al. (142) prepared purified [M.sub.2] receptors bound to methacholine (a-methyl acetylcholine), and determined the C[alpha]-C[beta] bond of methacholine bound to [M.sub.2] receptors by measuring the transferred nuclear Overhauser effect (TRNOE). Unexpectedly, the angle was estimated to be 60[degrees], which indicates the gauche form rather than the trans one. We interpreted these results as indicating that ACh binds with [M.sub.2] receptors in its gauche form and undergoes a conformational change from the gauche to the trans form when ACh-bound [M.sub.2] is activated and bound with G proteins (Fig. 10 (a)). [M.sub.2] receptors are known to show high affinity for agonists when reconstituted with G proteins in the absence of guanine nucleotides, whereas they show low affinity for agonists in the absence of G proteins (74,77) (Fig. 3). Thus, we assume that the low and high affinity ACh binding of [M.sub.2] receptors represent the ACh (gauche)--[M.sub.2] and ACh (trans)--[M.sub.2]-Gi interactions, respectively. It should be noted that the ACh-[M.sub.2]-Gi complex represents the transition state for the reaction of [M.sub.2] receptors and Gi. This assumption, however, has not been proved yet, but it has not been disproved either. It remains to be determined by TRNOE measurement if the C[alpha]-C[beta] bond angle of ACh bound to [M.sub.2]-Gi complex is trans or gauche.
We tentatively put ACh instead of QNB in the crystal structure of the QNB-[M.sub.2] complex, finding that the space occupied by QNB is too large to be occupied by ACh, and we could not determine which form of ACh, trans or gauche, fits the [M.sub.2] structure (Fig. 10 (b)). The conformational change in muscarinic receptors during their activation has been discussed, (143) based on the crystal structure of agonist-[[beta].sub.2] adrenergic receptor-Gs complex, (144) but no information nor discussion has been presented so far regarding a possible conformational change of muscarinic or adrenergic agonists during activation of their receptors. Determination of the crystal structure of the ACh-[M.sub.2]-Gi complex could give a crucial answer for or against the assumption that the conformational change of [M.sub.2] receptors is accompanied by the conformational change of ACh.
6.2. Positive allosteric modulators as possible therapeutic agents. Muscarinic receptors are assumed to be targets of drugs for various diseases including Alzheimer disease, schizophrenia, Parkinson disease, and peptic ulcer. (145) These drugs are expected to be subtype-specific in order to avoid side effects due to their actions on other subtypes. Several subtype-specific orthosteric ligands have been developed, but their specificity for one subtype is limited mostly because of the conservation of related amino acids among the five receptors. (6) On the other hand, subtype-specific allosteric ligands have been developed (see recent reviews refs. 146-148). Particularly, [M.sub.1]-and [M.sub.4]-specific positive allosteric modulators have been extensively developed for treatment of Alzheimer's disease and schizophrenia, respectively. These ligands are expected to enhance the activity of endogenous ACh by increasing its potency for a specific muscarinic receptor subtype. The merits of these allosteric modulators as compared with orthosteric agonists are that (1) their effects are subtype-specific and (2) they enhance only the receptors bound with ACh. Thus, they are expected to enhance specific muscarinic receptors when and where they are activated by ACh released under physiological conditions.
The structures of allosteric sites have been reported for [M.sub.2] and [M.sub.3] receptors, and are expected to be reported for [M.sub.1],[M.sub.4] and [M.sub.5] receptors in the near future. We look forward to developing positive allosteric modulators for Alzheimer's disease, schizophrenia, and other diseases based on these structures.
Abbreviations: ABT: aminobenztropine (muscarinic antagonist used for affinity chromatography); ACh: acetylcholine; cDNA: complementary DNA; DTT: dithiothreitol; E2 loop: the second extracellular loop; I3 loop: the third intracellular loop; [M.sub.1] ([M.sub.2], [M.sub.3], [M.sub.4], [M.sub.5]) receptors: muscarinic acetylcholine receptor [M.sub.1] ([M.sub.2], [M.sub.3], [M.sub.4], [M.sub.5]) subtypes; NMS: N-methylscopolamine (muscarinic antagonist); PrBCM: propylbenzylcholine mustard (irreversible muscarinic antagonist); QNB: quinuclidinyl benzylate (muscarinic antagonist); SDS-PAGE: sodium dodecyl sulfatepolyacrylamide gel electrophoresis; T[M.sub.1] (2-7): transmembrane segment 1 (2-7).
The author expresses his sincere thanks to referees for reading this manuscript and giving a number of useful comments, and to the colleagues who worked on muscarinic receptors, Prof. Ichiyama, A., and Drs. Nukada, T., Kajimura, M., Nishiyama, T., Uchiyama, H., Berstein, G., Ohara, K. and Ikegaya, T. in Hamamatsu University School of Medicine; Drs. Shiozaki, K., Kameyama, K., Nukada, T., Saffen, D.W., Nakamura, F., Nakata, H., Rinken, A., Hayashi, K.M., Tsuga, H., Guo, Z.D., Mieda, M., Fukuzaki, A., Furukawa, H., Jakubik, J. and Guo, Q. in Tokyo University; Drs. Takeda, S., Ichiyama, A. and Hashimoto, Y. in Gakushuin University; Prof. Burgen, A.S.V., Drs. Birdsall, N.J.M. and Hulme, E.C. in National Institute for Medical Research in London; Profs. Numa, S., Iwata, S., Drs. Kubo, T. and Kobayashi, T. in Kyoto University; Prof. Matsuo, H. and Dr. Kangawa, K. in Miyazaki University School of Medicine; Prof. Ui, M., Drs. Katada, T. and Kurose, H. in Hokkaido University; Prof. Kobilka, B. and Dr. Kruse, A.C. in Stanford University. And last but not least the author greatly appreciates the collaboration of Dr. Kazuko Haga for more then 30 years.
(1) Dale, H.H. (1914) The action of certain esters and ethers of choline, and their relation to muscarine. J. Pharmacol. Exp. Ther. 6, 147-190.
(2) Dale, H. (1935) Pharmacology and nerve-endings. Proc. R. Soc. Med. 28, 319-332.
(3) Nieuwenhuys, R. (1985) Chemoarchitecture of the Brain, Springer-Verlag, Berlin Heidelberg, pp. 7-11.
(4) Yamamura, H.I. and Snyder, S.H. (1974) Muscar inic cholinergic binding in rat brain. Proc. Natl. Acad. Sci. U.S.A. 171, 1725-1729.
(5) Birdsall, N.J.M. and Hulme, E.C. (1976) Biochem ical studies on muscarinic acetylcholine receptors. J. Neurochem. 27,7-16.
(6) Alexander, S.P.H., Mathie, A. and Peters, J.A. (2008) Guide to Receptors and Channels (GRAC), 3rd ed. Br. J. Pharmacol. 153 (Suppl. 2), S9-S10.
(7) Yamamura, H.I., Kuhar, M.J., Greenberg, D. and Snyder, S.H. (1974) Muscarinic cholinergic receptor binding: regional distribution in monkey brain. Brain Res. 66, 541-546.
(8) Haga, T. and Noda, H. (1973) Choline uptake systems of rat brain synaptosomes. Biochim. Biophys. Acta 291, 564-575.
(9) Kuhar, M.J., Sethy, V.H., Roth, R.H. and Aghajanian, G.K. (1973) Choline: selective accumulation by central cholinergic neurons. J. Neurochem. 20, 581-593.
(10) Hammer, R., Berrie, C.P., Birdsall, N.J.M., Burgen, A.S.V. and Hulme, E.C. (1980) Pirenzepine distinguishes between different subclasses of muscarinic receptors. Nature 283,90-92.
(11) Birdsall, N.J.M. and Hulme, E.C. (1983) Muscarinic receptor subclasses. Trends Pharmacol. Sci. 4, 459-463.
(12) Birdsall, N.J.M., Burgen, A.S.V. and Hulme, E.C. (1979) The binding of agonists to brain muscarinic receptors. Mol. Pharmacol. 14, 723-736.
(13) Beld, A.J. and Ariens, E.J. (1974) Stereospecific binding as a tool in attempts to localize and isolate muscarinic receptors. Part II. Binding of (D)-benzetimide, (-)-benzetimide and atropine to a fraction from bovine tracheal smooth muscle and to bovine caudate nucleus. Eur. J. Pharmacol. 25, 203-209.
(14) Caron, M.G. and Lefkowitz, R.J. (1976) Solubilization and characterization of the beta-adrenergic receptor binding sites of frog erythrocytes. J. Biol. Chem. 251, 2374-2384.
(15) Berrie, C.P., Birdsall, N.J.M., Haga, K., Haga, T. and Hulme, E.C. (1984) Hydrodynamic properties of muscarinic acetylcholine receptors solubilized from rat forebrain. Br. J. Pharmacol. 82, 839-851.
(16) Haga, T. (1980) Molecular size of muscarinic acetylcholine receptors of rat brain. FEBS Lett. 113,68-72.
(17) Birdsall, N.J.M., Burgen, A.S.V. and Hulme, E.C. (1979) A study of the muscarinic receptor by gel electrophoresis. Br. J. Pharmacol. 66, 337-342.
(18) Haga, K. and Haga, T. (1984) Affinity chromatography of muscarinic acetylcholine receptors. Seitai no Kagaku 35, 296-303 (in Japanese).
(19) Haga, K. and Haga, T. (1983) Affinity chromatography of the muscarinic acetylcholine receptor. J. Biol. Chem. 258, 13575-13579.
(20) Haga, K. and Haga, T. (1985) Purification of the muscarinic acetylcholine receptor from porcine brain. J. Biol. Chem. 260, 7927-7935.
(21) Peterson, G.L., Herron, G.S., Yamaki, M., Fullerton, D.S. and Schimerlik, M.I. (1984) Purification of the muscarinic acetylcholine receptor from porcine atria. Proc. Natl. Acad. Sci. U.S.A. 81, 4993-4997.
(22) Birdsall, N.J.M., Hulme, E.C. and Keen, M. (1986) The binding of pirenzepine to digitonin-solubilized muscarinic acetylcholine receptors from the rat myocardium. Br. J. Pharm. 87, 307-316.
(23) Nishiyama, T., Berstein, G., Ikegaya, T., Haga, T., Ichiyama, A., Kobayashi, A. and Yamazaki, N. (1989) Comparison between purified cerebral and atrial muscarinic acetylcholine receptors: Pirenzepine binding and the effect of sulfhydryl reagents. Biomed. Res. 10, 251-260.
(24) Berstein, G., Haga, T. and Ichiyama, A. (1989) Effect of the lipid environment on the differential affinity of purified cerebral and atrial muscarinic acetylcholine receptors for pirenzepine. Mol. Pharmacol. 36, 601-607.
(25) Berstein, G., Haga, K., Haga, T. and Ichiyama, A. (1988) Agonist and antagonist binding of muscarinic acetylcholine receptors purified from porcine brain: interconversion of high- and low-affinity sites by sulfhydryl reagents. J. Neurochem. 50, 1687-1694.
(26) Kubo, T., Fukuda, K., Mikami, A., Maeda, A., Takahashi, H., Mishina, M., Haga, T., Haga, K., Ichiyama, A., Kangawa, K., Kojima, M., Matsuo, H., Hirose, T. and Numa, S. (1986) Cloning, sequencing and expression of complementary DNA encoding the muscarinic acetylcholine receptor. Nature 323, 411-416.
(27) Kubo, T., Maeda, A., Sugimoto, K., Akiba, I., Fukuda, K., Mikami, A., Takahashi, H., Haga, T., Haga, K., Ichiyama, A., Kangawa, K., Matsuo, H., Hirose, T. and Numa, S. (1986) Primary structure of porcine cardiac muscarinic acetylcholine receptor deduced from the cDNA sequence. FEBS Lett. 209, 367-372.
(28) Peralta, E.G., Winslow, J.W., Peterson, G.L., Smith, D.H., Ashkenazi, A., Ramachandran, J., Schimerlik, M.I. and Capon, D.J. (1987) Primary structure and biochemical properties of an [M.sub.2] muscarinic receptor. Science 236, 600-605.
(29) Liang, M., Martin, M.W. and Harden, T.K. (1987) [[sup.3]H]Propylbenzilylcholine mustard-labeling of muscarinic cholinergic receptors that selectively couple to phospholipase C or adenylate cyclase in two cultured cell lines. Mol. Pharmacol. 32, 443-449.
(30) Ohara, K., Uchiyama, H., Ohara, K., Haga, T. and Ichiyama, A. (1990) Interaction of deglycosylated muscarinic receptors with ligands and G proteins. Eur. J. Pharmacol. 189, 341-346.
(31) Uchiyama, H., Ohara, K., Haga, K., Haga, T. and Ichiyama, A. (1990) Location in muscarinic acetylcholine receptors of sites for propylbenzilylcholine mustard binding and for phosphorylation with protein kinase C. J. Neurochem. 54, 1870-1881.
(32) Kurtenbach, E., Curtis, C.A.M., Pedder, E.K., Aitken, A., Harris, A.C.M. and Hulme, E.C. (1990) Muscarinic acetylcholine receptors: peptide sequencing identifies residues involved in antagonist binding and disulfide bond formation. J. Biol. Chem. 264, 489-495.
(33) Peralta, E.G., Ashkenazi, A., Winslow, J.W., Smith, D.H., Ramachandran, J. and Capon, D.J. (1987) Distinct primary structures, ligand-binding properties and tissue-specific expression of four human muscarinic acetylcholine receptors. EMBO J. 6, 3923-3929.
(34) Bonner, T.I. (1989) The molecular basis of muscarinic receptor diversity. Trends Neurosci. 12, 148-151.
(35) Berstein, G. and Haga, T. (1990) Molecular aspects of muscarinic receptors. In Current aspects of the neurosciences (ed. Osborne, N.N.), Macmillan Press, London, pp. 245-284.
(36) Kajimura, M., Haga, T., Ichiyama, A., Kaneko, E. and Honda, N. (1990) Carbachol-induced potentiation and inhibition of acid secretion by guinea pig gastric gland. Eur. J. Pharmacol. 178,59-69.
(37) Ovchinnikov, YuA. (1982) Rhodopsin and bacter iorhodopsin: structure-function relationships. FEBS Lett. 148, 179-191.
(38) Dixon, R.A., Kobilka, B.K., Strader, D.J., Benovic, J.L., Dohlman, H.G., Frielle, T., Bolanowski, M.A., Bennett, C.D., Rands, E., Diehl, R.E., Mumford, R.A., Slater, E.E., Sigal, I.S., Caron, M.G., Lefkowitz, R.J. and Strader, C.D. (1986) Cloning of the gene and cDNA for mammalian beta-adrenergic receptor and homology with rhodopsin. Nature 321,75-79.
(39) Takeda, S., Kadowaki, S., Haga, T., Takaesu, H. and Mitaku, S. (2002) Identification of G protein-coupled receptor genes from the human genome sequence. FEBS Lett. 520,97-101.
(40) Shukla, A.K., Xiao, K. and Lefkowitz, R.J. (2011) Emerging paradigms of [beta]-arrestin-dependent seven transmembrane receptor signaling. Trends Biochem. Sci. 36, 457-469.
(41) Rall, T.W. and Sutherland, E.W. (1958) Formation of a cyclic adenosine ribonucleotide by tissue particles. J. Biol. Chem. 232, 1065-1076.
(42) Rodbell, M., Birnbaumer, L., Posh, S.L. and Krans, H.M. (1971) The glucagon-sensitive adenylylcyclase system in plasma membranes of rat liver. V. An obligatory role of guanylnucleotides in glucagon action. J. Biol. Chem. 246, 1877-1882.
(43) Maguire, M.E., Van Arsdale, P.M. and Gilman, A.G. (1976) An agonist-specificeffect of guanine nucleotides on binding to the beta adrenergic receptor. Mol. Pharmacol. 12, 335-339.
(44) Chinkers, M., Garbers, D., Chang, M.-S., Lowe, D.G., Chin, H., Goeddel, D.V. and Schultz, S. (1989) A membrane form of guanylate cyclase is an atrial natriuretic peptide receptor. Nature 338, 78-83.
(45) Haga, T., Haga, K. and Gilman, A.G. (1977) Hydrodynamic properties of the beta-adrenergic receptor and adenylate cyclase from wild type and varient S49 lymphoma cells. J. Biol. Chem. 252, 5776-5782.
(46) Pfeuffer, T. (1977) GTP-binding proteins in mem branes and the control of adenylate cyclase activity. J. Biol. Chem. 252, 7224-7234.
(47) Cassel, D. and Selinger, Z. (1977) Mechanism of adenylate cyclase activation by cholera toxin: Inhibition of GTP hydrolysis at the regulatory site. Proc. Natl. Acad. Sci. U.S.A. 74, 3307-3311.
(48) Cassel, D. and Pfeuffer, T. (1978) Mechanism of cholera toxin action: Covalent modification of the guanyl nucleotide-binding protein of the adenylate cyclase system. Proc. Natl. Acad. Sci. U.S.A. 75, 2669-2773.
(49) Bourne, H.R., Coffino, P. and Tomkins, G.M. (1975) Selection of a variant lymphoma cell deficient in adenylate cyclase. Science 187, 750-752.
(50) Haga, T., Ross, E.M., Anderson, H.J. and Gilman, A.G. (1977) Adenylate cyclase permanently uncoupled from hormone receptors in a novel variant of S49 mouse lymphoma cells. Proc. Natl. Acad. Sci. U.S.A. 74, 2016-2020.
(51) Ross, E.M. and Gilman, A.G. (1977) Reconstitution of catecholamine-sensitive adenylate cyclase activity: Interactions of solubilized components with receptor-replete membranes. Proc. Natl. Acad. Sci. U.S.A. 74, 3715-3719.
(52) Ross, E.M. and Gilman, A.G. (1980) Biochemical properties of hormone sensitive adenylate cyclase. Annu. Rev. Biochem. 49, 533-546.
(53) Vauquelin, G., Geynert, P., Hanoune, J. and Strosberg, A.D. (1977) Isolation of adenylate cyclase-free, beta-adrenergic receptor from turkey erythrocyte membranes by affinity chromatography. Proc. Natl. Acad. Sci. U.S.A. 74, 3710-3714.
(54) Pederson, S.E. and Ross, E.M. (1982) Functional reconstitution of beta-adrenergic receptors and the stimulatory GTP-binding protein of adenylate cyclase. Proc. Natl. Acad. Sci. U.S.A. 79, 7228-7232.
(55) Murad, F., Chi, Y.-M., Rall, T.W. and Sutherland, E.W. (1962) Adenyl cyclase III. The effect of catecholamines and choline esters on the formation of adenosine 3',5'-phosphate by preparations from cardiac muscle and liver. J. Biol. Chem. 237, 1233-1238.
(56) Yamamura, H., Lad, P.M. and Rodbell, M. (1977) GTP stimulates and inhibits adenylate cyclase in fat cell membrnes through distinct regulatory processes. J. Biol. Chem. 252, 7964-7966.
(57) Rodbell, M. (1980) The role of hormone receptors and GTP-regulatory proteins in membrane trans duction. Nature 284,17-22.
(58) Katada, T. and Ui, M. (1982) Direct modification of the membrane adenylate cyclase system by islet-activating protein due to ADP-ribosylation of a membrane protein. Proc. Natl. Acad. Sci. U.S.A. 79, 3129-3133.
(59) Kurose, H., Katada, T., Amano, T. and Ui, M. (1983) Specific uncoupling by islet-activating protein, pertussis toxin, of negative signal transduction via alpha-adrenergic, cholinergic, and opiate receptors in neuroblastoma x glioma hybrid cells. J. Biol. Chem. 258, 4870-4875.
(60) Bokoch, G.M., Katada, T., Northup, J.K., Hewlett, E.L. and Gilman, A.G. (1983) Identification of the predominant substrate for ADP-ribosylation by islet activating protein. J. Biol. Chem. 258, 2072-2075.
(61) Sternweis, P.C. and Robishaw, J.D. (1984) Isolation of two proteins with high affinity for guanine nucleotides from membranes of bovine brain. J. Biol. Chem. 259, 13806-13813.
(62) Stryer, L., Hurley, J.B. and Fung, B.K.-K. (1981) Transducin: an amplifier protein in vision. Trends Biochem. Sci. 6, 245-247.
(63) Tang, W.J. and Gilman, A.G. (1991) Type-specific regulation of adenylate cyclase by G protein O. subunits. Science 254, 1500-1503.
(64) Logothesis, D.E., Kurachi, Y., Galper, J., Neer, E.J. and Clapham, D.E. (1987) The O. subunits of GTP-binding proteins activate the muscarinic KD channel in heart. Nature 325, 321-326.
(65) Tanabe, T., Nukada, T., Nishikawa, Y., Sugimoto, K., Suzuki, H., Takahashi, H., Noda, M., Haga, T., Ichiyama, A., Kangawa, K., Minamino, N., Matsuo, H. and Numa, S. (1985) Primary structure of the alpha-subunit of transducin and its relationship to ras proteins. Nature 315, 242-245.
(66) Nukada, T., Tanabe, T., Takahashi, H., Noda, M., Haga, K., Haga, T., Ichiyama, A., Kangawa, K., Hiranaga, M., Matsuo, H. and Numa, S. (1986) Primary structure of the alpha-subunit of bovine adenylate cyclase-inhibiting G-protein deduced from the cDNA sequence. FEBS Lett. 197, 305-310.
(67) Nukada, T., Mishina, M. and Numa, S. (1987) Functional expression of cloned cDNA encoding the alpha-subunit of adenylate cyclase-stimulat ing G-protein. FEBS Lett. 211,5-9.
(68) Gilman, A.G. (1987) G proteins: transducers of receptor-generated signals. Annu. Rev. Biochem. 56, 615-649.
(69) Tang, W.J. and Gilman, A.G. (1992) Adenylate cyclases. Cell 70, 869-872.
(70) Hanoune, J. and Defer, N. (2001) Regulation and role of adenylyl cyclase isoforms. Annu. Rev. Pharmacol. Toxicol. 41, 145-174.
(71) Berrie, C., Birdsall, N.J.M., Burgen, A.S.V. and Hulme, E.C. (1979) Guanine nucleotides modulate muscarinic receptor binding in the heart. Biochem. Biophys. Res. Commun. 87, 1000-1005.
(72) Nukada, T., Haga, T. and Ichiyama, A. (1983) Muscarinic receptors in porcine caudate nucleus. II. Different effects of N-ethylmaleimide on cismethyldioxolane binding to heat-labile (guanyl nucleotide-sensitive) sites and heat-stable (guanyl nucleotide-insensitive) sites. Mol. Pharmacol. 24, 374-379.
(73) Haga, K., Haga, T., Ichiyama, A., Katada, T., Kurose, H. and Ui, M. (1985) Functional reconstitution of purified muscarinic receptors and inhibitory guanine nucleotide regulatory protein. Nature 316, 731-733.
(74) Haga, K., Haga, T. and Ichiyama, A. (1986) Reconstitution of the muscarinic acetylcholine receptor. Guanine nucleotide-sensitive high affinity binding of agonists to purified muscarinic receptors reconstituted with GTP-binding proteins (Gi and Go). J. Biol. Chem. 261, 10133-10140.
(75) Haga, K., Uchiyama, H., Haga, T., Ichiyama, A., Kangawa, K. and Matsuo, H. (1989) Cerebral muscarinic acetylcholine receptors interact with three kinds of GTP-binding proteins in a reconstitution system of purified components. Mol. Pharmacol. 35, 286-294.
(76) Haga, T. and Haga, K. (1987) Interaction of the muscarinic acetylcholine receptor and GTP-binding proteins. Biomed. Res. 8, 149-156.
(77) Ikegaya, T., Nishiyama, T., Haga, K., Haga, T., Ichiyama, A., Kobayashi, A. and Yamazaki, N. (1990) Interaction of atrial muscarinic receptors with three kinds of GTP-binding proteins. J. Mol. Cell. Cardiol. 22, 343-351.
(78) Shiozaki, K. and Haga, T. (1992) Effects of magnesium on the interaction of atrial muscarinic acetylcholine receptors and GTP-binding regulatory proteins. Biochemistry 31, 10634-10642.
(79) Zhang, Q., Okamura, M., Guo, Z.-D., Niwa, S. and Haga, T. (2004) Effects of partial agonists and [Mg.sup.2+] ions on the interaction of [M.sub.2] muscarinic acetylcholine receptor and G protein Gai1 subunit in the [M.sub.2]-Gai1 fusion protein. J. Biochem. 135, 319-329.
(80) Takeda, S., Yamamoto, A. and Haga, T. (2002) Identification of a G protein-coupled receptor for 5-oxo-eicosatetraenoic acid. Biomed. Res. 23, 101-108.
(81) Suga, H. and Haga, T. (2007) Ligand screening system using fusion proteins of G protein-coupled receptors with G protein [alpha] subunits. Neurochem. Int. 51, 140-164.
(82) Hokin, L.E. and Hokin, M.R. (1955) Effects of acetylcholine on the turnover of phosphoryl units in individual phospholipids of pancreas slices and brain cortex slices. Biochim. Biophys. Acta 18, 102-110.
(83) Berridge, M.J. and Irvine, R.F. (1984) Inositol triphosphate, a novel second messenger in cellular signal transduction. Nature 312, 315-321.
(84) Nakamura, F., Ogata, K., Shiozaki, K., Kameyama, K., Ohara, K., Haga, T. and Nukada, T. (1991) Identification of two novel GTP-binding protein alpha-subunits that lack apparent ADP-ribosylation sites for pertussis toxin. J. Biol. Chem. 266, 12676-12681.
(85) Strathmann, M. and Simon, M.I. (1990) G protein diversity: A distinct class of a subunits is present in vertebrates and invertebrates. Proc. Natl. Acad. Sci. U.S.A. 87, 9113-9117.
(86) Nakamura, F., Kato, M., Kameyama, K., Nukada, T., Haga, T., Kato, H., Takenawa, T. and Kikkawa, U. (1995) Characterization of Gq family G proteins [G.sub.L1]alpha ([G.sub.14]alpha), [G.sub.L2]alpha ([G.sub.11]alpha), and [G.sub.q]alpha expressed in the baculovirus insect cell system. J. Biol. Chem. 270, 6246-6253.
(87) Lowenstein, C.J., Dineman, J.L. and Snyder, S.H. (1994) Nitric oxide: a physiologic messenger. Ann. Intern. Med. 120, 227-237.
(88) Siehler, S. (2009) Regulation of RhoGEF protein by G12/13-coupled receptors. Br. J. Pharmacol. 158, 41-49.
(89) Haga, K., Haga, T. and Ichiyama, A. (1990) Phosphorylation by protein kinase C of the muscarinic acetylcholine receptor. J. Neurochem. 54, 1639-1644.
(90) Haga, K., Kameyama, K., Haga, T., Kikkawa, U., Shiozaki, K. and Uchiyama, H. (1996) Phosphorylation of human m1 muscarinic acetylcholine receptors by G protein-coupled receptor kinase 2 and protein kinase C. J. Biol. Chem. 271, 2776-2782.
(91) Kuhn, H. (1978) Light-dependent binding of rho dopsin kinase and other proteins to cattle photoreceptor membranes. Biochemistry 17, 4389-4395.
(92) Benovic, J.L., Strasser, R.H., Caron, M.G. and Lefkowitz, R.J. (1986) Beta-adrenergic receptor kinase: Identification of a novel protein kinase that phosphorylates the agonist-occupied form of the receptor. Proc. Natl. Acad. Sci. U.S.A. 83, 2797-2801.
(93) Haga, T., Haga, K. and Kameyama, K. (1994) G protein-coupled receptor kinases. J. Neurochem. 63, 400-412.
(94) Kwatra, M.M. and Hosey, M.M. (1986) Phospho rylation of the cardiac muscarinic receptor in intact chick heart and its regulation by a muscarinic agonist. J. Biol. Chem. 261, 12429-12432.
(95) Haga, K. and Haga, T. (1989) Agonist-dependent phosphorylation of cerebral and atrial muscarinic receptors: blockade of the phosphorylation by GTP-binding regulatory proteins and its reversal by guanine nucleotides. Biomed. Res. 10, 293-299.
(96) Haga, K. and Haga, T. (1990) Dual regulation by G proteins of agonist-dependent phosphorylation of muscarinic acetylcholine receptors. FEBS Lett. 268,43-47.
(97) Haga, K. and Haga, T. (1992) Activation by G protein beta gamma subunits of agonist- or light-dependent phosphorylation of muscarinic acetylcholine receptors and rhodopsin. J. Biol. Chem. 267, 2222-2227.
(98) Benovic, J.L., Mayor, F. Jr., Somers, R.L., Caron, M.G. and Lefkowitz, R.J. (1986) Light-dependent phosphorylation of rhodopsin by beta-adrenergic receptor kinase. Nature 321, 869-872.
(99) Kwatra, M.M., Benovic, J.L., Caron, M.G., Lefkowitz, R.J. and Hosey, M.M. (1989) Phosphorylation of chick heart muscarinic cholinergic receptors by the [beta]-adrenergic receptor kinase. Biochemistry 28, 4543-4547.
(100) Kameyama, K., Haga, K., Haga, T., Kotani, K., Katada, T. and Fukada, Y. (1993) Activation by G protein beta-gamma subunits of beta adrenergic and muscarinic receptor kinase. J. Biol. Chem. 268, 7753-7758.
(101) Lodowski, D.T., Pitcher, J.A., Capel, W.D., Lefkowitz, R.J. and Tesmer, J.J.G. (2003) Keeping G proteins at bay: a complex between G protein-coupled receptor kinase 2 and GO.. Science 300, 1256-1262.
(102) Nakata, H., Kameyama, K., Haga, K. and Haga, T. (1994) Location of agonist-dependent-phosphorylation sites in the third intracellular loop of muscarinic acetylcholine receptors (m2 subtype). Eur. J. Biochem. 220,29-36.
(103) Haga, K., Kameyama, K. and Haga, T. (1994) Synergistic activation of a G protein-coupled receptor kinase by G protein beta gamma subunits and mastoparan or related peptides. J. Biol. Chem. 269, 12594-12599.
(104) Kameyama, K., Haga, K., Haga, T., Moro, O. and Sadee, W. (1994) Activation of a GTP-binding protein and a GTP-binding-protein-coupled receptor kinase ([beta]-adrenergic-receptor kinase-1) by a muscarinic receptor m2 mutant lacking phosphorylation sites. Eur. J. Biochem. 226, 267-276.
(105) Tsuga, H., Kameyama, K., Haga, T., Kurose, H. and Nagao, T. (1994) Sequestration of muscarinic acetylcholine receptor m2 subtypes: Facilitation by G protein-coupled receptor kinase (GRK2) and attenuation by a dominant-negative mutant of GRK2. J. Biol. Chem. 269, 32522-32527.
(106) Tsuga, H., Kameyama, K., Haga, T., Honma, T., Lameh, J. and Sadee, W. (1998) Internalization and down-regulation of human muscarinic acetylcholine receptor m2 subtypes: Role of third intracellular m2 loop and G protein-coupled receptor kinase 2. J. Biol. Chem. 273, 5323-5330.
(107) Tsuga, H., Okuno, E., Kameyama, K. and Haga, T. (1998) Sequestration of human muscarinic acetylcholine receptor hm1-hm5 subtypes: Effect of G protein-coupled receptor kinases GRK2, GRK4, GRK5 and GRK6. J. Pharmacol. Exp. Ther. 284, 1218-1226.
(108) Ichiyama, S., Oka, Y., Haga, K., Kojima, S., Tateishi, Y., Shirakawa, M. and Haga, T. (2006) The structure of the third intracellular loop of the muscarinic acetylcholine receptor [M.sub.2] subtype. FEBS Lett. 580,23-26.
(109) Hashimmoto, Y., Morisawa, K., Saito, H., Jojima, E., Yoshida, N. and Haga, T. (2008) Muscarinic [M.sub.4] receptor recycling requires a motif in the third intracellular loop. J. Pharmacol. Exp. Ther. 325, 947-953.
(110) Haga, K., Tsuga, H. and Haga, T. (1997) [Ca.sup.2+] dependent inhibition of G protein-coupled receptor kinase 2 by calmodulin. Biochemistry 36, 1315-1321.
(111) Haga, K., Ogawa, H., Haga, T. and Murofushi, H. (1998) GTP-binding-protein-coupled receptor kinase 2 (GRK2) binds and phosphorylates tubulin. Eur. J. Biochem. 255, 363-368.
(112) Yoshida, N., Haga, K. and Haga, T. (2003) Identification of sites of phosphorylation by G-protein-coupled receptor kinase 2 in [beta]-tubulin. Eur. J. Biochem. 270, 1154-1163.
(113) Haga, T., Abe, T. and Kurokawa, M. (1974) Polymerization and depolymerization of microtubules in vitro as studied by flow birefringence. FEBS Lett. 39, 291-295.
(114) Ribas, C., Penela, P., Murga, C., Salcedo, A., Garcia-Hoz, C., Juradao-Pueyo, M., Aymerich, I. and Mayor, F. Jr. (2007) The G protein-coupled receptor kinase (GRK) interactome: role of GRKs in GPCR regulation and signaling. Biochim. Biophys. Acta 1768, 913-922.
(115) Hayashi, M.K. and Haga, T. (1997) Palmitoylation of muscarinic acetylcholine receptor m2 subtypes: Reduction in their ability to activate G proteins by mutation of a putative palmitoylation site, cysteine 457, in the carboxyl-terminal tail. Arch. Biochem. Biophys. 340, 376-382.
(116) Mieda, M., Haga, T. and Saffen, D.W. (1996) Promoter region of the rat m4 muscarinic acetylcholine receptor gene contains a cell type-specific silencer element. J. Biol. Chem. 271, 5177-5182.
(117) Mieda, M., Haga, T. and Saffen, D.W. (1997) Expression of the rat m4 muscarinic acetylcholine receptor gene is regulated by the neuron-restrictive silencer element/repressor element 1. J. Biol. Chem. 272, 5854-5860.
(118) Hayashi, M.K. and Haga, T. (1996) Purification and functional reconstitution with GTP-binding regulatory proteins of hexahistidine-tagged muscarinic acetylcholine receptors (m2 subtype). J. Biochem. 120, 1232-1238.
(119) Haga, T. (1980) Solubilization of muscarinic acetyl choline receptors by L-[alpha]-lysophosphatidylcholine. Biomed. Res. 1, 265-268.
(120) Peterson, G.L., Rosenbaum, L.C. and Schimerlik, M.I. (1988) Solubilization and hydrodynamic properties of pig atrial muscarinic acetylcholine receptor in dodecyl [beta]-D-maltoside. Biochem. J. 255, 553-560.
(121) Rinken, A. and Haga, T. (1983) Solubilization and characterization of atrial muscarinic acetylcholine receptors in sucrose monolaurate. Arch. Biochem. Biophys. 301, 158-164.
(122) Haga, T., Nukada, T. and Haga, K. (1982) Solubilization of the muscarinic acetylcholine receptor by sodium cholate: Stabilization of the receptor by muscarinic ligands. Biomed. Res. 3, 695-698.
(123) Palczewski, K., Kumasaka, T., Hori, T., Behnke, C.A., Motoshima, H., Fox, B.A., Trong, I.L., Teller, D.C., Okada, T., Stenkamp, R.E., Yamamoto, M. and Miyano, M. (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science 289, 739-745.
(124) Rosenbaum, D.M., Cherezov, V., Hanson, M.H., Rasmussen, S.G.F., Thian, F.S., Kobilka, T.S., Choi, H.-J., Yao, X.-J., Weis, W.I., Stevens, R.C. and Kobilka, B.K. (2007) GPCR engineering yields high-resolution structural insights into [[beta].sub.2]-adrenergic receptor function. Science 318, 1266-1273.
(125) Landau, E.M. and Rosenbusch, J.P. (1996) Lipidic cubic phases: A novel concept for the crystallization of membrane proteins. Proc. Natl. Acad. Sci. U.S.A. 93, 14532-14535.
(126) Caffrey, M. and Cherezov, V. (2009) Crystallizing membrane proteins using lipidic mesophases. Nat. Protoc. 4, 706-731.
(127) Haga, K., Kruse, A.C., Asada, H., Yurugi Kobayashi, T., Shiroishi, M., Zhang, C., Weis, W.I., Okada, T., Kobilka, B.K., Haga, T. and Kobayashi, T. (2012) Structure of the human [M.sub.2] muscarinic acetylcholine receptor bound to an antagonist. Nature 482, 547-551.
(128) Kruse, A.C., Hu, J., Pan, A.C., Arlow, D.H., Rosenbaum, D.M., Rosemond, E., Green, H.F., Liu, T., Chae, P.S., Dror, R.O., Shaw, D.E., Weis, W.I., Wess, J. and Kobilka, B. (2012) Structure and dynamics of the [M.sub.3] muscarinic acetylcholine receptor. Nature 482, 552-559.
(129) Magnik, A., Kruse, A.C., Kobilka, T.S., Thian, F.S., Mathiesen, J.M., Sunahara, R.K., Pardo, L., Weis, W.I., Kobilka, B.K. and Granier, S. (2012) Crystal structure of the y-opioid receptor bound to a morphinian anatagonist. Nature 485, 321-326.
(130) Hulme, E.C., Lu, Z.L. and Bee, M.S. (2003) Scanning mutagenesis studies of the [M.sub.1] muscarinic acetylcholine receptor. Receptors Channels 9, 215-228.
(131) Goodwin, J.A., Hulme, E.C., Jangmead, C.J. and Tehen, B.G. (2007) Roof and floor of the muscarinic binding pocket: variations in the binding modes of orthosteric ligands. Mol. Pharmacol. 72, 1484-1496.
(132) Heitz, F., Holzwarth, J.A., Gies, J.P., Pruss, R.M., Trumpp-Kallmeyer, S., Hibert, M.F. and Guenet, C. (1999) Site-directed mutagenesis of the putative human muscarinic [M.sub.2] receptor binding site. Eur. J. Pharmacol. 380, 183-185.
(133) Wess, J. (1993) Mutational analysis of muscarinic acetylcholine receptors: structural basis of ligand/ receptor/G protein interactions. Life Sci. 53, 1447-1463.
(134) Birdsall, N.J.M. and Lazareno, S. (2005) Allosterism at muscarinic receptors: ligands and mechanisms. Mini-Reviews. Med. Chem. 5, 523-543.
(135) Proska, J. and Tucek, S. (1994) Mechanisms of steric and cooperative actions of alcuronium on cardiac muscarinic acetylcholine receptors. Mol. Pharmacol. 45, 709-717.
(136) Guo, Z.D., Haga, T., Itokawa, H. and Mizobe, F. (1995) Allosteric binding of 9-methoxy-alphalapachone and alcuronium to the muscarinic acetylcholine m2 subtype. Biomed. Res. 16, 327-335.
(137) Jones, C.K., Brady, A.E., Davis, A.A., Xiang, A., Bubser, M., Tanatawy, M.N., Kane, A.S., Bridges, T.M., Kennedy, J.P., Bradley, S.R., Peterson, T.E., Ansari, M.S., Baldwin, R.M., Kessler, R.M., Deutch, A.Y., Lah, J.J., Levey, A.I., Lindsley, C.W. and Conn, P.J. (2008) Novel selective allosteric activator of the [M.sub.1] muscarinic acetylcholine receptor regulates amyloid processing and produces antipsychotic-like activity in rats. J. Neurosci. 28, 10422-10433.
(138) May, L.T., Avlani, V.A., Langmead, C.J., Herdon, H.J., Wood, M.D., Sexton, P.M. and Christopoulos, A. (2007) Structure-function studies of allosteric agonism at [M.sub.2] muscarinic acetylcholine receptors. Mol. Pharmacol. 72, 463-476.
(139) Jager, D., Schmalenbach, C., Prilla, S., Schrobang, J., Kebig, A., Sennwitz, M., Heller, E., Trankle, C., Holzgrabe, U., Holtje, H.-D. and Mohr, K. (2007) Allosteric small molecules unveil a role of an extracellular E2/transmemebrane helix 7 junction for G protein-coupled receptor activation. J. Biol. Chem. 282, 34968-34976.
(140) Gregory, K.J., Hall, N.E., Tobin, A.B., Sexton, P.M. and Christopoulos, A. (2010) Identification of orthosteric and allosteric site mutations in [M.sub.2] muscarinic acetylcholine receptors that contribute to ligand-selective signaling bias. J. Biol. Chem. 285, 7459-7474.
(141) Taylor, P. and Insel, P.A. (1990) Molecular basis of pharmacologic selectivity. In Principle of Drug Action (eds. Pratt, W.B. and Palmer, T.), 3rd ed., Churchill Livingstone Inc., New York, pp. 1-74.
(142) Furukawa, H., Hamada, T., Hayashi, M.K., Haga, T., Muto, Y., Hirota, H., Yokoyama, S., Nagasawa, K. and Ishiguro, M. (2002) Conformation of ligands bound to the muscarinic acetylcholine receptor. Mol. Pharmacol. 62, 778-787.
(143) Hulme, E.C. (2013) GPCR activation: a mutagenic spotlight on crystal structures. Trends Pharmacol. Sci. 34,67-84.
(144) Rasmussen, S.G.F., DeVree, B.T., Zou, Y., Kruse, A.C., Chung, K.Y., Kobilka, T.C., Thian, F.S., Chae, P.S., Pardon, E., Calinski, D., Mathiesen, J.M., Shah, S.T., Lyons, J.A., Caffrey, M., Gellman, S.H., Streyaert, J., Skiniotis, G., Weis, W.I., Sunahara, R.K. and Kobilka, B.K. (2012) Crystal structure of the [[beta].sub.2] adrenergic receptor-Gs protein complex. Nature 477, 549-555.
(145) Wess, J., Eglen, R.M. and Gautan, D. (2007) Muscarinic acetylcholine receptors: mutant mice provide new insights for drug development. Nat. Rev. Drug Discov. 6, 721-733.
(146) Conn, P.J., Jones, C.K. and Lindsley, C.W. (2009) Subtype-selective allosteric modulators of muscarinic receptors for the treatment of CNS disorders. Trends Pharmacol. Sci. 30, 148-155.
(147) Jakubik, J. and El-Fakahany, E. (2010) Allosteric modulation of muscarinic acetylcholine receptors. Pharmaceuticals 3, 2838-2860.
(148) Keov, P., Sexton, P.M. and Christopoulos, A. (2011) Allosteric modulation of G protein-coupled receptors: a pharmacological perspective. Neuro pharm. 60,24-35.
(Received Feb. 24, 2013; accepted Apr. 30, 2013)
By Tatsuya HAGA * , ([dagger])
(Communicated by Masanori OTSUKA, M.J.A.)
*  Professor Emeritus, the University of Tokyo.
([dagger]) Correspondence should be addressed: T. Haga, Numama 2-3-1-411, Zushi, Kanagawa 249-0004, Japan (e-mail: tatsuya. firstname.lastname@example.org).
* (1) Jojima, E., Saito, H., Yoshida, N. and Haga, T. unpublished results.
* (2) Morisawa, K., Saito, K., Kakegai, R., Hashimoto, Y. and Haga, T. unpublished results.
Abbreviations: ABT: aminobenztropine (muscarinic antagonist used for affinity chromatography); ACh: acetylcholine; cDNA: complementary DNA; DTT: dithiothreitol; E2 loop: the second extracellular loop; I3 loop: the third intracellular loop; [M.sub.1] ([M.sub.2], [M.sub.3], [M.sub.4], [M.sub.5]) receptors: muscarinic acetylcholine receptor [M.sub.1] ([M.sub.2], [M.sub.3], [M.sub.4], [M.sub.5]) subtypes; NMS: N-methylscopolamine (muscarinic antagonist); PrBCM: propylbenzylcholine mustard (irreversible muscarinic antagonist); QNB: quinuclidinyl benzylate (muscarinic antagonist); SDS-PAGE: sodium dodecyl sulfatepolyacrylamide gel electrophoresis; T[M.sub.1] (2-7): transmembrane segment 1 (2-7).
Tatsuya Haga was born in 1941, graduated from Univ. of Tokyo, Fac. Sci., Dept. Biophys. Biochem. in 1963, and obtained Ph.D. from the same Faculty (Supervisor, Prof. Noda, H.) by studies on "Synthesis and release of [sup.C]]acetylcholine in synaptosomes" in 1969. He was Assist. Prof. in laboratories of Prof. Noda, H. and Kurokawa, M in Univ. of Tokyo (1969-1974), Assoc. Prof. in laboratory of Prof. Ichiyama, A. in Hamamatsu Univ. Sch. Med. (1974-1988), Prof. of Univ. of Tokyo, Fac. Med. (1988-2001) and of Gakushuin Univ. Fac. Sci. (2001-2011), and Prof. Emeritus of Univ. of Tokyo (2001-). He also worked in laboratories of Prof. Gilman, A.G. in Virginia Univ. (1975-1977), of Prof. Burgen, A.S.V. in Nat. Inst. Med. Res. (1981-1982), of Prof. Barnard, E. in Cambridge Univ. (1992). He was mainly engaged in biochemical studies of cholinergic neurons and G protein-coupled receptors (GPCRs), particularly on muscarinic acetylcholine receptors and high-affinity choline uptake.
Table 1. Muscarinic ligands with an amino group and their affinity for muscarinic receptors in cerebral membranes and in solubilized form 18) [IC.sub.50] ([micro]M) for inhibition of QNB binding Chemical structure Membrane Solubilized receptor receptor 1 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.06 0.6 2 [FORMULA NOT REPRODUCIBLE IN ASCII] 1 3 3 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.3 2 4 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.4 3 5 [FORMULA NOT REPRODUCIBLE IN ASCII] 2 10 6 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.8 3 7 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.5 3 8 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.2 2 9 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.003 0.05 10 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.02 0.25 11 [FORMULA NOT REPRODUCIBLE IN ASCII] 0.004 0.03 A model compound (1, BP), its derivatives with an amino group in different positions (2-8), derivatives of compound 8 with tropin (9) and quinuclidinol (10), and APCPP (11) were synthesized, and their affinities for muscarinic receptors were estimated from their effects on [[sup.3]H]QNB binding of membrane-bound and solubilized muscarinic receptors.
|Printer friendly Cite/link Email Feedback|
|Publication:||Japan Academy Proceedings Series B: Physical and Biological Sciences|
|Date:||Jun 1, 2013|
|Previous Article:||Structure and function studies on enzymes with a catalytic carboxyl group(s): from ribonuclease [T.sub.1] to carboxyl peptidases.|
|Next Article:||My career as an immunoglycobiologist.|