Meeting Report for: The 5th International Symposium on Subtypes of Muscarinic Receptors

Muscarinic receptor subtypes in the airways appear to subserve different physiological functions. M1-receptors facilitate neurotransmission through parasympathetic ganglia and enhance cholinergic reflexes, but are also localized to alveolar walls. M2-receptors act as autoreceptors on post-ganglionic cholinergic nerves and inhibit acetylcholine release. There is some evidence that they may be defective in asthma (as a consequence of airway inflammation?) and this may enhance cholinergic reflexes and account for beta-blocker-induced asthma. M2-receptors in airway smooth muscle may also counteract the bronchodilator action of beta-agonists. M3-receptors mediate contractile responses in airway smooth muscle via phosphoinositide hydrolysis, and are the predominant receptors on submucosal glands and airway vascular endothelium. M4- and M5-receptors have not been identified in human airways, but in rabbit lung M4-receptors are expressed on alveolar walls and smooth muscle. Anticholinergic drugs which selectively block M3 and M1-receptors may have an advantage over currently used non-selective antagonists in the treatment of airway obstruction.

The regions of muscarinic receptors that specify G-protein-coupling and ligand-binding have been defined in several recent studies. Overall, these studies have shown that amino acids within the third cytoplasmic loop of the receptors define their selectivity for different G-proteins, and that multiple, discontinuous epitopes contribute to their selectivities for different ligands. In fact, several competitive and allosteric antagonists can be classified into groups based on which of these epitopes contribute to their subtype selectivity. Site-directed mutagenesis, combined with covalent-labeling studies have suggested that ligands bind to a hydrophobic core of the receptors that is formed by multiple transmembrane (TM) domains. An aspartic acid located in TM3 is likely to bind to the ammonium headgroup of muscarinic ligands, and multiple hydroxyl-containing amino acids contribute to agonist but not antagonist binding. These data are discussed in the context of a computational model of a muscarinic receptor. Our model is based on a sequence alignment with bacteriorhodopsin, a seven TM protein for which a higher resolution structure is available. Most of the mutagenic data can be rationalized in the context of this model, and predict testable hypotheses concerning the mechanism by which ligands control the activity of muscarinic receptors.

The use of anticholinergics in antiobstructive therapy is well established in pulmonary medicine. We sought to improve the duration of action of inhaled antimuscarinics. A newly developed compound, Ba 679 BR (abbreviated Ba 679) proved to be a highly potent muscarinic antagonist in guinea pig tracheal rings. Its binding to human receptors (Hm1, Hm2, Hm3) was characterized by KD-values in the 10(-10) M concentration range. Assessment of the dissociation rate of complexes of labelled Ba 679 and human muscarinic receptors revealed very slow dissociation in comparison to ipratropium. The half-lives in hours were: Ba 679-Hm3: 34.7, -Hm1: 14.6, -Hm2: 3.6; ipratropium-Hm3: 0.26, -Hm1: 0.11, -Hm2: 0.035. The duration of action in vivo was determined by means of acetylcholine-induced bronchospasms in dogs following inhalation of the drugs. Ba 679 demonstrated a significantly longer duration of protection than an equipotent dose of ipratropium. The plasma levels following inhalation in dogs declined rapidly and are unlikely to reflect the duration of the pharmacological activity. In summary, Ba 679 represents a novel type of antimuscarinic bronchodilator with a long duration of action, most likely due to its slow dissociation from Hm3-receptors. In addition, the drug showed "kinetic receptor subtype selectivity" by having a more rapid dissociation from Hm2 than from Hm1 and Hm3 receptors.

Clinical trials with muscarinic agonists or acetylcholine esterase inhibitors for the treatment of Alzheimer's dementia have shown disappointing or equivocal results. An alternative treatment of this disease is the development of drugs which enhance the release of acetylcholine. It is believed, that of the five muscarinic receptor subtypes so far identified in the brain, M2 receptors are located presynaptically in the cortex and hippocampus and upon stimulation inhibit the release of acetylcholine. Based on this hypothesis, we initiated a drug discovery program with the aim of identifying selective and centrally active M2 antagonists which are capable of enhancing cholinergic transmission. These efforts resulted in the successful design and synthesis of novel muscarinic antagonists able to cross the blood brain barrier. Moreover, these compounds show few peripheral effects and possess a superior M2 versus M1 selectivity. The prototype of this novel class of M2 selective compounds, BIBN 99, could be a valuable tool to test the hypothesis that lipophilic M2 antagonists show beneficial effects in the treatment of cognitive disorders.

The ability of WAL 2014 to elicit muscarinic responses was investigated in various in vitro and in vivo models. In CHO cells transfected with human m1- or m3- receptor genes, WAL 2014 was clearly more effective in stimulating the M1-mediated PI response. In isolated tissue preparations, WAL 2014 exhibited full agonist properties in the rabbit vas deferens (putative M1 receptor) and behaved like a partial agonist at M2 receptors in the atrium and M3 receptors in the ileum of guinea-pigs. In the pithed rat, in which the increase in blood pressure is mediated through a stimulation of M1 receptors in sympathetic ganglia, WAL 2014 produced a full dose response curve, whereas the reference compounds RS 86 and arecoline exhibited a bell-shaped behaviour. This is in accord with the view that WAL 2014 selectively activates M1 receptors in sympathetic ganglia, whereas conventional agonists in the same dose range stimulate both ganglionic M1 and vascular M3 receptors. The preferential neuron-stimulating properties were confirmed by EEG recording in the rabbit, in which muscarinic activation occurred at doses similar to those for ganglionic stimulation in the pithed rat. On the other hand, higher doses of WAL 2014 were needed to elicit muscarinic effects in peripheral effector organs, i.e. bradycardia, urinary bladder contraction and increase in airway resistance. It is concluded that WAL 2014 due to its preferential neuronal activity is a promising candidate for a cholinergic substitution therapy in Alzheimer's disease.

The thiolactone analogue of pilocarpine, SDZ ENS 163, acts in vitro and in vivo as a partial agonist at M1/M3 and as an antagonist at M2 muscarinic receptors. In vitro, the properties of SDZ ENS 163 have been investigated in several functional models for muscarinic receptors: it is a full agonist at M1 (rat superior cervical ganglion, carbachol = 100%) and a partial agonist at M3 receptors (guinea pig ileum). However, the drug shows antagonistic properties at M2 receptors (rat atria). Radioligand binding studies with 3H-N-methylscopolamine (3H-NMS) using CHO cells expressing m1 or m3 receptors indicate that SDZ ENS 163 does not discriminate between m1 and m3 receptors (Ki 1.5 and 2.4 microM respectively). Regarding phosphoinositide (PI) turnover in A9L cells, SDZ ENS 163 is a partial agonist at m1 receptors. In ex vivo neurochemical studies in rats SDZ ENS 163 displays effects characteristic of muscarinic antagonists regarding the turnover of ACh which is increased in the brain. At a similar dose-range SDZ ENS 163 accelerates PI metabolism in the rat brain in vivo and increases the energy of the low frequency band (2-5 Hz) in the rat hippocampal EEG. These effects observed in vivo are consistent with postsynaptic M1 agonistic and presynaptic M2 antagonistic activities. Since SDZ ENS 163 at centrally active doses exerts no peripheral cholinergic effects, it may be useful for the symptomatic treatment of Alzheimer's disease.

Antagonist/agonist binding ratios (NMS/Oxo-M ratio) were used as an index of the efficacy of novel compounds acting at muscarinic receptors. These binding ratios have been used with a range of functional pharmacological assays to investigate the effects of varying the efficacy of muscarinic agonists. This strategy has been used as a means of obtaining functional receptor selectivity by exploiting differences in effective receptor reserves. The oxadiazole and pyrazine muscarinic agonists L-670,548 (NMS/Oxo-M ratio 1100) and L-680,648 (NMS/Oxo-M ratio 690) are amongst some of the most potent and efficacious agonists known. Decreasing the efficacy of compounds from these series, resulted in compounds with functional selectivity. The chloropyrazine L-689,660 (NMS/Oxo-M ratio 28) was an agonist on the rat superior cervical ganglion (M1), a partial agonist on the guinea-pig ileum (M3), but was an antagonist in the guinea-pig atria (M2). Synthesis of compounds with even lower predicted efficacy, such as the cyclopropyloxadiazole L-687,306 (NMS/Oxo-M ratio 15), maintained agonist activity in the ganglion, but showed antagonist activity in the M3 ileal, as well as the M2 atrial preparations. When tested in vivo these compounds did not produce many of the side effects associated with more efficacious agonists, particularly those associated with the cardiovascular system. However, they were active in reversing scopolamine-induced deficits in a variety of behavioural paradigms. This approach shows how functional selectivity for muscarinic receptor subtypes can be achieved in vitro, that in vivo reduces the dose-limiting side effects normally associated with muscarinic agonists.

Acetylcholine released from vagal nerve endings constricts airways by stimulating M3 muscarinic receptors on the airway smooth muscle. At the same time, released acetylcholine feeds back onto inhibitory M2 muscarinic autoreceptors on the nerve endings, limiting further release of acetylcholine. Loss of function of these M2 receptors increases vagally-mediated bronchoconstriction after viral airway infections, exposure to ozone, or antigen inhalation. Viral infections may decrease M2 receptor function by inducing inflammation or via direct damage to the receptors as a result of cleavage of sialic acid residues by viral neuraminidase. Inflammation appears to be critical in the loss of M2 receptor function after ozone exposure. Antigen-induced loss of M2 receptor function can be reversed acutely by administering the poly-anionic substances heparin or poly-l-glutamate, possibly by binding and neutralizing positively charged eosinophil proteins. Such positively charged eosinophil proteins, particularly major basic protein, may be acting as endogenous inhibitors at the M2 receptors, as can be demonstrated in in vitro ligand binding studies.

The regulation of expression and function of the muscarinic acetylcholine receptor has been studied using several different systems. The role of glycosylation of the m2 receptor was examined by removal of glycosylation sites using site-directed mutagenesis followed by expression in stably transfected cells. The results demonstrated that glycosylation was not required for the synthesis and appearance of the receptors on the cell surface or for the coupling of the receptors to inhibition of adenylyl cyclase activity. Site-directed mutagenesis also was used to demonstrate that the single cysteine in the carboxy terminal domain of the m2 receptor was not required for receptor function, thus rendering unlikely a model suggesting a requirement for palmitoylation of this cysteine in receptor function. The muscarinic receptors expressed in embryonic chick heart were identified by molecular cloning. Two genes were initially identified which are expressed in chick heart and correspond to the chick m2 and m4 receptors. Experiments using the polymerase chain reaction to identify low abundance mRNAs indicate that at least one addition receptor gene is expressed in chick heart. In cell culture, activation of the muscarinic receptors decreases the levels of mRNA encoding the cm2 and cm4 receptors. This probably results from decreased gene transcription due to both mAChR-mediated inhibition of adenylyl cyclase and mAChR-mediated stimulation of phospholipase C. The elucidation of the factors which regulate the expression and function of muscarinic acetylcholine receptors (mAChR) is of obvious importance in understanding the mechanisms underlying cholinergic transmission. In this chapter, we will describe studies on the expression and function of wild type and mutant muscarinic receptors, the molecular characterization of mAChR expressed in chick heart, and the regulation of mAChR gene expression in response to muscarinic receptor activation.

Effects of G proteins on the phosphorylation of muscarinic receptors (mAChRs) have been examined. Cerebral but not atrial mAChRs were phosphorylated by any one of three types of protein kinase C and 4-6 mol of phosphate were incorporated per mol of mAChR, mostly in the 12-14 kDa from the carboxyterminus. Atrial mAChRs were better substrates of cAMP-dependent protein kinase than cerebral mAChRs. Phosphorylation of mAChRs by protein kinase C or cAMP-dependent protein kinase was not dependent on the presence of agonists and G proteins except that a slight inhibition by G proteins was observed probably because G proteins were also substrates of the two kinases. Agonist-dependent phosphorylation of atrial mAChRs or recombinant human mAChRs (m2 subtype) by a kinase (mAChR kinase), which is the same or very similar to beta adrenergic receptor kinase (beta ARK), was found to be regulated by the G proteins in a dual manner; stimulation by G protein beta gamma subunits and inhibition by G protein alpha beta gamma trimer. The inhibition by the G protein trimer is restored by addition of guanine nucleotides and is considered to be due to the formation of a ternary complex of agonist, mAChR and guanine nucleotide free G proteins. The stimulation by G protein beta gamma subunits was also observed for the light- or agonist-dependent phosphorylation of rhodopsin and beta AR by the mAChR kinase but not for the light-dependent phosphorylation of rhodopsin by rhodopsin kinase. The phosphorylation by beta ARK 1 was also found to be stimulated by G protein beta gamma subunits. The beta gamma subunit is considered to interact with the extra 130 amino acid residue carboxyterminal tail of beta ARK, which does not exist in rhodopsin kinase, and the interaction results in the activation of the kinase. We may assume that the G protein coupled receptor kinase is an effector of G protein beta gamma subunits and that one of the functions of beta gamma subunits is to stimulate the phosphorylation of G protein coupled receptors thereby facilitating their desensitization.

The discovery of five muscarinic receptor subtypes by molecular genetic techniques has resulted in new approaches to understanding their function. This involves the expression of the individual genes encoding each receptor subtype in isolation, such that their effects and mechanisms of action can be studied. The coupling of the receptors with G-proteins and ion channels is the subject of this review and emphasis is placed upon the assignment of genetically defined receptor subtypes with a given physiological function. Activation of inwardly rectifying potassium conductances by m2 and m4 and inhibition by m1, as well as stimulation of calcium-dependent conductances by m1, m3 and m5 are discussed.

The muscarinic pharmacology of two novel agonists related to McN-A-343, 4-F-PyMcN and 4-F-PyMcN+, has been studied by the use of pharmacological and radioligand binding techniques. Both compounds were potent agonists at M1 receptors in rabbit vas deferens (pEC50 = 6.24 and 6.96) and rat duodenum (pEC50 = 5.47 and 6.38), but very weak partial agonists or competitive antagonists at guinea-pig cardiac M2 and ileal M3 receptors. There was no receptor reserve for 4-F-PyMcN in rabbit vas deferens, for which the potency (pEC50 = 6.24) and apparent affinity (pKA = 5.99 and 6.21) were similar. 4-F-PyMcN+ showed only limited binding selectivity between four muscarinic receptor binding assays with apparent affinity constants (pKi) of 5.8, 5.2, 5.6 and 5.7 for M1, M2, M3 and M4 muscarinic receptor subtypes. The two novel functionally M1-selective agonists may provide useful tools with which to study muscarinic receptor mechanisms. The non-quaternary compound, 4-F-PyMcN, might become a starting point for the development of drugs that selectively affect M1 receptors involved in central cholinergic function.

We have studied muscarinic agonist stimulated [35S]GTP gamma S binding and [gamma 32P]GTP hydrolysis (GTPase) in membranes from CHO cells stably transfected with human muscarinic m1-m4 receptors. 'Full' agonists were at least 10-fold more potent at m2 & m4 receptors than at m1 & m3. This pattern was less marked with 'partial' agonists, which had a greater maximal effect at m2 & m4 than at m1 & m3. McN-A343 uniquely was more potent and efficacious at m4 than at m2 receptors. Antagonist affinity constants were estimated by fitting the data from inhibition curves directly to the Schild model. Antagonist affinity estimates were very similar to those measured earlier in binding studies using animal tissues, and confirmed a small degree of m4 selectivity for tropicamide and secoverine. The receptor subtypes activated more than one G-protein subtype; m2 & m4 receptors activated only pertussis (PTX) sensitive G-proteins, while m1 & m3 coupled to both PTX sensitive and insensitive G-proteins. Acetylcholine (ACh) was more potent in stimulating guanine nucleotide exchange in PTX-treated m1 cells than in controls.

Knowledge of the distributions and functions of native m1-m5 muscarinic acetylcholine receptors in tissues is limited. To characterize the family of m1-m5 proteins directly, a panel of subtype-selective antibodies was generated against divergent i3 loop-fusion proteins. Each antibody was shown to bind a single cloned receptor specifically. In peripheral tissues and brain, four receptor proteins (m1-m4) were found to account for the vast majority of the muscarinic binding sites using immunoprecipitation studies with the subtype-specific antibodies. The subtypes were differentially distributed, although most tissues were comprised of a complex mixture of receptors. Moreover, within tissues there were major differences in the precise localization of the subtypes, as determined by immunocytochemistry. The immunological methods described offer a novel approach with exquisite sensitivity and specificity for delineating the distribution of m1-m5 receptors in animal and human tissues.

The venom of the Eastern green mamba from Africa, Dendroaspis angusticeps, contains a number of toxins which block the binding of 3H-antagonists to genetically-defined m1 and m4 muscarinic acetylcholine receptors. Most of the anti-muscarinic activity of the venom is due to the presence of a newly-isolated toxin, "m1-toxin", which has 64 amino acids and a molecular mass of 7361 Daltons. At present m1-toxin is the only ligand which is known to be capable of fully blocking m1 receptors without affecting m2-m5 receptors. It binds very rapidly, specifically and pseudoirreversibly to the extracellular face of m1 receptors on cells, in membranes or in solution, whether or not the primary receptor site is occupied by an antagonist. Bound toxin can either prevent the binding and action of agonists or antagonists, or prevent the dissociation of antagonists. The toxin is useful for identifying m1 receptors during anatomical and functional studies, for recognizing and stabilizing receptor complexes, and for occluding m1 receptors so that other receptors are more readily studied.

M1 muscarinic cholinergic receptors, G1 and G11 (Gq/11), and phospholipase C-beta 1 were highly purified from both natural sources and cells that express the appropriate cDNA's. When the proteins were co-reconstituted into phospholipid vesicles, the receptor efficiently and selectively promoted the activation of Gq/11, leading to marked stimulation of PLC activity in the presence of GTP gamma S. No stimulation was observed in the presence of GTP, however, which led to the finding that PLC-beta 1 stimulates the hydrolysis of GQ/11-bound GTP at least 50-fold. Thus, PLC-beta 1 is a GTPase activating protein, a GAP, for its physiologic regulator Gq/11. We discuss the implications of PLC-beta 1's GAP activity on the M1 muscarinic cholinergic signaling pathway.

Using recombinant CHO cells that express Hm1-Hm5 receptors, reference muscarinic agonists have been characterized with respect to their activity in receptor binding and second messenger assays. In whole cell [3H]-N-methyl scopolamine binding, no agonist was found to be truly subtype selective, although some showed marked differences between several of the subtypes (e.g. m1 vs. m2). As a functional index of receptor activation, phosphatidyl-inositol (PI) turnover was measured for m1, m3, and m5 receptors while inhibition of forskolin-stimulated cAMP accumulation was measured for m2 and m4 receptors. Both full and partial agonists were delineated in PI turnover, but all agonists showed similar responses on cAMP. Alkylation studies with propylbenzylcholine mustard showed that both efficacy and potency were markedly affected in the functional assays by the number of free receptors. Thus, receptor reserve appears to play a major role in the determination of subtype selectivity for agonists using functional measures. Even with these limitations, however, the use of transformed cell lines is playing a pivotal role in the discovery of selective agonists.

The synthesis of a series of potent and efficacious 1-azabicyclo[2.2.1]heptan-3-one oxime muscarinic agonists is described. The oximes have extended appendages designed to span the cavity defined by the seven transmembrane helices of the muscarinic receptor. Some members of the series are selective for receptors of the m1 subtype. One such oxime, 31, shows affinity and functional selectivity for m1 over m2, m3, and m4 muscarinic receptor types.