Cholinergic Transduction

Elliott Richelson

There are two classes of acetylcholine receptors, muscarinic and nicotinic (see Molecular Biology, Pharmacology, and Brain Distribution of Subtypes of the Muscarinic Receptor). These receptors play major roles in the function of the body. In brain, evidence suggests a role for muscarinic receptors in memory function (28, 71) and in the pathophysiology of affective illness (46, 47, 79) and schizophrenia (24, 50, 85, 86). Because of their putative role in cognitive function, muscarinic receptors have been a focus of research in Alzheimer's disease (see Experimental Therapeutics).

Researchers defined the two classes of cholinergic receptors early in this century on the basis of tissue responses to certain agonists and antagonists (22). A response caused by muscarine, an alkaloid derived from a poisonous mushroom (Amanita muscaria), and antagonism of a response caused by atropine, an alkaloid from the deadly nightshade (Atropa belladonna—a favorite of poisoners in past centuries), defined muscarinic receptors in a tissue. On the other hand, response to nicotine and blockade by d-tubocurarine (one of the alkaloids from the South American arrow poisons) defined nicotinic receptors. These definitions for the two types of cholinergic receptors are still used today.

Muscarinic receptors are widely distributed throughout the body. They mediate various types of responses in many different tissues—such as cardiac tissue, smooth muscle, and exocrine glands—and in cells throughout the peripheral and central nervous system. Within the nervous system, muscarinic receptors are present on some axon endings (heteroreceptors and autoreceptors), regulating neurotransmitter release (1, 67, 68, 90). These receptors are also on the soma and dendrites of many types of neurons, including cholinergic and noncholinergic neurons (67, 95).

In the past several years, discoveries by molecular biologists have made the field of muscarinic receptors immensely interesting and complex. At a time when pharmacologists could identify at most three subtypes of muscarinic receptors, molecular biologists showed that at least five different muscarinic receptors exist on the basis of molecular cloning experiments (15, 16, 54, 55, 63). Much knowledge about these receptors has come from the use of transfected cell lines that have incorporated into their genome and express the gene encoding the specific muscarinic receptor.

This chapter will focus mainly on the biochemical events associated with activation of muscarinic receptors. The concepts of receptors, their effectors, and second messengers will be presented along with a general discussion of the types of membrane-bound receptors. Specific aspects of transduction at cholinergic receptors will be presented along with a brief discussion of the neuropsychopharmacology of muscarinic receptors.

The concept of a receptor was put forth in the latter part of the nineteenth century and the early part of this century by Professors J. N. Langley and Paul Ehrlich. Professor Ehrlich has stated "the toxin must unite . . . with `receptors' in order for the toxin to have its effects" (30). Professor Langley used the term "receptive substance" (56).

A receptor is a highly specialized protein, which, in most cases, is a transmembrane spanning protein on the cellular surface (receptors for steroid hormones are intracellular). This receptor protein has the unique ability to recognize (bind) specific molecules. All neurotransmitters, neuromodulators, and hormones have specific receptors. Many neurotransmitters are known to have more than one type of receptor to which they bind.

Currently, receptors may be classified into four groups based upon their signal transduction mechanisms: (i) receptors that are ligand-gated ion channels [e.g., the nicotinic acetylcholine (21, 65) and gamma-aminobutyric acid (GABA) receptors (13, 75)]; (ii) receptors that are enzymes [e.g., the insulin receptor, which is a tyrosine kinase (42); and the atrial natriuretic peptide receptor, which is a particulate form of guanylate cyclase (78)]; (iii) receptors that couple to guanosine triphosphate (GTP)-binding proteins [e.g., muscarinic acetylcholine receptors and a multitude of others (44, 81)]; and (iv) receptors with unknown signal transduction mechanisms (e.g., the sigma receptor).

The term signal transduction refers to the mechanism used by the first messenger (the neurotransmitter, neuromodulator, or hormone) of the transmitting cell to convert its information into a second messenger within the receiving cell. Signal transduction will involve a receptor for the first messenger and may involve both transducers and effectors. In the field of receptors, a transducer may be defined as a molecule that translates one form of "energy" (e.g., the neurotransmitter) into another form, the second messenger. Effector is a molecule that mediates a specific effect (e.g., an ion channel).

The classical example of a transducer is a GTP-binding protein or G protein (70), of which there are many (36, 58, 81). Even light and olfaction make use of G proteins in the transduction mechanisms for these sensations. G proteins consist of three different proteins (a heterotrimer) labeled as Ga, Gb, and Gg, in the order of decreasing molecular weight. Biochemical studies as well as molecular cloning experiments have identified a multitude (nearly 20) of Ga subtypes and several (at least four each) Gb and Gg subtypes. Among the Ga subtypes are those involved with stimulating adenylate cyclase (Gas) and those involved with inhibiting (Gai) this enzyme. Recent evidence suggests that the bg subunits may also be involved in activation or inhibition of adenylate cyclase, depending on the type of adenylate cyclase (87).

G proteins can be broadly classified into four groups, according to their sensitivity to two different bacterial toxins—cholera toxin and pertussis toxin. Both toxins, which are enzymes, adenosine triphosphate (ADP)ribosylate specific sites on the G proteins. The four classes are as follows: (i) cholera-toxin-sensitive; (ii) pertussis-toxin-sensitive; (iii) pertussis- and choleratoxin-sensitive; and (iv) pertussis- and cholera-toxininsensitive (58, 81). The pertussis-toxin-sensitive Gao is the most abundant in brain, comprising about 1–2% of brain membrane protein.

The neurotransmitter (or hormone, etc.), its receptor, and a G protein, to which is bound guanosine diphosphate (GDP), form a ternary complex. This complex binds the neurotransmitter with high affinity. When GTP displaces GDP, the complex dissociates into its components, including a form of the receptor that binds neurotransmitter with low affinity. Support for this mechanism comes from radioligand binding studies showing that GTP (or an analogue) added to the binding reaction shifts the equilibrium dissociation constant for the agonist to a lower affinity. This lower-affinity binding results kinetically from GTP, causing an acceleration of the dissociation of the agonist from its binding site. Compounds with lower affinity dissociate more rapidly from their binding sites than do compounds with higher affinity.

The prime example of an effector is adenylate cyclase, the enzyme that synthesizes the second messenger cyclic adenosine 3¢,5¢-monophosphate (cyclic AMP) from adenosine triphosphate (ATP) (84). For example, norepinephrine causes the second messenger cyclic AMP to form within the cell by the following mechanism, involving the binding to a b-adrenoceptor. First, norepinephrine binds to its receptor. The G protein in the ternary complex (norepinephrine–receptor–G protein) is then released when GDP is exchanged for GTP. The G protein dissociates to Gas-GTP and the bg dimer. Gas-GTP is then available to activate adenylate cyclase and increase intracellular levels of cyclic AMP.

There are several subtypes of the muscarinic receptor, and there are many different types of cellular responses which depend upon the subtype of muscarinic receptor and the cells in which the receptor resides. Most, if not all, of these responses are dependent on G proteins. On the other hand, signal transduction is relatively simple at nicotinic acetylcholine receptors. The nicotinic receptor, consisting of five subunits surrounding an internal channel, is its own ligand-gated ion channel (21, 65). After binding two molecules of acetylcholine, the nicotinic receptor channel opens to allow the flow of sodium ions. As such, these sodium ions are the second messengers for nicotinic cholinergic neurotransmission.

In general, pharmacologists define receptors in an experimental system (for example, measurement of agonist-mediated second messenger synthesis by cultured cells) on the basis of selectivity of agonists and antagonists. However, antagonists are more selective for receptor subtypes than are agonists (51). For the muscarinic cholinergic receptor, the agonist muscarine and antagonist atropine have defined this receptor for most of this century.

More recently, pharmacologists had divided muscarinic receptors into two major groups, M1 and M2, with an agonist labeled McN-A-343 being selective for the M1 subtype (38).  1* However, only when the muscarinic antagonist and antiulcer drug pirenzepine (29) became available did researchers have a convincing pharmacologic tool to prove the existence of more than one type of muscarinic receptor. Unlike the classical muscarinic antagonist atropine, pirenzepine blocked different muscarinic receptor effects with different potencies (41). Blockade by pirenzepine at low concentrations (high potency) defined the muscarinic M1 receptor.

Even more recently, pharmacologists suggested that they could subdivide muscarinic receptors into four groupings (M1, M2, M3, and M4) based upon selective antagonists (91). On the other hand, molecular biologists have so far identified five subtypes of the muscarinic receptor (m1–m5)1 ().

The "brain" (m1) (55) and "heart" (m2) (54, 63) receptors were molecularly cloned from the partial amino acid sequences of the purified proteins. The remaining receptors (m3, m4, m5) were cloned by screening for homologous DNA sequences in DNA libraries (15, 16). All five receptors from humans and rats have been molecularly cloned. Studies with these cloned receptors expressed in transfected cell lines have added to our understanding of the selectivity of drugs for these subtypes and the transduction mechanisms associated with these subtypes.

With the use of oligonucleotide probes selective for each of the five subtypes of muscarinic receptors, researchers can show the locations of the cell bodies that synthesize the various receptor subtypes in brain. In particular, messenger RNA (mRNA) for the m1, m3, and m4 subtypes are abundantly and broadly detected in rat brain, including the cerebral cortex, striatum, and hippocampus (18; see also Molecular Biology, Pharmacology, and Brain Distribution of Subtypes of the Muscarinic Receptor). Next in abundance is mRNA for the m2 receptor. The m5 receptor message is least abundant in brain (57, 89). Heart is a rich source for the m2 receptor, and exocrine glands and smooth muscle are rich in the m3 subtype ().

All five muscarinic receptors are homologous proteins consisting of between 460 and 590 amino acids for the human receptors (15, 62). There is a very high degree of homology at the amino acid level for each receptor across species (for example, human m1 has 98.9% identity with porcine m1) (62). Homology between the muscarinic receptor, other receptors (specifically, the b-adrenoceptor), and opsin (55) led to the conclusion that the muscarinic receptors are in the family of seven-transmembranespanning receptors that link to G proteins. The number of proteins in this family numbers greater than 100 at this time. Included among these proteins are receptors for many biogenic amine neurotransmitters (e.g., norepinephrine, dopamine, serotonin) and for many neuropeptides (e.g., substance P, neurotensin).

The fact that all five muscarinic receptors are in this superfamily of G-protein-coupled receptors is important for understanding cholinergic transduction at these receptors. Thus, muscarinic cholinergic transduction at all five muscarinic receptor subtypes depends mostly on the coupling of each receptor to G proteins. This coupling is thought to involve the third intracellular loop of the protein () (see Molecular Biology, Pharmacology, and Brain Distribution of Subtypes of the Muscarinic Receptor and Signal Transduction Pathways for Catecholamine Receptors).

To discuss cholinergic transduction at specific muscarinic receptor subtypes and to interpret studies with animal tissues, one must first have knowledge of the pharmacological identification of these receptors. The corollary question of that posed above is: are there drugs that specifically or selectively block or stimulate muscarinic receptor subtypes?

A compound that is specific for a receptor binds to only one receptor. A compound that is selective for a receptor binds to one receptor with a higher affinity than that for its binding to another receptor. It is generally considered that a compound shows selectivity for a receptor when its affinity for one receptor is at least 10-fold greater than that for another receptor.

The definitive answer to the question posed above has come from studies with different transfected cell lines, each expressing a different subtype (m1–m5) of the muscarinic receptor. Currently, there is no selective or specific antagonist or agonist of the five muscarinic receptors (14, 27, 96). Thus, pirenzepine, the "M1-selective" antagonist, is not sufficiently selective to distinguish m1 from m4 receptors, although it will distinguish m1 from m2, m3, and m5 receptors (14, 27) (). The "M1-selective" agonist McN-A-343 has effects on all five receptors, with perhaps its best effects on m1 and m4 receptors (96). Until the chemistry and pharmacology catches up with the molecular biology, a combination of pharmacology and molecular biology must be used to classify a particular muscarinic response with a specific receptor subtype (see Molecular Biology, Pharmacology, and Brain Distribution of Subtypes of the Muscarinic Receptor). Therefore, this review will emphasize results with the molecularly cloned muscarinic receptors expressed in host cells.

There are caveats to consider when generalizing results with experiments involving cloned receptors to what occurs with receptors in their natural setting. In these experiments the host cell provides the machinery necessary for the signal transduction to occur. In this case the host cell may not have the necessary apparatus (as discussed below for muscarinic-receptor-mediated cyclic GMP synthesis) or may have components that are not present in those cells in vivo expressing these receptors. In addition, expression of these receptors (measured as Bmax in a radioligand binding assay) can be orders of magnitude greater than that found in the natural state. As a result, these receptors can couple to transduction mechanisms that would not occur if the receptors were expressed at lower levels. High levels of receptors can also increase the efficacy of agonists for these receptors (51).

About 40 years ago, Hokin and Hokin (43) first reported the "phospholipid labeling effect" for acetylcholine at muscarinic receptors. More specifically, these researchers showed that acetylcholine and the cholinergic agonist carbamylcholine, which cause enzyme secretion in the pancreas, markedly increase the incorporation of 32P into phospholipids in tissue slices of pigeon pancreas. The effects of these agonists were blocked by the muscarinic antagonist atropine, establishing the muscarinic nature of this effect.

This phospholipid labeling effect of muscarinic agonists is the result of the breakdown of a membranal phospholipid, phosphatidylinositol 4,5-bisphosphate (PIP2), caused by muscarinic receptor (m1, m3, or m5) activation of the enzyme phospholipase C (PLC-b1 isozyme) through a G protein called Gq (). This activation of PLC leads to a cascade of events. As illustrated in , these events, with a time course measured in seconds, begin with the breakdown of PIP2 into 1,2-diacyl-sn-glycerol (diacylglycerol, DAG) and D-myo-inositol 1,4,5-trisphosphate (IP3) (10, 33).

DAG is recycled, in a series of reactions, back to PIP2 (). Additionally, DAG becomes available to activate the enzyme protein kinase C (PKC), of which there are several subtypes (34). In its activated form, PKC can then phosphorylate proteins, thereby regulating their functions. For example, activation of PKC by phorbol esters can inhibit muscarinic-receptor-mediated release of inositol phosphates in cultured murine neuroblastoma cells (49). These results suggest a feedback role for DAG in signal transduction through PKC. DAG can also be degraded by diglyceride lipase to form arachidonic acid, the pivotal substrate for the synthesis of prostaglandins, leukotrienes, epoxides, and related compounds.

Lithium ion inhibits the enzyme inositol 1-monophosphatase, which hydrolyzes IP1 into inositol (40). Thus, in vitro lithium ion is used to amplify the signal resulting from receptor-mediated release of inositol phosphates (9). It is uncertain whether this action of lithium ion plays a role in its therapeutic effects in treating affective disorders. However, it may play a role in some of its adverse effects (11).

IP3 has receptors on smooth endoplasmic reticulum (23), which release stored calcium ions after the IP3 receptor is activated (8) (). Increased intracellular levels of calcium ions then cause a myriad of events to occur within the cell, including the potentiation of the effects of PKC.

Receptor-mediated activation of PLC is by no means exclusive to muscarinic subtypes m1, m3, and m5, or to muscarinic receptors in general. The odd-numbered muscarinic receptors are most efficiently coupled to this response and give the most robust responses, compared to those of the even-numbered receptors (3, 64). In addition, the odd-numbered muscarinic receptors activate PKC (b1 isozyme) by a pertussis-toxin-insensitive G protein (Gq). The other two receptors mediate this activation by a pertussis-toxin-sensitive G protein, likely Gai (25). There is an extensive list of receptors that mediate the release of inositol phosphates. A few examples of these are the a1-adrenergic, histamine H1, serotonin 5-HT2, and neurotensin receptors. Finally, there are receptors (e.g., dopamine D1) that inhibit receptor-mediated inositol phosphate release, including that at muscarinic receptors (94).

Phospholipases of the D type are enzymes that cleave choline from the membranal phospholipid, phosphatidylcholine (lecithin), yielding phosphatidic acid and choline, an acetylcholine precursor (19, 31). A phosphatidatephosphohydrolase acting on phosphatidic acid yields DAG. Thus, PLD acting on phosphatidylcholine provides another source of DAG to activate PKC, which is known to activate PLD (39). In fact, phosphatidylcholine, being much more abundant, provides a greater and more sustained source of DAG than does PIP2.

All receptors that stimulate the breakdown of PIP2 may also stimulate the breakdown of phosphatidylcholine. Muscarinic receptors activate PLD by a mechanism that appears to involve G proteins (66, 72, 74). As with the coupling to PLC, the odd-numbered muscarinic receptors are more efficiently and robustly coupled to PLD (73). This type of signal transduction has been much less well-characterized than that involving the phosphatidylinositol-specific PLC. However, this coupling of muscarinic receptors to PLD may be of importance to cholinergic transmission in general, by providing the precursor of acetylcholine and by providing other compounds that may affect other signal transduction mechanisms.

Arachidonic acid is a bioactive compound that serves as a precursor of many compounds derived from the action of lipoxygenases, epoxygenases, and cyclooxygenases. Additionally, arachidonic acid is involved with the activation of PKC (see Arachidonic Acid).

Several different receptors, including the odd-numbered muscarinic receptors (20, 32, 80, 83), can mediate the release of arachidonic acid. Thus, by itself or through a metabolite, arachidonic acid is a second messenger.

The principal effector involved in this receptormediated arachidonic acid release is PLA2 (7). This enzyme acting on a variety of phospholipids at the 2-position releases arachidonic acid with the formation of lysophospholipids. Another enzyme that can release arachidonic acid is diglyceride lipase, which releases this compound from DAG, produced by the action of PLC on phosphatidylinositol 4,5-bisphosphate (26). There is good evidence to suggest a role for G proteins in receptor-mediated PLA2 activation (7, 53), although other mechanisms have been suggested (17, 32). The exact subtype of G protein involved with PLA2 activation is unknown.

Cyclic guanosine 3¢,5¢-monophosphate (cyclic GMP) and the enzymes (soluble and particulate) that synthesize this cyclic nucleotide are widely distributed in brain and elsewhere in the body (37, 93). Cyclic GMP formation mediated by acetylcholine was first reported over two decades ago with rat heart (35). Shortly thereafter, muscarinic responses in other tissues were reported. The dependence of the response on calcium ions was established early (77). All the major target organs of parasympathetic cholinergic fibers contain muscarinic receptors that mediate an increase in cyclic GMP (37). Muscarinic receptors in sympathetic ganglia and in brain also mediate cyclic GMP synthesis.

Many different established cell lines of nervous system origin have muscarinic receptors that mediate cyclic GMP synthesis. A widely studied example is murine neuroblastoma clone N1E-115 cells (2, 59, 69). In this model system, the muscarinic receptor and six others that increase intracellular levels of the second messenger cyclic GMP in a calcium-dependent manner mediate the release of inositol phosphates (69). Calcium-dependent, receptor-mediated cyclic GMP synthesis occurs only with intact cells. Calcium ions and nitric oxide (NO) (60, 61) are very likely involved in this receptor-mediated stimulation of soluble guanylate cyclase. NO, which is synthesized from L-arginine by the enzyme NO synthase in a calcium/ calmodulin-dependent manner (61), directly stimulates soluble guanylate cyclase.

NO, a free radical, is a novel second messenger of muscarinic and other receptors. It may be involved with both intracellular and intercellular communication of neurons (see Nitric Oxide and Related Substance as Neural Messengers).

The role of G proteins in muscarinic receptor-mediated cyclic GMP synthesis has not been defined. However, it may be that synthesis of NO and, subsequently, cyclic GMP following receptor activation is secondary to the increase in intracellular calcium ions, resulting from the release of IP3 from PIP2 by the action of PLC.

The human muscarinic receptors expressed in Chinese hamster ovary cells, despite robust release of inositol phosphates, do not mediate cyclic GMP synthesis. These cells lack NO synthase and guanylate cyclase. These facts likely explain the absence of muscarinic-receptor-mediated cyclic GMP synthesis in Chinese hamster ovary cells.

With clone N1E-115 cells, there is a very close association between the activation of muscarinic M1 (and other) receptors and the formation of inositol phosphates and cyclic GMP. These results suggest that the two events are linked to one another and that one (cyclic GMP response) could be dependent on the other (inositol phosphate release). However, with rat brain tissue, there is evidence to suggest that different subtypes of the muscarinic receptor mediate these responses independently (52, 88). Although more research needs to be done on this topic, these data could be explained in part by the absence of NO synthase in certain cells that have muscarinic receptors (see Nitric Oxide and Related Substance as Neural Messengers).

Another biochemical property established early for muscarinic receptors is their "negative coupling" to adenylate cyclase (37), the enzyme that synthesizes cyclic AMP from ATP. Negative coupling means the inhibition by the muscarinic receptor of adenylate cyclase activation mediated by a second receptor. Other types of receptors negatively couple to adenylate cyclase (e.g., a2-adrenoceptors, delta-opioid receptors). It is mediated by the Gai subunit of the G protein, although the bg subunits may also be involved (87). The even-numbered muscarinic receptors, m2 and m4, are the ones that use this type of signal transduction mechanism.

Muscarinic receptor activation can reduce cyclic AMP levels by another mechanism (45), involving the activation of cyclic AMP phosphodiesterase, the enzyme that degrades cyclic AMP. This effect appears to be secondary to mobilization of intracellular calcium ions by the action of PLC and the subsequent activation of a calcium/ calmodulin-dependent phosphodiesterase. So far, this muscarinic response has been shown to occur only with a human astrocytoma cell line. This cell line has the adenosine receptor that inhibits adenylate cyclase activity by a pertussis-toxin-sensitive mechanism, indicating a requirement for Gai. These results suggest that, unlike the molecularly cloned muscarinic receptors, the muscarinic receptor in these tumor cells is incapable of coupling to this type of G protein.

Activation of muscarinic receptors leads to a diverse array of electrical responses within cells. The type of response depends upon the subtype of muscarinic receptor and the type of cell involved. Some examples of these responses are inhibition and stimulation of inward rectifier currents (see Electrophysiology). As might be expected from the preceding, studies with the molecularly cloned receptors appear to show that the odd-numbered muscarinic receptors evoke similar electrical responses and that these responses are different from those of the even-numbered receptors. All these electrical responses appear to require coupling to G proteins. The reader is referred to a recent publication (48) for an in-depth review of this topic.

Researchers have published hypotheses and supporting data implicating the muscarinic cholinergic system in affective disorders (46, 47, 79) and in schizophrenia (24, 50, 85, 86). Only one study, involving manic patients, has looked at muscarinic receptor transduction (76). Other studies in the literature involving muscarinic transduction used brain material from patients who died from Alzheimer's disease. One group suggests that muscarinic receptor coupling to G proteins is altered in brains of Alzheimer's patients (97; see also Experimental Therapeutics).

Animal studies suggest that lithium ion inhibits the coupling of muscarinic and b-adrenergic receptors to G proteins (5). More specifically, these studies showed that agonist-induced increases in the binding of [3H]GTP to pertussis-toxin-sensitive (muscarinic receptor) and cholera-toxin-sensitive (b-adrenergic receptor) G proteins in cerebral cortex was blocked by treatment of rats for 12–21 days with lithium carbonate. This effect is reversed in vitro by magnesium ions (4). Consistent with these data, this treatment abolished the GTP shift of the agonist to a lower-affinity state as measured in binding assays with a nonselective muscarinic radioligand. The pertussis toxin-sensitivity of this effect suggests that Gi or Go is involved and not Gq.

Evidence was presented to suggest that the muscarinic receptor involved is the M1 subtype (6). More work needs to be done to prove this point (for example, obtaining equilibrium dissociation constants for a series of antagonists and not IC50 values) because this conclusion does not fit with the molecular biology and molecular pharmacology. It is more likely that the m4 receptor is involved (see Molecular Biology, Pharmacology, and Brain Distribution of Subtypes of the Muscarinic Receptor).

These animal studies suggested that receptor coupling at muscarinic and b-adrenergic receptors is awry in bipolar disorder, the illness effectively treated with lithium salts. Therefore, the same researchers sought clinical data to test this idea (76). The source of receptors was mononuclear leukocytes. These were obtained from untreated bipolar patients in the manic phase of their illness, from euthymic patients treated with lithium salts, and from healthy volunteers. Similar to the results found in the cerebral cortex of rat, agonists of the two receptors stimulated the binding of an analogue of GTP to membranes from the leukocytes. In addition, this stimulation was abolished at muscarinic and b-adrenergic receptors by pertussis toxin and cholera toxin, respectively.

Interestingly, there was a much more robust stimulation of the binding at both receptors for leukocytes from untreated manic patients compared to that for controls. In addition, the effects of agonists with the membranes of cells from the treated, euthymic bipolar patients were no different from those of controls. Although these studies need to be replicated, it does suggest that a defect exists in certain types of G proteins in bipolar patients.

Antidepressants, neuroleptics, and antiparkinsonism drugs antagonize muscarinic receptors with varying degrees of potency and selectivity. Interruption of muscarinic cholinergic transduction is effectively achieved by this muscarinic receptor blockade.

We have studied many compounds within these three classes of drugs at the five cloned human muscarinic receptors expressed in Chinese hamster ovary cells (14, 82) (). No drug is specific or selective (as defined above) for the five receptors. However, some compounds are relatively potent and relatively selective. For example, the classical tricyclic compound amitriptyline is the most potent antidepressant at all five muscarinic subtypes. The neuroleptic clozapine is very potent and relatively selective for the m1 subtype. This selectivity may explain clozapine's unusual efficacy in refractory schizophrenic patients and its low incidence of extrapyramidal side effects. However, because most other atypical neuroleptics studied lacked high affinity and selectivity at muscarinic receptor subtypes, it is likely that other mechanisms are involved as well. Potent blockade of muscarinic receptors can also cause a number of adverse effects such as memory dysfunction (possibly m1), sinus tachycardia (m2), and dry mouth (possibly m3).

The existence of muscarinic receptors has been known for most of this century. We enter the next century with the knowledge that there are at least five different muscarinic receptors that can play a whole host of "instruments" to control many aspects of the functioning of the organism. In the total picture, however, muscarinic receptors are several among the perhaps hundreds of receptors for neurotransmitters, neurohormones, and neuromodulators. There are many interactions that are known between muscarinic and other receptors. Many more are yet to be found.

There are suspicions based upon experimental data that a few more muscarinic receptors are waiting to be found. In the future, these other muscarinic receptor subtypes may be molecularly cloned. If this were the case, muscarinic receptor pharmacology would become even more complex. As it is we must await breakthroughs from the chemists, collaborating with molecular biologists and pharmacologists to obtain the truly specific or truly selective muscarinic receptor ligands. However, these receptors are so highly homologous, especially in regions where they are thought to bind agonists and antagonists, that some think we may never obtain the desired compounds.

How else will we be able to assign function to these receptors in the brain as we await the specific and selective compounds? One approach to consider is antisense technology. This technique can selectively inhibit the expression of a gene of interest. By observing function or behavior and after knocking out the gene (92), we may be able to assign function to each of the five muscarinic subtypes. Whether antisense technology will provide new muscarinic receptor "antagonists" is the subject for some exciting future research.

The writing of this chapter was supported in part by the Mayo Foundation and U.S.P.H.S. Grant MH27692.

published 2000