INTRODUCTION

Acetylcholine (ACh) receptors in the mammalian CNS can be divided into muscarinic (mAChR) and nicotinic (nAChR) subtypes based on the ability of the natural alkaloids, muscarine and nicotine, to mimic the effects of ACh as a neurotransmitter. Until recently, studies of the neuropsychopharmacological effects of ACh have focused on mAChRs, while nAChRs have been evaluated primarily for their role in mediating neuromuscular and autonomic transmission.  However, over the last decade, this trend has changed following preclinical and clinical studies indicating that neuronal nAChRs may have a substantial role in mediating antinociception, cognitive performance, modulating affect, and enhancing the release of other neurotransmitters.

The majority of evidence defining potential therapeutic targets involving neuronal nAChRs has resulted from studies on the effects of (-)-nicotine in a variety of preclinical and, to a lesser extent, clinical models.  And, while a significant number of neuronal nAChR receptor subtypes have been potentially identified based on subunit structure at the molecular level, little is known regarding the physiological role of most of these receptor subtypes beyond what can be deduced from their discrete localization within brain tissue. The development of receptor subtype selective ligands, especially antagonists, may be anticipated to facilitate the definition of receptor subtype function.  In this chapter the molecular biology of neuronal nAChRs is discussed within the context of the pharmacology of agonists, antagonists and allosteric modulators.  In addition, the potential CNS therapeutic targets for nAChRs are reviewed.

Classification of nAChRs

Workers at the Salk Institute and at the Institut Pasteur have established that the muscle nAChR is a ligand-gated ion channel (LGIC) receptor composed of 5 subunits—two a1 subunits, and one each of b1, d and g (or e, depending on the stage of development) [14].  Several genes have been identified in rat and chick neural or sensory tissue that encode for neuronal nAChR subunits that are distinct from those in the muscle nAChR, providing for a multitude of potential subtypes of neuronal nAChRs.  The wide distribution of the some of these transcripts in mammalian brain indicates that neuronal nAChRs represent a major neurotransmitter receptor superfamily related to other LGICs including serotonin (5HT₃), GABA_A, N-methyl-D-aspartate (NMDA), and glycine.  However, in contrast to these other LGICs where established pharmacology rapidly segued into the molecular biology, the pharmacology of neuronal nAChRs has only started to emerge as a result of the rapid advances in the molecular biology of the nAChR family.

Receptor nomenclature in the nAChR area has been derived from classical pharmacology approaches, including receptor sensitivity to snake toxins.  Following from Dale's conceptualization of ACh receptor subtypes over 80 years ago, Barlow and Ing and Paton and Zaimis showed that the antagonist decamethonium (C10) was more effective than hexamethonium (C6) in blocking muscle nAChRs, whereas C6 was effective in autonomic ganglia [83].  This led to the description of 'C10 ' (muscle) and 'C6' (neuronal) receptors.  Muscle nAChRs are selectively activated by phenyltrimethylammonium and are pseudo-irreversibly blocked by a-bungarotoxin (a-BgT).  Ganglionic nAChRs are preferentially activated by 1,1-dimethyl-4-phenylpiperazinium (DMPP), competitively blocked by trimethaphan, and are resistant to snake a-toxins, yet sensitive to neuronal bungarotoxin (n-BgT: also known as k-BgT, a-BgT 3.1, or toxin F).

In mammalian brain, two major neuronal nAChR subclasses can be delineated using radioligand binding [17]: those that bind a-BgT with high affinity (a-BgT-sensitive nAChRs (Kd ~ 0.5 nM using [¹²⁵I]a-BgT) and those that do not (a-BgT-insensitive nAChRs).  a-BgT-sensitive nAChRs have low affinity for (-)-nicotine, whereas a-BgT-insensitive nAChRs have high affinity (Kd = 0.5 - 5 nM) for [³H](-)-nicotine, [³H]ACh, [³H]methylcarbamylcholine (MCC), and [³H]cytisine).  All four of these [³H]agonist ligands are thought to interact with the same ACh binding sites on the nAChR.

Molecular Diversity of nAChRs  

nAChRs are encoded by a family of related but distinct genes that share a common origin and have a long phylogenetic history.  To date eleven neuronal nAChR subunits have been described;  Eight (a2 - a9) code for a subunits (see [7,14] for reviews) based on the presence of adjacent cysteine residues in the predicted protein sequences, in a region homologous to the putative agonist binding site of the muscle a subunit (a1) while three are referred to as  non-a or b-subunits  (b2-b4).  Each of the nAChR subunits displays a characteristic phenotype of structural features extending from the N-terminus to the C-terminus (Figure 1): (1) a large (~200 amino acids) N-terminal hydrophilic domain  containing the multiple loops of the neurotransmitter binding site; (2) the highly variable C-terminal hydrophilic domain that faces the cytoplasm, where it can be phosphorylated; and (3) a set of four closely spaced transmembrane domains—termed M1-M4—immediately following the large extracellular domain.  The M2 domain is believed to form the wall of the ion channel.  The amino acid sequences of the neuronal nAChR genes except a7 and a8 are between 40 - 60% similar [7,49].  The a7 and a8 genes share approximately 70% similarity, with a much lower level (<30%) of similarity to the other nAChR genes.  The a9 subunit is the most unique of the a-subunits displaying less than 50% similarity to a2-a8 [26].  Comparison of the nAChR sequences also reveals that several pairs of subunits display homologies that are instructive in considering their function.  For example the b2 and b4 are highly homologous, consistent with their ability to substitute for one another in forming functional channels when paired with a2, a3, or a4 subunits (see below).  The similarities in sequence between a6 and a3 may have caused misidentification of a6 subunits as a3 subunits in early localization studies on the a3 subunit ([41], but see [87]).

The chromosomal localizations of many of the human neuronal nAChR genes have been determined [34]. The human a4 gene is located on chromosome 20 and is associated with an autosomal dominant nocturnal frontal lobe epilepsy [44,49]. The human a7 gene has been mapped to chromosome 15 and has been associated with some forms of schizophrenia [46].  Additionally, the human a3, a5 and b4 are tightly linked on chromosome 15 while the a2 and b3 genes are found on chromosome 8 [34].   

Although all neuronal nAChRs are pentamers, the structure and stoichiometry of many neuronal nAChRs remains to be fully defined.  The a7, a8 and a9 subunits can form functional pentameric homo-oligomers while the stoichiometry of a4b2 nAChRs consists of two a4 and three b2 subunits [7].  It is believed that pairwise expression of a2, a3, and perhaps a6 subunits with either b2 or b4 subunits results in similar stoichiometries.  More complex combinations containing three (a3b4a5) or four subunits (a3b2b4a5) have been immunoisolated from brain [49].  Site-directed mutagenesis of affinity labelled residues in the channel and binding site domains have demonstrated that mutations of single amino acids can modify multiple functions of the nAChR [14].  For instance, mutations in the M2 channel lining domain of the a7 subunit can produce either a significant increase in apparent affinity for agonists or a loss of desensitization and a conversion of competitive antagonists to agonists [14]. 

CNS Expression of Neuronal nAChRs

Both a-BgT-sensitive and a-BgT-insensitive nAChRs have been extensively mapped in rodent brain and, to some extent, in human brain.  Using radioligand binding and in situ hybridization, the topographical distribution of nAChRs corresponds well with the effects elicited by (-)-nicotine and the known functions associated with each brain region. 

In rat, high affinity nAChR sites revealed by [³H](-)-nicotine, are abundant in selective areas of the cerebral cortex (predominantly layers III & IV ), thalamus, interpeduncular nucleus and the superior colliculus, but are of low to moderate abundance in the hippocampus and hypothalamus (Table 1).  The second class of sites, labeled by [¹²⁵I] a-BgT, are enriched in the hippocampus, hypothalamus and layers I and VI of the cerebral cortex [17].

In situ hybridization assays have demonstrated that the a4 subunit mRNA is expressed strongly in a number of areas including the thalamus, deeper layers of the cerebral cortex, ventral tegmental area (VTA), the medial habenula and substantia nigra (SN) pars compacta [44] and that a3 is expressed strongly in the locus coeruleus (LC), habenula, and interpeduncular nucleus [87].  These subunits are also expressed in peripheral tissue, with a3 being expressed in autonomic and sensory ganglia and a4 being expressed in the trigeminal ganglion [28].  The a2 subunit has a much more restricted pattern of expression, with mRNA being expressed at high levels only in parts of the interpeduncular nucleus [44].  a7 message is particularly high in the hippocampal formation, which correlates well with the high level of a-BgT binding this region [44].  a8 has not been found in rat and the distribution of a9 appears restricted to skin and sensory tissue.  The presence of a9 subunits in nAChRs in cholinergically-innervated outer hair cells of the cochlea may explain the unusual cholinergic pharmacology found in this tissue [59].  Among the b subunits, b2 is widely expressed, whereas expression of b4 and b3 is more variable [44].  b4 message is highly expressed in a number of areas in which substantial a3 message is found, including the habenula, area postrema, and LC [87].  One notable exception to this pattern is that substantial expression of a3 message is found in DA neurons in the SN, whereas b4 message is absent [87].  b3 message appears to be prominent in brainstem catecholaminergic areas such as SN and LC [44].  Similarities in the distributions of a6 and b3 message suggest that these two subunits might be important in the nAChR-mediated effects on catecholamine release [44], although the lack of pharmacological tools selective for nAChR subtypes containing these subunits has restricted investigation of the pharmacology of this effect (see below). 

Localization of mRNA or binding sites to particular brain nuclei does not reveal which cell types in the identified regions express nAChR subunits.  Interest in this more fine-grained analysis has accelerated of late, although information is still somewhat sparse.  Dopamine neurons in the SN appear to express the a3 subunit but not the b4 subunit, since neurons labeled by an antibody for tyrosine hydoxylase in the SN express mRNA for a3, but not b4.  Moreover, tyrosine hydroxylase-positive neurons in the SN and the VTA can also immunostained by an antibody for the a6 subunit [32].  Similarly, double labeling experiments suggest that serotonergic neurons in the brainstem express the a4 subunit [6].

Much less is known about the expression of nAChR subunit genes in human brain (for review see [34]).  In humans high densities of high affinity nicotine binding sites are found in some areas that are also rich in a4 message in rats, such as SN (pars compacta) and thalamus; and a7 message and a-BgT binding are prominent in the hippocampal formation of both rats and humans [8,34].  Differences do exist in the regional distribution of these subunits across species, but differences in the post-mortem handling of tissue from these species may account for some of these discrepancies [8].  Whether interspecies  differences in subunit distribution translate into differences in pharmacological properties across species will require further investigation.    

Functional Analysis of nAChRs in Heterologous Expression Systems Has Revealed Pharmacological Diversity

The co-existence of different nAChR subunits in the same central or peripheral nervous system pathway (and even in the same neuron) has made the in vivo study of the properties of individual  nAChR subtypes extremely difficult.  For this reason, heterologous expression studies in either Xenopus oocytes or mammalian cell lines have provided considerable information on the electrophysiological and pharmacological properties of different subunit combinations. 

Functional responses occur in oocytes or cell lines transfected with pairwise combinations of rodent, avian or human a and b subunits, confirming biochemical findings which suggest that many native nAChRs consist of a/b heteromers [15,34].   These heterologous expression studies have demonstrated that both the a and b subunits determine the functional and pharmacological properties of  the subunit combination.  For example, when expressed with the b2 subunit, the a2, a3 and a4 subunits all form functional channels that differ in their channel open times, single channel conductance and agonist/antagonist sensitivity. However, not all subunit combinations form functional nAChRs.   The b3 and a5 subunits are unable to form functional nAChRs when expressed alone or in pairs with other subunits but instead appear to act as “regulatory” subunits [34].  For example, in combination with the human a3 and b2 or b4 subunits, a5 increases the desensitization rate, and in the case of the a3b2 combination, a5 significantly alters the EC₅₀ values for nicotine and ACh.  It may be that the a5 subunit occupies a position equivalent to the b1 subunit of muscle nAChRs so that it interfaces with the surfaces of adjacent subunits that are incapable of forming ACh-binding sites [86]. 

Studies of the single channel properties of neuronal nAChRs expressed in oocytes indicate considerable diversity among heterologously expressed subunit combinations [62]. For example, two distinct populations of open channel conductances can be observed after injection of either a2b2 or a3b2 subunits into oocytes. In contrast, the a4b2 subunit combination generates only a single type of channel.  Of the b2 - containing receptors, the current of the a3b2 receptors is the most sustained while the a2b2 combination gives the greatest peak current. nAChRs containing the b2 subunit are thus likely to generate brief synaptic currents in vivo, creating the potential for rapid signal processing.  In contrast, currents for the a3b4 subunit combination are of a smaller conductance but do not desensitize as rapidly.  Accordingly, if a3b4 receptors predominate at synapses, responses may be prolonged, providing more time to organize/integrate a cellular response.

a7, a8  and a9 gene products differ from other members of the nAChR superfamily in that they can form functional receptors in oocytes when expressed as homo-oligomers [7,34]. The most striking pharmacological characteristic of the a7 homo-oligomeric receptor is its marked permeability to calcium ions [1].  This finding, coupled with the unique distribution pattern of this receptor in brain [17] has led to heightened interest in the a7 subtype as a potential therapeutic target.  It is noteworthy that most of the structures with the highest abundance of a7 transcripts in rodents are major components of the limbic system.  A role for this receptor subtype in the regulation of neurite outgrowth and survival has also been suggested [25].  

Considerable progress has been made in the understanding of the pharmacological diversity of nAChRs based on the availability of  herterologous expression systems expressing various a/b subunit combinations.  The rank order of potency of four nAChR activators, ACh, (-)-nicotine,   (-)-cytisine, and DMPP on receptors formed from human b2 or b4 subunits in combination with a2, a3, or a4 subunits in Xenopus oocytes, has established the importance of both a - and b-subunits in defining the pharmacological properties of the nAChR with these distinct subunit combinations ([15]; Table 2).  Cytisine was the least efficacious agonist at AChRs containing b2 subunits but it displayed significant activity at b4 containing subunits.  ACh was the most efficacious agonist at all AChRs except a3b2 but was among the least potent of the agonists at all nAChR subunit combinations.  While studies using cell lines stably expressing these same human nAChR subunit combinations have revealed a similar rank order of potency for these agonists [33,78], some important differences in either potency and/or efficacy were observed.  The pharmacological differences detected in Xenopus oocytes and cell lines stably expressing defined human nAChR subunit combinations may result from differences in the methods used to analyze function (electrophysiology versus ion flux), differences in subunit stoichiometries, or differences between the two expression systems.  Additionally, significant pharmacological differences in the agonist potencies and sensitivities between the oocyte-expressed a3b2 and a3b4 human subtypes and the same rat and/or chick subtype have been observed [34].  For example, (-)-nicotine is more potent than ACh at the human a3b2 subtype but it is less potent than ACh at the rat subtype.  These species differences suggest that ascribing functional/behavioral effects of nAChR ligands to a specific rodent subtype(s) based on data generated using human nAChRs expressed in heterologous expression systems may not be valid.

The homomeric a7- a9 subunits expressed as homo- oligomers in oocytes also form functional cation channels gated by nicotinic agonists with differing pharmacological properties. The agonist sensitivities of chick a7 and a8 expressed in oocytes show that the a8 homomers exhibit higher affinity for nicotinic agonists as compared to a7 homomers.  The order of potency for a8 was (-)-nicotine = (-)-cytisine (EC₅₀ = 1 µM) ~ ACh > DMPP > tetramethylammonium.  In  contrast, DMPP is a very weak partial agonist for a7 and tetramethylammonium has no effect. a9 nAChRs expressed in vitro exhibit the most unique pharmacological properties identified to date.  (-)-Nicotine, atropine, strychine, d-tubocurare and muscarine all behave as antagonists while DMPP acts as an agonist [26].

While studies in oocytes and cell lines have yielded some clues as to the physiological/pharmacological roles of the different nAChR subunits in vivo, some caution is required in interpreting these findings because of the atypical nature of the host cell environment.  In addition, the expression of multiple nicotinic genes in central and peripheral tissues suggests that some nAChR subtypes in these areas may contain more than two types of subunit [84], a finding that would explain why some of the pharmacological properties of receptors formed by injection of a single subunit (a7) or the pairwise combination of a/b subunits do not correlate well with the properties of receptors found in neurons [18,50].  Despite the lack of precise correspondence between nAChRs in these in vitro expression systems and native receptors, pharmacological data derived from oocyte studies have been useful in characterizing native nAChRs.  For example, three distinct hippocampal nAChRs identified electrophysiologically were designated as a7-containing, b2-containing, and b4-containing based on the similarities between the characteristics of these native receptors and characteristics of a7-containing, b2-containing, and b4-containing nAChRs expressed in oocytes [1].  The accuracy of this designation has recently been supported by the demonstration that only the current originally believed to be mediated by a b2-containing receptor is absent in b2-knockout mice, and only the current believed to be mediated by an a7-containing receptor is missing in a7-knockout mice [61,89].

Information from heterologous expression studies combined with in situ hybridization studies has identified intact in vitro  model systems in which the pharmacology of compounds acting at putative subtypes of neuronal nAChRs can be evaluated using electrophysiological and biochemical techniques.  For example, the a4b2 subunit isoform appears to modulate the flux of monovalent ions as measured by efflux of [⁸⁶Rb⁺] from thalamic synaptosomes [54] and is believed to play a role in the release of ACh from rat hippocampus (see below).

^{Thus, heterologous expression studies have started to refine the combinatorial diversity of neuronal nAChR subunits suggested by the initial cloning and expression studies.  A significant challenge in the years to come will be to relate the properties of nAChRs recorded in vivo at the level of a particular neuronal pathway, with a defined homo-oligomeric or hetero-oligomeric subunit combination.  Such an understanding may pave the way to “circuit-targeted” molecular pharmacology and drug discovery [14].}

Functionally Distinct Transition States of nAChRs

In addition to structurally distinct subtypes of nAChRs, there are functionally distinct transition states for an individual nAChR.  Current evidence regarding the states of activation and desensitization of nAChRs derives primarily from work on the muscle and a7 nAChRs [14].  Distinct ligand binding sites, some sensitive to ACh and (-)-nicotine and others involving distinct classes of allosteric modulator sites on, and between, the various receptor subunits, can cooperatively modify, either positively or negatively, the equilibrium between the receptor states affecting the proportion of receptors existing in each state but not significantly altering the intrinsic binding and physiological properties of the states themselves.  Thus the nAChR functions within the context of the classical allosteric “concerted scheme” [12,14] for oligomeric proteins that incorporates the multiple state concept originally proposed by Katz and Thesleff for the nAChR.

The allosteric transition state model considers a minimum of four interconvertible states with differing rates of interconversion: a resting state (R); an activated state (A) with the channel opening in the ms to ms timescale and having low affinity (mM to mM) for agonists; and two 'desensitized' closed channel states (I or D) that are refractory to activation on a ms (I) to minute (D) timescale but exhibit a high affinity (pM - nM) for nAChR agonists and some antagonists.  nAChR ligands may therefore be considered to differentially stabilize the conformational states to which they preferentially bind. 

A more persistent modulation of nAChR function can occur by phosphorylation of the receptor protein [85].  While little is known regarding phosphorylation of neuronal nAChRs, the sites of phosphorylation and the associated protein kinases have been well characterized in the Torpedo  receptor.  At least four kinases differentially phosphorylate muscle and Torpedo nAChR subunits: cAMP-dependent kinase (PKA); protein kinase C (PKC), which also phosphorylates the neuronal receptor; a tyrosine kinase; and a Ca²⁺ -calmodulin kinase.  Phosphorylation can enhance the rate of nAChR desensitization and increase the frequency of spontaneous channel openings.

Sites of nAChR-Ligand Interaction

Evidence is rapidly emerging to indicate that the nAChR channel may be activated through sites distinct from the classical ACh binding sites and suggests that “cholinergic channel modulators” (ChCMs) may be a more appropriate, and all encompassing, classification for those compounds that activate, inhibit or desensitize nAChRs [10].

The ACh binding site.  Binding site(s) for cholinergic ligands on the nAChR were initially thought to reside solely on the a subunit. More recently, site - directed mutagenesis studies have shown that binding sites for cholinergic ligands on nAChR are located at the interfaces between the a and b subunits in heteromeric receptors and between a subunits in homomeric receptors [14].  For example, a4-containing and a3-containing neuronal nAChRs differ dramatically in their sensitivity to nicotinic agonists and antagonists.  Analysis of chimeric subunits consisting of portions of these two a subunits have indicated that the region from the amino terminus to position 84 is important in determining sensitivity to the agonists, ACh and   (-)-nicotine, positions 84 to 121 and from position 121 to 181 contain amino acid residues important in determining n-BgT - sensitivity, while the sequence segment from position 195 to 215 is important for both agonist and antagonist sensitivity.  In particular, the amino acid residue at position 198 (glutamine in a3 and proline in a2) is believed to be important in determining the sensitivity of neuronal nAChRs. In the case of the a4 subunit, amino acids 151-155 and 183-191 have been found to confer the physiological and pharmacological properties typical of the a4b2 receptor [19].

Alternative  Channel "Activator" sites.  Neuronal nAChR function may also be enhanced via ligand binding sites distinct from those at which ACh or (-)-nicotine interact. These sites are thought to be present at the level of the a subunit and are not subject to the same desensitization mechanisms described for (-)-nicotine.  Compounds that interact with this novel site to increase neuronal nAChR mediated ion conductance have been termed “channel activators.”  The cholinesterase inhibitors physostigmine and galanthamine, (+)-2-methylpiperidine, and the antihelminthic agent, ivermectin, are examples of compounds that act as channel activators at this site which is distinct from the (-)-nicotine site [1,12].   Ivermectin (30 mM) has been shown to enhance ACh-evoked current in chick or human a7 nAChRs.  The concomitant increase in apparent affinity and cooperativity of the ACh dose-response curve suggests that ivermectin acts as a positive allosteric effector of this subtype [12]. 

Alternative ligand-binding sites that modulate nAChR function.  Based primarily on work from the muscle nAChR, and supported by preliminary work from the neuronal nAChR, there is evidence to indicate that there are a number of other ligand-binding sites that can modulate neuronal nAChR function.

Non-competitive (negative allosteric modulators) blockers.  A number of chemically diverse molecules, including mecamylamine (Fig. 2), histrionicotoxin, chlorpromazine, phencyclidine (PCP), MK 801, local anesthetics, lipophilic agents such as detergents, fatty acids, barbiturates, volatile anesthetics, and n-alcohols can modify the properties of the nAChR without interacting with the ACh binding site, or directly affecting the binding of ACh.

These non-competitive blockers (NCBs) interact with at least two distinct sites that differ from those of the competitive blockers. The first site binds ligands in the low micromolar range, is found within the pore and composed of amino acids in the M2 segments of the five subunits. Binding of NCBs is facilitated by agonist binding.  Single channel experiments suggest that interaction at this site causes either a rapid reversible channel blockade or simply shortens channel opening times in a voltage-sensitive manner [43]. A second low affinity site has a distinct pharmacology in that NCBs accelerate desensitization of the nAChR by shifting the equilibrium towards the desensitized state [43]. Since the ligands to these sites are generally lipophilic and the number of sites calculated per receptor in reconstitution experiments depends on the lipid-to-protein ratio, it has been suggested that these sites lie at the interface between the nAChR protein and membrane lipids.

Steroid  binding sites. Steroids can modulate neuronal nAChRs expressed in oocytes, chromaffin cells and in brain.  This is not surprising considering the clinical effect of the steroid-like, neuromuscular blocking agent, pancuronium.  Steroids are thought to act at an allosteric site distinct from both the ACh binding site and the ion channel.  Progesterone and testosterone, but not cholesterol or pregnenolone, inhibit in a voltage-insensitive manner, the a4b2, a3b2 and a7 nAChRs [12]. In chromaffin cells, dexamethasone, hydrocortisone and prednisolone behave as non-competitive inhibitors of the nAChRs, and in vivo  there is an intriguing association between circulating corticosteroids, [¹²⁵I]a-BgT binding proteins, and behavioral sensitivity to (-)-nicotine [63].  Adrenalectomy results in corticosterone-reversible increases in the sensitivity to   (-)-nicotine in a variety of behavioral and physiological tests in mice, and chronic corticosterone selectively reduces the density of  [¹²⁵I]a-BgTnAChRs.  In vitro  corticosterone (high mM concentrations) inhibited binding of [¹²⁵I]a-BgT to rat brain membranes and reduced the affinity of (-)-nicotine for this binding site, which is consistent with a negative allosteric interaction. Estradiol has been found to differentially modulate different nAChRs subtypes.  While this agent potentiates the responses to ACh at a4b2 nAChRs expressed in oocytes, it inhibits the effects of ACh at the a3b2 subtype [12].  The subtype-specific ability to activate or inhibit nAChR function with a single allosteric modulator suggests the potential of targeting the steroid site for drug discovery.

Calcium modulation.  The neuronal nAChR has substantial influence on Ca²⁺ dynamics neurons by virtue of the permeability of their associated channels to Ca²⁺.  Interestingly, nAChRs may also be a target at clinically relevant (low micromolar) concentrations of dihydropyridine Ca-blockers like nimodipine [51] and are also influenced directly by extracellular Ca²⁺.  This latter effect appears to be mediated by a binding site on the nAChR complex that is distinct from the agonist binding site [1].  Thus, compounds that affect the dynamics of Ca²⁺ flux may also indirectly affect nAChR function. 

Antagonists.  Neurotoxins are commonly used to distinguish between neuronal nAChR receptor subunit combinations [52].  Lophotoxins are a family of related neurotoxins isolated from marine soft coral that non-discriminantly inhibit both neuronal and muscle subtypes of nAChRs.  Neosurugatoxin (NSTX), isolated from the Japanese ivory mollusc (Babyloni japonica) exerts potent blocking action in autonomic ganglia, antagonizes (-)-nicotine-induced antinociception in mice, inhibits (-)-nicotine-evoked release of [³H] dopamine from rat striatal synaptosomes, and blocks ACh- elicited currents in oocytes containing a2b2, a4b2, and a3b2, but not a7 and a1b1dg  nAChR subtypes.  The rat and chick a7 gene expressed as a homo-oligomer in oocytes is highly sensitive to a-BgT and ACh-gated currents can be completely blocked by nanomolar concentrations of this toxin.  Neuronal bungarotoxin (n-BgT) completely blocks ACh-induced currents in oocytes injected with a3b2 and partially blocks the a4b2 subtype but does not modulate a2b2 and a3b4 function.  The alkaloids dihydro-b-erythroidine (Fig. 2) and erysodine are competitive nAChR antagonists that appear to display some selectivity for b2 containing nAChRs particularly the a4b2 subtype [21].  Purified from the Conus magus snail venom, a-Conotoxin-MII inhibits the a3b2 subtype with an IC₅₀ of 0.5 nM, whereas it is from two to four orders of magnitude less potent at other nAChR subtypes [13].  Methyllycaconitine (MLA, Fig. 2), isolated from the plant, Delphinium brownii, is a very potent (K_i= 1 nM) inhibitor of [¹²⁵I] a-BgT binding in rat forebrain preparations, produces a potent reversible blockade of a7, is >30-fold less potent at the a3b2 or a4b2, and is inactive at the muscle nAChR [88].  Thus, MLA clearly differentiates between BgT sensitive sites on neuronal and muscle nAChRs.  It is anticipated that highly selective antagonists for the different nAChR subtypes that can readily penetrate the blood brain barrier will become available in the next few years.  Such agents will greatly aid the ability to correlate in vitro functional selectivity and in vivo activity.

Agonists.  The last few years have seen a flurry of medicinal chemistry activity targeted towards the identification of compounds that activate different nAChR subtypes for the potential treatment of a  variety of disorders (see below).  While agonists have been identified that display subtype selectivity in radioligand binding studies, there are currently no agonist ligands that potently and selectively activate the  major nAChR subtypes in in vitro functional assays.  A more complete overview of the “classical” agonists (e.g. nicotine and cytisine) can be found in several recent reviews [10,22,34,35].  The following discussion is focused on the newer agonists that have appeared in the literature in recent years (see Fig. 3).

Epibatidine, a chloropyridine natural product isolated from the venom of the “poison arrow” frog (Epipedobates tricolor), is among the most potent nAChR ligands identified to date [3], with a Ki value of 40 pM at the a4b2 nAChR and 20 nM at the a7 subtype.  Both of the isomers of epibatidine are potent full agonists of a4b2, a3b2, a3b4, a7 and a8 nAChRs.  Epibatidine also exhibits high potency and efficacy in activating muscle and ganglionic (a3b4a5?) type nAChRs.  The observation that epibatidine has potent analgesic properties prompted considerable interest in identifying cholinergic channel modulators lacking the side-effect liabilities of epibatidine as analgesic agents [3].   

ABT-594, a 3-pyridyl ether, is a centrally acting nAChR agonist with potent antinociceptive and anxiolytic-like effects in rodent models [23].  In vitro functional assays for the human a4b2, a7 and abdg nAChR subtypes indicate that ABT-594 is a full agonist at these subtypes but displays enhanced selectivity for the a4b2 subtype relative to the a7 and abdg subtypes relative to epibatidine.  This enhanced in vitro selectivity supports in vivo studies demonstrating an improved separation between analgesic effects and side-effects for ABT-594 compared to epibatidine.

DBO-83, a 3,8-diazabicyclo[3.2.1]octane derivative, is a novel analgesic agent that like epibatidine is a full agonist at a4b2 and ganglionic nAChRs but unlike epibatidine DBO-83 lacks any appreciable activity at neuromuscular junction nAChRs [30].

SIB-1508Y is a pyridine-modified nicotine analog that is more potent and selective than nicotine at the human a4b2 relative to the human a2b4 nAChR subtype but is significantly less potent than nicotine at the human a7 and a3b4 subtypes.  The enhanced functional subtype selectivity of SIB-1508Y relative to nicotine is supported by neurotransmitter release studies demonstrating differential effects of SIB-1508Y and nicotine on dopamine and norepinephrine release.  SIB-1508Y is active in preclinical rodent and primate models of Parkinson’s disease and is in clinical development for this indication [56].

RJR-2403 (transmetanicotine) is as potent and efficacious as nicotine in stimulating cation efflux from rat thalamic synaptosomes (thought to reflect activation of the a4b2 nAChR) but is 10-30-fold less potent than nicotine in stimulating the release of dopamine from rat striatal slices.  In vivo, RJR-2403 possesses cognitive enhancement activity at least comparable to that of nicotine but is 10-30-fold less potent in eliciting changes in cardiovascular parameters and locomotor activity [5].

GTS-21 (4-dimethylaminocinnamylidene anabaseine; DMXB) appears to act as a potent partial agonist at the rodent a7 subtype but is a very weak partial agonist at the rodent a4b2 subtype, blocking the effects of ACh in a non-competitive manner [57].  In contrast, the compound appears to have very weak (12% efficacy of nicotine) and negligible agonist activity at the human a7 and a4b2 subtypes, respectively.  GTS-21 exhibits cytoprotective effects and improves cognitive performance in preclinical models.

A-85380 is a 3-pyridyl ether that displays marked binding potency (Ki = 50 pM) and selectivity for the a4b2 nAChR subtype relative to the a7 (Ki = 148 nM) and abdg (Ki  = 314 nM) nAChR subtypes.  In functional models, A-85380 is more potent than nicotine in activating a number of rodent and human nAChR subtypes [81].  Thus, A-85380 retains the high potency of epibatidine towards the a4b2 nAChR but displays a selectivity for this subtype not observed with epibatidine.  These features have made this compound a very useful tool to probe the structure and function of this subtype.

NEURONAL NICOTINIC RECEPTOR PHARMACOLOGY

The multiplicity of CNS actions of (-)-nicotine in vivo may be related to the subunit combination on the nAChR (i.e. receptor subtype) activated, the neuronal system affected (e.g. dopaminergic vs. noradrenergic) in a brain region mediating a specific behavior, and the intrinisic channel properties of the subtype activated (e.g. ion selectivities & channel conductance properties).

Functional And Behavioral Effects Of (-)-Nicotine

Activation of nAChRs produces a variety of behavioral and physiological effects in experimental animals, including effects on cognitive performance, vigilance, locomotor activity, body temperature, respiration, cardiovascular function, EEG activity, cortical blood flow, and pain perception.  Many of these same actions have also been observed in humans.  The diverse and often profound effects produced by compounds acting at nAChRs are somewhat surprising given the relative scarcity of nAChRs in the brain.  Moreover, it has been difficult to demonstrate that these CNS effects are mediated by nicotinic actions similar to the fast excitation observed in autonomic ganglia and in striated muscle.  An alternative view of central nicotinic cholinergic transmission holds that the actions of (-)-nicotine and other nAChR activators are mediated through modulation of other neurotransmitter systems [70].  (-)-Nicotine interacts with presynaptic nAChRs to facilitate the release of a variety of neurotransmitters, including ACh, dopamine (DA), norepinephrine (NE), serotonin (5-HT), g-aminobutyric acid (GABA) and glutamate [37], many of which have been implicated in mediating / modulating a number of behaviors.

Because the addictive properties of (-)-nicotine have been tentatively linked to interactions with the DA system (see below), the mechanisms of nAChR-mediated dopamine release have been most extensively studied.  (-)-Nicotine can induce release of DA through actions in either cell body or terminal regions of DA neurons [37].  For example, several nAChR agonists evoke release from both striatal slices and synaptosomal preparations in vitro [16,37,39].  The observation of release from synaptosomes suggests that at least a part of this release is mediated by presynaptic nAChRs located directly on DA neuron terminals.  The identity of the nAChRs involved is uncertain, with data available that would support the involvement of either a4- or a3-containing receptors.  The complex pharmacology of this effect has been confirmed by the recent finding that the a3b2-selective a-conotoxin MII, blocks approximately 35-50% of the release from striatal synaptosomes but only 25% of the release from slices [37,39].  Neither the a7 selective a-conotoxin ImI nor a-bungarotoxin affects DA-release from striatal synaptosomes, so it is unlikely that these effects are mediated by presynaptic a7 nAChRs on dopaminergic terminals [39].  Thus, it appears that a3b2-containing receptors in the terminal field play a role in DA release, along with at least one other nAChR subtype.  The likelihood that an additional b2-containing nAChR is involved is suggested by the inability of (-)-nicotine to effect measurable DA release in vivo in genetically altered mice lacking the b2 subunit [64].  The difference between the effect of a-conotoxin MII in synaptosomes and slices indicates that nAChRs that are not located on dopamine terminals likely also play a role, an interpretation consistent with the observation that the NMDA receptor antagonist kynurenic acid attentuates nAChR-mediated DA release from striatal slices but not from striatal synaptosomes [37]. 

Evaluation of the mechanisms underlying nAChR-mediated DA release reaches new levels of complexity when actions in the DA cell body regions are considered.  For example, administration of mecamylamine directly into the VTA blocks the ability of systemically-administered (-)-nicotine to increase DA release in the n. accumbens [37].  Conversely, DA release in the accumbens produced by systemic (-)-nicotine can be mimicked by administration of (-)-nicotine into the VTA [37], and DA neurons in VTA slices respond to application of (-)-nicotine, an effect that is not observed in b2 knockout mice [64].  However, it is not likely that all of the effects of (-)-nicotine in the VTA are mediated through direct activation of DA neurons.  There is at least a component of the effect that can be blocked by NMDA receptor antagonist administration into the VTA, suggesting that nAChR-mediated glutamate release may be involved [76].

Examination of nAChR-mediated release of other neurotransmitters has been less extensive, but there is pharmacological evidence that nAChR subtypes involved in the release of NE may be distinct from those involved in DA release [16,39].  a-Conotoxin MII, for example, is much less effective in blocking (-)-nicotine-induced NE release from hippocampal synaptosomes than it is in blocking DA release from striatal synaptosomes [39].  Thus, it appears that NE release in the hippocampus is not modulated by presynaptic a3b2-containing nAChRs.  Instead, there is evidence a3b4 receptors may be more important for synaptosomal NE release than for DA release [16,53].  Hippocampal NE release can also be increased by direct injection of nAChR agonists into the locus coeruleus (LC).  Interestingly, however, the NE nuclei A1 and A2, which project to the paraventricular nucleus (PVN) of the hypothalamus, are even more sensitive to (-)-nicotine [55].  NE released in the PVN by even low doses of (-)-nicotine stimulates release of corticotropin-releasing hormone (CRH).  Thus, nAChR-mediated NE release in the CNS has important influences on the hypothalamic-pituitary-adrenal (HPA) axis [55].  A still different pharmacology has emerged for nAChR-mediated feedforward release of ACh that implicates a4b2-and a7-containing nAChRs in ACh release [37,70,82].

There is also evidence for nACh-mediated release of the major inhibitory and excitatory neurotransmitters, GABA and glutamate.  Because release of glutamate has been difficult to detect, most of the evidence for nAChR-mediated glutamate release in the CNS comes from analysis of electrophysiological data.  Glutamate is the likely transmitter in the connection between the medial habenula and the intrapeduncular nucleus, and application of (-)-nicotine potentiates glutamatergic transmission at intrapeduncular synapses with a pattern consistent with enhanced release of glutamate from the presynaptic element [70].  Electrophysiological evidence of nAChR-mediated glutamate release has also been obtained in the hippocampus [1,67].  The pharmacology of these effects is consistent with actions mediated through a7-containing nAChRs located on glutamatergic terminals.  Similar evidence for nAChR-mediated GABA release has been observed, with at least a portion of this release being mediated by a7 nAChRs [29].  However, an important role for b2-containing nAChRs (perhaps a4b2) in mediating GABA release has also been demonstrated [1,45].

Although the data clearly indicate that activation of nAChRs influences the release of a multitude of neurotransmitters, it is not entirely clear what role cholinergic regulation of transmitter release plays in normal brain function.  However, the capacity to modulate neurotransmission and the HPA axis so broadly clearly provides a mechanism for amplifying the impact of manipulations of the cholinergic system.  Moreover, the emerging evidence for subtype selective regulation of neurotransmitter release suggests that it may be possible to develop selective nAChR agents that would target specific neurotransmitters or even specific anatomical subsystems involved in neurological and psychiatric disease.

Therapeutic Targets

Clinical research with nAChR agonists has, to date, been limited primarily to the pharmacological evaluation of (-)- nicotine; and although preclinical research continues to focus on (-)-nicotine and other related, naturally occurring, alkaloids, a number of novel nAChR agonists have been synthesized in the last few years that may have therapeutic potential in a number of neurological and psychiatric conditions.

Cognition Enhancement.  Studies with both humans and experimental animals suggest that    (-)-nicotine has cognition-enhancing properties, although positive effects are not observed in all models [47,48,60].  Activation of nAChRs can enhance release of several neurotransmitters believed to be important modulators of learning and memory, and it appears that the effects of nAChR activators on cognitive function may be mediated through influences on some of these neurotransmitter systems (see [10,22]).  Moreover, nAChRs may play a more direct role in information storage through modulation of glutamatergic neurotransmission and resulting effects on synaptic plasticity as exemplified by long term potentiation [14].  It is likely that different features of (-)-nicotine’s cognitive effects are mediated by distinct effects on different neurotransmitter systems and that multiple nAChR subunits are involved in the cholinergic influences on cognitive function.  DHbE, n-BTX, and MLA can all disrupt performance when injected directly into the brain (see [10,48]), potentially supporting the involvement of several nAChR subtypes; but initial studies with b2 knockout mice did not reveal gross memory deficits.  The b2-knockout mice, however, were insensitive to the memory-enhancing effects of (-)-nicotine [14], suggesting that the effects of (-)-nicotine on this task require activation of b2-containing nAChRs but that these receptors are not required for normal cognitive function.  However, these mice do develop cognitive deficits relative to wild type mice as they age, perhaps because they eventually lose the capacity to compensate for the lack of b2-containing nAChRs [14].

The major target disease for a cognition enhancer is Alzheimer’s disease (AD).  In AD brain tissue, cortical nAChRs are markedly reduced [77], reflecting the cholinergic deficits associated with AD.  Pilot trials using nicotine patches have demonstrated improved attention in AD patients [48].  Moreover, pharmacoepidemiological studies have shown a reduced incidence of AD in populations of individuals who have previously smoked [42].  The potential protective effects of (-)-nicotine in this neurodegenerative disease may be related to neuroprotective properties observed with nicotine and other nAChR activators in in vitro and in vivo experimental studies [25].

In an effort to improve on the separation between the adverse and cognition-enhancing effects of (-)-nicotine, several nAChR agonists have recently been synthesized, some of which have entered clinical development for the possible treatment of AD.  GTS-21, ABT-418, SIB-1553A, and RJR-2403 have shown promise in preclinical cognition models [2,5,56,58].

Attention-Deficit Disorder (ADD).  Attention-deficit disorder, with or without hyperactivity, is a behavioral disorder characterized by distractibility and impulsiveness.  ADD is currently treated with stimulants like amphetamine, methylphenidate, and pemoline, all of which are thought to act via augmentation of DA neurotransmission.  Given that nAChR agonists can enhance DA release and appear to improve cognitive function, including attention, nAChR-targeted compounds may represent a useful acute treatment for the deficits in attention seen in ADD.  Small, clinical studies have demonstrated that (-)-nicotine patches produce significant improvements in adults with ADD [48].   It is likely, however, that compounds more selective than (-)-nicotine and with improved separation between efficacy and side effect liability will be required if this approach is to be of widespread utility, particularly since the predominant use of medication for ADD is in children.

Parkinson’s Disease.  Parkinson’s disease (PD) is primarily a motor disorder, characterized by tremors at rest, rigidity, bradykinesia, and impaired postural reflexes--effects resulting from loss of DA cells in the substantia nigra--and is typically treated with L-DOPA, which enhances DA transmission in the nigrostriatal pathway.  The neuroprotective and DA-releasing properties of nAChR activators suggest the possible use of this approach to treating PD.  (-)-Nicotine can attenuate the loss of DA neurons in the substantia nigra in rats with lesions of the nigrostriatal pathway, suggesting that nAChR agonists may have the potential to protect against degeneration of this system.  Acute symptom relief can be obtained with (-)-nicotine administration, which would be consistent with the ability of this compound to increase dopamine release [4].  Promising preclinical evidence has been obtained with SIB-1508Y, a novel nAChR agonist that increases striatal DA release in rats more effectively than (-)-nicotine.  This compound appears to have an improved preclinical safety profile relative to (-)-nicotine and significantly potentiates the effects of L-DOPA on motor and cognitive function in an primate MPTP model of PD [56].  Moreover, since cognitive decline is a common feature of PD, the potential cognitive-enhancing properties of nAChR-targeted compounds may be of additional benefit [60].

Schizophrenia. Attentional deficits and increased sensitivity to auditory stimuli in schizophrenics and their immediate relatives may be related to a diminished gating of an auditory evoked potential wave designated as P₅₀ in humans and N₄₀ in rats [46]. In normal subjects, paired presentation of auditory stimuli results in a diminished response to the second stimulus.  In schizophrenics this auditory gating is impaired.  In rodents, the N₄₀ wave originates in the hippocampal CA3 region; and auditory gating is disrupted by fimbria-fornix lesions that disrupt hippocampal cholinergic input and by a-BgT, but not by mecamylamine.  Interestingly, hippocampal tissue from schizophrenics is deficient in a-BgT binding sites and in a7 mRNA.  Among psychiatric patients, those with schizophrenia are more likely to be smokers than those with other psychiatric diagnoses [24].  Administration of (-)-nicotine to non-smoking relatives of schizophrenics can restore the deficient P₅₀ sensory gating, although this is a short - lived effect, possibly due to nAChR desensitization.  In an animal models of sensory gating deficits, (-)-nicotine and ABT-418 have short-lived effects, whereas GTS-21, a partial agonist at a7-containing nAChRs, is effective upon repeated administration [46,79].  Moreover, since currently used antipsychotics target the DA system, it is possible that nAChR-mediated modulation of DA neurotransmission may have therapeutic potential.  The possibility that nAChR agonist effects on cognitive function might be of benefit in schizophrenia is also worth exploring.

Tourette's Syndrome.  Classical neuroleptics are used to treat Tourette's syndrome, a condition characterized by uncontrolled spontaneous motor and verbal tics, but are limited in their usefulness due to sedation, exacerbation of learning difficulties, and potential tardive dyskinesia liability. (-)- Nicotine can potentiate the behavioral effects of neuroleptics like haloperidol in a number of preclinical models of behavior and thus may be useful in potentiating the beneficial actions of neuroleptics while diminishing their side effect profile [74].  Pilot clinical trials have indicated that both (-)- nicotine gum and patches can ameliorate the symptoms of Tourette's syndrome in non-smoking adolescents who are not satisfactorily controlled with neuroleptics [74].  It is perhaps surprising the (-)-nicotine, which releases DA, would have effects mimicking those of DA antagonists.  One possible explanation of this apparent anomaly is that prolonged exposure to (-)-nicotine, as would be the case with administration via a patch, might actually act by desensitizing nAChRs.  This interpretation could account for the observation that the ameliorative effects of (-)-nicotine patches in this condition often last long after patch removal and is supported by the more recent finding that similar improvements can be produced by the nAChR antagonist, mecamylamine [75].

Smoking Cessation. Tobacco smoke contains a large variety of substances; however, the addictive nature of smoking is attributable to the actions of nicotine [80].  Nicotine addiction is a complex phenomenon involving cognition enhancement, psychological conditioning, stress adaptation, reinforcing properties, and relief from the withdrawal syndrome.  The mesolimbic dopaminergic system appears to play a major role in the reinforcing properties of (-)-nicotine.  As is true of other addictive drugs, such as cocaine, morphine, and amphetamine, (-)-nicotine increases glucose utilization and releases DA in the rat nucleus accumbens [66], a region believed to be an important component of the reward system of the brain.  Moreover, concentrations of    (-)-nicotine similar to the plasma concentrations found in smokers increase activity in DA neurons in VTA slices [65], whereas nicotine withdrawal in rats is accompanied by an attenuation of brain reward mechanisms [27].  DA also appears to be involved in the reinforcing properties of (-)-nicotine in smokers, although DA antagonists increase smoking behavior in humans but decrease self-administration in rats, an apparent discrepancy that may be related to an attempt in humans to overcome the effects of the DA blocker [72].

The severe health liabilities and high mortality rates associated with tobacco usage, have resulted in major efforts to identify therapeutic treatments, most notably that of nicotine replacement therapy.  Nicotine gum and nicotine patches have been developed as aids in smoking cessation.  The initial optimism of a “cure” for smoking via nicotine replacement therapy in the form of gum and patches has been dampened by patient disillusionment due to the inability of either nicotine formulation to replace the nicotine provided in cigarettes as well as overcome the psychological cues associated with smoking, e.g. smoke inhalation and oral and hand cues.  Nonetheless, second generation nicotine replacement therapy is focused on increasing the amount of (-)- nicotine being delivered by gum or patch and on alternative delivery systems (e.g., nasal spray, inhalers) that more closely resemble the kinetics of nicotine administration produced by smoking, although preliminary data do not demonstrate a clear advantage for these alternative delivery systems [40].

Alternative therapies under development are the  “non-nicotine” nAChR agonists and partial agonists with reduced side effect liability, as well as combined agonist/antagonist treatment.  (-)-Lobeline, a nAChR ligand with full agonist, partial agonist and full antagonist properties depending on the test paradigm examined, is in Phase III clinical trials for smoking cessation [31].  The use of partial agonists in drug dependence therapy combines both substitution (agonist) and blockade of reinforcement (antagonist) in a single molecule, a concept that has been argued to “insulate" the addicted individual from reinforcement while preventing withdrawal symptoms.  This combined agonist/antagonist concept has been validated in a recent randomized, double-blind, placebo-controlled trial that evaluated concurrent orally administered mecamylamine with (-)-nicotine skin patch treatment for smoking cessation [71].

Anxiety Disorders. (-)-Nicotine has anxiolytic actions in man and some, but not all, preclinical models of anxiety [4].  Evaluation of the human data are difficult, however, because these studies are typically conducted with smokers and the reported anxiolytic actions of (-)-nicotine may be confounded by relief of withdrawal-induced anxiety.  Of course, the observation that (-)-nicotine-induced withdrawal produces what might be characterized as “rebound anxiety” may be taken as evidence that (-)-nicotine does indeed have anxiolytic effects.  One interesting explanation of the anxiolytic effects of (-)-nicotine holds that (-)-nicotine acts by desensitizing the stress response.  The acute physiological effects of (-)-nicotine resemble those produced by stress, but these effects are altered by prolonged exposure [4].  Thus, (-)-nicotine effects on stress and anxiety may be the result of receptor desensitization, which would be consistent with the observation that continuous infusion of (-)-nicotine at doses that would be expected to desensitize nAChRs still has anxiolytic-like effects in experimental animals [11].

Depression. Clinical studies have demonstrated a positive correlation between nicotine-dependence and major depression. This is apparently not a causal relationship but results from shared predispositions involving genetic or environmental factors [9]. One interpretation of these data is that people with major depression use (-)-nicotine as a form of self medication, which is consistent with the increased likelihood of depressive episodes observed during attempts to stop smoking.  There are also pilot clinical data suggesting that transdermal administration of (-)-nicotine produces antidepressant effects in nonsmokers [73].  Effects of (-)-nicotine on 5-HT and NE might provide a mechanism by which antidepressant effects could be mediated, and the proposed stress-reducing effects of (-)-nicotine and influences on the HPA axis could also play an important role [4,55].  Additional investigation of this area is clearly needed to evaluate the potential of nAChR-targeted compounds as antidepressants.

Analgesia. (-)-Nicotine has long been know to have antinociceptive actions in experimental animals and in man, but the relatively short duration and modest efficacy of this effect, coupled with the side effect profile of (-)-nicotine have discouraged development of this approach to analgesia.  However, the discovery that epibatidine, a potent and highly efficacious antinociceptive agent in rodent, produces its effects through actions at nAChRs has sparked considerable interest in the potential of ChCMs as analgesics.  With little separation between antinociceptive and toxic doses in rodent models, epibatidine lacks the selectivity that would make it a clinical candidate [3].  The potential to develop safer antinociceptive nAChR agonists, however, is exemplified by ABT-594.  This compound displays the broad spectrum of antinociceptive activity and the full efficacy of epibatidine in preclinical models but with an improved safety profile [23]. 

Further refinements in the ChCM approach to analgesia will require an improved understanding of the mechanisms involved in nAChR-mediated antinociception.  It is likely that activation of a variety of other neurotransmitter systems with inhibitory influences on pain signaling plays an important role in the antinociceptive effects of nAChR agonists.  Intrathecal administration of  mAChR, 5-HT, a-adrenergic, but not opioid, antagonists can attenuate the antinociceptive effects of nAChR activation [36,69].  Similarly, lesions that deplete NE or 5HT attenuate nAChR-mediated antinociception [23,68].  Given that interference with any single one of these other neurotransmitters is insufficient to produce complete blockade of the effect [36,69], it appears that release of several neurotransmitters contribute in parallel rather than in series.  Since intrathecal mecamylamine only modestly attenuates the antinociceptive effects of systemic (-)-nicotine and direct injection of any of several nAChR agonists into the brainstem can produce antinociception [23,36,69], it appears likely that activation of descending inhibition originating in brainstem sites, such as the nucleus raphe magnus, plays an important role in nAChR-mediated antinociception.

Relatively little is known about the nAChR subtypes that might be involved in antinociception.  The ability of DHbE to block nAChR-mediated antinociceptive effects suggests that b2-containing (perhaps a4b2) nAChR subtypes may play a role, whereas the lack of effect with lower doses of MLA makes it likely that a7-containing nAChRs do not [20,38,68].  Adding to the complexity is the finding that nAChR agonists such as epibatidine can produce behavioral signs of both irritation and antinociception when they are injected intrathecally and that the pharmacology of these two actions are distinct [38].  This raises the possibility that nAChR activation induces release of both nociceptive neurotransmitters (e.g., glutamate) and antinociceptive neurotransmitters (e.g., 5-HT and NE).  Since these effects are likely mediated by actions at different nAChR subtypes, it may be possible to improve on the efficacy of epibatidine by developing compounds that selectively increase the release of inhibitors of pain signaling.

Future Aspects

The pentameric structure of the neuronal nAChR and the considerable molecular diversity in subunits offers the possibility of a large number of nAChR subtypes, which, based on pharmacological precedent, may serve a variety of discrete functions within the CNS and thus represent novel targets for therapeutic agents. To capitalize on this opportunity, given the paucity and conflicting nature of therapeutic data to date, will require characterization of functionally relevant subunit combinations with respect to their localization within the CNS and identification of selective ligands that modulate receptor function as both direct and allosteric agonists and antagonists.  An increase in the number of pharmacophores active at nAChRs over the last few years provides a wealth of interesting tools to complement the molecular diversity of the receptor.  This development should increase our knowledge of the functional importance of nAChR subtype diversity.  Consequently, the potential for developing nAChR ligands for use in AD, PD, smoking cessation, anxiety, depression and schizophrenia, as well for use as novel analgesics, appears high.  Thus, nAChR pharmacology provides a challenge and an emerging therapeutic opportunity that is comparable in many ways to the identification and development of selective ligands with demonstrated therapeutic utility for the ever expanding serotonin receptor superfamily.