Abstract
Neuronal nicotinic acetylcholine receptors are distributed extensively throughout the central and peripheral nervous systems. Currently, there is great interest in determining the structural and functional diversity of these receptors, and in developing subtype-selective agonists that have potential as therapeutic agents for neuropathology and disease. However, relatively little attention has been focused on the development of subtype-selective nicotinic receptor antagonists. Such antagonists would be beneficial for establishing the role of specific nicotinic receptor subtypes in physiological function and for unraveling the complexities of neuronal nicotinic receptor function. Furthermore, these subtype-selective antagonists may also prove to be beneficial in the treatment of neuropathology and disease. The current perspective summarizes the research that has been carried out with both classical competitive antagonists and more recently developed competitive nicotinic receptor antagonists.
Neuronal Nicotinic Acetylcholine Receptors
Nicotinic acetylcholine receptors (nAChRs) consist of transmembrane proteins of pentameric structure and are members of a superfamily of ligand-gated ion channels, which include 5-hydroxytryptamine3, glycine, and several γ-aminobutyric acid receptors. nAChRs are widely distributed throughout the body, modulating the function of the CNS, peripheral nervous system, cardiovascular and immune systems (Sargent, 1993). In the CNS, nAChRs are located mainly presynaptically, modulating synaptic activity by regulation of neurotransmitter release (Wonnacott, 1997). Remarkable subtype diversity exists among nAChRs, although the exact subunit composition, stoichiometry and arrangement of native nAChRs remains to be conclusively determined (Lukas et al., 1999). The most abundant native nAChRs in the CNS are believed to be α4β2 heteromeric receptors and α7 homomeric receptors. Results from [3H]nicotine binding, affinity labeling, immunoprecipitation, in situ hybridization, and gene knockout studies indicate that the α4β2 subtype represents the high-affinity nicotine binding site in the CNS (Marks et al., 1986; Whiting and Lindstrom, 1988; Wada et al., 1989; Picciotto et al., 1998). The α4β2 receptor subtype contains two ligand binding domains located at the interface of each α/β subunit (Lippiello, 1989). The β subunit is also believed to be an important contributor to nAChR antagonist sensitivity (Cachelin and Rust, 1995). The α7 subtype contains five ligand binding domains with potentially different affinities, and represents the major α-bungarotoxin (αBTX) binding site in the CNS (Orr-Urtreger et al., 1997; Rakhilin et al., 1999). The α3β2* receptor subtype is less prevalent in brain and is at least in part responsible for mediating DA release, although other receptor subtypes may also be involved (Crooks et al., 1995; Kulak et al., 1997;Charpantier et al., 1998).
A significant limitation in drug discovery and pharmacophore development is the current putative assignment of the specific subunit composition of native nAChRs. The pharmacological specificity for each receptor subunit combination most likely results from a unique topography at the ligand binding site. Thus, topographical variations arising from differences in amino acid sequence and from specific subunit combinations result in receptor diversity, affording the opportunity for subtype selectivity for both nAChR agonists and antagonists. Recently, exciting developments in drug discovery have focused on nAChR agonists as therapeutic agents for diseases such as cognitive dysfunction, neurodegeneration, and other CNS pathologies (Holladay et al., 1997; Lloyd and Williams, 2000). However, relatively little attention has focused on the development of nAChR antagonists as potential drug candidates (Dwoskin et al., 2000).
Nicotinic Receptor Antagonists
The development of selective nAChR antagonists is a relatively new field, requiring the use of functional assays to distinguish between agonist and antagonist activity and radioligand binding assays. To characterize a drug candidate as a nAChR antagonist, the drug must by definition have no intrinsic activity, i.e., no agonist efficacy at the targeted nAChR. However, receptor binding affinity and lack of intrinsic efficacy are not sufficient to establish an antagonist mechanism of action. Moreover, the drug must inhibit the response elicited by an appropriate nAChR agonist. Depending upon the functional assay used, the apparent affinity of the drug for a specific nAChR may not reflect its true affinity for the binding site on the respective nAChR protein. Competitive antagonists interact with the acetylcholine (ACh) binding site at the α/β subunit interface of heteromeric nAChRs, or at the α/α subunit interface of homomeric nAChRs, preventing allosteric transition to the open channel state. Variations in the topography and geometry of the nAChR subunit interfaces of the various subunit combinations constitute the basis for development of competitive, subtype-selective ligands.
In comparison with structure-activity relationships (SARs) of nAChR agonists, correlations of the SAR of nAChR antagonists have not been carried out, and structural similarities between many well known nAChR antagonists are not apparent. Thus, pharmacophore geometries for antagonist binding sites are ill-defined. In addition, most of the known antagonists are not nAChR subtype-selective. Previous hypotheses proposed that competitive antagonists interact with the same binding site as agonists, but via a different mode of interaction, i.e., agonists and antagonists bind differently to the different quaternary states of the nAChR receptor protein (Sheridan et al., 1986). Sheridan proposed that antagonists, although containing the appropriate agonist pharmacophore geometry, are usually large molecules that interact with the agonist binding site and extend outside the normal agonist volume. Thus, this extended area of antagonist interaction prevents the agonist-induced conformational change in the receptor protein that produces channel opening. However, the latter hypothesis is based on a simplified model involving a single pharmacophore. The complexity arising from the existence of multiple nAChR subtypes and their different physiological functions necessitates the development of subtype-selective antagonists as effective therapeutic agents with minimal untoward side effects. The elucidation of the molecular structure and subunit composition of the binding sites of native nAChRs will be crucial for the rational development of nAChR subtype-selective antagonists.
Thus, the current limited understanding of native nAChR subtype composition is a major caveat in the drug development process. High-throughput technology in drug discovery is now routine in facilitating rapid drug development in the nAChR field. However, this technology is often used without knowledge of the specific subunit composition of the native nAChR target. In many cases, recombinant receptor expression assays are utilized for high-throughput evaluation, because of their obvious practical advantages. However, these advantages are offset by the possibility that the pharmacology of the drug candidate from nAChR expression assays may not be consistent with that obtained using native receptor assays. Thus, the selection of potential clinical candidates developed using expressed nAChRs should be considered only subsequent to their assessment in native nAChR assays.
Competitive Nicotinic Receptor Antagonists
Dihydo-β-erythroidine (DHβE) and Related Compounds.
DHβE (Fig. 1, 1 ), an alkaloid found in seeds of Erythrina, has been widely used as a classical, nonselective, competitive nAChR antagonist. DHβE has nanomolar affinity at both α4β2 and α3β2* receptor subtypes, inhibiting [3H]nicotine binding to and DA release from rat striatal slices (IC50 = 30 nM in each assay;Crooks et al., 1995). DHβE blocks the agonist response at rat recombinant α4β2 receptors with an IC50 value of 370 nM (Harvey and Luetje, 1996). Furthermore, DHβE is a relatively weak antagonist at α3β4* (ganglionic-like) and (α1)2β1γδ (muscle-type) receptors expressed in PC12 and TE671 cells, respectively, and has ∼10-fold greater selectivity for human α4β4 than for human α4β2 nAChRs (Chavez-Noriega et al., 1997). Rat α4β4 and α3β4 nAChRs differ in sensitivity to DHβE by 120-fold (IC50 = 0.19 and 23.1 μM, respectively; Harvey and Luetje, 1996). Erysodine (Fig.1, 2 ), another Erythrina alkaloid structurally related to DHβE, is a more potent inhibitor (Ki = 5 nM) of α4β2 nAChRs compared with DHβE (Ki = 35 nM) in the [3H]cytisine binding assay (Decker et al., 1995a). Erysodine and DHβE exhibit similar low affinity (Ki = 4 and 9 μM, respectively) for the α7 subtype, inhibiting 125I-αBTX binding to rat brain membranes. Erysodine is a competitive, reversible antagonist at α3β2* nAChRs, inhibiting nicotine-evoked (100 nM) DA release from superfused rat striatal slices (IC50= 58 nM, equipotent with DHβE; Decker et al., 1995a). Schild analysis revealed that erysodine interacts competitively and at a single site on this nAChR. Erysodine also inhibits nicotine-evoked86Rb+ efflux from α3β4* nAChRs expressed by IMR32 cells (IC50 = 7 nM), and is 10-fold more potent than DHβE. Thus, both DHβE and erysodine are potent nAChR antagonists but are not subtype-selective.
α-Lobeline (Lobeline).
Lobeline (Fig. 1, 3 ), a lipophilic alkaloid from Lobelia, has many nicotine-like effects, and until recently was considered to be an agonist at nAChRs (Decker et al., 1995b). However, unlike nicotine, chronic lobeline administration does not up-regulate nAChRs (Bhat et al., 1991). Lobeline binds to α4β2 nAChRs in the low nanomolar range and is a potent inhibitor of [3H]DHβE binding (Williams and Robinson, 1984; Teng et al., 1997). Lobeline also inhibits [3H]methyllycaconitine ([3H]MLA) binding (Ki = 7 μM) (L. P. Dwoskin and P. A. Crooks, unpublished results) to rat brain membranes, indicating an interaction with the α7 nAChR subtype. More recently, lobeline has been shown to inhibit nicotine-evoked (1 μM)86Rb+ efflux from rat thalamic synaptosomes (IC50 = 700 nM) and to competitively inhibit nicotine-evoked (10 μM) [3H]DA release (IC50 = 200 nM) from superfused rat striatal slices (Miller et al., 2000). At high concentrations (>1.0 μM), lobeline releases DA from its presynaptic terminals in a manner insensitive to mecamylamine (i.e., not nAChR-mediated; Teng et al., 1997). Results from these more recent studies clearly indicate that lobeline acts as a competitive, nonselective antagonist at α4β2 and α3β2* nAChRs.
MLA.
MLA (Fig. 1, 4 ), a tertiary diterpenoid alkaloid isolated from Delphinium brownii, is one of the most potent (Ki = 1 nM), selective, competitive nonpeptide antagonists at the α7 nAChR subtype (Bergmeier et al., 1999; Davies et al., 1999). Ring E analogs of MLA also exhibit similar affinity for the α7 subtype (Bergmeier et al., 1999). Recent SAR reveals that the methylsuccinimidobenzoyl portion of the molecule is necessary for high-affinity interaction with the α7 subtype, and removal of the methyl group, the methylsuccino, or the entire substituted benzoyl moiety results in a 20-, 1000-, or 2000-fold decrease, respectively, in affinity (Hardick et al., 1995). Other structural changes in the diterpenoid portion of the molecule produce little effect. MLA has 200-fold greater selectivity for α7 compared with (α1)2β1γδ receptors, and 50- to 100-fold greater selectivity compared with α3β4, α4β2 and α3β2 subtypes when expressed in oocytes (Drasdo et al., 1992;Vijayaraghavan et al., 1992; Yum et al., 1996). MLA is 1000-fold more selective for rat α7 compared with rat α4β2 subtypes (Alkondon et al., 1992; Quik et al., 1996) and has little ability to inhibit nicotine-evoked DA release from striatal synaptosomes (Drasdo et al., 1992). However, using the more intact, striatal slice preparation, support exists for α7* receptor involvement in nicotinic agonist-evoked DA release. Specifically, MLA has been reported to partially inhibit anatoxin-a-evoked DA release from striatal slices via circuitry including glutamatergic terminal innervation (Kaiser and Wonnacott, 2000).
d-Tubocurarine.
d-Tubocurarine (Fig.1, 5 ), isolated from the Chondodendron tomentosumplant is a well known antagonist of (α1)2β1γδ nAChRs, and also competitively inhibits ACh-evoked responses in rat α2β2 and α3β2 nAChRs expressed in Xenopus oocytes (Cachelin and Rust, 1995). More recently, d-tubocurarine was shown to inhibit within a 20-fold potency range, agonist response in Xenopus oocytes expressing human nAChRs, including α2β2, α3β2, α4β2, α2β4, α3β4, α4β4, and α7 nAChRs (Chavez-Noriega et al., 1997). The α4β4 subtype was found to be the most sensitive, followed by the α2β2 subtype, which had a 6-fold lower sensitivity.d-Tubocurarine has also been reported to inhibit [3H]nicotine binding to mouse striatal membranes (Ki = 67 μM) and to inhibit nicotine-evoked DA release from superfused mouse synaptosomes (>100 μM) (Grady et al., 1992), demonstrating nonselective inhibition at native α4β2 and α3β2* nAChR subtypes.
Bungarotoxin.
Two very potent and selective antagonists, αBTX and n-bungarotoxin (nBTX), have been isolated from the venom of Bungarus multicinctus, the Formosan banded krait. αBTX selectively inhibits α7, α8, and α9 homomeric nAChRs, but not heteromeric nAChRs (McGehee and Role, 1995). αBTX is a 75 amino acid peptide, with high affinity at (α1)2β1γδ nAChRs and at the α7 nAChR (Marks et al., 1986; Kd = 1 nM). In this respect, nicotine binds to the α7 nAChR with micromolar affinity. α7 mRNA distribution in rat brain is directly correlated with 125I-αBTX binding (Clarke et al., 1985). The pharmacology of recombinant α7 homomeric receptors resembles that of native α7* αBTX-sensitive receptors (Barrantes et al., 1994;Alkondon and Albuquerque, 1995; Gopalakrishnan et al., 1995; Quik et al., 1996), supporting the hypothesis that native α7* receptors represent homomeric α7 nAChRs; however, the channel properties of native and expressed α7* receptors are not identical, such that the subunit composition of native α7* receptors currently remains putative (Albuquerque et al., 1995). nBTX has received less attention than αBTX, probably due to its lack of availability. Historically characterized as an α3β4* antagonist, nBTX potently inhibits nicotine-evoked DA release from rodent striatal tissue at α3β2* nAChRs (Grady et al., 1992) and inhibits binding of [3H]epibatidine at non-α4β2 nAChRs in rat brain membranes (Houghtling et al., 1995). In Xenopus oocyte expression studies, nBTX has been shown to inhibit the α3β2 nAChR subtype, but also partially inhibits the α4β2 subtype (Luetje et al., 1990; Gotti et al., 1997); and thus, nBTX is not a subtype-selective antagonist.
Conotoxins.
Venoms isolated in minute amounts fromConus snails, including α-, β-, and ω-conotoxins, represent another group of peptide neurotoxins that target nAChRs, muscle sodium channels, and neuronal calcium channels, respectively. The α-conotoxins (α-conotoxin MII, α-conotoxin ImI, and similar homologous peptides) are small peptides of about 14 to 17 amino acids in length (McIntosh et al., 1999). α-Conotoxin MII partially inhibits nicotine-evoked DA release from rat striatal synaptosomes. α-Conotoxin MII has 4 orders of magnitude greater affinity for the α3β2* nAChR compared with the α3β4* subtype, i.e., specifically inhibiting α3β2* receptors at low concentrations (0.1–10 nM), but inhibiting additional nAChR subtypes at higher concentrations (100 nM–1 mM). Selectivity for α3β2* nAChRs is indicated by the lack of α-conotoxin MII-induced inhibition of K+-stimulated DA release from striatal synaptosomes, and lack of inhibition of nicotine-evoked [3H]norepinephrine release from rat hippocampal synaptosomes. In Xenopus oocyte expression studies, α-conotoxin MII blocks the α3β2 subtype with an affinity 2 to 4 orders of magnitude higher than at other nAChR subtype combinations. The following order of potency (IC50 value) for α-conotoxin MII has been reported: α3β2 (0.5 nM) > α7 (0.1 μM) > α4β2 (0.4 μM) = α2β2 (1 μM) = α3β4 (1.1 μM) > (α1)2β1γδ (4 μM) > α4β4 (>4.0 μM) > α2β2 (>4.0 μM) (Cartier et al., 1996). The structurally related α-conotoxin ImI and α-conotoxin MI are selective for α7- and α1-containing receptors, respectively, and moreover, do not block nicotine-evoked DA release from rat striatal synaptosomes. α-Conotoxin AuIA, AuIB, and AuIC, isolated from Conus aulicus, inhibit ACh-evoked electrophysiological currents in rat α3β4-expressingXenopus oocytes. At relatively low concentrations (0.3 and 1.0 μM), these conotoxins are selective and inhibit only α3β4 subtypes. Other conotoxin peptides, including FAT-MII, PnIA, PnIB, PnIA A10L and Pn1AN11S, as well as modified or non-natural amino acids, are also being investigated for selective inhibition of nAChR subtypes.
Other Neurotoxins.
Over 100 additional neurotoxins have been identified that act at nAChRs, including cobra toxin from the cobraNaja naja, erabutoxin from the sea snake Laticauda semifasciata, histrionicotoxin from the frog Dendrobates histrionicus, lophotoxin from coral of the genusLophogorgia, neosurugatoxin from the Japanese ivory molluscBabylonia japonica, and nereistoxin from the marine annelidLumbriconereis heteropoda (Adams and Swanson, 1994). These neurotoxins have been shown to be both competitive (e.g., nereistoxin) and noncompetitive (e.g., neosurugatoxin) nAChR antagonists. Although these toxins have proven valuable for in vitro studies, most demonstrate little subtype selectivity, and their limited availability diminishes their usefulness.
N-Alkylnicotinium Iodides.
N-Alkylation of the pyridino-N atom ofS(−)-nicotine has afforded a number ofN-alkylnicotinium iodides (Fig.2, 6 ), which competitively inhibit native nAChRs (Crooks et al., 1995; Dwoskin et al., 1999,2000). At low concentrations, these analogs do not have intrinsic activity in the DA release or86Rb+ efflux assays using rat striatum (i.e., α3β2* receptor assay) and thalamus (α4β2 receptor assay), and moreover, act as competitive antagonists at these nAChR subtypes. The following order of potency (IC50) was obtained in the nicotine-evoked (10 μM) DA release assay: NDDNI (9 nM; 6d ) > NNNI (210 nM; 6b ) > NONI (620 nM; 6a ) > NDNI (>100 μM; 6c ). In the α4β2 subtype assay assessed via inhibition of [3H]nicotine binding in rat striatal membranes, the following order of potency (Ki) was obtained: NDNI (93 nM) > NDDNI (140 nM) > NNNI (840 nM) > NONI (20 μM). Antagonist activity at the α4β2 subtype was assessed by inhibition of nicotine-evoked (1 μM)86Rb+ efflux from rat thalamic synaptosomes, and the following order of potency (IC50) was obtained: NDNI (14 nM) = NNNI (9 nM) > NDDNI (40 nM) > NONI (10 μM). Thus, NONI and NDNI are selective for α3β2* and α4β2 receptors, respectively, whereas NNNI and NDDNI have inhibitory activity at both α3β2* and α4β2 receptors. The N-n-alkylnicotinium iodide series is of considerable interest, because of the high potency and selectivity at native α3β2* or α4β2 nAChR subtypes. Alkylation of the pyridyl-N atom, which converts the nicotine molecule from an agonist into an antagonist at nAChRs, alters the physicochemical properties of the molecule causing a 2-unit drop in the pKa value of the pyrrolidineN (from 8.4 to 6.4). Consequently, the roles of the pharmacophoric nitrogen-containing moieties may be reversed, such that these analogs interact with the nAChR in a unique manner, i.e., the quaternary pyridinium center of the antagonist molecule interacts with the binding site that normally accommodates the protonated pyrrolidine nitrogen in agonist binding, and the unprotonated pyrrolidine nitrogen of the antagonist molecule substitutes for the pyridine nitrogen of the agonist at the hydrogen bonding site.
Two series of conformationally restrained analogs of theN-n-alkylnicotinium iodides have been developed (Crooks et al., 2000). One series incorporates an ethano bridge between the 2-pyridyl carbon and the 3′-pyrrolidino carbon of the nicotinium moiety (Fig. 2, 7 ). A second series incorporates an ethano bridge between the 4-pyridyl carbon and the 3′-pyrrolidino carbon of the nicotinium moiety (Fig. 2, 8 ). Both of these isomeric tricyclic diazaheterocycle systems contain a B/C ring junction withcis-stereochemistry. None of these analogs showed affinity for the α4β2 nAChR subtype assessed in the [3H]nicotine binding assay, or for the α7 subtype assessed in the [3H]MLA binding assay. Increasing the N-n-alkyl chain length from C8 to C11 enhanced the potency (IC50 range = 50–700 nM) of inhibition of nicotine-evoked (10 μM) [3H]DA release from superfused striatal slices, indicating antagonist activity at α3β2* nAChRs. The interesting exception was theN-n-dodecyl analog, 7e , which exhibited an 18-fold lower potency compared with 8e . Conformational restriction of the C12 analog NDDNI ( 6d ) reduced affinity by 1 to 2 orders of magnitude at the α3β2* nAChR subtype, but increased selectivity for this subtype. Thus, restricting the conformation ofN-alkylnicotinium iodides results in elimination of activity at the α4β2 and α7 receptor subtype and affords ligands with high affinity and selectivity for the α3β2* subtype.
N-Alkylpyridinium Halides.
SimpleN-n-alkyl pyridinium halides (Fig. 2, 9 ) have also been evaluated for antagonist activity at nicotinic receptors (Crooks et al., 2000). The C6, C8, and C10 alkylpyridinium halides had low affinity (Ki = 20–250 μM) for the [3H]nicotine binding site using rat striatal membranes, and had little or no affinity at the [3H]MLA binding site. However, in the nicotine-evoked [3H]DA release assay using superfused striatal slices, two alkylpyridinium analogs, i.e.,N-n-decylpyridinium iodide (NDPI) andN-n-dodecylpyridinium iodide (NDDPI) were relatively potent (IC50 = 110 and 90 nM, respectively), resembling the conformationally restricted nicotinium iodides, and exhibiting a potency 1 order of magnitude less than NDDNI (IC50 = 9 nM). Thus, the minimal structural requirements that govern affinity and selectivity for the α3β2* receptor subtype may be simply the presence of a quaternary pyridinium moiety and a N-n-alkyl group with optimal chain length of between 10 and 12 carbons.
Analogs of Nicotine and Nornicotine.
Interest has recently been focused on the design of novel nicotinoids that are structurally related to the tobacco alkaloids nicotine ( 10a ) and nornicotine ( 10b ). These tobacco alkaloids have been shown to have high affinity for nAChRs and to act as agonists in most receptor assays (Dwoskin et al., 1993, 1995; Teng et al., 1997). Numerous structurally related compounds (see Fig.3) act as antagonists at nicotinic receptors; however, in most cases, their IC50values and mechanism of inhibition (i.e., whether competitive or noncompetitive) have not been determined. These structurally related compounds have been reported primarily in the patent literature, which has been reviewed recently (Dwoskin et al., 2000).
A large number of compounds in the patent literature have been reported by Abbott Laboratories constituting two structural groups, i.e., analogs in which a 2-furo[3.2-b]pyridyl moiety has replaced the 3-pyridyl moiety in both nicotine and nornicotine; and analogs in which the pyrrolidino moiety has been replaced with a hexahydro-1H-pyrrolizine moiety. Antagonist activity was determined by inhibition of nicotine-evoked (100 μM)86Rb+ efflux from IMR-32 human neuroblastoma clonal cells expressing α3β4* nAChRs and from K177 cells expressing α4β2* receptors. The most interesting compound in the first group was compound 11 (10 μM), which acts nonselectively as an antagonist at both α4β2* and α3β4* nAChRs (55% inhibition and 6% intrinsic activity in the K177 cell assay, with 38% inhibition and 7% intrinsic activity in the IMR-32 assay). In the second group, compound 12 (10 μM) afforded 60% inhibition but exhibited 14% intrinsic activity in the K177 cell assay, indicating partial agonist activity at the α4β2* nAChR. The 2-substituted hexahydro-1H-pyrrolizine enantiomers, 13 and 14 (30 μM) afforded 55 and 50% inhibition, respectively, in the IMR-32 assay; however, intrinsic activity for R-enantiomer 13 was not reported, and enantiomer 14 had no intrinsic activity. Thus, these compounds may represent promising leads for the development of antagonists of the human α3β4* nAChR.
Another group of nicotine analogs disclosed by Abbott Laboratories in which a methyleneoxy or ethyleneoxy bridge has been inserted between the 2′-pyrrolidinyl and the 3-pyridyl positions have been examined only in the IMR-32 assay described above, in order to assess antagonist activity at the α3β4* nAChR. Compounds 15a , 15b , and 16a produced 50% inhibition but 15b elicited 15% intrinsic activity. Compound 16a (10 μM) exhibited 60% inhibition, whereas the dichloro analog 16b (1–10 μM) produced 20–90% inhibition, with little intrinsic activity. Analogs 17a and 17b in the 3-(1-methyl-2-pyrrolidinylethoxy)pyridine series had no intrinsic agonist activity, and exhibited IC50 values of 75 and 6 μM, respectively. These data indicate that introduction of a methyl group into the C-6 position of the pyridyl ring in this series increased inhibitory potency 2-fold, whereas introducing a chloro-substituent increased the potency by 30-fold. Thus, several compounds in this series are pure antagonists at the α3β4* nAChR. The azetidine compound 18 (1 μM) afforded 55% inhibition and demonstrated no intrinsic agonist efficacy, also indicating antagonist activity at the α3β4* nAChR. Introducing a 5-ethynyl substituent into the molecule affords compounds of general structure 19 ; these compounds produced ∼50% inhibition in the IMR-32 assay at 1 μM. Compound 19a (1 μM) exhibited 66% inhibition, and 19b (10 μM) produced complete inhibition, indicating that such compounds may be relatively potent α3β4* nAChR antagonists. Analogs containing an ethenyl moiety at the pyridino C-5 position, i.e., 20a and 20b , produced 90% inhibition at 10 μM; 20b had an IC50 of 2 μM. A large group of analogs containing a substituted 5-phenyl group, which include compounds 21a and 21b produced ∼50% inhibition at 1 μM. Other compounds in this series were reported to produce complete inhibition. Interestingly, the R- andS-enantiomers of 21c had similar inhibitory potencies (IC50 = 9 μM), demonstrating a lack of stereoselectivity at the α3β4* nAChR. The 5-[2-(4-ethenyl)-pyridyl]-substituted analog, 22 , also exhibited 50% inhibition at 1 μM. It is important to note that intrinsic activity for compounds 19 to 22 has not been reported. Thus, the classification of these analogs as either antagonists or partial agonists at the α3β4* nAChR cannot be determined.
Analogs of Anabaseine and Anabasine.
The minor tobacco alkaloids, anabaseine ( 23 ) and anabasine ( 24 ), have high affinity for nAChRs and act as agonists in most receptor assays (Dwoskin et al., 1995). Structurally related compounds (see Fig.3) have been reported to act as antagonists at nAChRs; however, in most cases, their IC50 values and mechanism of inhibition (i.e., whether competitive or noncompetitive) have not been determined. These structurally related compounds are in the recently reviewed patent literature (Dwoskin et al., 2000).
Several 3-trans-cinnamylidene derivatives ( 25a – 25c ) of anabaseine ( 23 ) exhibit partial agonist activity and subtype-selectivity for the α7 subtype in Xenopus oocyte preparations and are devoid of agonist activity at α4β2, α3β4* nAChRs in PC12 cells, and in the muscle contractility assay (Kem et al., 1994). However, whether these compounds exhibit antagonist properties at the α7 subtype is not clear, since the reported IC50 values are in the same range of concentration as the EC50 values (5–15 μM), indicating that these compounds are partial agonists, rather than antagonists. Furthermore, the rapid desensitization of the α7 receptor, which occurs before the peak analog concentration can be obtained during the superfusion of oocytes, may be a significant factor in explaining the observed antagonism/partial agonist effects.
The racemic conformationally restricted anabasine analog 26, in which a 2-(1-azabicyclo[2.2.2]octane) moiety has replaced the 2-piperidinyl moiety, is a potent antagonist (IC50 = 154 nM) of nicotine-evoked (10 μM)86Rb+ efflux in the rat thalamic synaptosomal assay and has high affinity for the [3H]nicotine binding site using rat brain membranes (Ki = 1 nM), indicating antagonist activity at the α4β2 receptor subtype (Crooks et al., 1998). However, 26 has potent partial agonist activity (EC50 = 2 nM,Emax = 40%) at α3β2* nAChRs mediating DA release from rat striatal synaptosomes and at α3β4* (EC50 = 1100 nM) and (α1)2β1γδ (EC50 = 59 nM) nAChRs (Bencherif et al., 1998). Thus, 26 has mixed agonist/antagonist activity, with an antagonist profile at α4β2 receptors and an agonist profile at other nicotinic receptor subtypes. The enantiomers of 26 have also been evaluated (Caldwell, 1999). The R-isomer was slightly more potent than theS-enantiomer, with the R-isomer exhibiting aKi value of 0.5 nM in the [3H]nicotine binding assay, an EC50 value of 9 nM, and anEmax of 55% in the α3β2* assay (i.e., partial agonist activity), an EC50 value of 375 nM in the α3β4* nAChR assay, and an EC50 value of 50 nM in the (α1)2β1γδ nAChR assay. Thus, modest enantioselectivity was observed.
The 1-aza-2-(3-pyridyl)tricyclo-[3.3.1.13,7]decanes ( 27a and 27b ) have been reported by R. J. Reynolds Tobacco Company and the University of Kentucky in the patent literature (see review, Dwoskin et al., 2000). These compounds exhibited IC50 values of 695 and 846 nM, respectively, in the nicotine-evoked DA release assay using rat striatal synaptosomes while showing no intrinisic agonist activity, demonstrating antagonism of α3β2* nAChRs. Compound 27b exhibited an IC50 value of 630 nM, with no intrinsic activity in the nicotine-evoked (100 μM)86Rb+efflux assay using rat thalamic synaptosomes, demonstrating antagonist activity at α4β2 nAChRs. Thus, 27b is a high-affinity, nonselective antagonist, which inhibits α3β2* and α4β2 nAChR subtypes.
Compounds 28 and 29 , analogs of 26 , in which either a CH2 or an unsaturated CH bridge, respectively, has been inserted between the 3-pyridyl and 2′-quinuclidinyl carbons, have recently been described by R. J. Reynolds Tobacco Company as antagonists of α4β2 nAChRs (Ki = 37 and 73 nM, respectively) in the [3H]nicotine binding assay using rat cortical membranes. Compounds 28 and 29 were not agonists at (α1)2β1γδ nAChRs assessed in the TE671/RD cell assay, at α4β2 receptors assessed in the86Rb+ efflux assay using rat thalamic synaptosomes or at α3β4* nAChRs assessed using PC12 cells (Jeff Schmidt, personal communication). No information is available on the ability of these compounds to inhibit agonist-evoked response in the 86Rb+efflux assay, which is necessary for the unequivocal characterization of these compounds as nAChR antagonists at the α4β2 nAChR subtype.
Therapeutic Applications of Nicotinic Receptor Antagonists
Antagonists acting at peripheral nicotinic receptors have been used historically as clinically useful muscle relaxants during surgery and as anti-hypertensive agents. However, until very recently, there have been no examples of clinically useful nAChR antagonists. Recently, mecamylamine, a noncompetitive and nonselective channel blocker of nAChRs, as well as lobeline, a competitive nicotinic receptor antagonist, have been demonstrated to have utility as tobacco smoking cessation agents (Schneider and Olssen, 1996; Rose et al., 1998). Furthermore, preclinical results suggest that lobeline may be a potential treatment for psychostimulant abuse (Miller et al., 2000;Harrod et al., 2001). Recently, a structural variant of a nAChR subtype has been suggested to account for some forms of epilepsy (Steinlein, 1996). Treatment with nAChR antagonists may prove beneficial for such pathologies (Holladay et al., 1997). Currently, drug discovery is focusing on nAChRs as novel targets for the development of therapeutic agents for a wide variety of CNS diseases including drug addiction, neuroendocrine, neuropsychiatric, and neurological diseases, memory and learning disabilities, eating disorders, and the control of pain, as well as cardiovascular and gastrointestinal disorders. In this respect, nAChR antagonists have good potential as therapeutic agents, since they offer another means of modulating nAChR receptor function. Nicotinic agonists rapidly desensitize nAChRs, essentially inhibiting their function. Thus, one perspective is that inhibition of receptor function may be the action that confers clinical utility, indicating that nAChR antagonists could also be beneficial in the treatment of diseases for which nAChR agonists are currently being developed. For example, schizophrenia and drug abuse have both been associated with hyperactivity of CNS dopaminergic systems, and inhibition of nAChRs may be advantageous in reducing such hyperactivity. Furthermore, the availability of subtype-selective nAChR antagonists will be invaluable agents in both basic and clinical research, with regard to both the treatment and diagnosis of disease. Finally, subtype-selective antagonists will define the role of specific nAChR subtypes in both physiological function and disease states.
Conclusions
A vast diversity in the structure of nAChR antagonists exist. Some nAChR antagonists are structurally complex alkaloids (e.g., MLA,d-tubocurarine, and DHβE). Others are analogs of the structurally less complex tobacco alkaloids (e.g., nicotine and anabasine). In addition, several potent antagonists are small peptides (e.g., the bungarotoxins and the α-conotoxins). This structural diversity affords good opportunities for SAR development. One of the main goals is to achieve subtype-selective antagonists in order to reduce possible untoward side effects. Based upon current findings, it appears that subtype-selective antagonism can be achieved by appropriate modification of agonist molecules, such as nicotine or anabasine, or by varying the amino acid sequences in active peptide molecules, such as bungarotoxin and α-conotoxin. Thus, there are exciting opportunities in SAR for the further development of more potent and selective antagonists. Future prospects look bright for the clinical development of this new class of therapeutic agent.
Acknowledgments
We acknowledge the technical assistance of Dr. Rui Xu and Joshua Ayers.
Footnotes
-
This work was supported in part by Grants DA00399, DA10934, and DA13519 from the National Institutes of Health.
- Abbreviations:
- nAChRs
- neuronal nicotinic acetylcholine receptors
- CNS
- central nervous system
- αBTX
- α-bungarotoxin
- ACh
- acetylcholine
- SARs
- structure-activity relationships
- DHβE
- dihydro-β-erythroidine
- DA
- dopamine
- MLA
- methyllycaconitine
- nBTX
- n-bungarotoxin
- NDNI
- N-n-decylnicotinium
- NDDNI
- N-n-dodecylnicotinium
- NNNI
- N-n-nonylnicotinium
- NONI
- N-n-octylnicotinium
- α3β2*
- * indicates putative receptor subtype assignment
- Received January 26, 2001.
- Accepted April 3, 2001.
- The American Society for Pharmacology and Experimental Therapeutics