High Affinity Binding of Epibatidine to Serotonin Type 3 Receptors*

Epibatidine and mecamylamine are ligands used widely in the study of nicotinic acetylcholine receptors (nAChRs) in the central and peripheral nervous systems. In the present study, we find that nicotine blocks only 75% of 125I-epibatidine binding to rat brain membranes, whereas ligands specific for serotonin type 3 receptors (5-HT3Rs) block the remaining 25%. 125I-Epibatidine binds with a high affinity to native 5-HT3Rs of N1E-115 cells and to receptors composed of only 5-HT3A subunits expressed in HEK cells. In these cells, serotonin, the 5-HT3R-specific antagonist MDL72222, and the 5-HT3R agonist chlorophenylbiguanide readily competed with 125I-epibatidine binding to 5-HT3Rs. Nicotine was a poor competitor for 125I-epibatidine binding to 5-HT3Rs. However, the noncompetitive nAChR antagonist mecamylamine acted as a potent competitive inhibitor of 125I-epibatidine binding to 5-HT3Rs. Epibatidine inhibited serotonin-induced currents mediated by endogenous 5-HT3Rs in neuroblastoma cell lines and 5-HT3ARs expressed in HEK cells in a competitive manner. Our results demonstrate that 5-HT3Rs are previously uncharacterized high affinity epibatidine binding sites in the brain and indicate that epibatidine and mecamylamine act as 5-HT3R antagonists. Previous studies that depended on epibatidine and mecamylamine as nAChR-specific ligands, in particular studies of analgesic properties of epibatidine, may need to be reinterpreted with respect to the potential role of 5-HT3Rs.

Epibatidine is an alkaloid originally isolated from Ecuadorian poison frog skin (1). Electrophysiological and ligand binding studies demonstrate that epibatidine is an agonist that binds with high affinity to a subset of nAChRs 2 throughout the central and peripheral nervous systems (2)(3)(4)(5)(6). Epibatidine administration has toxic effects, such as seizures and hypertension (7)(8)(9). However, at similar doses, epibatidine has antinociceptive effects that are blocked by nAChR antagonists, such as mecamylamine, consistent with the analgesic activity of epibatidine being mediated through nAChRs (2,10,11). Administration of mecamylamine alone also has pronounced physiological effects that alter pain perception and blood pressure (7,12,13). 5-HT 3 Rs and nAChRs are members of the same ionotropic neurotransmitter receptor family that also includes GABA A and glycine receptors (14,15). Within this neurotransmitter receptor family, 5-HT 3 Rs are most similar to nAChRs, sharing up to 30% sequence homology (16,17). Physiological studies indicate that 5-HT 3 Rs and nAChRs regulate many of the same pathways (18 -21). Co-expression of these two classes of receptors occurs in regions of the nervous system associated with pain processing, such as the dorsal horn of the spinal cord, nucleus of the solitary tract, raphe magnus nucleus, and nucleus dorsalis (22)(23)(24)(25). Local or systemic administration of nAChR agonists reduces the pain response in rodents (2, 20, 26 -30). 5-HT 3 R ligands appear to be both pro-and antinociceptive, depending on the pain stimulus and route of administration (19,(31)(32)(33)(34)(35). 5-HT 3 Rs and nAChRs also contribute to the modulation of blood pressure. Activation of nAChRs increases pressor and heart rate in normotensive and spontaneously hypertensive animals (18, 36 -38). In spontaneously hypertensive rats, chronic inhibition of 5-HT 3 Rs lowers blood pressure (39), whereas in normal rats, activation of 5-HT 3 Rs decreases arterial pressure (40). Inhibition of 5-HT 3 Rs likewise produced a significant decrease in blood pressure in obese rats prone to stress-induced hypertension (41). In addition to these effects, drug-induced and epilepsy-associated convulsive activity display an nAChR (42-44) and 5-HT 3 R (45-47) sensitivity. These observations link 5-HT 3 R function with responses associated with administration of some nAChR-specific ligands.
Experimental evidence indicates that select nAChR ligands cross-react with 5-HT 3 Rs as either agonists or antagonists (48 -52). In this study, using a combination of neuroblastoma cell lines, heterologous expression of 5-HT 3A Rs, and rat brain preparations, we show that epibatidine binds with high affinity to 5-HT 3 Rs. In addition, both epibatidine and mecamylamine function as antagonists when bound to 5-HT 3 Rs. Thus, 5-HT 3 Rs may participate in physiological effects previously thought to be nAChR-specific.

EXPERIMENTAL PROCEDURES
Cell Lines and Transfection-The human kidney epithelial cell line tsA201 (HEK cells) was maintained in Dulbecco's modified Eagle's medium supplemented with 10% calf serum (Hyclone, Logan, UT). Hemagglutinin epitope-tagged mouse 5HT 3A was subcloned into the pcDNA3.1 vector using standard methodology. HEK cells were transiently transfected with the subunit cDNA constructs using a calcium phosphate protocol (53). In some experiments, for comparison with mouse 5-HT 3A receptors, HEK cells were transfected with cDNA encoding the human 5-HT 3A receptor subunit. Mouse neuroblastoma N1E-115 cells (ATCC) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C, 5% CO 2 . Mouse neuroblastoma NB41A3 cells (ATCC) were maintained in F-12 medium supplemented with 15% horse serum, 5% fetal calf serum. Cells were plated onto 35-mm culture dishes (Fisher) for use in electrophysiological experiments. All medium was supplemented with penicillin (100 IU/ml) and streptomycin (100 g/ml).
Rat Brain Membranes-Membranes were prepared as described previously (54). Briefly, adult rats were decapitated, and the entire brain was dissected and placed in ice-cold 50 mM NaPO 4 , pH 7.4, 50 mM NaCl, 2 mM EDTA, and 2 mM EGTA plus protease inhibitors (2 mM phenylmethylsulfonyl fluoride, 2 mM N-ethylmaleimide, chymostatin, pepstatin, 1-chloro-3-tosylamido-7-amino-2-heptanone, and leupeptin at 10 g/ml). All chemicals were obtained from Sigma unless otherwise specified. The brain tissue was minced and homogenized in a Teflon-glass homogenizer. Homogenates were centrifuged at 100,000 ϫ g for 1 h. Pellets were taken through one more cycle of homogenization and centrifugation. The resulting pellets were resuspended in 150 mM NaCl, 5 mM EDTA, 50 mM Tris, pH 7.4, 0.02% NaN 3 plus protease inhibitors and frozen at Ϫ80°C until needed.
Ligand Binding Studies-Unlabeled epibatidine in phosphate-buffered saline was added to 125 I-epibatidine (2200 Ci/mmol) or [ 3 H]epibatidine (47.3 Ci/mmol) (PerkinElmer Life Sciences) to a total epibatidine concentration of 1.7 and 10 M, respectively. Equilibrium binding to whole cell membrane fractions was measured by incubating at room temperature in phosphate-buffered saline with increasing amounts of radiolabeled epibatidine on a table top rotator for 20 min. For competitive binding experiments, cells were preincubated with different serotonergic or nicotinic ligands at the specified concentrations, followed by the addition of 50 nM 125 I-epibatidine. Nonspecific binding was determined in the presence of 1 mM epibatidine, 100 M MDL72222 or by performing binding on cells transfected with empty vector. Samples were washed three times rapidly with ice-cold phosphate-buffered saline in a Brandel apparatus. 125  Electrophysiological Recording-The whole cell configuration of the patch clamp technique was used to record currents from cells expressing either recombinant 5-HT 3A receptors (HEK cells) or native 5-HT 3 receptors (N1E-115 and NB41A3 cells). The electrode solution contained 140 mM CsCl, 2 mM MgCl 2 , 0.1 mM CaCl 2 , 1.1 mM EGTA, and 10 mM HEPES (pH 7.4 with CsOH). The extracellular solution contained 140 mM NaCl, 2.8 mM KCl, 2 mM MgCl 2 , 1 mM CaCl 2 , 10 mM HEPES, and 10 mM glucose (pH 7.4 with NaOH). Cells were voltageclamped at an electrode potential of Ϫ60 mV. 5-HT 3 receptors were activated by locally applying agonists to the cell by pressure (10 p.s.i.) ejection (Picospritzer II; General Valve Corp.) from modified patch pipettes. The recording chamber was continuously perfused with extracellular solution (5 ml/min). Experiments took place at 20 -22°C. Currents were monitored using an Axopatch 200B, low pass-filtered at 2 kHz, digitized at 10 KHz using a Digidata 1320A interface, and acquired using pCLAMP8 software (all from Axon Instruments) onto the hard drive of a personal computer for off-line analysis. The peak amplitudes of agonist-activated currents were measured using pCLAMP8 software. Concentration-response relationships were fitted with modified logistic functions to determine EC 50 , IC 50 , and Hill slope values, as described previously (56).

RESULTS
High Affinity Epibatidine Binding in Rat Brain to Sites Other than nAChRs-When relatively high concentrations of 125 Iepibatidine (50 nM) were used to measure binding to membranes prepared from rat brain homogenates, a component of the 125 I-epibatidine binding was insensitive to nicotine competition (Fig. 1A). Although nicotine blocked 75% of the binding to rat brain membranes, 25% of the binding remained intact even at nicotine concentrations as high as 1 mM. Nicotine blocked 125 I-epibatidine binding with a K I of 2 nM, which was derived from the fit of the Hill equation to the data in Fig. 1A. The K I value for nicotine is similar to other measurements using epibatidine binding to rat brain synaptosomal membranes (55) and rat cerebral cortex (57). These data indicate that 25% of the 125 I-epibatidine binding is to sites other than nAChRs. Since epibatidine has been shown to block serotonininduced currents mediated by 5-HT3 A Rs in oocytes (51), and nAChRs and 5-HT 3 Rs share a high degree of sequence similarity (16), we tested the effects 5-HT 3 R-specific ligands on 125 Iepibatidine binding to rat brain fractions. As shown in Fig. 1B, increasing concentrations of the 5-HT 3 R-specific antagonist, MDL72222, blocked 25% of the 125 I-epibatidine binding in experiments parallel to those in Fig. 1A using the rat brain membranes. We obtained similar data using a 5-HT 3 R-specific agonist, chlorophenylbiguanide (CPBG) (data not shown). The effects of nicotine and the 5-HT 3 R-specific ligands were additive, as shown in Fig. 1C. Preincubation of the rat brain membranes with saturating concentrations of either nicotine and MDL72222 or nicotine and CPBG blocked all specific 125 I-epibatidine binding to the membrane. These results indicate that 125 I-epibatidine binds to two distinct subpopulations of receptors in rat brain: 1) a nicotine-sensitive subpopulation consisting of nAChRs and 2) a nicotine-insensitive subpopulation consisting of 5-HT 3 Rs.
Epibatidine Binding to Native and Heterologous 5-HT 3 Rs-To further test whether there are 125 I-epibatidine binding sites on 5-HT 3 Rs, we performed binding assays on preparations in which native and heterologously expressed 5-HT 3 Rs were present in the absence of any significant nAChR expression. To study native 5-HT 3 Rs, we used undifferentiated mouse neuroblastoma N1E-115 cells. These cells express endogenous 5-HT 3A and 5-HT 3B subunits (58) but little to no nicotinic receptors (59). Consistent with an absence of nAChR high affinity sites, we detected no specific binding to membranes from undifferentiated N1E-115 cells at 125 I-epibatidine concentrations of 0.5 nM. At higher concentrations, we observed an increasing level of specific, high affinity 125 I-epibatidine binding to the neuroblastoma cell 5-HT 3 Rs with an estimated K D of 27 Ϯ 6 nM ( Fig. 2A). We obtained similar results (an estimated K D of 22 Ϯ 5 nM) with [ 3 H]epibatidine binding to neuroblastoma cell 5-HT 3 Rs (Fig. 2A). Serotonin, MDL72222, and CPBG were highly potent inhibitors of 125 Iepibatidine binding to N1E-115 cell membranes as compared with nicotine in competition binding experiments (Fig. 2B). Surprisingly, mecamylamine, a noncompetitive antagonist of different nAChRs, blocked 125 I-epibatidine binding to 5-HT 3 Rs. The relative potency in competing for 125 I-epibatidine binding sites in N1E-115 cell membranes was MDL72222 Ͼ epibatidine Ͼ 5-HT ϭ CPBG Ͼ mecamylamine nicotine (Table 1). 5-HT 3 Rs native to neuroblastoma cells have functional properties similar to those of homomeric 5-HT 3A receptors (60). To compare 5-HT 3 Rs of neuroblastoma cells with those of known composition, we performed experiments on membranes from HEK cells expressing mouse 5-HT 3A subunits. As shown in Fig.  2C, 5-HT 3A Rs bound 125 I-epibatidine with a K D of 14 Ϯ 4 nM, which is approximately half of the K D observed for native receptors in N1E-115 cells. Competition binding experiments were also performed on membranes from HEK cells expressing 5-HT 3A Rs. The relative potency in competing for 125 I-epibatidine binding sites from the HEK cell membranes was epibatidine Ͼ MDL72222 Ͼ 5-HT ϭ mecamylamine Ͼ CPBG nicotine ( Table 1).
D-Tubocurarine, a plant alkaloid that inhibits nAChR function, also inhibits 5-HT 3 Rs (50). The potency of D-tubocurarine as an antagonist of 5-HT 3 R-mediated currents is highly speciesdependent. Human 5-HT 3A Rs are Ͼ1000 times less sensitive to inhibition by D-tubocurarine than are rodent 5-HT 3A Rs. Therefore, we examined whether the antagonist potency of epibatidine also varies between mouse and human 5-HT 3A Rs. 5-HT (30 M)-activated currents recorded from HEK cells expressing recombinant human 5-HT 3A Rs were inhibited by epibatidine with an IC 50 of 7.6 Ϯ 0.7 M (n ϭ 5), a value that is only slightly higher than the IC 50 of epibatidine as an inhibitor of mouse 5-HT 3A Rs (Table 2).

DISCUSSION
In this study, we identify high affinity 125 I-epibatidine binding sites in the brain with the pharmacological characteristics of 5-HT 3 Rs. These sites are distinctly serotonergic, since the 5-HT 3 R agonist (CPBG) and antagonist (MDL72222) effectively block 125 I-epibatidine binding, whereas nicotine is ineffective at blocking binding. Nicotine is also a poor competitor against 125 I-epibatidine binding to 5-HT 3A R homomers expressed heterologously in HEK cells and endogenous 5-HT 3 Rs expressed by N1E-115 mouse neuroblastoma cells. In these cells, epibatidine blocks serotonin-induced currents in a competitive manner with IC 50 values similar to that observed for the inhibition of recombinant 5-HT 3 R-mediated currents in oocytes (51). Thus, in addition to being an agonist at nAChRs, we conclude that epibatidine is a competitive antagonist of 5-HT 3 Rs.
The affinity of epibatidine for 5-HT 3 R sites is approximately 3 orders of magnitude lower than for nAChR subtypes that bind epibatidine with high affinity. Epibatidine binds to these  Table 1. Each point represents the mean Ϯ S.D. of three determinations.
nAChRs with K D values in the 10 -100 pM range (57,62,63). This variability might arise from ligand depletion during binding studies that can result in underestimation of the affinity of epibatidine for nAChRs. In studies that avoided ligand depletion, epibatidine binds to a single population of brain nAChRs with K D values in the 1-8 pM range (4,55). Because of its extremely high affinity for nAChRs, virtually all studies measuring equilibrium binding have used epibatidine concentrations less than 1 nM, which would result in little, if any, binding to 5-HT 3 Rs. Thus, the failure to observe epibatidine binding to 5-HT 3 Rs in earlier studies is most likely due to the relatively low concentrations of radiolabeled epibatidine used to measure binding to nAChRs. When epibatidine concentrations much higher than 1 nM were used in radiolabeled epibatidine binding studies, a second population of binding sites with lower affinity for epibatidine was observed (4,64,65). Based on the works of Marks et al. (65), using nAChR subunit null mice, at least some of the lower affinity epibatidine sites consist of nAChRs. However, data presented in this study suggest that a portion of the lower affinity sites may also consist of 5-HT 3 Rs.
Our results are consistent with earlier studies that have established significant pharmacological crossover between 5-HT 3 Rs and nAChRs. For example, the nAChR antagonist, D-tubocurarine, and 5-HT 3 R-specific antagonists, such as MDL72222 and GR65630, bind to 5-HT 3 Rs with comparable affinities. Furthermore, D-tubocurarine blocks 5-HT 3 R-mediated currents (49,50). Similar actions are observed for the nAChR channel blockers, chlorpromazine and QX222 (48). Other studies also show that tropisetron, which is a selective 5-HT 3 R antagonist, binds with high affinity and is a potent partial agonist at the ␣7 nAChR (66), whereas the ␣7 agonist, PSAB-OFP, acts as a potent agonist at 5-HT 3 Rs (67). Relevant to our results, the nAChR agonists, anabaseine and epibatidine, are shown to block serotonin-activated currents in oocytes expressing 5-HT 3A Rs with IC 50 values of 14 and 8 M, respectively (51,52). Similarly, we find that epibatidine effectively blocks rodent 5-HT 3 R-mediated currents in a competitive manner with IC 50 values of 3.3 and 4.1 M in neuroblastoma cells and HEK cells, respectively.
The mechanisms underlying the physiological effects of epibatidine have not been fully characterized. The concentrations of epibatidine required for analgesia are in the micromolar range (27,29,30,68,69) and also produce adverse symptoms,   including hypothermia, hypertension, ataxia, and seizures (8,70). These concentrations are similar to the concentrations that we found antagonize 5-HT 3 Rs and are orders of magnitude higher than what is needed to saturate the high affinity binding site on nAChRs but in line with the epibatidine concentrations that activate nAChRs. As an agonist, epibatidine (10 -100 nM) increases neurotransmitter release from hippocampal slice preparations (70,71) and current responses in hippocampal and retinal neurons (72,73). Using heterologous expression of ␣4␤2 nAChRs, Buisson et al. found a similar dose dependence for epibatidine as an nAChR agonist (74). Further evidence that agonist activity of epibatidine underlies its physiological role is that the nicotinic antagonist mecamylamine blocks both the functional activity and antinociceptive effects of epibatidine (2,10,11). However, the fact that mecamylamine has pronociceptive properties (12) and our data showing that mecamylamine competes with the epibatidine binding to 5-HT 3 Rs (Table 1) indicate that the previous findings with mecamylamine are also consistent with these ligands interacting with 5-HT 3 R sites. Our findings that epibatidine binds with high affinity to and antagonizes 5-HT 3 Rs raise questions about where epibatidine elicits its physiological actions and indicate that 5-HT 3 Rs should be considered as a potential target.