Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The nicotinic acetylcholine (ACh) receptor (AChR) and 5-hydroxytryptamine (5-HT)₃ receptor (5-HT₃R) belong to a superfamily of the ligand-gated ion channels, which mediate fast synaptic transmission in the nervous system (reviewed in Derkach et al., 1989; Unwin, 1993; Jackson and Yakel, 1995). Each of the members of the ligand-gated ion channel family exhibits a high degree of sequence similarity, with hydropathy plots of the subunit sequences predicting four transmembrane domains, a large extracellular N-terminal domain, an intracellular loop between the third and the fourth membrane-spanning segments, and a short extracellular C-terminal domain. The AChR and 5-HT₃R are both integral membrane glycoproteins composed of transmembrane subunits arranged as a pentamer around a central ion channel pore (reviewed in Bertrand and Changeux, 1995; Hucho et al., 1996). Despite clear differences in channel properties, including ion selectivities, there is a rapidly growing body of evidence indicating that these receptors share a significant degree of structural homology. For example, a chimeric receptor consisting of the extracellular, agonist-binding, N-terminal domain of the neuronal 7 AChR and the corresponding C-terminal region of the 5-HT₃R forms functional channels that are gated by ACh and have permeability properties similar to those of the 5-HT₃R (Eisele et al., 1993; Kriegler et al., 1999).

One method for comparing and contrasting the structures of the different ligand-gated ion channel family members is to examine ligands that bind to more than one receptor. Along these lines, the competitive antagonist d-tubocurarine has been shown to bind with nanomolar affinity to both the AChR and 5-HT₃R (Peters et al., 1990; Yan et al., 1998). Furthermore, structural analogs of d-tubocurarine bind to each receptor with similar affinity and site-selectivity, suggesting that the ligand-binding sites for these two receptors share common structural features (Yan et al., 1998). Noncompetitive antagonists (NCAs), compounds that antagonize receptor function by binding to a site that is distinct from the agonist-binding site, are ligands that can also be used to compare structural features of the different ligand-gated ion channel members. For the AChR, affinity labeling studies with photoreactive NCAs such as chlorpromazine (CPZ) (Giraudat et al., 1986), triphenylmethylphosphonium (TPMP) (Hucho et al., 1986), and 3-trifluoromethyl-3-(m[¹²⁵I]iodophenyl)diazirine ([¹²⁵I]TID) (White and Cohen, 1992) have been instrumental in identifying the membrane-spanning segment M2 as forming the lining of the pore of the channel. A large number of structurally diverse compounds have been identified as NCAs of both muscle- and neuronal-type AChRs. Under equilibrium binding conditions, the majority of AChR NCAs bind preferentially to the desensitized conformation of the AChR [e.g., CPZ, TPMP, phencyclidine (PCP), and meproadifen]. There are, however, examples of NCAs that either bind preferentially to the resting receptor (tetracaine) (Cohen et al., 1985) or bind equally to both the resting and desensitized receptor states (TID) (White et al., 1991). In contrast, for the 5-HT₃R, very few NCAs have been identified or characterized in any detail (Fan, 1994, 1995).

The compound 3,4,5-trimethoxybenzoic acid 8-(diethylamino)octyl ester (TMB-8), which is used widely as a calcium channel antagonist, has also been shown to be an NCA of both muscle- and neuronal-type AChRs (Bencherif et al., 1995) and, more recently, as an open-channel blocker of 42 neuronal AChRs (Buisson and Bertrand, 1998). In the present study, we further examined the interaction of TMB-8 with the AChR. We also assessed the interaction of TMB-8 with the 5-HT₃R to determine whether TMB-8 is an effective 5-HT₃R NCA. Using a combination of photoaffinity labeling and radioligand-binding assays using AChR NCAs ([¹²⁵I]TID, [³H]PCP) and the agonist [³H]nicotine, we found that like TID, TMB-8 is a potent NCA of the AChR, binding with micromolar affinity to the resting and desensitized AChR channel. In contrast, using both electrophysiological and radioligand-binding methods, we show that TMB-8 acts as a competitive antagonist of the mouse 5-HT₃R. Given the sequence similarities of the channel-lining M2 segments of the 5-HT₃R and AChR subunits and the emerging evidence of structural homology of each channel (Xu and Akabas, 1996), these results demonstrate that there are important differences in the structure of the channels of these two homologous receptors.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Torpedo californica electric organ was obtained frozen from Aquatic Research Consultants (San Pedro, CA). [¹²⁵I]TID (10 Ci/mmol) was obtained from Amersham Corp. (Arlington Heights, IL). [³H]PCP (52 Ci/mmol) and [³H]GR65630 (60 Ci/mmol) were obtained from New England Nuclear (Boston, MA). [³H]Nicotine (64 Ci/mmol) was from American Radiolabeled Chemicals (St. Louis, MO). TMB-8, proadifen, and carbamylcholine were obtained from Sigma Chemical Co. (St. Louis, MO). -Bungarotoxin was from Research Biochemicals Inc. (St. Louis, MO).

AChR-Rich Membranes. AChR-rich membranes were isolated from the electric organ of T. californica according to the procedure of Sobel et al., (1977) with the modifications described previously (Pedersen et al., 1986). The final membrane suspensions in ~38% sucrose/0.02% NaN₃ were stored at 80°C.

[¹²⁵I]TID Labeling of AChR-Rich Membranes. For labeling experiments, T. californica AChR-rich membranes [1 mg/ml in vesicle dialysis buffer, 10 mM 3-(N-morpholino)propanesulfonic acid, 100 mM NaCl, 0.1 mM EDTA, and 0.02% NaN₃, pH 7.5] were incubated for 2 h at room temperature with [¹²⁵I]TID (1.25 µM) in the absence or presence of 250 µM carbamylcholine or TMB-8 (Fig. 1) at various concentrations. In some experiments, AChR-rich membranes were preequilibrated with 10 µM -bungarotoxin for 20 min before the addition of TMB-8. Incubations were performed in 10 × 75-mm glass culture tubes in the dark. The samples were then irradiated with a 365-nm UV lamp (Spectroline EN-280L) for 7 min at a distance of less than 1 cm, and the membrane suspensions were transferred to 1.5-ml microcentrifuge tubes and centrifuged at 39,000g for 1 h. Pellets were solubilized in electrophoresis sample buffer and subjected to SDS-polyacrylamide gel electrophoresis.

View larger version (6K): [in this window] [in a new window]   Fig. 1.   Chemical structure of TMB-8.

[³H]PCP and [³H]Nicotine Radioligand Binding Assays. The equilibrium binding of the AChR NCA [³H]PCP and the agonist [³H]nicotine with T. californica AChR-rich membranes was assayed by centrifugation. For [³H]PCP assays, single 500-µl aliquots of membrane suspensions (0.5 mg protein/ml in vesicle dialysis buffer, ~0.6 µM AChR) were equilibrated in the presence of 250 µM carbamylcholine with [³H]PCP (6 nM) and increasing concentrations of TMB-8 for 2.5 h. For [³H]nicotine assays, membrane suspensions were equilibrated with [³H]nicotine (19.4 nM) and increasing concentrations of TMB-8 for 3.5 h. For both assays, membrane suspensions were equilibrated at room temperature in 10 × 75-mm disposable culture tubes (Corning) and then transferred to 1.5-ml plastic microcentrifuge tubes and pelleted by centrifugation at 39,000g for 1 h (Beckman JA-20 rotor). After removal of the supernatants, the membrane pellets were solubilized in 100 µl of 10% SDS, and the pellet ³H was determined by liquid scintillation counting. Nonspecific binding of [³H]PCP was measured in the presence of the NCA proadifen (0.2 mM; Cohen et al., 1985), that of [³H]nicotine in the presence of 0.2 mM carbamylcholine.

Electrophysiological Recordings. Mouse and human 5-HT₃R cDNAs were subcloned into pCR-Script Amp SK(+) (Stratagene, La Jolla, CA). The cDNAs were prepared and then transcribed with T3 mMESSAGE mMACHINE (Ambion, Austin, TX) by following manufacturer's instructions. Xenopus laevis oocytes were prepared and injected as described previously (Machu et al., 1998).

[³H]GR65630 Binding Assay. Full-length mouse 5-HT₃R cDNA was cloned into the baculovirus transfer vector pBacHis-3. Production of recombinant baculovirus and viral infections was conducted by using the media and protocols described in the BacPAK baculovirus expression system (ClonTech, Palo Alto, CA). Briefly, the host cell line Sf21 was infected with recombinant baculovirus. Recombinant baculovirus was then plaque purified, propagated, and used to infect Sf21 insect cells growing in Grace's insect cell medium containing 10% FBS.

Data Analysis. GraphPAD Prism was used to calculate EC₅₀ values, Hill coefficients, and K_i values and to generate the Schild plot. IC₅₀ and EC₅₀ values are reported as both arithmetic and geometric mean values.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Effect of TMB-8 on [¹²⁵I]TID Labeling of Resting AChR Channel. In this study, we wanted to further characterize the interaction of TMB-8 (Fig. 1) with the AChR. We examined the effect of TMB-8 on the binding of two well characterized NCAs of the AChR, both of which bind within the pore of the receptor ion channel. The uncharged photoreactive compound [¹²⁵I]TID is a potent NCA of the AChR, binding with micromolar affinity to both the resting and desensitized states of the receptor (White et al., 1991; Wu et al., 1994). In the resting state, [¹²⁵I]TID specifically photolabels homologous aliphatic residues at positions 9 and 13 in each channel-lining M2 segment (e.g., Leu-265 and Val-269; White and Cohen, 1992). Although TID binds with equal affinity to both the resting and desensitized channel, [¹²⁵I]TID photoincorporates into the resting channel ~10-fold more efficiently than into the desensitized channel. In addition, in the absence of agonist, the vast majority of [¹²⁵I]TID incorporation into individual receptor subunits reflects labeling of the resting channel. The addition of AChR agonists or NCAs such as tetracaine reduces by more than 75% the incorporation of [¹²⁵I]TID into receptor subunits (White and Cohen, 1992; Moore and McCarthy, 1994).

View larger version (48K): [in this window] [in a new window]   Fig. 2.   Effects of carbamylcholine or TMB-8 on the photoincorporation of [125I]TID into AChR-rich membranes. AChR-rich membranes were equilibrated (2 h) with [125I]TID (1.25 µM) in the absence (lanes 1-5) and in the presence (lane 6) of 250 µM carbamylcholine or in the presence of 0.33, 3.3, 16.5, or 33 µM TMB-8 (lanes 2-5). AChR-rich membranes were then irradiated at 365 nm for 7 min, and polypeptides were resolved by SDS-polyacrylamide gel electrophoresis. A, corresponding autoradiograph with the positions of the AChR subunits indicated on the left. B, individual AChR subunit bands were excised from the gel, and the amount of [125I]TID incorporated into each subunit was determined by gamma counting (). Shown are data for the AChR  subunit. The dashed line corresponds to the amount of [125I]TID incorporated in the AChR  subunit in the presence of carbamylcholine and represents the amount of nonspecific incorporation (see Results). The solid line represents the nonlinear least-squares fit of the binding data [IC50 = 3.1 ± 0.6 (S.E.M.) µM; nH = 0.86 ± 0.12 (S.E.M.)].

Effect of TMB-8 on [³H]PCP binding to Desensitized AChR. In the presence of agonist, [³H]PCP binds with high affinity (K_eq = 1 µM) to a single site per AChR (Heidmann et al., 1983); the results of other studies also strongly support a binding site for PCP in the desensitized channel (Eaton et al., 1997; Blanton et al., 1998a). The effect of TMB-8 on the equilibrium binding of [³H]PCP to the desensitized state of the AChR was determined using a simple centrifugation assay. As Fig. 3 shows, in the presence of agonist, the addition of TMB-8 results in a concentration-dependent reduction in the amount of [³H]PCP bound to the AChR. TMB-8 completely eliminates all the specifically bound [³H]PCP (99.4%) with a calculated IC₅₀ value of 2.4 ± 0.1 µM. A repeat of the [³H]PCP-binding assay yielded a nearly identical IC₅₀ value of 2.24 ± 0.14. 

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of TMB-8 on the binding of [3H]PCP to AChR-rich membranes in the presence of carbamylcholine. AChR-rich membranes (0.5 mg/ml, 0.6 µM ACh-binding sites) containing 6 nM [3H]PCP and 250 µM carbamylcholine were equilibrated for 2.5 h with increasing concentrations of TMB-8. Bound tritiated ligand () was determined by centrifugation (see Experimental Procedures). The dashed line indicates nonspecific bound tritiated ligand in the presence of the AChR NCA proadifen (200 µM; ) at 0 and 200 µM TMB-8. The solid line represents the nonlinear least-squares fit of the binding data [IC50 = 2.4 ± 0.1 (S.E.M.) µM, nH = 0.93 ± 0.03 (S.E.M.)].

Inhibitory Effect of TMB-8 on 5-HT-Evoked Currents in X. laevis Oocytes Expressing Mouse and Human 5-HT₃Rs. TMB-8 was found to markedly reduce currents elicited by 5-HT in oocytes expressing 5-HT₃Rs. As indicated in the representative tracings shown in Fig. 4, 10 µM TMB-8 inhibited 5-HT-evoked currents by ~50% in both mouse (Fig. 4A) and human (data not shown) 5-HT₃Rs. The washout of TMB-8 inhibition was observed with the next application of 5-HT 5 min later. Concentration-response curves were generated for oocytes expressing human or mouse 5-HT₃Rs (Fig. 4B). The concentrations of 5-HT used (0.5 and 0.75 µM for mouse and human 5-HT₃Rs, respectively) represent effective concentrations that produce an average of 10% of the maximal response (EC₁₀). Mouse and human 5-HT₃Rs have similar sensitivities to TMB-8. An IC₅₀ value of 8.4 ±0.96 µM (geometric mean = 8.5 ± 2.6 µM) was obtained for mouse 5-HT₃Rs, whereas an IC₅₀ value of 11.76 ± 0.01 µM (geometric mean = 12 ± 4.9 µM) was obtained for human 5-HT₃Rs. Hill coefficients were 1.21 ± 0.12 (geometric mean = 1.58 ± 0.23) and 0.96 ± 0.13 (geometric mean = 1.08 ± 0.09) for mouse and human 5-HT₃Rs, respectively.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Effect of TMB-8 on 5-HT-mediated responses in oocytes expressing mouse 5-HT3Rs. Representative tracings of currents produced in a single oocyte expressing mouse (A) 5-HT3Rs. Serotonin (0.5 µM and 0.75 µM) was applied for 30 s to obtain a baseline current for mouse and human 5-HT3Rs, respectively. TMB-8 (10 µM) and 5-HT were coapplied, and an inhibition of ~50% of the baseline current was observed. Washout of TMB-8 inhibition was observed with the next application of 5-HT 5 min later. B, concentration-response curves for TMB-8 were generated in oocytes expressing mouse or human 5-HT3Rs. TMB-8 (0-1000 µM) was coapplied with 5-HT for 30 s. An IC50 value of 11.76 ± 0.01 µM (geometric mean = 12 ± 4.9 µM) was obtained for human 5-HT3R. For mouse 5-HT3Rs, the IC50 value was 8.4 ± 0.96 µM (geometric mean = 8.5 ± 2.6 µM). Hill coefficients were 0.96 ± 0.13 (geometric mean = 1.08 ± 0.09) and 1.21 ± 0.12 (geometric mean = 1.58 ± 0.23) for human and mouse 5-HT3Rs, respectively.

TMB-8 Is a Competitive Antagonist of 5-HT₃R. To determine the nature of inhibition of TMB-8 at the mouse 5-HT₃R, 5-HT concentration-response curves were generated in the absence or presence of 5, 50, 75, or 100 µM TMB-8 (Fig. 5A). Parallel shifts in the 5-HT concentration-response curves were obtained as the TMB-8 concentration was increased. The inhibition produced by each TMB-8 concentration was completely overcome by high concentrations of 5-HT. As the TMB-8 concentration increased, the EC₅₀ values for 5-HT increased, from 0.998 ± 0.002, 1.381 ± 0.006, 8.684 ± 0.1412, and 31.65 ± 0.15 to 75.13 ± 0.30 µM (respective geometric mean values are 0.98 ± 0.05, 1.48 ± 0.12, 11.14 ± 3.7, 31.72 ± 5.3, and 77.84 ± 5.9 µM) in the presence of 0, 5, 50, 75, and 1000 µM TMB-8, respectively. To fully characterize the competitive nature of TMB-8 inhibition at 5-HT₃Rs, a Schild plot was generated (Fig. 5B). The Schild plot had a slope of 2.22 ± 0.21, which is significantly different from zero, suggesting that TMB-8 is a competitive antagonist of 5-HT₃Rs. The pA2 was 4.78, and the K_i of TMB-8 at the 5-HT₃R was 16.38 µM. These results rule out a noncompetitive form of TMB-8 inhibition at 5-HT₃Rs over the range of TMB-8 concentrations tested.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   TMB-8 shifts the EC50 values of 5-HT for mouse 5-HT3Rs. A, 5-HT concentration-response curves were generated in the presence or absence of TMB-8 (n = 2-6). The EC50 values generated for each of the curves were 0.998 ± 0.002, 1.381 ± 0.006, 8.684 ± 0.1412, 31.65 ± 0.15, and 75.13 ± 0.30 µM (respective geometric means are 0.98 ± 0.05, 1.48 ± 0.12, 11.14 ± 3.7, 31.72 ± 5.3, and 77.84 ± 5.9 µM) in the presence of 0, 5, 50, 75, and 100 µM TMB-8, respectively. B, the apparent competitive nature of TMB-8 inhibition for mouse 5-HT3R was confirmed by Schild plot analysis. The pA2 value is 4.78, and the Ki value is 16.38 µM.

Binding Studies Confirm Competitive Nature of TMB-8 Inhibition of 5-HT₃R. The competitive antagonism of 5-HT₃Rs by TMB-8 was further confirmed by a competitive binding study. TMB-8 was used to displace binding of the 5-HT₃R competitive antagonist [³H]GR65630 to membranes of Sf21 insect cells infected with recombinant baculovirus containing the mouse 5-HT₃R. Competition binding analysis indicated that TMB-8 displaced the binding of [³H]GR65630 to 5-HT₃Rs in a concentration-dependent manner (Fig. 6). In the presence of 0.4 nM [³H]GR65630, specific binding was completely displaced with TMB-8 concentrations above 10 µM, whereas in the presence of 1 nM [³H]GR65630, specific binding was completely displaced with TMB-8 concentrations above 316 µM. The IC₅₀ values for TMB-8 were 0.19 ± 0.005 µM (geometric mean = 0.194 ± 0.05 µM) and 7.6 ± 0.64 µM (geometric mean = 7.9 ± 0.09 µM) in the presence of 0.4 and 1 nM [³H]GR65630, respectively. The calculated K_i value was 2.5 ± 0.21 µM (geometric mean = 2.25 ± 0.026 µM; specific binding = 85%).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   TMB-8 reduces ligand affinity of [3H]GR65630. Crude cell membrane proteins isolated from Sf 21 insect cells infected with recombinant baculovirus containing a full-length mouse 5-HT3R cDNA were used. Binding assays were conducted as described in Experimental Procedures with 0.4 nM () or 1 nM () [3H]GR65630, respectively. All analyses were performed in duplicate. The IC50 values for TMB-8 inhibition were 0.19 ± 0.005 µM (geometric mean = 0.194 ± 0.05 µM) and 7.6 ± 0.64 µM (geometric mean = 7.9 ± 0.09 µM) in the presence of 0.4 and 1 nM [3H]GR65630, respectively. The calculated Ki value was 2.5 ± 0.21 µM (geometric mean = 2.25 ± 0.026 µM; specific binding was 85%).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

One of the goals of the present work was to further characterize the interaction of TMB-8 with the nicotinic AChR. In this regard, the principal findings were that 1) TMB-8 is a potent inhibitor of the incorporation of the photoreactive NCA [¹²⁵I]TID into the resting AChR channel, and 2) TMB-8 is also an effective inhibitor of the high-affinity equilibrium binding of [³H]PCP to the desensitized receptor. TMB-8 has also been shown to be an effective open-channel blocker of human 42 neuronal nicotinic AChRs (Buisson and Bertrand, 1998); therefore, TMB-8 binds with high affinity to the resting, open, and desensitized AChR channel. Under equilibrium binding conditions, the majority of AChR NCAs bind preferentially to the desensitized state. There are far fewer examples of NCAs, such as dl-perhydrohistrionicotoxin (Blanchard et al., 1979), TID (White et al., 1991), and diazofluorene (Blanton et al., 1998a), which bind with approximately equal affinity to both the resting and desensitized states of the receptor. Finally, there are only a few known examples of compounds that bind preferentially to the resting state of the receptor (e.g., tetracaine; Cohen et al., 1985; adiphenine, Boyd and Cohen, 1984). Because of the tremendous structural diversity among NCAs, when examined as a whole, it has not been possible to identify a unique structural motif (Barrantes et al., 1997). Furthermore, it is clear, at least for the desensitized receptor, that NCAs bind to unique regions of the channel (Galzi and Changeux, 1995; Blanton et al., 1998b). It is therefore likely that there are different structural requirements for binding to different regions of the channel. In contrast, the structural requirements for binding to the AChR channel in the resting state appear to be much more stringent than those for binding to the desensitized state. A detailed examination of the structures of TMB-8, tetracaine, and so on might then provide insight into the structural requirements for binding to the resting AChR channel. TMB-8 and tetracaine both contain a tertiary amine group; however, it is unlikely that this group is solely responsible for the affinity of the compounds for the resting AChR channel because NCAs such as proadifen that also contain a tertiary amine group bind preferentially to the desensitized receptor (Boyd and Cohen, 1984). On the other hand, we determined that the compound trimethoxy benzoic acid, as well as trimethoxy benzyl alcohol, does not inhibit [¹²⁵I]TID labeling of the resting channel. These results indicate that the trimethoxy benzyl portion of TMB-8 is, by itself, not sufficient to account for the affinity of the compound for the resting channel. It is clear that the moieties at both ends of the TMB-8 molecule are important for binding to the resting channel. We are currently examining the effects of a series of TMB-8 derivatives on the incorporation of [¹²⁵I]TID into the AChR to elucidate the structural requirements for TMB-8 interaction with the resting AChR channel.

The other major goal of this study was to assess the interaction of TMB-8 with the structurally homologous 5-HT₃R. Given the sequence similarity of the M2 segments of both the 5-HT₃R and AChR subunits (Fig. 7) and the emerging evidence of homology in the channel structures (Xu and Akabas, 1996), it was anticipated that TMB-8 would also bind to the pore of the 5-HT₃R channel and act as an NCA. In the present study, we found that TMB-8 antagonized both mouse and human 5HT₃Rs with similar potency (IC₅₀ = 8.5 and 12 µM, respectively) (Fig. 4B). The 5-HT concentration-response curves generated in the absence or presence of TMB-8, however, indicate that TMB-8 is a competitive antagonist of the mouse 5-HT₃R (Fig. 5A). The fact that increasing concentrations of 5-HT completely overcome the TMB-8 antagonism is also important in that it indicates that there are not two components to the antagonism by TMB-8. In other words, if in addition to acting as a competitive antagonist TMB-8 was also acting as a channel blocker, only partial recovery of the maximum 5-HT-evoked currents would be observed. The competitive nature of TMB-8 inhibition of the 5-HT₃Rs was further confirmed with a Schild plot (Fig. 5B) and competition binding analysis (Fig. 6). The conclusion that TMB-8 acts as a competitive antagonist of the 5-HT₃R was not so surprising given that TMB-8 is not the only example of a AChR NCA that was found to be a competitive antagonist of 5-HT₃R. Chlorpromazine and the lidocaine derivative QX-222 are both channel blockers of AChRs but also competitive antagonists of 5-HT₃Rs (Sepulveda et al., 1994). Furthermore, at higher concentrations, most AChR NCAs also interact with the agonist-binding site (Heidmann et al., 1983; Moore and McCarthy, 1994). It was, however, surprising that there was no evidence to indicate any interaction of TMB-8 with the 5-HT₃R channel because the AChR and 5-HT₃R channels possess significant homology (Fig. 7; Xu and Akabas, 1996). More specifically, the competitive nature of TMB-8 and either [¹²⁵I]TID or [³H]PCP binding suggests common binding loci in the resting and desensitized AChR channels. In the resting state, [¹²⁵I]TID specifically labels homologous aliphatic residues at positions 9 and 13 in each M2 segment (e.g., Leu-265 and Val-269; White and Cohen, 1992). Sequence alignments of the M2 regions of different AChR subunits and 5-HT₃Rs (Fig. 7) show that Leu-265 (position 9) and Val-269 (position 13) are conserved between 5-HT₃Rs and AChRs. These two residues are clearly involved in forming the binding site for TID, and presumably for TMB-8, in the resting AChR channel. It remains to be determined whether additional residues in the M2 segments of AChR subunits contribute to form the intact TID/TMB-8-binding site in the resting channel or, alternatively, whether residues in the 5-HT₃R act to disrupt the formation of a binding site for TMB-8 in the 5-HT₃R channel. In the desensitized AChR channel, residues at position 6 and 10 (e.g., Ser-262 and Ala-266) are implicated in forming the binding loci for PCP (Eaton et al., 1997, 1998; Blanton et al., 1998a). In the 5-HT₃R, a threonine residue is present at position 6 and a glycine residue is present at position 10 (Fig. 7). In the T. californica AChR subunits, a serine residue is present at position 6, and either an alanine or a serine residue is present at position 10. We are presently constructing site-directed mutants of the 5-HT₃R at position 6 (Thr to Ser) and position 10 (Gly to Ala) in an attempt to create a binding site for TMB-8 in the 5-HT₃R channel. Finally, as was mentioned earlier, we are continuing to examine the effects of a series of TMB-8 structural analogs on [¹²⁵I]TID and [³H]PCP interaction with the AChR to elucidate the structural requirements of the NCA-binding site in the resting and desensitized channel. Clearly, TMB-8 provides a valuable tool for examining the channel structures of both the nicotinic AChR and 5-HT₃R.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Alignments of the amino acid sequences of the TM2 region of the T. californica AChR subunits with those of 5-HT3Rs from different species. Residues that are photoaffinity labeled with [125I]TID in the AChR , , , and  subunits are denoted by  and . Corresponding residues at positions 9 and 13 in the 5-HT3R M2 segments are identical with those in the AChR subunits.