Interactions between 3-(Trifluoromethyl)-3-(m-[¹²⁵I]iodophenyl)diazirine and Tetracaine, Phencyclidine, or Histrionicotoxin in theTorpedo Species Nicotinic Acetylcholine Receptor Ion Channel

Experimental Procedures

Materials.

nAChR-rich membranes were isolated from electric organs of Torpedo species (Winkler Enterprises, San Pedro, CA) as described elsewhere (Pedersen et al., 1986). The membranes used contained 1 to 2 nmol of [³H]ACh binding sites per milligram of protein. Endoproteinase Lys-C (EndoLys-C) was from Roche Molecular Biochemicals (Indianapolis, IN), 1-tosylamido-2-phenylethychlormethyketone-treated trypsin was from Worthington Biochemical (Freehold, NJ), and Staphylococcus aureus glutamyl endopeptidase (V8 protease) was from ICN Biomedical Inc. (Cosa Mesa, CA). Prestained low molecular mass gel standards were from Life Technologies (Gaithersburg, MD): ovalbumin (43 kDa), carbonic anhydrase (29 kDa), β-lactoglobulin (18 kDa), lysozyme (14 kDa), bovine trypsin inhibitor (6 kDa), and insulin (2.8 kDa). Genapol C-100 (10%) and trifluoroacetic acid were from Pierce (Rockford, IL). 1-Azidopyrene was bought from Molecular Probes (Eugene, OR). [¹²⁵I]TID (10 Ci/mmol) was obtained from Amersham Pharmacia Biotech (Piscataway, NJ), and [³H]tetracaine (48 Ci/mmol) was prepared by tritium reduction (PerkinElmer Life Science Products, Boston, MA) of 3,5-dibromotetracaine as described previously (Gallagher and Cohen, 1999). PCP was from Applied Science (State College, PA) and HTX was kindly provided by Dr. Y. Kishi (Harvard University, Cambridge, MA). Nonradioactive TID was a gift from Dr. S. Husain (Massachusetts General Hospital, Boston, MA) and was repurified by silica gel chromatography (100% hexane elution) and stored in 100% ethanol after the hexane was removed by rotary evaporation. Purity and concentration of the TID was assessed by silica thin-layer chromatography and UV absorption at 365 nm (Brunner and Semenza, 1981). Structures of the drugs are shown in Fig. 1.

[³H]Tetracaine Binding.

The equilibrium binding of [³H]tetracaine (10 nM) to nAChR-rich membranes in Torpedo species physiological saline (250 mM NaCl, 5 mM KCl, 3 mM CaCl₂, 2 mM MgCl₂, 5 mM NaPi, pH 7.0) was measured by a filtration assay (Middleton et al., 1999). Membrane suspensions (400 nM ACh sites) were incubated in the dark for 30 min at room temperature with [³H]tetracaine and various concentrations of nonradioactive TID before filtration, and nonspecific binding was defined as the amount of [³H]tetracaine bound in the presence of 100 μM nonradioactive tetracaine.

[¹²⁵I]TID Photolabeling.

Photolabeling on an analytical scale was used to quantify [¹²⁵I]TID photoincorporation into nAChR subunits in the presence of tetracaine at concentrations from 1 μM to 1 mM. nAChR-rich membranes (100 μl, 2 mg of protein/ml) were incubated in the dark in microtiter plates with [¹²⁵I]TID (6.5 μM) and nonradioactive tetracaine for 30 min. The suspensions were then photolyzed for 30 min at 365 nm (model EB-280C; Spectroline, Westbury, NY) at a distance of 6 cm. Sample buffer was then added directly to the membrane suspensions, which were then separated by SDS-PAGE gel on an 8% acrylamide gel. The polypeptides on the gel were visualized by Coomassie blue stain, and the [¹²⁵I]TID incorporation in the gel was detected by autoradiography or by use of a Storm PhosphorImager (Molecular Dynamics, Sunnyvale, CA). The nAChR γ-subunit¹²⁵I incorporation was quantified using the ImageQuaNT software, with background levels calculated locally.

Preparative [¹²⁵I]TID photolabeling was performed essentially as described previously (White and Cohen, 1992;Blanton and Cohen, 1994). nAChR-rich membranes (5–7 mg at 2 mg of protein/ml in Torpedo species physiological saline) were equilibrated for 30 min with 6.5 μM [¹²⁵I]TID and in the absence or presence of an amine NCA and then photolyzed at 365 nm under one of two conditions. The samples were either photolyzed for 30 min in open glass containers at a distance of 6 cm (White and Cohen, 1992) or irradiated in closed glass vials for 10 min at a distance of 1 cm (Blanton and Cohen, 1994). Because the closed vials more effectively contained volatilized [¹²⁵I]TID and because the pattern of [¹²⁵I]TID labeling within the nAChR M2 segments did not vary between the two methods, the closed vial method was used routinely.

To facilitate the isolation of subunits and proteolytic fragments, the [¹²⁵I]TID-labeled membranes were further photolabeled with the fluorescent, hydrophobic probe 1-azidopyrene (1-AP) as described previously (Blanton and Cohen, 1994). The membrane suspensions were then pelleted, resuspended in sample buffer, and electrophoresed on 1.5-mm thick slab gels.

After electrophoresis, the unstained gels were illuminated on a 365-nm UV light box and the nAChR subunits were visualized by 1-AP fluorescence. The bands containing the nAChR β-, γ-, and δ-subunits were excised from the gels, diced, and eluted in 10 ml of elution buffer (0.1 M NH₄HCO₃/0.1% SDS, pH 7.8), whereas the α-subunit was subjected to an “in-gel” S. aureus glutamyl endopeptidase (V8 protease) digestion (Cleveland et al., 1977; Pedersen et al., 1986). The α-subunit fragment of ∼20 kDa (αV8-20) was visualized by 1-AP fluorescence (Blanton and Cohen, 1992), excised from the gel, and eluted as described above.

After 4 days of elution, the samples were concentrated in Centriprep-10 (αV8-20) or Centriprep-30 (β-,γ-, and δ-subunits) Microconcentrators (Amicon, Beverly, MA) to approximately 400 μl and precipitated in 1.6 ml of acetone at −20°C to remove excess SDS. The αV8-20 fragments and δ-subunits were resuspended in 125 μl of 15 mM Tris, pH 8.1/0.1% SDS, and the β-subunits were resuspended in 150 μl of 0.1 M NH₄HCO₃, pH 7.8/0.5% Genapol/0.02% SDS. Protein concentrations of the samples were measured using a bicinchoninic acid-based protein assay (Micro BCA Protein Assay; Pierce), and the amount of [¹²⁵I]TID incorporation into each subunit was determined by gamma counting.

Proteolytic Digestion and Fragment Purification.

The goal in these experiments was to characterize the effect of noncompetitive antagonists on [¹²⁵I]TID incorporation into the M2 hydrophobic segments. Therefore, the techniques used to generate the [¹²⁵I]TID-labeled fragments used the digestion conditions previously optimized (White and Cohen, 1992; Gallagher and Cohen, 1999) to produce fragments beginning at the N termini of the M2 segments. Briefly, solutions (1 mg of protein/ml) of αV8-20 fragments and intact δ-subunit in 15 mM Tris, pH 8.1, 0.1% SDS were digested for 3 to 5 days with 4 U/ml of EndoLys-C, whereas the β-subunit (approximately 1 mg/ml) in 100 mM NH₄HCO₃, pH 7.8/0.02% SDS/0.5% Genapol C-100 was digested with trypsin [3:2 (w/w), protein/trypsin] for 3 to 7 days. The αV8-20 and δ-subunit digests were fractionated on 16.5% T/3% C Tricine gels (Schagger and von Jagow, 1987), and the β-subunit digests were resolved on a 16.5%T/6%C Tricine gel to obtain higher resolution. Regions of the gels that were expected to contain peptides beginning at the N termini of the M2 segments were identified based upon the migration of prestained molecular mass markers and the pattern of 1-AP fluorescence (Gallagher and Cohen, 1999). Nonfluorescent regions at 8 and 7 kDa were excised from the αV8-20 and β-subunit digests, respectively, and a fluorescent band at 10 kDa was excised from the δ-subunit digest. Material was eluted from the bands, and the eluates were concentrated to 0.3 ml in Centricon-3 microconcentrators (Amicon). The concentrates were then fractionated by reversed phase HPLC as a second purification step that also removed the majority of the SDS. Based upon the distribution of ¹²⁵I, fractions were pooled, dried by vacuum centrifugation, and then resuspended in 25 μl of 100 mM NH₄HCO₃/0.05% SDS.

Sequence Analysis.

The purified samples were applied to chemically modified glass fiber filters (Beckman, Palo Alto, CA) that were first washed with distilled water and methanol to remove impurities that inhibit binding of hydrophobic peptides. The purified peptides were fixed to the filters by delivery of gas trifluoroacetic acid for 4 min and then the SDS was removed from the filters by washing them for 5 min with ethyl acetate. The peptides were sequenced on an ABI 477 protein sequencer (Applied Biosystems, Foster City, CA) with gas phase sequencing cycles. During Edman degradation, two thirds of each sample was sent to a fraction collector for¹²⁵I gamma counting and one third was injected on an amino acid analyzer (ABI 120A) to identify the PTH-amino acid. Picomoles of each PTH amino acid were calculated from the peak height on the analyzer chromatogram (ABI 610A Data Analysis Program version 1.2.1). The background-subtracted yields (excluding serines and cysteines) of each sequencing cycle were fit to the equationI(n) = I₀ ×Rⁿ, where I(n) is picomoles of PTH amino acid at cycle n,I₀ is the initial peptide quantity in picomoles, and R is the sequencer repetitive yield. The efficiency of [¹²⁵I]TID incorporation into residue n, expressed as cpm_n per picomole (I_n), was calculated by [cpm_n − cpm_{n − 1}]/2 × I_n with the factor of 2 in the denominator because at each cycle the volume counted was twice the volume injected into the amino acid analyzer. We employed a single factor ANOVA test (Microsoft Excel) to determine whether tetracaine, PCP, or HTX induced significant differences in the percentage of change of [¹²⁵I]TID labeling efficiencies of αM2-9, αM2-13, δM2-9, and δM2-13. We also used a two-tailed ttest (Microsoft Excel; Microsoft, Redmond, WA) to determine whether there were significant differences in the ratios of¹²⁵I incorporation in M2-9 and M2-13 within a single subunit.

Results

In initial experiments, we examined the effects of tetracaine on the photoincorporation of [¹²⁵I]TID into nAChR-rich membranes as well as the effects of TID on the reversible binding of [³H]tetracaine (Fig.2). nAChR-rich membranes (1.5 μM ACh sites) were equilibrated with 6.5 μM [¹²⁵I]TID and various concentrations of tetracaine, and after photolysis the polypeptides were separated by SDS-PAGE. The gel was stained with Coomassie blue to allow identification of nAChR subunits and to verify that the lanes contained similar amounts of protein, and the distribution of¹²⁵I was determined by PhosphorImaging (Fig. 2A). Consistent with previous results (White and Cohen, 1988), [¹²⁵I]TID was photoincorporated into each nAChR subunit, with the γ-subunit labeled most efficiently. (A prominently labeled γ-subunit fragment migrating near 45 kDa was also noted.) Tetracaine reduced ¹²⁵I incorporation in each nAChR subunit in a dose-dependent manner, whereas it had no effect on the ¹²⁵I incorporation in the α-subunit of the Na⁺/K⁺-ATPase, which was labeled at lower efficiency than any of the nAChR subunits in the absence of tetracaine. When the concentration dependence of tetracaine inhibition of ¹²⁵I incorporation in nAChR γ-subunit was quantified (Fig. 2B), the data were fit by a single site model with an IC₅₀ value of 3 μM, and tetracaine reduced subunit labeling by ∼90%. The IC₅₀ value was the same as for its inhibition of [³H]HTX binding and, as expected, for tetracaine binding to its high-affinity site in the nAChR M2 ion channel domain (Gallagher and Cohen, 1999; Middleton et al., 1999). Nonradioactive TID also inhibited the reversible binding of [³H]tetracaine (Fig. 2B) with a concentration dependence characterized by an IC₅₀ value of 3 μM. At high concentrations, TID inhibited at least 95% of specifically bound [³H]tetracaine. These results contrasted with the interactions between TID and PCP or HTX (White and Cohen, 1988; White et al., 1991). TID did not inhibit [³H]PCP binding in the absence of agonist, and PCP reduced [¹²⁵I]TID photolabeling of nAChR γ-subunit by less than 30%.

Mapping [¹²⁵I]TID Photoincorporation in nAChR M2 Segments.

Suspensions of nAChR-rich membranes (5 mg, 2 mg/ml) were incubated with [¹²⁵I]TID in the absence of additional ligand and in the presence of 30 μM tetracaine, 50 μM PCP, or 30 μM HTX. The suspensions were photolyzed, and the nAChR β-, γ-, and δ-subunits as well as the αV8-20 fragment were isolated as described under Experimental Procedures. Typically, 100 to 200 μg of each subunit or of αV8-20 were recovered. Based upon previous studies (Gallagher and Cohen, 1999), digestion of αV8-20 or δ-subunit with EndoLys-C would produce fragments of ∼10 kDa beginning at the N termini of the M2 segments that can be purified by Tricine SDS-PAGE and HPLC. For the β-subunit, digestion with trypsin produces a similar fragment of ∼7 kDa (White and Cohen, 1992). For nAChRs labeled with [¹²⁵I]TID in the absence or presence of amine NCAs, we also found that when aliquots of the subunit digests were fractionated by Tricine SDS-PAGE, for both αV8-20 and the δ-subunit, the predominant [¹²⁵I]TID-labeled fragments had mobilities of ∼10 kDa (Fig.3). The presence of tetracaine or HTX, but not PCP, reduced ¹²⁵I incorporation in the α- and δ-subunit fragments without the appearance of any novel labeled bands. Trypsin digestion of labeled β-subunit produced a specifically labeled band of ∼7 kDa (data not shown). The bulk of these digests was fractionated by Tricine SDS-PAGE on a preparative scale, and the material eluted from the gel bands was further purified by reversed phase HPLC (Fig. 4), from which the fractions containing the peak of ¹²⁵I were pooled for N-terminal sequence analysis.

[¹²⁵I]TID Photoincorporation in the Presence of Tetracaine.

Subunit fragments were isolated from nAChRs labeled with [¹²⁵I]TID in the absence and presence of 30 μM tetracaine. Sequence analysis of the purified α-, β-, and δ-subunit fragments each revealed a single sequence beginning at the N terminus of the M2 segments (Fig. 5). For each of the subunit fragments isolated from nAChRs that were labeled in the absence of tetracaine, there was¹²⁵I release at cycles 9 and 13 of Edman degradation, corresponding to incorporation at αLeu-251 (αM2-9) and αVal-255 (αM2-13), βLeu-257 (βM2-9), and βLeu-261 (βM2-13), or δLeu-265 (δM2-9) and δLeu-269 (δM2-13). These results agree with the previous characterizations of amino acids in the nAChR M2 domain that are specifically labeled by [¹²⁵I]TID (Blanton and Cohen, 1992; White and Cohen, 1992). As was seen previously, within β-subunit [¹²⁵I]TID was incorporated with greater efficiency at βM2-9 than at βM2-13, whereas for α- and δ-subunits, the labeling at M2-13 was more prominent. Sequence analysis of the fragments isolated from nAChRs labeled in the presence of tetracaine similarly were characterized by¹²⁵I release in the M2 segments only at positions 9 and 13 (Fig. 5), but for each subunit the efficiency of incorporation at M2-9 and M2-13 was reduced by 90% (Table1A). When the data were combined from multiple labeling experiments, tetracaine at 30 μM reduced the efficiency of [¹²⁵I]TID labeling of αM2-9, αM2-13, βM2–9, βM2-13, δM2-9, and δM2-13 by 88 ± 4% (Table 2). There was no significant difference in the relative reduction of labeling in any of the residues (ANOVA, p = 0.76). Furthermore, there were no significant differences in the ratios of ¹²⁵I incorporation in δM2-9 to δM2-13 (p = 0.77) or αM2-9 to αM2-13 (p = 0.89) in the presence and absence of tetracaine. Thus, in the presence of tetracaine there was no evidence that [¹²⁵I]TID was shifted within the M2 ion channel domain to be in contact with amino acids other than M2-9 and M2-13, and the proportionate reduction of¹²⁵I incorporation at all amino acids was consistent with mutually exclusive binding of either [¹²⁵I]TID or tetracaine.

[¹²⁵I]TID Photoincorporation in the Presence of PCP and HTX.

nAChR-rich membranes were incubated with [¹²⁵I]TID in the absence of additional ligand and in the presence of either 50 μM PCP or 30 μM HTX. These concentrations of NCAs were chosen to maximize binding to the high-affinity site while avoiding binding to the agonist site or to other low-affinity binding sites (Heidmann et al., 1983). At these concentrations, PCP and HTX each caused maximal change in of [¹²⁵I]TID incorporation into the nAChR as determined at the level of individual subunits (White and Cohen, 1988). The suspensions were photolyzed, and the nAChR β-, γ-, and δ-subunits as well as the αV8-20 fragment were isolated. HTX reduced ¹²⁵I incorporation in αV8-20 by 80%, β-subunit by 70%, γ-subunit by 80%, and δ-subunit by 50%. PCP reduced labeling in the β-subunit by 20% and in the αV8-20 fragment by 60%, but it increased labeling in the δ-subunit by 20% and in the γ-subunit by 10%. These results were generally consistent with previous studies (White and Cohen, 1988). Fragments of the α-, β-, and δ-subunits that begin at the N termini of the M2 segments were generated, purified, and sequenced (Figs. 6 and 7). For the fragments isolated from nAChRs labeled in the presence of PCP (Fig. 6) or HTX (Fig. 7) there were peaks of ¹²⁵I release in cycles 9 and 13, without the appearance of any novel labeled amino acids within the M2 segments. In contrast to tetracaine, PCP altered¹²⁵I incorporation into each residue differently (Fig. 6; Table 1B). The addition of 50 μM PCP enhanced the labeling of δM2-9 by 2-fold while reducing incorporation by 30 to 70% at the equivalent position in αM2 and βM2. In the presence of PCP there was no change in the labeling of δM2-13. When the data were combined from multiple experiments (Table 2) these changes in the [¹²⁵I]TID labeling pattern were statistically significant by ANOVA (p = 0.007). Furthermore, the ratios of ¹²⁵I labeling in M2-9 to M2-13 were significantly different in the presence and absence of PCP in both δM2 (p < 0.001) as well as αM2 (p= 0.015).

HTX (30 μM) reduced [¹²⁵I]TID incorporation into each M2 residue (Fig. 7), with the percentage reduction differing among the labeled residues: αM2-9 (93%), αM2-13 (94%), βM2-9 (54%), δM2-9 (67%), and δM2-13 (86%) (Table 1C). When the results from independent experiments were averaged (Table 2), the variances in percentage of reduction of ¹²⁵I incorporation (Table 1) were such that there was no statistically significant difference in the alterations in labeling efficiencies (ANOVA p = 0.14). Nonetheless, there were consistently larger reductions at δM2-13 than δM2-9. Consequently, HTX caused statistically significantly changes in the ratios of incorporation in M2-9 to M2-13 in the δ-subunit (0.51, −HTX; 1.1, +HTX,p ≤ 0.001). There was no significant change in the ratio of incorporation in αM2 (p = 0.17).

Discussion

In these studies we examined the effects of tetracaine, PCP, and HTX on the pattern of photolabeling of [¹²⁵I]TID in the nAChR M2 ion channel domain. Previously [¹²⁵I]TID and [³H]tetracaine had each been shown by direct photolabeling to bind within the ion channel at the level of M2-9 and M2-13 in the closed state. Our results now demonstrate that tetracaine, the dimethylaminoethyl ester of p-butylaminobenzoic acid, which is charged at physiological pH, binds in a mutually exclusive manner with [¹²⁵I]TID, an uncharged benzene derivative. In contrast, for PCP, a three-ring aromatic tertiary amine, and HTX, a bulky two-ring secondary amine (Fig. 1), we found no evidence for simple, competitive interactions with [¹²⁵I]TID within the ion channel in the closed state. In the presence of PCP or HTX, there was no evidence that [¹²⁵I]TID was shifted to another locus within the ion channel domain, because [¹²⁵I]TID photolabeling was still restricted to amino acids M2-9 and M2-13. In the presence of HTX or especially PCP there was an alteration of the pattern of [¹²⁵I]TID labeling at M2-9 and M2-13 inconsistent with mutually exclusive binding. The simplest interpretation of the data is that in the absence of agonist, PCP, or HTX binds at another site, presumably in the ion channel domain, and the binding results in a subtle shift of the orientation of [¹²⁵I]TID bound at M2-9/13.

Tetracaine.

In the analytical labeling experiments, high concentrations of tetracaine reduced [¹²⁵I]TID photoincorporation into the nAChR γ-subunit by 92 ± 1% (Fig.1). This amount of inhibition would be expected for a drug that prevented [¹²⁵I]TID photolabeling within the M2 ion channel domain either by preventing the binding of [¹²⁵I]TID or by stabilizing the desensitized state of the nAChR (White and Cohen, 1988; 1992). However, tetracaine at concentrations up to 100 μM does not desensitize the nAChR (Boyd and Cohen, 1984; Middleton et al., 1999), and the IC₅₀ value of 3 μM for tetracaine inhibition is consistent with its binding to the M2 ion channel domain in the absence of agonist. In our studies, nonradioactive TID reduced the equilibrium binding of [³H]tetracaine by 95 ± 3%, a result consistent with competitive inhibition, although in view of the uncertainties, noncompetitive inhibition cannot be excluded.

We characterized [¹²⁵I]TID photoincorporation in the M2 ion channel domain in the absence and presence of 30 μM tetracaine. Sequence analysis through the M2 segments from the α-, β-, and δ-subunits showed that tetracaine reduced the efficiency of labeling at the positions normally labeled by [¹²⁵I]TID in the absence of tetracaine (M2-9 and M2-13) and did not shift [¹²⁵I]TID labeling to any other amino acids in the M2 segments. Tetracaine at 30 μM was not expected to produce maximal inhibition of [¹²⁵I]TID photolabeling. Based upon the IC₅₀ value from the analytical photolabeling experiments (Fig. 2), if tetracaine and [¹²⁵I]TID bound in a formally competitive manner then 30 μM tetracaine would reduce [¹²⁵I]TID's occupancy of the ion channel by 90% and as a result, there would be a 90% reduction in the labeling at M2-9 and M2-13. Because that is the value seen experimentally (88%), we conclude that the remaining [¹²⁵I]TID photolabeling at M2-9 and M2-13 originates from the nAChRs that have not bound tetracaine.

PCP and HTX.

In the absence of agonist PCP binding to its high-affinity site in the nAChR causes an allosteric inhibition of [¹²⁵I]TID photolabeling of the α-, β-, and γ-subunits and an increased photolabeling of the δ-subunit (White and Cohen, 1988; Ryan et al., 2001). HTX binding to its high-affinity site causes an allosteric inhibition of [¹²⁵I]TID photoincorporation into each nAChR subunit (White and Cohen, 1988). In our study, concentrations of PCP (50 μM) and HTX (30 μM) were used that in the previous studies were shown to cause maximal inhibition (or potentiation) of [¹²⁵I]TID photoincorporation into the nAChR as determined at the level of individual subunits. When we characterized by sequence analysis PCP's or HTX's effects on [¹²⁵I]TID incorporation in the M2 segments from nAChR α-, β-, and δ-subunits, we found that neither PCP nor HTX resulted in the labeling by [¹²⁵I]TID of any amino acids within M2 not labeled in the absence of the drugs. PCP (50 μM) enhanced the labeling of δM2-9 by 2- to 3-fold and altered the efficiency of [¹²⁵I]TID incorporation into the other residues by significantly different percentages (Fig. 6; Table 1). Furthermore, the ratios of¹²⁵I incorporation in αM2-9 to αM2-13 and δM2-9 to δM2-13 were significantly different from those observed in the absence of PCP (Table 2). This alteration in the pattern of labeling at M2-9/M2-13 suggests that PCP changes the orientation of [¹²⁵I]TID's diazirine with respect to the labeled residues, a result consistent with an allosteric interaction between PCP and TID. Similarly, the change in the ratio of [¹²⁵I]TID incorporation in δM2-9 to δM2-13 seen in the presence of HTX indicates that HTX also alters allosterically the structure of the TID binding domain.

Is it possible that PCP or HTX might bind simultaneously with [¹²⁵I]TID at the level of M2-9/M2-13? We have shown that tetracaine, which by direct photolabeling studies is known to bind in the ion channel at the level of M2-9/M2-13 (Gallagher and Cohen, 1999), inhibits competitively [¹²⁵I]TID photolabeling in the ion channel. Because tetracaine is a small aromatic amine containing only a single benzene ring, it is highly unlikely that a bulky, three-ringed compound such as PCP could bind within the pore at the level of the TID site and have the effects that it does on [¹²⁵I]TID labeling of M2-9 and M2-13. The enhancement of [¹²⁵I]TID photoincorporation at δM2-13 might be rationalized if the binding of PCP shifted [¹²⁵I]TID toward that position, but there was no evidence that PCP completely shielded [¹²⁵I]TID from contact with the other residues at M2-9/13. Similarly, it is improbable that HTX, a bulky two-ringed compound, could bind simultaneously with [¹²⁵I]TID at M2-9/13 and cause only a ∼70% allosteric reduction in [¹²⁵I]TID labeling, whereas the binding of tetracaine and [¹²⁵I]TID is mutually exclusive. Therefore, we believe that PCP and HTX do not bind at the level of M2-9/M2-13 in the closed channel. In the desensitized nAChR aromatic amines NCAs, including [³H]chlorpromazine and [³H]triphenylmethylphoshponium bind at a site centered at M2-6 (Giraudat et al., 1986, 1987, 1989; Hucho et al., 1986; Revah et al., 1990), whereas [³H]meproadifen mustard seems to bind near M2-20 at the extracellular end of the channel (Pedersen et al., 1992). Although we would be surprised if PCP or HTX binds near M2-6 in the resting state, perhaps either (or both) bind toward the extracellular end of the channel domain. Alternatively, it is possible that they may bind to a site not involving contributions from the M2 segments, as has been suggested for quinacrine in the open channel state (DiPaola et al., 1990).

Because TID and tetracaine label many of the same residues within the M2 segments and interact in a formally competitive manner, it is surprising that PCP and HTX interact allosterically with TID but competitively with tetracaine. In its extended conformation tetracaine is a longer molecule than TID. Perhaps PCP and HTX bind within the ion channel at a site that can overlap with tetracaine, but not TID. We consider it more likely that the ion channel cannot accommodate two compounds that are positively charged (such as PCP and tetracaine or HTX and tetracaine) even though they bind at different levels within the channel. However, the ion channel seems to be able to accommodate two compounds if one is positively charged and one is neutral (such as TID), as long as they bind at different loci within the ion channel.

Because PCP binds with higher affinity to the nAChR in the desensitized state than in the resting state (Heidmann et al., 1983), it is possible that the PCP-induced conformational changes in the [¹²⁵I]TID binding domain may reflect stabilization by PCP of the nAChR-desensitized state. For nAChRs desensitized by the agonist carbamylcholine, the efficiency of [¹²⁵I]TID incorporation in M2-9/13 is reduced by 90%, and there is also labeling at M2-2 and M2-6 that is not seen in the absence of agonist (White and Cohen, 1992). The effects of PCP on [¹²⁵I]TID photolabeling were clearly different than those of carbamylcholine, but it will be important to also examine the effects of PCP on [¹²⁵I]TID photoincorporation in the presence of agonist. If the alteration of the pattern of [¹²⁵I]TID photolabeling we see at M2-9/13 occurs because PCP is shifting the nAChR to the desensitized state, we would expect to see a similar labeling pattern in the presence of agonist. It would also be of interest to determine in the presence of agonist the effects on [¹²⁵I]TID photolabeling of chlorpromazine or meproadifen, strongly desensitizing NCAs whose binding sites are known.

Given the effects of PCP and HTX on [¹²⁵I]TID incorporation observed here, it will be of interest to directly identify their binding sites in the closed channel state. Preliminary studies have examined the ability of [³H]HTX and [³H]azido PCP to photoincorporate into theTorpedo species nAChR (Oswald and Changeux, 1981; Mosckovitz et al., 1987). Our data indicate that in the closed channel state [³H]HTX and [³H]azido PCP do not bind at the level of M2-9/13, but direct photoaffinity labeling studies will be required to determine where they do bind.