Background

Allosteric modulation of membrane receptors is increasingly recognized as a common mechanism used to control cellular signal transduction [1, 2]. In general, allosteric modulator binding causes changes in the response of the receptor to the "native ligand", presumably by altering the energetic barrier between resting and activated conformations. In most cases the modulator does not activate the target receptor in the absence of agonist. While there has been substantial progress in identifying the binding sites for many allosteric modulators (for example, [3]), the mechanisms by which modulators induce their effects remain poorly defined.

Some of the best examples of allosteric modulation involve members of the Cys-loop family of ligand-gated ion channels that includes nicotinic AChRs as well as the GABAA, glycine, and 5-hydroxytryptamine-3 receptors [4]. Cys-loop receptors transduce the energy of agonist binding into conformational changes that lead to channel opening [5]. All family members share a similar structure: they are transmembrane proteins assembled from five homologous or identical subunits. Each of these subunits is comprised of a large amino terminal extracellular domain (ECD), a large intracellular loop, and a four α-helix bundle forming a transmembrane domain (TMD). Recent studies aimed at identifying the structural basis for ligand gating have focused on the "transition zone" [6] a region of the receptor at the boundary between the ECD the TMD. The transition zone includes structural elements thought to link the TMD and the ligand binding site [7–9]. While the evidence for this linkage is preliminary, a number of experimental approaches have unequivocally mapped the site for ligand binding to the interface of adjacent subunit ECDs [5]. More recently, the crystal structures of ACh-binding proteins (AChBPs) from Lymnea, Aplysia, and Bulinus [10–12] have provided a structural context for these biochemical and functional studies. The AChBPs are soluble proteins that act as ACh buffers in invertebrates [13]; they share both sequence and functional homology to the ECD of Cys-loop receptors. Earlier this year another structure, that of a homologous bacterial ligand-gated ion channel, was added to the structural database of Cys-loop receptors [14]. These crystal structures have been used to develop and refine homology models of Cys-loop receptors [15, 16]. Our goal is to use these refined models to test specific mechanistic hypotheses that attempt to explain the dynamics of both ligand-induced receptor activation and allosteric modulation [7–9, 17].

Many neuronal nicotinic AChRs exhibit positive allosteric modulation by physiological concentrations of Ca2+ [18, 19]. In α7 nAChRs (but not other neuronal AChRs) Ba2+ or Sr2+ can elicit effects similar to Ca2+ [18, 20, 21]. This modulation consists of an increase in both the efficacy and the potency of ACh. The functional effects of divalents are similar to those caused by an emerging class of nicotinic modulating drugs collectively referred to as PAMs (positive allosteric modulators; [22]). Thus one rationale for a mechanistic characterization of divalent modulation of α7 AChRs is to serve as a model for studies of drugs such as PAMs developed to elicit a similar effect.

Previous studies demonstrated that the modulation of α7 AChRs by divalent cations is independent of divalent cation permeation, suggesting that the binding site for modulation is extracellular [21]. In addition several studies have demonstrated the importance of conserved ECD glutamate residues (E44 and E172 in chick α7) in divalent cation modulation, and it has been suggested that these may form the allosteric modulation binding site [15, 20, 23]. In this paper, we tested the hypothesis that conformational changes evoked by divalent cation modulators of the α7 AChR are similar to those evoked by ACh. In addition, we examined whether E44 and E172 are required for divalent cation-evoked conformational changes. We found some similarities between Ba2+ evoked conformational changes and those caused by ACh. Surprisingly, we also found that the effects Ba2+ on modification rates did not require E44 or E172, suggesting that these residues do not form the divalent cation binding site.

Results

The substituted cysteine accessibility method (SCAM) is an established experimental approach to examine protein conformational dynamics [24]. We previously used this approach to scan regions of the chick α7 AChR and identified residues where the rates of thiol-specific modification by MTSEA were altered by ACh [23, 25]. We consider two alternative mechanisms for ACh-dependent effects on modification rates. If the substituted cysteine is at a position that is part of the agonist binding site [5], then the effect of ACh could be due to steric occlusion. Alternatively, if the substituted cysteine is not near the binding site, then we infer that a change in modification rates is a result of conformational or electrostatic change induced by agonist-dependent activation. In this way, these residues serve as reporters of intramolecular changes during receptor activation.

SCAM can be used in the same way to identify conformational changes caused by allosteric modulators. Figure 1 shows a representation of the region of the α7 ECD targeted in this study. A discrete region of the inner β sheet, including M37, M40, and N52 was initially chosen to examine the effects of the divalent cation Ba2+ on MTSEA modification rates. We also examined the effects of Ba2+ at transition zone residues previously implicated in modulation by divalent cations, including E44, E172, as well as an adjacent position N170. All of the cysteine replacements at these residues have previously been shown to exhibit agonist-sensitive MTSEA modification rates, allowing us a basis for comparison for the effects of Ba2+.

Figure 1
figure 1

A model of the α7 AChR extracellular domain. Ribbon cartoon showing two of the five subunits viewed from the outside. In the subunit to the left of the central interface (yellow), the outer β sheet in is highlighted in orange, the transition zone E44 residue is orange, and the W148 residue is shown in gray to identify the ACh binding pocket (Zhong et. al., 1998). The subunit to the right shows a view of the inner sheet (teal), and other residues targeted in this study. The sequence surround mutants characterized in this study is shown beneath the cartoon: M37, M40 cyan; N52 green; N170 blue; E172 purple.

Full size image

For these studies we began by confirming the modulatory effects of Ba2+ on ACh-dependent activation of our parental phenotype, α7 C115A/L247T. Wild-type α7 AChRs exhibit a complex positive modulation by divalent cations such as Ca2+ or Ba2+ that includes increases in both efficacy and potency [20]. In contrast, receptors with the L247T phenotype typically exhibit a simplified modulatory response consisting only of a 5- to 10-fold left shift in the ACh dose-response.

Figure 2A shows the effect of 10 mM Ba2+ on the α7 C115A/L247T receptor. There was a leftward shift in the dose-response curve corresponding to a ~10 fold decrease in EC50 (increase in potency). Of note, we do not see an effect of Ba2+ on efficacy in the parental background. We suspect this is due to the higher gating constant of receptors with the L247T mutation. Figure 2B shows that modulation by Ba2+ was eliminated in the E44C mutant, confirming that this conserved glutamate is required for Ba2+ binding or allosteric coupling of Ba2+ binding to ACh-dependent activation. This result is similar to the effect of an E44Q mutant described in wild-type and L247T α7 AChRs [20, 21]. Table 1 provides a compilation of EC50's, modulatory effects of 10 mM Ba2+, and the maximal responses of the mutants described in this report. Neither of the transition zone glutamate mutants (E44C and E172C) exhibited a positive modulation, while the N52C mutant displayed a high partial agonism by Ba2+ in the absence of ACh.

Figure 2
figure 2

Positive allosteric modulation by divalent cations requires E44. ACh dose-response curves for the parental C115A/L247T (A) and the E44C mutant (B) in the absence (open squares) and presence (filled squares) of 10 mM BaCl2. Data are fitted to the Hill equation (solid lines). The positive allosteric modulation (leftward shift in the dose response curve) typically exhibited by α7 AChRs (A) is lost in the E44C mutant (B). Data are mean values (± SEM) from three determinations, normalized to the maximal value of the Hill equation fit of each data set. Hill coefficients for C115A/L247T (A): 2.5 ± 0.2 (open squares, -Ba2+), 1.9 ± 0.4 (filled squares, +Ba2+); and for the E44C mutant (B) 1.7 ± 0.2 (open squares, -Ba2+), 2.9 ± 0.7 (filled squares, +Ba2+).

Full size image
Table 1 Effects of Ba2+ on ACh evoked currents.
Full size table

Previously, we measured the effects of ACh on reactivity of cysteine mutants in the inner β sheet of the chick α7 AChR [26]. Several residues (M37C, M40C, and N52C) exhibited a change MTSEA reaction rates in the presence of ACh. We interpret differences in modification rates in the absence or presence of ACh to reflect differences in the apparent accessibility of the introduced cysteine between the unliganded and liganded states. To test if these sites could also be used as reporters of allosteric modulator-induced conformational change we examined the effects of Ba2+ on rates of MTSEA modification.

Figure 3 shows an example of the protocol used to measure the thiol modification rate of the receptors with the M37C mutation. ACh-evoked current amplitudes decreased following brief, repeated exposure to a limiting concentration of MTSEA (5 μM, 15 seconds). To ensure that the modification reactions were complete, all rate measurements included a final prolonged application of ~100-fold higher concentrations of MTSEA (Fig. 3A and 3B, right). Currents measured following this application represent the endpoint of the reaction between MTSEA and receptors. When the same protocol included Ba2+ pretreatment and co-application with MTSEA (see Methods), the decreases in current amplitudes were slowed but the same endpoint was obtained (Fig. 3B). Normalized current amplitudes were plotted as a function of the cumulative time of exposure to MTSEA, and pseudo first-order rates were extracted from the single-exponential fits (Fig. 3C). We observe a significant decrease in the MTSEA modification rate for M37C (Fig. 4) in the presence of 10 mM Ba2+, demonstrating that this modulator caused changes in the conformation or electrostatic environment around the M37C side chain.

Figure 3
figure 3

Barium slows the rate of MTSEA modification at M37C. Example of experimental paradigm used to assess Ba2+ effects on modification rates. (A) Successive ACh-evoked current traces recorded before and after repeated exposures to MTSEA (5 μM, 15 seconds), showing a decrement in responses to 30 μM ACh. Endpoints of MTSEA modification are determined by prolonged application of 500 μM MTSEA (right). (B) The same protocol, including Ba2+ pretreatment and co-application with MTSEA. Current traces are truncated in both (A) and (B) between consecutive MTSEA applications; in all cases the currents were allowed to return to baseline prior to the next application of MTSEA ± ACh. (C) Peak current amplitudes from (A) and (B) are normalized and plotted versus total MTSEA exposure time. Data from this single experiment (no error bars) are fitted to a single-exponential decay (solid line) to extract an apparent pseudo first-order rate constant. The pseudo first-order rate constants calculated in this experiment were 0.011 s-1 and 0.0019 s-1 for control (A) and +Ba2+ (B) measurements, respectively. Second-order rate constants are calculated from these values (Figures 4–6; Table 2).

Full size image
Figure 4
figure 4

Barium alters the rate of MTSEA modification at inner β sheet residues. (A) Using the protocol described in Figure 3, we determined second-order rate constants for three reporter residues in the α7 AChR inner β sheet (M37C, M40C, and N52C). Mean values for second-order rate constants for modification by MTSEA alone (control), MTSEA + ACh, and MTSEA + Ba2+ are shown. * Rate was significantly different from control (P < 0.05). (B) A plot of the ratios of second-order rate constants. Ba2+ and ACh both slowed the rates of modification of M37C and M40C. At N52C, however, the rate of modification in the presence of Ba2+ was not significantly different from control, while ACh accelerated the modification rate. See Table 2 for summary including (n) for each condition.

Full size image

Using the protocol described in Figure 3, we determined second-order rate constants for modification of the three reporter residues in the inner β sheet (M37C, M40C, and N52C). Figure 4A shows mean values of the rate constants measured in the presence of MTSEA alone, MTSEA plus ACh, and MTSEA plus Ba2+. We observed significant decreases in reaction rates of both M37C and M40C in the presence of 10 mM Ba2+. The effects are quantitatively similar to those measured in the presence of ACh [26], consistent with the idea that Ba2+ causes conformational changes similar to those induced by agonist in this region of the α7 AChR. To more directly compare the effects of ACh and Ba2+ on reaction rates, we plot rate constants as ratios in Figure 4B. This figure highlights the differences in MTSEA rates under different conditions at these three positions. In contrast to M37C and M40C, we observed no significant effect of Ba2+ on MTSEA modification rate of N52C. This result parallels our previous study in which we found the effect of ACh on MTSEA modification rate of N52C was also different from that of neighboring residues M37C and M40C. Collectively, these results suggest that divalent cations such as Ba2+ act to promote some, but not all, of the conformational or electrostatic changes elicited by ACh. While ACh acts to stabilize the open state, Ba2+ acts to stabilize a state (or states) that are energetic intermediates between closed and open channels.

Modulation by divalent cations is known to require the conserved acidic residues at E44 and E172 (23, 24). We next tested whether Ba2+ could cause changes in the rates of MTSEA modification at E44C, N170C, and E172C. Similar to residues in the inner β sheet, each of these mutants has been shown to be a reporter of conformational or electrostatic changes induced by ACh [23, 25]. Figure 5A shows mean values of second order rate constants measured in the presence of MTSEA alone, MTSEA plus ACh, and MTSEA plus Ba2+. At N170C, a mutant that showed allosteric modulation (Table 1), the rate of MTSEA modification in the presence of Ba2+ was the same as that measured in MTSEA alone, but was different from that measured in the presence of ACh. This observation suggests that conformational or electrostatic changes induced by modulators at this residue are distinct from those induced by ACh. Differences between the effects of ACh and Ba2+ were most pronounced at E44C; at this residue, the modification rate was ~10 fold higher in the presence of Ba2+ compared to that measured in the presence of ACh (Figure 5B). Surprisingly, despite the fact that both E44C and E172C show no positive allosteric modulation of ACh currents by Ba2+ (Fig. 2, Table 2), both exhibited significant Ba2+-dependent decreases in MTSEA modification rate.

Figure 5
figure 5

Barium alters the rate of MTSEA modification at residues required for modulation by divalent cations. Second-order rate constants were measured for three residues in the "transition zone" of the α7 AChR (E44C, N170C, and E172C). (A) Mean values for second-order rate constants for modification by MTSEA alone, MTSEA + ACh, and MTSEA + Ba2+. Ba2+ caused a significant decrease in MTSEA modification rates of both E44C and E172C, despite the loss of divalent cation-dependent modulation. Ba2+ did not have a significant effect on the modification rate of N170C, although ACh significantly increased the rate of modification of this residue. *Rate was significantly different from control (P < 0.05). ‡Rate was significantly different from that obtained in presence of ACh (P < 0.05). The plot of rate constant ratios (B) shows that the effect of Ba2+ on the rate of modification of E44C was significantly less than the effect of ACh. See Table 2 for summary including (n) for each condition.

Full size image
Table 2 Summary of MTSEA modification data.
Full size table

The hypothesized requirement for E44 and E172 in divalent cation modulation was based upon studies of charge neutralization mutants (E44Q, E172Q) in which modulation is lost. From these and other studies, both residues were proposed to be participants in a binding site which mediates the divalent cation allosterism [20, 21]. The loss of Ba2+ dependent modulation in E44C or E172C mutations (also charge neutralization mutations) is consistent with this proposal, but the effects of Ba2+ on MTSEA modification rates are not. One possible explanation for these observations is that Cys replacements at E44 and E172 do not prevent Ba2+ binding, but cause an uncoupling of binding and allosteric modulation. If Ba2+ binds near E44C and E172C, the slowed modification rate at E44C or E172C would be explained by physical occlusion of the thiol side-chain by bound Ba2+. Alternatively, the glutamates could be a required component in the transduction pathway between Ba2+ binding and receptor modulation, but are not direct participants in the binding site. In this case divalent cations bind at a different site and elicit conformational or electrostatic changes (detected as changes in E44C and E172C modification rates), but binding does not lead to modulation. To test this possibility we examined MTSEA modification rates at M40C α7 AChRs in which a second, charge-neutralizing mutation (E172Q) was introduced. We reasoned that if E172 is required for binding of divalent cations, then the modification rate of M40C should be insensitive to Ba2+. If, however, Ba2+ binds to the receptor and causes conformational changes, despite the mutation at E172, this would be reflected by changes in the rate of modification of M40C.

We measured dose-response relationships for the M40C/E172Q double mutant in the absence and presence of 10 mM Ba2+ and confirmed that it was not positively modulated by Ba2+ (Table 1). When the rates of MTSEA modification of the M40C in this background were measured, we found that Ba2+ caused a significant slowing of MTSEA modification rate (Figure 6). The modification rates of M40C were independent of the E172 mutation. This result suggests that the binding site for Ba2+ modulation is somewhere other than a site which includes E44 and E172 in the α7 AChR transition zone [15].

Figure 6
figure 6

Charge neutralization at E172 does not alter the rate of modification M40C by MTSEA. Mean values for second-order modification rate constants for M40C (left, data from Fig. 3) compared to those obtained in receptors containing the E172Q mutation (M40C/E172Q). *Rates were significantly different from control (P < 0.05). See Table 2 for summary, including (n) for each condition.

Full size image

Discussion

The mechanisms of protein allosterism have been the subject of exhaustive modeling and model refinement since studies of Monod, Wyman and Changeux [27] and those of Koshland and colleagues [28]. In the nicotinic receptors, several different types of allosteric behavior have been described. The concerted, or MWC model, for example, refers to the allosteric effect of ligand binding on channel opening; this model suggested that binding of multiple agonists acted in concert to yield their "at a distance" effect. Experimental tests of the MWC model with combinations of agonists and antagonists suggest that a stepwise process more accurately describes the activation process [29, 30].

Another type of allosterism seen in some nicotinic receptors is the positive allosteric modulation by divalents such as Ca2+ or Ba2+ [18, 20]. For this type of allosterism, a fundamental question is whether the modulation alters the conformational "pathway" from closed to open states or simply modifies the kinetics of an agonist-dependent closed to open transition. Few studies have attempted to address this question, but a recent report does examine the conformational effects of positive allosteric modulators (benzodiazepines) in GABAA receptors [31]. This study used SCAM to show that a prominent effect of benzodiazepines is to increase the access of GABA to its binding site, reducing the energetic barrier to the initial step in receptor activation, GABA association. Divalent cation effects on α7 AChRs provide a similar paradigm in which to examine the conformational changes evoked by allosteric modulators.

In an earlier report we described an α7 AChR mutant with a pair of cysteine substitutions positioned to introduce a disulfide bond in the outer β sheet [26]. In our parental background this mutant was fully activated by divalent cations in the absence of ACh, but when expressed in an α7 AChR without the L247T mutation it required both ACh and divalent cations for channel activation. If we assume that activation of this mutant occurs as a result of Ba2+ interaction with the divalent cation allosteric site, then it suggests that Ba2+ and ACh promote two overlapping but distinct sets conformational changes. The experiments presented in this report provide further evidence that conformational effects of divalent cations are similar to those of elicited by ACh. Thus the simplest interpretation of our data is that divalent cations act by enhancing transitions in an ACh-dependent activation pathway without substantial effect on the final closed to open transition.

Le Novere and colleagues [15] proposed a model for a divalent cation binding site that was based on earlier experiments, homology between the α7 AChR ECD and the Lymnea AChBP, and the known database of divalent cation binding proteins [10, 20]. The focus of this model was a cluster of 4 negatively charged residues in the transition zone: D41, D43, E44, and E172. Mutational analysis suggested that the glutamate residues were critical, since charge neutralization at either of the aspartate residues had only modest effects on divalent modulation. While the geometry of these residues in models of the ECD is consistent with their proposed model, our results with Ba2+-induced conformational changes are not. A mutation of either E44 or E172 to cysteine eliminates the modulation, but not the conformational changes associated with Ba2+modulation. This strongly suggests that the allosteric effects of Ba2+ are "transmitted" in a conformational pathway that requires these glutamates for some role other than divalent cation binding. It is unlikely that the Cys substitution is able to act as a functional substitute for Glu in a divalent cation site: a survey of all known Ca2+ binding sites found that Cys was never a contributor to a Ca2+ co-ordination site, while it often plays this role in Zn2+ binding sites [32]. Other possible candidates for a divalent cation modulation site in the α7 ECD include acidic residues in β6 and β8, which may combine with neighboring carbonyl groups to form a site for divalent cation binding. Alternatively, the recent work of Horn and colleagues [33] has demonstrated that aromatic residues may provide the negative electrostatic environment required for formation of a physiologically relevant divalent cation binding site through the π-cation-type interactions. This is the same structural motif that has been shown to provide the negative electrostatic environment in the cholinergic agonist binding site [34].

Methods

Reagents

MTSEA (2-aminoethylmethane thiosulfonate) was obtained from Toronto Research Chemicals (Toronto, Canada). Gentamicin was from Invitrogen (Carlsbad, CA). All other reagents were obtained from Sigma-Aldrich (St. Louis, MO).

Site-directed mutagenesis

A cDNA clone of the chick α7 receptor containing two mutations (C115A, L247T) was used as the parental phenotype for mutations described in this study. We mutated the lone unpaired cysteine in the extracellular domain (C115) to alanine to allow for a more straightforward interpretation of thiol modification experiments. We observed no functional effect of this mutation on receptor expression or ACh response. We included the mutation of leucine 247 in the M2 transmembrane domain (L247T; L9'T) because of its large current amplitudes and non-desensitizing kinetics compared to wild-type α7 receptors [35]. These receptors exhibit a higher "gating constant" than wild-type α7 AChRs [36], suggesting that the closed-to-open equilibrium of liganded C115A/L247T receptors favors the open state. Mutation at the L247 position enhanced the ability to measure modification rates for cysteine replacements in which the ACh-evoked current amplitudes are attenuated [37]. In preliminary experiments, modification rates of cysteines introduced into wild type α7 AChRs were similar to those in L247T-containing receptors, suggesting that conformational changes in the ECD of L247T-containing receptors are similar to those in wild-type receptors (not shown). All mutations were introduced by site-directed mutagenesis using the QuikChange method (Agilent Technologies, La Jolla, CA) as described previously [21], and were confirmed by DNA sequencing.

Xenopus oocyte maintenance and expression

cRNA was prepared using the T7 RNA polymerase and mMessage mMachine kit as described by the manufacturer (Applied Biosystems, Austin, TX). Oocytes were surgically removed and prepared from female Xenopus laevis in accordance with UNC Institutional Animal Care and Use Committee guidelines. Oocytes were injected with 20 ng of cRNA and incubated at 18°C in ND96 (96 mM NaCl, 2 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 5 mM Na pyruvate, 50 μg/ml gentamicin, 5 mM HEPES, pH 7.5) for 2–5 days before use. For some mutants we co-injected cRNA encoding the human RIC-3 [38], a protein shown to enhance expression of α7 AChRs in both mammalian cells and oocytes [39]. This co-injection (at a 1:1 ratio, 20 ng per oocyte) enhanced maximal current responses without significant effect on ACh EC50 (not shown).

Data collection and analysis

Oocytes were superfused in normal extracellular solution containing a reduced Ca2+ concentration (ESLC; 96 mM NaCl, 2 mM KCl, 1 mM MgCl2, 0.1 mM CaCl2, and 10 mM HEPES, pH 7.5). This solution minimized Ca2+ influx and eliminated Ca2+-activated chloride currents. Two-electrode voltage clamp was performed with a GeneClamp 500B controlled by pCLAMP8 software (Molecular Devices, Sunnyvale, CA). Electrodes were filled with 3 M KCl contacting Ag-AgCl wires and had resistances of 0.5 to 2.0 MΩ. Currents were recorded at a constant holding potential of -60 mV. Currents were low pass filtered at 50 Hz and sampled at 100 Hz. Agonist dose-response curves were obtained as described previously [21], and data were fit to the Hill equation using Origin software (Microcal Software, Northampton, MA).

Expression and modification rates

Each mutant was initially screened for functional expression over a range of ACh concentrations to generate a dose-response relationship and determine its EC50. To test for reactivity of introduced free thiols, we compared responses of each mutant to an ~EC50 ACh dose before and after exposure to high concentrations of MTSEA (0.5 – 1.0 mM) applied by continuous flow for 30 to 60 seconds. MTSEA was prepared daily in distilled water and stored on ice. Stock solution was diluted to the appropriate working concentration in ESLC immediately before each application. Rates were measured by determining a limiting dose of MTSEA (0.1–100 μM), then exposing oocytes to these low concentrations of MTSEA repeatedly for 15–30 seconds followed by a challenge with an ~EC50 concentration of ACh. The limiting dose, yielding 20–40% of the maximal MTSEA effect, was identified for each mutant. To measure the effect of ACh on modification rates we used the same protocol but included an ~EC110 ACh dose with the applied MTSEA. To measure the effect of Ba2+ on modification rates we pre-applied 10 mM Ba2+ for 30 seconds prior to co-application of 10 mM Ba2+ plus MTSEA. This concentration of Ba2+ is equivalent to an approximate EC110 for the modulatory effects in both parental and mutant AChRs. Kinetic data were analyzed as described previously [40]; rate data were fit to a single exponential to extract a pseudo-first order rate constant; this was divided by the MTSEA concentration used to determine the second order rate constant for thiol modification.

Statistical Analysis

Statistical analysis of EC50 values and second-order rate constants was conducted using a one-way analysis of variance, followed by a post hoc Tukey test. P values of < 0.05 were interpreted to indicate significant differences.

Structural models of α7

A model of the chick α7 nicotinic receptor extracellular domain, based on the coordinates of the Lymnea ACh Binding Protein [10] was constructed as described previously [23, 26]. Images of the model were generated with Pymol (DeLano Scientific, South San Francisco, CA).