Positional Scanning Mutagenesis of α-Conotoxin PeIA Identifies Critical Residues That Confer Potency and Selectivity for α6/α3β2β3 and α3β2 Nicotinic Acetylcholine Receptors*

Background: Ligands that selectively target α6β2* nAChRs are needed. Results: An analog of α-conotoxin PeIA was synthesized that was >15,000-fold more potent at inhibiting α6/α3β2β3 receptors than the closely related α3β2 subtype. Conclusion: PeIA analogs can be synthesized that distinguish between α6/α3β2β3 and α3β2 receptors. Significance: Selective ligands will facilitate the identification of α6β2* receptors in tissues that co-express α3β2* receptors. The nicotinic acetylcholine receptor (nAChR) subtype α6β2* (the asterisk denotes the possible presence of additional subunits) has been identified as an important molecular target for the pharmacotherapy of Parkinson disease and nicotine dependence. The α6 subunit is closely related to the α3 subunit, and this presents a problem in designing ligands that discriminate between α6β2* and α3β2* nAChRs. We used positional scanning mutagenesis of α-conotoxin PeIA, which targets both α6β2* and α3β2*, in combination with mutagenesis of the α6 and α3 subunits, to gain molecular insights into the interaction of PeIA with heterologously expressed α6/α3β2β3 and α3β2 receptors. Mutagenesis of PeIA revealed that Asn11 was located in an important position that interacts with the α6 and α3 subunits. Substitution of Asn11 with a positively charged amino acid essentially abolished the activity of PeIA for α3β2 but not for α6/α3β2β3 receptors. These results were used to synthesize a PeIA analog that was >15,000-fold more potent on α6/α3β2β3 than α3β2 receptors. Analogs with an N11R substitution were then used to show a critical interaction between the 11th position of PeIA and Glu152 of the α6 subunit and Lys152 of the α3 subunit. The results of these studies provide molecular insights into designing ligands that selectively target α6β2* nAChRs.

The nicotinic acetylcholine receptor (nAChR) subtype ␣6␤2* (the asterisk denotes the possible presence of additional subunits) has been identified as an important molecular target for the pharmacotherapy of Parkinson disease and nicotine dependence. The ␣6 subunit is closely related to the ␣3 subunit, and this presents a problem in designing ligands that discriminate between ␣6␤2* and ␣3␤2* nAChRs. We used positional scanning mutagenesis of ␣-conotoxin PeIA, which targets both ␣6␤2* and ␣3␤2*, in combination with mutagenesis of the ␣6 and ␣3 subunits, to gain molecular insights into the interaction of PeIA with heterologously expressed ␣6/␣3␤2␤3 and ␣3␤2 receptors. Mutagenesis of PeIA revealed that Asn 11 was located in an important position that interacts with the ␣6 and ␣3 subunits. Substitution of Asn 11 with a positively charged amino acid essentially abolished the activity of PeIA for ␣3␤2 but not for ␣6/␣3␤2␤3 receptors. These results were used to synthesize a PeIA analog that was >15,000-fold more potent on ␣6/␣3␤2␤3 than ␣3␤2 receptors. Analogs with an N11R substitution were then used to show a critical interaction between the 11th position of PeIA and Glu 152 of the ␣6 subunit and Lys 152 of the ␣3 subunit. The results of these studies provide molecular insights into designing ligands that selectively target ␣6␤2* nAChRs.
Nicotinic acetylcholine receptors are ligand-gated ion channels ubiquitously expressed throughout the nervous system. In vertebrates, there are nine ␣ (␣2-␣10) and three ␤ (␤2-␤4) subunits that assemble in various combinations to form pentameric channels (1). Some ␣ subunits, including ␣7, ␣8, and ␣9, can form homopentamers, but all other ␣ subunits, with the exception of ␣10 (which can form ␣9␣10 receptors), require a ␤ subunit for functional expression, such as those that contain ␣6 or ␣3 subunits. In the mammalian central nervous system, ␣6and ␣3-containing nAChRs 2 may be co-expressed in several regions, including the inferior and superior colliculi, optic nerve, ventrolateral geniculate, striatum, medial habenula, interpeduncular nucleus, and dorsal horn of the spinal cord (2)(3)(4)(5)(6). The ␣6␤2* subtype has been shown to play a critical role in the modulation of striatal dopamine release (7)(8)(9) and has been implicated in diseases of the basal ganglia, including Parkinson disease (10,11). More recently, ␣6␤2* receptors have been shown to modulate dopaminergic activity in reward centers of the brain and are thus likely to be key contributors to the reinforcing properties of nicotine (12)(13)(14)(15). The ability to pharmacologically distinguish between ␣6␤2* and ␣3␤2* in these areas has been hampered by the scarcity of ligands that can discriminate between these two subtypes. This lack of selective ligands is due in part to the high homology of the N-terminal, ligand-binding domains of ␣6 and ␣3 subunits; consequently, receptors that contain these subunits share similar pharmacological profiles.
Peptides identified in the venom of marine cone snails that target nAChRs are called ␣-conotoxins (␣-Ctxs) (16,17) and have proven to be highly valuable tools for distinguishing among the various nAChR subtypes. These peptides are usually 12-20 amino acids in length and are characterized by the presence of two disulfide bridges between cysteine residues that serve to keep the peptide in its biologically active configuration. Some of these ␣-Ctxs include the widely used MII as well as PnIA, OmIA, LtIA, and PeIA (18 -23). Unfortunately, these peptides all target both ␣6and ␣3-containing nAChRs with similar potencies and thus do not discriminate well between ␣6␤2* and ␣3␤2* nAChRs.
In this report, we used positional scanning mutagenesis of PeIA to identify residues that confer potency for ␣6/␣3␤2␤3 (the ␣6/␣3 subunit is a chimera where the extracellular ligand binding domain of ␣6 is spliced with the transmembrane domain of ␣3) and ␣3␤2 nAChRs. The results show that a single amino acid substitution in PeIA is sufficient to abolish activity for the ␣3␤2 subtype and confer selectivity for ␣6-containing nAChRs. Through successive substitutions of several amino acids, we generated a ligand that was Ͼ15,000-fold more potent at inhibiting the ␣6/␣3␤2␤3 over the ␣3␤2 subtype.

EXPERIMENTAL PROCEDURES
Materials and Methodologies-Rat ␣3, ␣4, ␣6, and ␣7 nAChR subunit clones were provided by S. Heinemann (Salk Institute, San Diego, CA), and C. Luetje (University of Miami, Miami, FL) provided the ␤2, ␤3, and ␤4 subunits in the high expressing pGEMHE vector. Construction of the ␣6/␣3 subunit chimera has been described previously and consists of an ␣3 subunit where the first 237 amino acids of the ligand binding domain were replaced with the corresponding ␣6 amino acids (18). This chimera was used because injection of non-chimeric ␣6 with ␤2 produces few functional receptors (25)(26)(27). However, injection of ␤2 and ␤3 cRNA in conjunction with the ␣6/␣3 chimera produces sufficient numbers of receptors for electrophysiological measurement. The ␤3 subunit is not believed to form part of the canonical agonist binding site and serves a structural role in the pentamer (28). However, most native ␣6␤2* receptors likely contain the ␤3 subunit because genetic deletion of ␤3 substantially reduces levels of ␣6␤2* nAChR expression (29,30). Point mutations in the ␣6/␣3 and ␣3 subunits were made by PCR as described previously (31). Acetylcholine chloride (ACh) and bovine serum albumin were obtained from Sigma-Aldrich. HEPES was purchased from Research Organics (Cleveland, OH).
Peptide Synthesis-Solid phase Fmoc peptide chemistry was used to generate the ␣-Ctx peptides either as described previously (21) or with an AAPPTec Apex 396 automated peptide synthesizer (Louisville, KY). The peptides were initially constructed on a preloaded Fmoc-Rink Amide MBHA resin (substitution: 0.4 mmol/g Ϫ1 ; Peptides International Inc.). All standard amino acids were purchased from AAPPTec except for N-␣-Fmoc-O-t-butyl-L-trans-4-hydroxyproline (O, Hyp), which was purchased from EMD Millipore (Billerica, MA). Side-chain protection for the following amino acids was as follows: Arg, 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl; Lys, tert-butyloxycarbonyl Hyp; His and Asn, trityl; Ser, tert-butyl. Cys residues were orthogonally protected by trityl for Cys 1 and Cys 3 and acetamidomethyl for Cys 2 and Cys 4 . The peptides were synthesized at 50 mol scale. Coupling activation was achieved with 1 eq of 0.4 M benzotriazol-1-yloxytripyrrolidinophosphonium hexafluorophosphate and 2 eq of 2 M N,N-diisopropylethyl amine in N-methyl-2-pyrrolidone as the solvent. For each coupling reaction, a 10-fold excess of amino acid was used, and the reaction was carried out for 60 min. Fmoc deprotection was performed for 20 min with 20% (v/v) piperidine in dimethylformamide. The peptides were cleaved from the resin using Reagent K, trifluoroacetic acid/ phenol/ethanedithiol/thioanisol/H 2 O (9:0.75:0.25:0.5:0.5 by volume), and a two-step oxidation protocol was used to selectively fold the peptides into the correct disulfide configuration. Briefly, the first disulfide bridge was closed using 20 mM potassium ferricyanide and 0.1 M Tris-HCl, pH 7.5. The solution was allowed to react for 45 min, and the monocyclic peptide was purified by reverse-phase HPLC. Simultaneous removal of the acetamidomethyl groups and closure of the second disulfide bridge was accomplished by iodine oxidation. The monocyclic peptide in HPLC eluent was dripped into an equal volume of 10 mM iodine in H 2 O/trifluoroacetic acid/acetonitrile (78:3:25 by volume) and allowed to react for 10 min. The reaction was terminated by the addition of ascorbic acid diluted 20-fold with 0.1% (v/v) trifluoroacetic acid, and the bicyclic product was purified by reverse-phase HPLC. The masses of the peptides were verified by matrix-assisted laser desorption ionization time-of-flight mass spectrometry at the Salk Institute for Biological Studies (San Diego, CA) under the direction of Dr. J. Rivier and are provided in supplemental Table 1.
Two-electrode Voltage Clamp Electrophysiology of Xenopus laevis Oocytes-Detailed methods for conducting electrophysiological experiments of nAChRs heterologously expressed in X. laevis oocytes have been described previously (32). Briefly, stage IV-V oocytes were injected at a 1:1 ratio with cRNA encoding cloned rat nAChR subunits ␣3, ␣4, ␣6/␣3, ␣7, ␤2, ␤3, and ␤4 and used 1-5 days after injection. The oocytes were clamped at a holding potential of Ϫ70 mV and continuously gravity-perfused with standard ND96 solution buffered with HEPES and stimulated with 1-s pulses of ACh once every minute. The concentrations of ACh used were 100 M for all receptor subtypes except ␣7, where 300 M ACh was used. These concentrations are submaximal and within ϳ3-fold of the EC 50 for all subtypes . The ACh concentrations and brief exposure time were chosen to avoid open channel block and long term desensitization of the receptors. The solution changes were controlled through a series of three-way solenoid valves interfaced with a personal computer via a CoolDrive valve driver (Neptune Research & Development, West Caldwell, NJ) and LabVIEW software (National Instruments, Austin, TX). The ACh-gated currents (I ACh ) were acquired using an Oocyte OC-725 series voltage clamp amplifier (Warner Instruments, Hamden, CT), filtered through a 5-Hz low pass Bessel filter (model F1B1; Frequency Devices, Ottawa, IL), and digitized at 50 Hz using a National Instruments USB-6009 digital to analog converter. The toxins were dissolved in ND96 and either perfusion-applied (for concentrations of Յ1 M) or applied in a static bath for 5 min (for concentrations of Ն10 M).
Data Analysis-Concentration-response curves for inhibition of I ACh were generated by fitting the data to the Hill equation, % response ϭ 100/(1 ϩ ([toxin]/IC 50 ) n H ), using GraphPad Prism software (La Jolla, CA).
Homology Binding Modeling-The 2.4 Å resolution crystal structure of the Aplysia californica acetylcholine-binding protein (AChBP) complexed with ␣-Ctx PnIA(A10L,D14K) (33) was used to model the binding of PeIA analogs to rat ␣6 and ␣3 subunits. The coordinates of the complex were rendered using PyMOL (version 1.2r3pre, Schrödinger, LLC), and changes in residues of the AChBP and PnIA(A10L,D14K) were accomplished using the mutagenesis function. The numbering system used to identify residues of the AChBP follow the alignment of rat ␣3 and ␣6 subunits (Fig. 7).

Alanine Scanning Mutagenesis Reveals Amino Acids Critical
for PeIA Activity-PeIA is a 16-amino acid peptide whose sequence was determined from a cDNA library of Conus pergrandis. It is a potent antagonist of rat ␣3␤2, ␣6/␣3␤2␤3, and ␣9␣10 nAChRs (23,24). This peptide shares a similar amino acid sequence with other 4/7 framework ␣-Ctxs that target ␣6/␣3␤2␤3 and ␣3␤2 nAChRs ( Fig. 1). We used positional scanning mutagenesis to identify residues of PeIA that were important for its binding to ␣6/␣3␤2␤3 and ␣3␤2 nAChRs. The Ala-substituted analogs were tested on cloned rat nAChRs expressed in X. laevis oocytes, and their activities were compared with that of native PeIA. Substitution of His 5 or Pro 6 in the first Cys loop with Ala or Hyp, respectively, significantly reduced the potency of PeIA for both receptor subtypes, whereas substitution of Ser 4 with Ala had no effect (Fig. 2, A and  B). In the second Cys loop, a slight increase in potency for both subtypes was observed with an Ala substitution of Ser 9 , whereas H12A significantly reduced potency (Fig. 2, C and D). Alanine substitution of Glu 14 or Leu 15 slightly reduced the potency for ␣3␤2 nAChRs, whereas of these, only the L15A substitution resulted in decreased potency for ␣6/␣3␤2␤3 nAChRs (Fig. 2, C and D). Substitutions of Asn 11 with Ala or of Pro 13 with either Ala or Hyp had little effect on the potency for either subtype. A summary of the changes in potencies of the Ala-and Hyp-substituted analogs is shown in Table 1.
A study by Pucci et al. (34) suggested that a positively charged amino acid in the second Cys loop might be an important determinant favoring binding to the ␣6 subunit while disfavoring binding to the ␣3 subunit. The sequence of ␣-Ctx TxIB was recently determined from a cDNA library of Conus textile (35). Remarkably, TxIB and PeIA only differ by 6 amino acids, yet TxIB is exclusively selective for ␣6/␣3␤2␤3 nAChRs. We note that TxIB has an Arg in the 9th position and a Lys in the 11th position instead of the Ser and Glu, respectively, found in PeIA (Fig. 1). Therefore, we substituted Ser 9 , Val 10 , and Glu 11 one at a time with Arg and evaluated the resulting analogs for changes in potency for ␣6/␣3␤2␤3 and ␣3␤2 nAChRs. Substitution of Ser 9 with Arg increased the potency for both subtypes, whereas substitution of Val 10 with Arg significantly reduced the potency for both subtypes (Fig. 3 (C and D) and Table 2). The N11R substitution had the effect of essentially abolishing the activity of PeIA for the ␣3␤2 subtype ( Fig. 3C and Table 2) while minimally affecting the potency for ␣6/␣3␤2␤3 receptors ( Fig. 3D and Table 2). Similarly, the N11K substitution dramatically increased the IC 50 for inhibition of ␣3␤2 nAChRs by more than 2,300-fold ( Fig. 3C and Table 2).

DISCUSSION
In this study, we used positional scanning mutagenesis of ␣-Ctx PeIA in conjunction with mutagenesis of ␣6 and ␣3 subunits to examine the interaction between PeIA and the nAChR subtypes ␣6/␣3␤2␤3 and ␣3␤2. Native PeIA has equal affinity for rat ␣6/␣3␤2␤3 and ␣3␤2 nAChRs; therefore, changes in potency for these receptor subtypes, resulting from mutations in the sequence of PeIA, can be used to identify functionally important residues. This strategy generated several PeIA analogs that were used to determine whether the differences in non-conserved amino acids of ␣6 and ␣3 subunits could be exploited to generate ligands that were selective for ␣6/␣3␤2␤3  over the ␣3␤2 subtype. We started with alanine scanning mutagenesis of PeIA. None of the three residues in the first Cys loop enhanced the peptide's ability to discriminate between the two nAChR subtypes, and substitution of His 5 and Pro 6 with Ala or Hyp, respectively, significantly reduced the potency of PeIA on both subtypes (Fig. 2, A and B, and Table 1). The amino acids Pro and His are thought to confer structural rigidity to the ␣-helical portion of ␣-Ctxs, and Pro in the 6th position is highly conserved among 4/7 ␣-Ctxs across species (Fig. 1). The fact that PeIA(H5N) retains activity whereas PeIA(P6A) and PeIA(P6O) do not suggests that Pro 6 , and not His 5 (Figs. 2 (A  and B) and 3 (A and B) and Tables 1 and 2), is critical for the ␣-helical structure of PeIA. Of the substitutions made in the first Cys loop, only H5N and A7V showed any discrimination between ␣6/␣3␤2␤3 and ␣3␤2 nAChRs (Fig. 3 (A and B) and Table 2). Substitutions of amino acids in the second Cys loop showed some similarities with those of the first Cys loop. Replacing His 12 with Ala essentially abolished activity for ␣6/␣3␤2␤3 and ␣3␤2 nAChRs (Fig. 2 (B and C) and Table 1), whereas replacing Pro 13 with Ala, Hyp, or Ser had no effect (Figs. 2 (B and C) and  3 (A and B) and Tables 1 and 2), suggesting that, in contrast to the His and Pro in the first Cys loop, His 12 is more important than Pro 13 for structural rigidity. Alanine substitution of Asn 11 and Glu 14 had no effect on ␣6/␣3␤2␤3 potency, whereas the S9A substitution increased potency (Fig. 3B and Table 2). For ␣3␤2 nAChRs, S9A increased potency, whereas N11A had little effect, and E14A decreased potency ( Fig. 2A and Table 1). Both ␣6/␣3␤2␤3 and ␣3␤2 nAChRs were inhibited less potently by PeIA(L15A) (Fig. 2, C and D, and Table 1). Non-Ala substitutions in the second Cys loop produced more significant changes in potency and selectivity. Substitution of Ser 9 with Arg increased the potency for both receptor subtypes, whereas V10R nearly abolished activity (Fig. 3 (C and D) and Table 2). Finally, substitution of Asn 11 with a positively charged amino acid, either Arg or Lys, abolished activity on ␣3␤2 nAChRs but produced little change in potency for ␣6/␣3␤2␤3 nAChRs (Fig. 3  (C and D) and Table 2). We note that the amino acid sequence FIGURE 7. Sequence alignment of rat nAChR subunits. A, sequence alignment of the N-terminal ligand binding domains of ␣6 and ␣3. A pairwise comparison of the sequences of the two subunits identified 137 (67%) identities and 32 (15%) similarities at the amino acid level. The arrows indicate the three amino acids of the ␣6 and ␣3 ligand binding domains that were examined in this study using mutagenesis. B, the first 210 amino acid pairs of the ␤2 and ␤3 subunits contained 81 (38%) identities and 22 (21%) similarities. The asterisks in B identify residues of the ␤2 subunit that have been shown previously to be important for ␣-Ctx binding (19,47).
of TxIB shows considerable homology with that of PeIA, differing by only six amino acids. TxIB is selective for ␣6/␣3␤2␤3 nAChRs (35), whereas PeIA shows no discrimination between ␣6/␣3␤2␤3 and ␣3␤2 nAChRs (IC 50 values of 11.1 and 9.70 nM, respectively) (24). Thus, some combination of these six differing residues must confer the selectivity of TxIB for the ␣6 subunit. TxIB has a Lys in the 11th position (Fig. 1), and this, taken together with the effects of the N11R substitution in PeIA, suggests that a positively charged amino acid in this position is important for conferring selectivity for the ␣6 subunit over the ␣3 subunit.
Although residue 152 appears to be a critical determinant of the FIGURE 8. The sensitivity of ␣6/␣3␤2␤3 and ␣3␤2 nAChRs to inhibition by PeIA(N11R) and PeIA(A7V,S9H,V10A,N11R,E14A) is determined by three residues in the ␣6/␣3 and ␣3 subunits. Xenopus oocytes expressing ␣6/␣3␤2␤3, ␣3␤2, and mutants of these receptors were subjected to TEVC as described under "Experimental Procedures," and the inhibition of the I ACh by PeIA(N11R) and PeIA(A7V,S9H,V10A,N11R,E14A) was determined. The IC 50 values are summarized in Table 5. Error bars, S.E. from 4 -5 oocytes for each experimental determination.  ability of PeIA(N11R) to discriminate between ␣6/␣3␤2␤3 and ␣3␤2 nAChRs, other residues of the receptor also appear to be involved in the interaction of PeIA and its analogs with the nicotinic receptors as evidenced by the larger shifts in potencies of PeIA(A7V,S9H,V10A,N11R,E14A) for the different receptor subtypes and their mutants. The IC 50 value for this analog for inhibition of ␣3K152E,E184D,Q195T/␤2 decreased by Ͼ2,000-fold and approximated the value obtained for ␣6/␣3␤2␤3 receptors ( Fig. 8B and Table 5). This suggests that these three residues account for a large portion of the high affinity binding of PeIA(A7V,S9H,V10A,N11R,E14A) to the ␣6-␤2 binding site. A homology binding model was generated to gain a better understanding of the interaction between PeIA and its analogs with the three residues of the ␣6 and ␣3 subunits that affected ligand binding (Fig. 9). Although it is known that residue 152 of the ␣ subunit can affect ␣-Ctx binding (31), this residue is located in the B-loop rather than the canonical ligand binding site composed of residues in the C-loop. However, as shown in Fig. 9, it appears that an interaction between residues of the B-loop, and in particular residue 152, may directly interact with ␣-Ctxs occupying the ACh binding pocket. Thus, when ␣-Ctxs have positively charged amino acids in the 11th position, such as TxIB or PeIA(N11R), binding is disfavored to ␣3-containing nAChRs potentially due to a repulsive charge-charge interaction with Lys 152 . The other two residues examined in this study, residues 184 and 195, are located in the C-loop but do not appear to interact directly with the ␣-Ctx because their side chains are oriented away from the ␣-Ctx. These residues are located in an area of the C-loop that has been proposed to act like a hinge allowing it to move toward the complementary subunit upon binding of agonists. Crystal structures of AChBPs complexed with different ␣-Ctxs show that the C-loop is even more open than its configuration in the absence of a bound ␣-Ctx (33,36). Subtle changes in the position of the C-loop may be required to accommodate ␣-Ctxs of different sizes and amino acid compositions. The smaller Asp 184 and Thr 195 , relative to Glu and Gln of the ␣3 subunit, respectively, may allow the C-loop of the ␣6 subunit to be more flexible and permit some ␣-Ctxs to bind more easily. Conversely, the C-loop of the ␣3 subunit may be more rigid, hindering ␣-Ctx binding. Other residues of the C-loop have been shown to play a more direct role in the binding of ␣-Ctxs. Beissner et al. (37) found that an Arg in position 185 of the ␣4 subunit prevents high affinity binding of several 4/7 ␣-Ctxs to the ␣4␤2 subtype, and mutation to Ile, the residue found in the homologous position of ␣6 and ␣3 subunits, increased binding affinity. Last, the differing residues of the various ␣ subunits may produce unique conformations of the C-loop itself, allowing ␣-Ctxs to bind to some receptor subtypes but not others, as suggested in a recent study examining the binding of ␣-Ctx BuIA to ␣6/␣3␤2␤3 nAChRs (38).
Current evidence suggests that, along with the ␣4␤2* subtype, ␣6␤2* receptors play an important role in modulating the release of dopamine in the nigrostriatal system, and loss of dopaminergic tone leads to disordered control of movement. In Parkinson disease and in animal models of Parkinson disease, the extent and severity of the disease correlates with a reduction of 125 I-MII binding, suggesting a loss of ␣6␤2* receptors (10,11,44). However, MII also potently binds to ␣3␤2* nAChRs, which have also been suggested to contribute, although to a much lesser degree, to nigrostriatal dopamine release. The differential contribution of ␣6␤2* and ␣3␤2* nAChRs to this release has yet to be unequivocally determined for lack of highly selective ligands that distinguish between these two subtypes. This situation has now been remedied with the availability of PeIA(A7V,S9H,V10A,N11R,E14A), and this new analog may be particularly useful for differentiating the role of ␣6␤2* from that of ␣3␤2* nAChRs in the nigrostriatal system. Additionally, converging evidence from multiple studies suggests that drugs that target ␣6␤2* may be useful in the pharmacotherapy of Parkinson disease and nicotine dependence (45,46). However, compounds that target ␣6␤2* would need to be devoid of activity on the ␣3␤2* subtype to avoid unwanted cardiovascular and enteric side effects. The results of the present study should be useful in guiding the further development of compounds that selectively target ␣6␤2* nAChRs.