Selective α3β4 Nicotinic Acetylcholine Receptor Ligand as a Potential Tracer for Drug Addiction

α3β4 Nicotinic acetylcholine receptor (nAChR) has been recognized as an emerging biomarker for the early detection of drug addiction. Herein, α3β4 nAChR ligands were designed and synthesized to improve the binding affinity and selectivity of two lead compounds, (S)-QND8 and (S)-T2, for the development of an α3β4 nAChR tracer. The structural modification was achieved by retaining the key features and expanding the molecular structure with a benzyloxy group to increase the lipophilicity for blood-brain barrier penetration and to extend the ligand-receptor interaction. The preserved key features are a fluorine atom for radiotracer development and a p-hydroxyl motif for ligand-receptor binding affinity. Four (R)- and (S)-quinuclidine-triazole (AK1-AK4) were synthesized and the binding affinity, together with selectivity to α3β4 nAChR subtype, were determined by competitive radioligand binding assay using [3H]epibatidine as a radioligand. Among all modified compounds, AK3 showed the highest binding affinity and selectivity to α3β4 nAChR with a Ki value of 3.18 nM, comparable to (S)-QND8 and (S)-T2 and 3069-fold higher affinity to α3β4 nAChR in comparison to α7 nAChR. The α3β4 nAChR selectivity of AK3 was considerably higher than those of (S)-QND8 (11.8-fold) and (S)-T2 (294-fold). AK3 was shown to be a promising α3β4 nAChR tracer for further development as a radiotracer for drug addiction.


Introduction
Nicotinic acetylcholine receptors (nAChRs) are ligand-gated ion channel receptors in the Cys-loop superfamily, which are expressed in the central nervous system (CNS) and in the peripheral nervous system (PNS) [1][2][3]. The subunits of nAChRs can be alpha (α 2 -α 10 ) and beta (β 2 -β 4 ) subunits surrounding an ion pore. The nAChR subtypes can be divided into two classes, according to subunit assembly: homopentameric receptors such as α 7 and α 9 and heteropentameric receptors such as α 3 β 4 and α 4 β 2 [3]. The different subunit combinations bring about the distinct pharmacological profile of nAChR [1]. For example, the α 7 and α 4 β 2 nAChR subtypes, which are expressed at high levels in the brain, are involved in neurological disorders (Alzheimer's disease, schizophrenia, Parkinson's disease, Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 3 of 15 Figure 1. Structures of (a) α3β4 nAChR ligands and (b) radio-imaging agents.

Results and Discussion
Four quinuclidine-triazole compounds (AK1-AK4) were modified by merging the structures of two lead compounds, (S)-QND8 and (S)-T2 [46], followed by single point modification ( Figure 2) to improve α3β4 nAChR binding affinity and selectivity profiles. Two pharmacophoric features, a chiral quinuclidine ring ((R) and (S)) acting as a cationic center and a triazole linker acting as a hydrogen bond acceptor [47], were kept in the modified structures. The key functional motifs in the hydrophobic part of (S)-QND8 and (S)-T2 were preserved: a hydroxyl group (-OH) for forming a hydrogen bond with the receptor [48] and a fluorine atom for further development as a positron-emitting radionuclide, fluorine-18. Besides retaining the fluorine atom and the hydroxyl group, the single-point modification was made by extending the hydrophobic part with a benzene ring at the hydroxyl function. The benefit of the added benzyloxy is that it can increase the lipophilicity of (S)-QND8 and (S)-T2 from log P of 1.87 and 2.03, respectively to log P of 4.03 and accordingly enhance brain penetration and subtype selectivity. The hydrophobic area has been reported to mediate nAChR subtype selectivity.
The binding affinities (Ki) of AK1-AK4 and two lead compounds were presented in Table 1. All compounds exhibited binding affinity to α3β4 nAChR, α7 nAChR, and α4β2 nAChR in the range of 2.28-9760 nM. Comparing (R)-and (S)-enantiomers, the (S)-isomers of both AK1 and AK3 preferably bound to α3β4 nAChR, whereas their corresponding (R)-isomers (AK2 and AK4) selectivity bound to α7 nAChR. The stereoselective binding of (S)-enantiomers to α3β4 nAChR and (R)-enantiomers to α7 nAChR agrees with previous reports [46,48]. This might lead to some limitations of this study. The data discussed stereoselectivity came from one core structure, which is a quinuclidine ring.

Results and Discussion
Four quinuclidine-triazole compounds (AK1-AK4) were modified by merging the structures of two lead compounds, (S)-QND8 and (S)-T2 [46], followed by single point modification ( Figure 2) to improve α 3 β 4 nAChR binding affinity and selectivity profiles. Two pharmacophoric features, a chiral quinuclidine ring ((R) and (S)) acting as a cationic center and a triazole linker acting as a hydrogen bond acceptor [47], were kept in the modified structures. The key functional motifs in the hydrophobic part of (S)-QND8 and (S)-T2 were preserved: a hydroxyl group (-OH) for forming a hydrogen bond with the receptor [48] and a fluorine atom for further development as a positron-emitting radionuclide, fluorine-18. Besides retaining the fluorine atom and the hydroxyl group, the single-point modification was made by extending the hydrophobic part with a benzene ring at the hydroxyl function. The benefit of the added benzyloxy is that it can increase the lipophilicity of (S)-QND8 and (S)-T2 from log P of 1.87 and 2.03, respectively to log P of 4.03 and accordingly enhance brain penetration and subtype selectivity. The hydrophobic area has been reported to mediate nAChR subtype selectivity.   The binding affinities (K i ) of AK1-AK4 and two lead compounds were presented in Table 1. All compounds exhibited binding affinity to α 3 β 4 nAChR, α 7 nAChR, and α 4 β 2 nAChR in the range of 2.28-9760 nM. Comparing (R)-and (S)-enantiomers, the (S)-isomers of both AK1 and AK3 preferably bound to α 3 β 4 nAChR, whereas their corresponding (R)-isomers (AK2 and AK4) selectivity bound to α 7 nAChR. The stereoselective binding of (S)-enantiomers to α 3 β 4 nAChR and (R)-enantiomers to α 7 nAChR agrees with previous reports [46,48]. This might lead to some limitations of this study. The data discussed stereoselectivity came from one core structure, which is a quinuclidine ring.             The binding affinities of new (S)-enantiomers (AK1 and AK3) to α 3 β 4 nAChR with K i values of 2.28 and 3.18 nM are high, almost the same as those of the lead compounds, (S)-QND8, (S)-T2, and AT-1001 (K i = 2.48, 2.25, and 2.60 nM, respectively) [46]. The selectivity of AK3 over the α 7 subtype is the highest with a 3069-fold higher affinity to the α 3 β 4 in comparison to the α 7 subtype, whereas for all compounds the selectivity for the α 4 β 2 subtype is much lower with a range of 10-220-fold. Molecular docking of AK1, AK3, and lead compounds to α 3 β 4 and α 7 nAChR homology models were performed to explain the remarkably high selectivity profile of AK3 to α 3 β 4 over α 7 nAChRs.
The molecular docking of new compounds and lead compounds to α 3 β 4 nAChR homology model showed that all compounds aligned and formed hydrogen bonds and π-π interactions with amino acid residues in an aromatic cage located at the interface between the α 3 subunit and β 4 subunit. The protonated quinuclidine ring of all (S)-enantiomers pointed to Asp173, a key amino acid determinant to form a salt bridge and hydrogen bond interaction in slightly shorter distances than their (R)-counterparts, particularly of 1.74 Å of AK3 vs. 2.14 Å of AK4 ( Figure 3, Table 2). In agreement with a previous report [48], the (S)-enantiomer of a quinuclidine ring allowed the docked pose to accommodate a salt bridge interaction with Asp173, the key determinant of the α 3 β 4 nAChR binding.
Other common key residues of α 3 β 4 nAChR for binding are Trp149 and Tyr190 in the α 3 subunit for which the interactions can be observed to Trp149 and/or Tyr190 in the modified compounds. For AK1, the protonated quinuclidine formed a salt bridge interaction with Asp173, and the phenolic −OH formed a hydrogen bond with Ser148. The additional interactions were cation-π interaction with Trp59, π-π interactions with Trp59, Trp149, Trp190, and Tyr197, and halogen bond with Trp149 ( Figure 4A and Table 2). Even though the number of hydrogen bonds and π-π interaction of AK3 with key residues was less than those of AK1, the binding affinity of AK3 to α 3 β 4 nAChR (K i = 3.18 nM) was found to be comparable to AK1 (K i = 2.28 nM). This result confirms that the extending benzyloxy group increased hydrophobic interactions as designed. These interactions from the expansion contributed to the similar binding affinity and ligand efficiency (LE), a parameter to assess the binding affinity of different molecular weights, of these two compounds (LE = −0.52 and −0.51 for AK1 and AK3) ( benzyloxy group in the AK3 structure, significantly enhanced the selectivity profile by the hydrophobic interactions with the residues in a β4-complementary subunit ( Figure 3B). Therefore, the added benzyloxy group is the key contributor to the enhanced α3β4-selectivity of AK3 by providing the hydrophobic interactions with the β4 subunit. MD simulation of AK3 to α3β4 and α7 nAChRs was performed to gain more insight into the high affinity and selectivity of AK3 to α3β4 nAChR (Ki = 3.18 nM, 3069-fold selectivity over α7 nAChR). The strong interaction from the salt bridge formation between protonated quinuclidine and the carboxylate of Asp173, which appeared to mediate high af-  The hydrogen bonds were analyzed and measured by AutoDock4.2, and the salt bridges, cation-π, and π-π interactions were analyzed by BIOVIA Discovery Studio Visualized. The interaction of amino acid residues with protonated quinuclidine is presented in bold. * LE is the ratio of Gibbs free energy (∆G) to the number of non-hydrogen atoms of the ligand. in AK3-α3β4 nAChR complex (Figure 6b) was stronger than the hydrogen bond interaction to Ser166 of α7 nAChR. The results from both molecular docking and MD simulation supported the higher affinity and selectivity of AK3 to α3β4 nAChR than α7 nAChR. For (R)-enantiomers, AK2 and AK4 showed lower binding affinities to α3β4 nAChR with Ki values of 601 and 112 nM, respectively than their (S)-counterparts and the lead compounds. From molecular docking results, the ligand-receptor interactions provided by quinuclidine and hydrophobic pharmacophores of AK2 and AK4 were not different from those of AK1 and AK3 ( Figure 3 and Table 2). Only the triazole ring pointed to a different angle leading to the lower number of π-π interactions between the triazole ring of AK2 and AK4 with α3β4 nAChR compared to their (S)-counterparts (AK1 and AK3) showing the important role of this moiety for the binding affinity ( Figure 4). However, the (R)-enantiomers AK2 and AK4 bound to α7 nAChR with Ki values of 4.49 and 53.6 nM, respectively, which are higher than the values of their (S)-counterparts ( Table 1). The molecular docking to the α7 nAChR homology model showed that the quinuclidine ring interacted with Trp149 and Tyr93, key determinants of the α7 subtype leading to the high affinity of AK2 and AK4 ( Figure 6). The key interactions of AK2 are strong cation-π to Trp149, hydrogen bonds to Tyr93 and Trp149, and π-π interactions to Trp55 and Trp149, In terms of selectivity, AK3 possesses a 3069-fold selectivity to α 3 β 4 nAChR over α 7 nAChR, whereas the selectivity of AK1 and the lead compounds (S)-QND8 and (S)-T2 was much lower. The α 7 nAChR binding poses of AK3 and (S)-QND8 showed that the quinuclidine ring of these two compounds is aligned in different directions leading to the higher number of hydrogen bond interactions of (S)-QND8 ( Figure 5B). Three hydrogen bond interactions of (S)-QND8 to α 7 nAChR are composed of the interaction of protonated quinuclidine to Asp164 and Ser166 and the interaction of -OH to Tyr93. Only one hydrogen bond interaction between protonated quinuclidine to Tyr93 was observed in AK3 resulting in less preference for the α 7 subtype. Hence, the extension approach, via the added benzyloxy group in the AK3 structure, significantly enhanced the selectivity profile by the hydrophobic interactions with the residues in a β 4 -complementary subunit ( Figure 3B). Therefore, the added benzyloxy group is the key contributor to the enhanced α 3 β 4 -selectivity of AK3 by providing the hydrophobic interactions with the β 4 subunit.  The hydrogen bonds were analyzed and measured by AutoDock4.2, and the salt bridges, cation-π, and π-π interactions were analyzed by BIOVIA Discovery Studio Visualized. The amino acid residues' interaction with protonated quinuclidine is presented in bold. * LE is the ratio of Gibbs free energy (ΔG) to the number of non-hydrogen atoms of the ligand. MD simulation of AK3 to α 3 β 4 and α 7 nAChRs was performed to gain more insight into the high affinity and selectivity of AK3 to α 3 β 4 nAChR (K i = 3.18 nM, 3069-fold selectivity over α 7 nAChR). The strong interaction from the salt bridge formation between protonated quinuclidine and the carboxylate of Asp173, which appeared to mediate high affinity and selectivity to α 3 β 4 nAChR, was detected by MD simulation. Besides the salt bridge formation, the major interactions observed in the MD simulation of AK3 to α 3 β 4 nAChR are four π-π interactions between both middle and terminal benzene rings of the hydrophobic part of AK3 and residues Tyr93, Trp149, Tyr190, and Tyr 197 (Figure 6a). When compared with molecular docking, most of the interacting amino acid residues were the same (Tyr93, Ser148, Trp149, Ser150, and Tyr197 in principal α 3 subunit and Ala42, Trp59, Ile113, Leu123, Pro125, and Asp173 in complementary β 4 subunit) but the binding interaction types were altered, due to the flexible and solvated receptor in MD simulation. For example, the halogen bond between a fluorine atom and Trp149 and π-π interactions of a triazole ring to Trp59 and Trp149 observed in molecular docking were replaced by additional π-π interactions of middle and terminal benzene rings to Tyr93, Tyr190, and Tyr 197 in MD simulation (Figure 6a). Although the number of main interactions in the complexes of AK3-α 3 β 4 nAChR and AK3-α 7 nAChR, were equal (four π-π interactions), the salt bridge interaction between the protonated quinuclidine and Asp173 in AK3-α 3 β 4 nAChR complex (Figure 6b) was stronger than the hydrogen bond interaction to Ser166 of α 7 nAChR. The results from both molecular docking and MD simulation supported the higher affinity and selectivity of AK3 to α 3 β 4 nAChR than α 7 nAChR. Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 9 of 15  In terms of the structure-activity relationship (SAR) of quinuclidine-triazole derivatives targeting α3β4 nAChR, three components are required for high affinity and selectivity: the (S)-enantiomer of a quinuclidine ring, a triazole ring and a large hydrophobic group (extended benzene ring). Among four synthesized compounds, AK3 showed the highest binding affinity and selectivity to α3β4 nAChR, qualifying this compound as a non- For (R)-enantiomers, AK2 and AK4 showed lower binding affinities to α 3 β 4 nAChR with K i values of 601 and 112 nM, respectively than their (S)-counterparts and the lead compounds. From molecular docking results, the ligand-receptor interactions provided by quinuclidine and hydrophobic pharmacophores of AK2 and AK4 were not different from those of AK1 and AK3 ( Figure 3 and Table 2). Only the triazole ring pointed to a different angle leading to the lower number of π-π interactions between the triazole ring of AK2 and AK4 with α 3 β 4 nAChR compared to their (S)-counterparts (AK1 and AK3) showing the important role of this moiety for the binding affinity ( Figure 4). However, the (R)-enantiomers AK2 and AK4 bound to α 7 nAChR with K i values of 4.49 and 53.6 nM, respectively, which are higher than the values of their (S)-counterparts ( Table 1). The molecular docking to the α 7 nAChR homology model showed that the quinuclidine ring interacted with Trp149 and Tyr93, key determinants of the α 7 subtype leading to the high affinity of AK2 and AK4 (Figure 6). The key interactions of AK2 are strong cation-π to Trp149, hydrogen bonds to Tyr93 and Trp149, and π-π interactions to Trp55 and Trp149, whereas AK4 bound to the receptor with the hydrogen bond to Tyr93 and π-π interaction to Trp149 in addition to halogen bonds with Ser148 and Trp149 (Figure 7, Table 3). The absence of a strong cation-π interaction of the quinuclidine ring together with the lack of -OH to form additional hydrogen bonds resulted in the lower binding affinity of AK4 compared to AK2. In addition, the extended benzyloxy group present in AK4 might have caused steric hindrance to the receptor, leading to a decrease in affinity. The higher binding affinities of AK2 than AK4 (K i of 53.6 and 4.49 nM) were found to agree with the ligand efficiency (LE) of these two compounds: AK4 bound to α 7 nAChR weaker than AK2 (LE = −0.47 and −0.52, respectively) ( Table 3).  In terms of the structure-activity relationship (SAR) of quinuclidine-triazole derivatives targeting α3β4 nAChR, three components are required for high affinity and selectivity: the (S)-enantiomer of a quinuclidine ring, a triazole ring and a large hydrophobic group (extended benzene ring). Among four synthesized compounds, AK3 showed the highest binding affinity and selectivity to α3β4 nAChR, qualifying this compound as a non-  The hydrogen bonds were analyzed and measured by AutoDock4.2, and the salt bridges, cation-π, and π-π interactions were analyzed by BIOVIA Discovery Studio Visualized. The amino acid residues' interaction with protonated quinuclidine is presented in bold. * LE is the ratio of Gibbs free energy (∆G) to the number of non-hydrogen atoms of the ligand.
In terms of the structure-activity relationship (SAR) of quinuclidine-triazole derivatives targeting α 3 β 4 nAChR, three components are required for high affinity and selectivity: the (S)-enantiomer of a quinuclidine ring, a triazole ring and a large hydrophobic group (extended benzene ring). Among four synthesized compounds, AK3 showed the highest binding affinity and selectivity to α 3 β 4 nAChR, qualifying this compound as a non-radioactive reference compound for α 3 β 4 nAChR. Therefore, AK3 is a high potential candidate for further development as α 3 β 4 nAChR PET tracer for monitoring drug addiction, after replacing the fluorine atom with the radioactive isotope, fluorine-18.

Synthesis
All chemicals and solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA), Merck (Darmstadt, Germany), AK Scientific (Union City, CA, USA), Oakwood (Estill, SC, USA), and Fisher Scientific (Waltham, MA, USA) and used without further purification. The NMR spectroscopic data ( 1 H, 13 C, COSY, HSQC, HMBC) were recorded with a Varian Mercury-300. High-Resolution Mass Spectra (HRMS) were recorded on the FT-ICR APEX II spectrometer using electrospray ionization (ESI) in positive ion mode.
AK1-AK4 were synthesized by the previously described methods [47]. In brief, the terminal alkyne and quinuclidine azide were prepared first. For the alkyne building block, 2-fluoro-4-iodophenol reacted with trimethylsilyl acetylene in base condition using PdCl 2 (PPh 3 ) 2 and CuI as catalysts under nitrogen atmosphere overnight and purified by SiO 2 column chromatography using 15%EtOAc in hexane as a mobile phase. The intermediate compound was desilylated with TBAF in THF at room temperature and purified by SiO 2 column chromatography using 15%EtOAc in hexane as a mobile phase to get the terminal alkyne (Scheme 1) for AK1 and AK2. For terminal alkyne of AK3 and AK4, 2-fluoro-4-iodophenol first reacted with benzyl bromide in base condition at room temperature overnight (Scheme 2) before performing a Sonogashira cross-coupling reaction. The crude product was purified by SiO 2 column chromatography using 5%Et 3 N, 10%MeOH in DCM as a mobile phase. radioactive reference compound for α3β4 nAChR. Therefore, AK3 is a high potential candidate for further development as α3β4 nAChR PET tracer for monitoring drug addiction, after replacing the fluorine atom with the radioactive isotope, fluorine-18.

Synthesis
All chemicals and solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA), Merck (Darmstadt, Germany), AK Scientific (Union City, CA, USA), Oakwood (Estill, SC, USA), and Fisher Scientific (Waltham, MA, USA) and used without further purification. The NMR spectroscopic data ( 1 H, 13 C, COSY, HSQC, HMBC) were recorded with a Varian Mercury-300. High-Resolution Mass Spectra (HRMS) were recorded on the FT-ICR APEX II spectrometer using electrospray ionization (ESI) in positive ion mode.
AK1-AK4 were synthesized by the previously described methods [47]. In brief, the terminal alkyne and quinuclidine azide were prepared first. For the alkyne building block, 2-fluoro-4-iodophenol reacted with trimethylsilyl acetylene in base condition using PdCl2(PPh3)2 and CuI as catalysts under nitrogen atmosphere overnight and purified by SiO2 column chromatography using 15%EtOAc in hexane as a mobile phase. The intermediate compound was desilylated with TBAF in THF at room temperature and purified by SiO2 column chromatography using 15%EtOAc in hexane as a mobile phase to get the terminal alkyne (Scheme 1) for AK1 and AK2 . For terminal alkyne of AK3 and AK4, 2fluoro-4-iodophenol first reacted with benzyl bromide in base condition at room temperature overnight (Scheme 2) before performing a Sonogashira cross-coupling reaction. The crude product was purified by SiO2 column chromatography using 5%Et3N, 10%MeOH in DCM as a mobile phase. Scheme 1. Sonogashira cross-coupling reaction and desilylation with TBAF in THF.
For the preparation of (R)-and (S)-quinuclidine azides, the reaction of trifluoromethanesulfonic anhydride and sodium azide was run in the mixture of water and toluene (1:1.5) at 0 °C for 2 h and the reaction mixture was extracted with toluene to yield trifluoromethanesulfonyl azide (TfN3). The freshly prepared TfN3 in toluene was added to (R)-or (S)-of 3-aminoquinuclidine, K2CO3, and CuSO4.5H2O in the mixture of water and methanol (1:2) at room temperature overnight (Scheme 3). The prepared azide building blocks were then used without further purification. after replacing the fluorine atom with the radioactive isotope, fluorine-18.

Synthesis
All chemicals and solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA), Merck (Darmstadt, Germany), AK Scientific (Union City, CA, USA), Oakwood (Estill, SC, USA), and Fisher Scientific (Waltham, MA, USA) and used without further purification. The NMR spectroscopic data ( 1 H, 13 C, COSY, HSQC, HMBC) were recorded with a Varian Mercury-300. High-Resolution Mass Spectra (HRMS) were recorded on the FT-ICR APEX II spectrometer using electrospray ionization (ESI) in positive ion mode.
AK1-AK4 were synthesized by the previously described methods [47]. In brief, the terminal alkyne and quinuclidine azide were prepared first. For the alkyne building block, 2-fluoro-4-iodophenol reacted with trimethylsilyl acetylene in base condition using PdCl2(PPh3)2 and CuI as catalysts under nitrogen atmosphere overnight and purified by SiO2 column chromatography using 15%EtOAc in hexane as a mobile phase. The intermediate compound was desilylated with TBAF in THF at room temperature and purified by SiO2 column chromatography using 15%EtOAc in hexane as a mobile phase to get the terminal alkyne (Scheme 1) for AK1 and AK2 . For terminal alkyne of AK3 and AK4, 2fluoro-4-iodophenol first reacted with benzyl bromide in base condition at room temperature overnight (Scheme 2) before performing a Sonogashira cross-coupling reaction. The crude product was purified by SiO2 column chromatography using 5%Et3N, 10%MeOH in DCM as a mobile phase. Scheme 1. Sonogashira cross-coupling reaction and desilylation with TBAF in THF.
For the preparation of (R)-and (S)-quinuclidine azides, the reaction of trifluoromethanesulfonic anhydride and sodium azide was run in the mixture of water and toluene (1:1.5) at 0 °C for 2 h and the reaction mixture was extracted with toluene to yield trifluoromethanesulfonyl azide (TfN3). The freshly prepared TfN3 in toluene was added to (R)-or (S)-of 3-aminoquinuclidine, K2CO3, and CuSO4.5H2O in the mixture of water and methanol (1:2) at room temperature overnight (Scheme 3). The prepared azide building blocks were then used without further purification. For the preparation of (R)-and (S)-quinuclidine azides, the reaction of trifluoromethanesulfonic anhydride and sodium azide was run in the mixture of water and toluene (1:1.5) at 0 • C for 2 h and the reaction mixture was extracted with toluene to yield trifluoromethanesulfonyl azide (TfN 3 ). The freshly prepared TfN 3 in toluene was added to (R)-or (S)-of 3-aminoquinuclidine, K 2 CO 3, and CuSO 4 .5H 2 O in the mixture of water and methanol (1:2) at room temperature overnight (Scheme 3). The prepared azide building blocks were then used without further purification.
after replacing the fluorine atom with the radioactive isotope, fluorine-18.

Synthesis
All chemicals and solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA), Merck (Darmstadt, Germany), AK Scientific (Union City, CA, USA), Oakwood (Estill, SC, USA), and Fisher Scientific (Waltham, MA, USA) and used without further purification. The NMR spectroscopic data ( 1 H, 13 C, COSY, HSQC, HMBC) were recorded with a Varian Mercury-300. High-Resolution Mass Spectra (HRMS) were recorded on the FT-ICR APEX II spectrometer using electrospray ionization (ESI) in positive ion mode.
AK1-AK4 were synthesized by the previously described methods [47]. In brief, the terminal alkyne and quinuclidine azide were prepared first. For the alkyne building block, 2-fluoro-4-iodophenol reacted with trimethylsilyl acetylene in base condition using PdCl2(PPh3)2 and CuI as catalysts under nitrogen atmosphere overnight and purified by SiO2 column chromatography using 15%EtOAc in hexane as a mobile phase. The intermediate compound was desilylated with TBAF in THF at room temperature and purified by SiO2 column chromatography using 15%EtOAc in hexane as a mobile phase to get the terminal alkyne (Scheme 1) for AK1 and AK2 . For terminal alkyne of AK3 and AK4, 2fluoro-4-iodophenol first reacted with benzyl bromide in base condition at room temperature overnight (Scheme 2) before performing a Sonogashira cross-coupling reaction. The crude product was purified by SiO2 column chromatography using 5%Et3N, 10%MeOH in DCM as a mobile phase. Scheme 1. Sonogashira cross-coupling reaction and desilylation with TBAF in THF.
For the preparation of (R)-and (S)-quinuclidine azides, the reaction of trifluoromethanesulfonic anhydride and sodium azide was run in the mixture of water and toluene (1:1.5) at 0 °C for 2 h and the reaction mixture was extracted with toluene to yield trifluoromethanesulfonyl azide (TfN3). The freshly prepared TfN3 in toluene was added to (R)-or (S)-of 3-aminoquinuclidine, K2CO3, and CuSO4.5H2O in the mixture of water and methanol (1:2) at room temperature overnight (Scheme 3). The prepared azide building blocks were then used without further purification. The prepared terminal alkynes and quinuclidine azides reacted via copper-catalyzed azide-alkyne cycloaddition (CuAAC) or click chemistry using 20 mol% of sodium ascorbate and 5 mol% of copper sulfate as catalysts to yield AK1-AK4 as quinuclidine triazole compounds (Scheme 4). The prepared terminal alkynes and quinuclidine azides reacted via copper-catalyzed azide-alkyne cycloaddition (CuAAC) or click chemistry using 20 mol% of sodium ascorbate and 5 mol% of copper sulfate as catalysts to yield AK1-AK4 as quinuclidine triazole compounds (Scheme 4).

Binding Affinity
The binding affinities of quinuclidine triazole compounds AK1-AK4 to nAChRs were determined by radioligand displacement assays [46]. SH-SY5Y cells stably transfected with human α 7 nAChR and HEK293 cells stably transfected with human α 4 β 2 nAChR or α 3 β 4 nAChR were used in the experiments. Cells were collected, sedimented (800 rpm, 3 min), diluted with 50 mM TRIS-HCl, pH 7.4, and stored at −25 • C until use. Frozen cell suspensions were thawed and homogenized by a 27-gauge needle and diluted with incubation buffer (50 mM TRIS-HCl, pH 7.4, 120 mM NaCl, and 5 mM KCl). The membrane suspension was incubated with (±)-[ 3 H]epibatidine (0.3 to 0.6 nM final concentration; molar activity 2.22 GBq/mmol). Nonspecific binding was determined by co-incubation with 300 µM (-)-nicotine tartrate. The incubation was performed at room temperature for 120 min and terminated by rapid filtration using Whatman GF/B glass-fiber filters presoaked in 0.3% polyethyleneimine and a 48-channel harvester (Biomedical Research and Development Laboratories, Gaithersburg, MD, USA) followed by 4 times washing with ice-cold 50 mM TRIS-HCl, pH 7.4. Filter-bound radioactivity was quantified by liquid scintillation counting. The 50% inhibition concentrations (IC 50 ) were estimated from the competition curves by nonlinear regression using GraphPad Prism software and the K i values were calculated according to the Cheng-Prusoff equation [49].

Molecular Docking
The homology models of α 3 β 4 and α 7 nAChRs were prepared as described in our previous study [48]. Briefly, the human amino acid sequences of α 3 β 4 and α 7 nAChRs were downloaded from UniProt for searching a proper protein template from Protein Data Bank (PDB) by Blast protein in Chimera 1.10.2 (Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, CA, USA). The amino acid sequences of α 3 β 4 nAChR or α 7 nAChR and acetylcholine binding protein (AChBP) (PDB ID 5AFH) used as a template were aligned by Clustal Omega (Conway Institute, University College Dublin, Dublin, Ireland) and the homology model of α 3 β 4 nAChR and α 7 nAChR was generated by Modeller 9.15 (University of California San Francisco, San Francisco, CA, USA). Several parameters i.e., the Discrete Optimized Protein Energy (DOPE) score, the GA341 score, and the Ramachandran plot were used to evaluate model quality. The structures of AK1-AK4 were drawn as protonated forms by Chem Draw Ultra 12.0 program (PerkinElmer, Waltham, MA, USA). The parameters for the molecular docking study with AutoDock4.2 included: 100 GA runs, a population size of 150, a maximum of 10,000,000 evaluations, and a maximum of 27,000 generations. The similar 3D conformations orientation within 2.0 Å were grouped as conformation clusters. The docked poses in the highest cluster as well as free binding energies (∆G binding) and ligand efficiency (LE) were analyzed. The binding interactions between ligand and target protein were visually analyzed by AutoDock4.2 in addition to BIOVIA Discovery Studio Visualized (Biovia, San Diego, CA, USA).

Molecular Dynamics (MD) Simulation
The molecular docking complexes of AK3 to a homology model of α 3 β 4 and α 7 nAChRs were optimized by MD simulation using NAMD software (University of Illinois at Urbana-Champaign, Urbana, IL, USA) [50] with CHARMM force field [51]. The complexes were solvated in the TIP3P model water box. The charge of the system was neutralized with an appropriate number of counter ions. Initially, the water box was minimized by the conjugate gradient method. Before the MD simulation, the system was equilibrated for 200 ps using an NPT ensemble at 310 K and 1 atm which was controlled by the Nosé-Hoover Langevin piston method [52] with 2 fs time steps and SHAKE algorithm. Periodic boundary conditions (PBC) and Particle Mesh Ewald (PME) method [52] were used for calculation. In the production steps, 100 ns of MD simulations were performed with trajectories saving every 2 ns for analysis. The complexes' stability was evaluated using root mean square deviation (RMSD) (Supplementary Figure S1). Finally, the complexes were analyzed by BIOVIA Discovery Studio 2020 (Biovia, San Diego, CA, USA) [53].

Statistical Analysis
Data are represented as mean derived from two independent experiments performed in triplicate. The mean differences were statistically analyzed by one-way ANOVA followed by Tukey's multiple comparison test using GraphPad Prism software.

Conclusions
The structural modification of α 3 β 4 nAChR ligands for the development of a PET tracer has been achieved. The newly designed quinuclidine-triazole derivative AK3 showed good binding affinity (K i = 3.18 nM) and a significantly enhanced selectivity α 7 nAChR (3069-fold). The structural features contributing to the significant improvement of affinity and selectivity profiles of AK3 are the (S)-enantiomer of the quinuclidine ring, the triazole linker, and the extended hydrophobic part for interaction with the β 4 subunit. The presence of a fluorine atom in AK3 provides the opportunity to develop AK3 as an α 3 β 4 nAChRtargeted PET tracer. Therefore, AK3 is a promising compound for further development as a drug-seeking behavior monitoring agent.