2034 Journal of Medicinal Chemistry, 2008, VoL. 51, No. 7 Xie et al (15 mL), drops of aqueous ammonia were added to adjust the ph ligand. The final conformation of the ligand was obtained after to 9, and the residue was extracted with CHCl3(20 mL x 3) 000 steps of energy minimization using the Tripos force field. The combined CHCl3 was dried with anhydrous Na2SO4 and Molecular docking was carried out using GOLD 3.0(CCDC graphed on silica gel and eluted with MeOH/CHCl3(0.5: 9.5)to conformations for the ligand. The active site was defined as all rovide 5c, d as an oil. Addition of dry HCl-ether to the solution atoms within a radius of 25 A around some specific residue atoms of 5c, d in dry ether, with pH adjusted to 3-4, gave the final OH of Tyr337 for mAChE and N2 of His438 for hBChE. We chose salt 5c, d' 2HCI as a powder more enlarged binding pockets here, concerned that smaller sites AN -(- e -b is- --n or- MlEP H drochloride:i e. catalytic and peripheral sites of AchE. The 600 genetic algorithm hydride(0. 20 g, 5.26 mmol), THF (30 mL), H2O(0.20 mL), 15% (GA)runs instead of the default 10 were erformed owing to the NaOH(0.20 mL), and H20 (0.60 mL)were used to produce 5c high flexibility of the ligand possessing a great many rotatable (0. 19 g, 36%0) Subsequent salt formation gave 5c. 2HCI(0.12 bonds. For each GA run, the default GA settings were used except 55%):mpl10-115°C;alb-51.96°(c0.092,MeOH);IR for the prohibited early termination and the permitted pyramidal 3254,2936,1600,1586,1448,1268cm-;HNMR(DMSO nitrogen inversion. d6)998,977(brs,4/3H,NH),9.56,9.54,9.43,9.42(s,2H Advanced combination approach of clustering and consensus Ar-OH).8.46(brs,2/3H,NH+),7.21-7.13(m,2H,Ar-H) scoring was used to guide the selection of the most reliable 685-665(m,6H,Ar-H),3.83(m,~2/3H,N-CH2),3.52(m, information from among a set of candidate conformations that 04/3 H. N-CH,). 3.36-3. 15(m, 10H. N-CH-)2.38(m, H. GOLD generated. All conformations were evaluated with four CH,). 2. 10-1.46(m, 19H. CH,0.49(t 6H, J=7.0 Hz, CH) available scoring functions, including three scoring functions MS(ESD)[M +H]+493.3, [M+2H)2+ 247.2 HRMS m/z caled (G-Score, D_Score, and ChemScore) from the CSCORE38module IR=7.45 min, 98.0% purity 1.2.1.The"rank-by-rank" "strategy reported by Wang et al. was NNN-(1,5-Pentylene)-bis-(-)-nor- MEP Hydrochloride (50 adopted to make consensus scoring. The final rank of a certain conformation was its average rank received from all four 2HCI). Compound 9d(0.55 g, 1.03 mmol), lithium aluminum functions. GOLD generated clusters based on different rmsd(root hydride(0.20 g, 5.26 mmol), THF (30 mL), H20(0.20 mL), 15% mean square deviation)criteria. A proper rmsd standard of clustering aOH(0.20 mL), and H20(0.60 mL) were used to produce 5d was important to the selection of representative clusters. Herein, (0.16 g, 31%. Subsequent salt formation of 5d( 0.10 g) gave rmsd values of 2.55 and 2.40 A were chosen as the criteria to cluster d·2HC1(0.06g,52%):mp80-82C;[alb-48.10°(c0.08 the docked conformations with mAChE and hBChE, respectively MeOH;IRv3417,2937,1600,1585,1447,1268cm-1;HNMR DMSO-d6)996,9.75(brs,5/4H,NH),9.57,9.52,9.44,9.42 Among the top 10 clusters for each ChE, the one with members possessing the minimum average"rerank" was identified as the (s,2H,Ar-OH,8.34(brs,3/4H,NH),7.19-7.12(m,2H representative cluster. The top"reranked"conformation in the (m, 5/4 H, N-CH2) 3.34-3.06('m, OH, N-CH2) 2.37( m, H. mode for the gander was selected as the representauve binding CH2),2.10-1.21(m,21H,CH2),0.48(m,6H,cHy);Ms(ESD)[M Inhibition of AChE-lnduced Ap Aggregation. Aliquots of 2 H 507.3,[M+ 2H2542 HRMS m/z caled for C33HsI N2O2 HL AB(1-40) peptide(Biosource), lyophilized from 2 mg/mL HFIP IM+H+,507.3945; found,507.3961.HPLC:lg=777min, solution and dissolved in DMSO, were incubated for 48 h at room temperature in 0.215 M sodium phosphate buffer(pH 8.0)at a final In Vitro AChE/BChE Inhibition Assay. Inhibitory activit concentration of 230 uM. For coincubation experiments, aliquots against AChE was evaluated by a modified Ellman's method. (16 AL) of human recombinant AChE (Sigma-Aldrich)(final Mice brain homogenate prepared in saline was used as a source of concentration of 2.3 uM, AB/AChE molar ratio of 100: 1)and AChE AChE; mice serum was the source of BChE. The AChE activity in the presence of 2 uL of the tested inhibitors were added. Blanks vas determined in a reaction mixture containing 200 uL of containing Af nd AB plus inhibitors at various concentra- solution of AChE(0.415 U/mL in 0. 1 M phosphate buffer, pH8.0), tions in 0.215 M sodium phosphate buffer(pH8.0)were prepared. 300 uL of a solution of 5, 5-dithio-bis(2-nitrobenzoic)acid(3.3 The final volume of each vial was 20 AL. Each assay was run in mM DTNB in 0.1 M phosphate buffered solution, pH 7.0, duplicate. Inhibitor stock solutions were prepared (c= 10 mM) containing NaHCO 6 mM), and 30 uL of a solution of the inhibitor and diluted in 0.215 M sodium phosphate buffer(pH 8.0)when (six to seven concentrations). After incubation for 20 min at 37 used. To quantify amyloid fibril formation, the thioflavin T oC, acetylthiocholine iodide(300 uL of 0.05 mM water solution) fluorescence method was then applied. 20 was added as the substrate, and AChe activity was determined by After incubation, the samples containing AB, AB plus AChE,or UV spectrophotometry from the absorbance changes at 412 nm fo AB plus AChE in the presence of inhibitors were diluted with 50 3 min at 25"C. The concentration of compound that produced 50% mM glycine-NaOH buffer(pH 8.5)containing 1.5 uM thioflavin inhibition of the AChE activity (ICso) was calculated by nonlinear T (Sigma-Aldrich) to a final volume of 2.0 mL. Fluorescence was regression of the response-concentration (log) curve. BChE monitored by PE LS45, with excitation at 446 nm and emission at inhibitory activity determinations were similarly carried out usin 490 nm. A time scan of fluorescence was performed, and the butyrylthiocholine iodide(0.05 mM)as the substrate. Results are intensity values reached at the plateau(around 300 s)were averaged reported as the mean+ SEM of ICso obtained from at least three after subtracting the background fluorescence from 1.5 uM thiola- vin T and AChE. The percent inhibition of the AChE-induced Molecular Docking. Molecular simulations were performed on aggregation due to the presence of increasing concentrations of the an R14000 SGI Fuel workstation with software package SYBYL inhibitor was calculated by the following expression: 100-(IF/ 6.9(Tripos Inc, St Louis, MO). Standard parameters were used IFo x 100), where IFi and IFo were the fluorescence intensities nless otherwise indicated. The X-ray crystallographic structures obtained of the mAChE complex with succinylcholine(PDB code 2HA2) inhibitor, respectively, after subtracting the fluorescence of respec and the native hBChE(IPoD)were retrieved from PDB. Het- tive blanks. Inhibition curves and linear regression parameters wer eroatoms and water molecules in the PDb files were removed obtained for each compound, and the ICso was extrapolated, when and all hydrogen atoms were subsequently added to the protein 3D coordinates of the ligand were generated by CORINA 3.0 MTT Assay of Cell Viability. The human neuroblastoma cell Molecular Networks GmbH, Erlangen, Germany, 2004).35Two ure Collection) cells were spN atoms were both protonated, and the Gasteiger-Huckel cultured in MEM/F-12(1: 1) (Invitrogen, Grand Island. partial charges 6, 37 were assigned to each atom of the resultant NY) supplemented with 10%o fetal calf serum(FCS, Invitrogen),(15 mL), drops of aqueous ammonia were added to adjust the pH to 9, and the residue was extracted with CHCl3 (20 mL × 3). The combined CHCl3 was dried with anhydrous Na2SO4 and concentrated in vacuo to give a residue, which was chromato￾graphed on silica gel and eluted with MeOH/CHCl3 (0.5:9.5) to provide 5c,d as an oil. Addition of dry HCl-ether to the solution of 5c,d in dry ether, with pH adjusted to 3-4, gave the final salt 5c,d · 2HCl as a powder. N,N′-(1′,4′-Butylene)-bis-(-)-nor-MEP Hydrochloride (5c · 2HCl). Compound 9c (0.56 g, 1.08 mmol), lithium aluminum hydride (0.20 g, 5.26 mmol), THF (30 mL), H2O (0.20 mL), 15% NaOH (0.20 mL), and H2O (0.60 mL) were used to produce 5c (0.19 g, 36%). Subsequent salt formation gave 5c · 2HCl (0.12 g, 55%): mp 110-115 °C; [R]D -51.96° (c 0.092, MeOH); IR ν 3254, 2936, 1600, 1586, 1448, 1268 cm-1 ; 1 H NMR (DMSO￾d6) 9.98, 9.77 (br s, 4/3 H, NH+), 9.56, 9.54, 9.43, 9.42 (s, 2H, Ar-OH), 8.46 (br s, 2/3 H, NH+), 7.21–7.13 (m, 2H, Ar-H), 6.85–6.65 (m, 6H, Ar-H), 3.83 (m, ∼2/3 H, N-CH2), 3.52 (m, ∼4/3 H, N-CH2), 3.36–3.15 (m, 10H, N-CH2), 2.38 (m, H, CH2), 2.10–1.46 (m, 19H, CH2), 0.49 (t, 6H, J ) 7.0 Hz, CH3); MS (ESI) [M + H]+ 493.3, [M + 2H]2+ 247.2. HRMS m/z calcd for C32H49N2O2 [M + H]+, 493.3791; found, 493.3789. HPLC: tR ) 7.45 min, 98.0% purity. N,N′-(1′,5′-Pentylene)-bis-(-)-nor-MEP Hydrochloride (5d · 2HCl). Compound 9d (0.55 g, 1.03 mmol), lithium aluminum hydride (0.20 g, 5.26 mmol), THF (30 mL), H2O (0.20 mL), 15% NaOH (0.20 mL), and H2O (0.60 mL) were used to produce 5d (0.16 g, 31%). Subsequent salt formation of 5d (0.10 g) gave 5d · 2HCl (0.06 g, 52%): mp 80-82 °C; [R]D -48.10° (c 0.084, MeOH); IR ν 3417, 2937, 1600, 1585, 1447, 1268 cm-1 ; 1 H NMR (DMSO-d6) 9.96, 9.75 (br s, 5/4 H, NH+), 9.57, 9.52, 9.44, 9.42 (s, 2H, Ar-OH), 8.34 (br s, 3/4 H, NH+), 7.19–7.12 (m, 2H, Ar-H), 6.83–6.64 (m, 6H, Ar-H), 3.82 (m, 3/4 H, N-CH2), 3.52 (m, 5/4 H, N-CH2), 3.34–3.06 (m, 10H, N-CH2), 2.37 (m, H, CH2), 2.10–1.21 (m, 21H, CH2), 0.48 (m, 6H, CH3); MS (ESI) [M + H]+ 507.3, [M + 2H]2+ 254.2. HRMS m/z calcd for C33H51N2O2 [M + H]+, 507.3945; found, 507.3961. HPLC: tR ) 7.77 min, 96.1% purity. In Vitro AChE/BChE Inhibition Assay. Inhibitory activity against AChE was evaluated by a modified Ellman’s method.34 Mice brain homogenate prepared in saline was used as a source of AChE; mice serum was the source of BChE. The AChE activity was determined in a reaction mixture containing 200 µL of a solution of AChE (0.415 U/mL in 0.1 M phosphate buffer, pH 8.0), 300 µL of a solution of 5,5′-dithio-bis(2-nitrobenzoic) acid (3.3 mM DTNB in 0.1 M phosphate buffered solution, pH 7.0, containing NaHCO3 6 mM), and 30 µL of a solution of the inhibitor (six to seven concentrations). After incubation for 20 min at 37 °C, acetylthiocholine iodide (300 µL of 0.05 mM water solution) was added as the substrate, and AChE activity was determined by UV spectrophotometry from the absorbance changes at 412 nm for 3 min at 25 °C. The concentration of compound that produced 50% inhibition of the AChE activity (IC50) was calculated by nonlinear regression of the response-concentration (log) curve. BChE inhibitory activity determinations were similarly carried out using butyrylthiocholine iodide (0.05 mM) as the substrate. Results are reported as the mean ( SEM of IC50 obtained from at least three independent measures. Molecular Docking. Molecular simulations were performed on an R14000 SGI Fuel workstation with software package SYBYL 6.9 (Tripos Inc., St. Louis, MO). Standard parameters were used unless otherwise indicated. The X-ray crystallographic structures of the mAChE complex with succinylcholine (PDB code 2HA2)32 and the native hBChE (1P0I)28were retrieved from PDB. Het￾eroatoms and water molecules in the PDB files were removed, and all hydrogen atoms were subsequently added to the protein. 3D coordinates of the ligand were generated by CORINA 3.0 (Molecular Networks GmbH, Erlangen, Germany, 2004).35 Two sp3 N atoms were both protonated, and the Gasteiger-Huckel partial charges36,37 were assigned to each atom of the resultant ligand. The final conformation of the ligand was obtained after 1000 steps of energy minimization using the Tripos force field. Molecular docking was carried out using GOLD 3.0 (CCDC, Cambridge, U.K., 2005)33 to generate an ensemble of docked conformations for the ligand. The active site was defined as all atoms within a radius of 25 Å around some specific residue atoms: OH of Tyr337 for mAChE and N
2 of His438 for hBChE. We chose more enlarged binding pockets here, concerned that smaller sites might neither accommodate a large bis-ligand nor include both the catalytic and peripheral sites of AChE. The 600 genetic algorithm (GA) runs instead of the default 10 were performed owing to the high flexibility of the ligand possessing a great many rotatable bonds. For each GA run, the default GA settings were used except for the prohibited early termination and the permitted pyramidal nitrogen inversion. Advanced combination approach of clustering and consensus scoring was used to guide the selection of the most reliable conformation from among a set of candidate conformations that GOLD generated. All conformations were evaluated with four available scoring functions, including three scoring functions (G_Score, D_Score, and ChemScore) from the CSCORE38 module in SYBYL and another stand-alone scoring function, X-SCORE 1.2.1.39 The “rank-by-rank” strategy reported by Wang et al.40 was adopted to make consensus scoring. The final rank of a certain conformation was its average rank received from all four scoring functions. GOLD generated clusters based on different rmsd (root mean square deviation) criteria. A proper rmsd standard of clustering was important to the selection of representative clusters. Herein, rmsd values of 2.55 and 2.40 Å were chosen as the criteria to cluster the docked conformations with mAChE and hBChE, respectively. Among the top 10 clusters for each ChE, the one with members possessing the minimum average “rerank” was identified as the representative cluster. The top “reranked” conformation in the representative cluster was selected as the representative binding mode for the ligand. Inhibition of AChE-Induced A
Aggregation. Aliquots of 2 µL A
(1–40) peptide (Biosource), lyophilized from 2 mg/mL HFIP solution and dissolved in DMSO, were incubated for 48 h at room temperature in 0.215 M sodium phosphate buffer (pH 8.0) at a final concentration of 230 µM. For coincubation experiments, aliquots (16 µL) of human recombinant AChE (Sigma-Aldrich) (final concentration of 2.3 µM, A
/AChE molar ratio of 100:1) and AChE in the presence of 2 µL of the tested inhibitors were added. Blanks containing A
, AChE, and A
plus inhibitors at various concentra￾tions in 0.215 M sodium phosphate buffer (pH 8.0) were prepared. The final volume of each vial was 20 µL. Each assay was run in duplicate. Inhibitor stock solutions were prepared (c ) 10 mM) and diluted in 0.215 M sodium phosphate buffer (pH 8.0) when used. To quantify amyloid fibril formation, the thioflavin T fluorescence method was then applied.20 After incubation, the samples containing A
, A
plus AChE, or A
plus AChE in the presence of inhibitors were diluted with 50 mM glycine-NaOH buffer (pH 8.5) containing 1.5 µM thioflavin T (Sigma-Aldrich) to a final volume of 2.0 mL. Fluorescence was monitored by PE LS45, with excitation at 446 nm and emission at 490 nm. A time scan of fluorescence was performed, and the intensity values reached at the plateau (around 300 s) were averaged after subtracting the background fluorescence from 1.5 µM thiofla￾vin T and AChE. The percent inhibition of the AChE-induced aggregation due to the presence of increasing concentrations of the inhibitor was calculated by the following expression: 100 - (IFi/ IFo × 100), where IFi and IFo were the fluorescence intensities obtained for A
plus AChE in the presence and in the absence of inhibitor, respectively, after subtracting the fluorescence of respec￾tive blanks. Inhibition curves and linear regression parameters were obtained for each compound, and the IC50 was extrapolated, when possible. MTT Assay of Cell Viability. The human neuroblastoma cell line SH-SY5Y (American Type Culture Collection) cells were cultured in MEM/F-12 (1:1) medium (Invitrogen, Grand Island, NY) supplemented with 10% fetal calf serum (FCS, Invitrogen), 2034 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 7 Xie et al