复旦大学：《药物设计学》课程教学资源（教学研究）基于靶点结构的药物设计_Novel piperazine derivative PMS1339 exhibits tri-functional properties and cognitive improvement in mice

al Journal of Neuropsychopharmacology(2009), 12, 1409-1419. Copyright G CINP 2009 ARTICLE /S1461145709000455 Novel piperazine derivative pms1339 exhibits O CINP tri-functional properties and cognitive improvement in mice Jean Marc Miezan Ezoulin*, Bi-yun Shao*, Zheng Xia*, Qiong Xie, Juan Li, Yong-yao Cui, Hao Wang, Chang-zhi Dong, Yan-xing Zhao, France Massicot, Zhui-bai Qiu, Francoise Heymans and Hong-zhuan Chen' i Department of Pharmacology, Institute of Medical Sciences, Shanghai Jiao Tong LIniversity School of Medicine, Shanghai, P.R. China Unite de Pharmacochimie Moleculare et Systemes Membranaires(EA2381), Laboratoire de plarmacochimie Moleculare Universite Paris 7-Denis diderot, Batiment Lavoisier, Paris, france Department of Medicinal Chemistry, School of Pharmacy, Fudan University, Shanghai, PR China 4 UNiversite Paris Descartes, Faculte des Sciences Pharmaceutiques et Biologiques, Laboratoire de Toxicologie, Paris, France Abstract Amyloid-B-induced neuroinflammation plays a central role in the extensive loss of cholinergic neurons nd cognitive decline in Alzheimers disease. The acetylcholinesterase(AChE)inhibitors are the first class of drugs used to enhance surviving cholinergic activities. However, their limited effectiveness following long-term treatment raises a need for new multi-target therapies. We report herein a novel piperazine derivative compound PMS1339 possesses multifunctional properties including anti-platelet-activating factor, AChE inhibition, AB aggregation cantly inhibit both mice brain AChE (ICs0=4.41+0. 63 uM) and sera butyrylcholinesterase(BuChE, ICso=1.09+0.20 uM). PMS1339 was also found to inhibit neuronal AChE secreted by SH-SY5Y cell line (Co=17.95+2.31 uM). Enzyme kinetics experiments performed on electric eel AChE indicated that PMS1339 acts as a mixed type competitive AChE inhibitor Molecular docking studies using the X-ray crystal structure of AChE from Torpedo californica elucidated the interactions between PMS1339 and ChE: PMS1339 is well buried inside the active-site gorge of AChE interacting with Trp84 at the bottom, Tyr121 halfway down and Trp279 at the peripheral anionic site(PAS). Thioflavin T-based fluorimetric assay revealed the ability of PMS1339 to inhibit AChE-induced Ap aggregation. In-vivo study indicated PMS1339(1 mg/kg i.p. reversed scopolamine-induced memory impairment in mice. Overall, these findings indicated that PMS1339 exhibits tri-functional properties in vitro and cognitive improvement Received 17 February 2009; Reviewed 12 March 2009; Revised 30 March 2009; Accepted 9 April 2009 First published online 22 May 2009 Key words: AChE inhibitor, Alzheimer's disease, amyloid B, PAF antagonist, PMS1339 Introduction Clinically, this neuropathology is characterized by an Alzheimers disease(AD) is a progressive neuro- sidious loss of memory and cognitive abilities in close connection with an impairment of brain cholin- degenerative disorder considered to be the leading ergic neurotransmission(Walsh Selkoe, 2004) cause of dementia affecting the elderly population. Despite enormous research efforts devoted to AD so Author for correspondence: Professor H-Z. Chen, Department of far, an effective treatment seems to be out of reach. The Pharmacology, Institute of Medical Sciences, Shanghai Jiao Tong use of acetylcholinesterase(AChE) inhibitors to en- University School of Medicine, 280 South Chongqing Road, hance surviving cholinergic activity is still the main Shanghai 200025, P.R. China. Tel.:+862164674721Fax:+862164674721 Email:hongzhuan_chen@hotmail.com(H-Z.Chen Email: heymans@ccr jussieu. fr(Professor F Heymans) improving cognition as well as behavioural and func- a These authors contributed equally to this work. tional daily activities but do not stop the diseases

Novel piperazine derivative PMS1339 exhibits tri-functional properties and cognitive improvement in mice Jean Marc Miezan Ezoulin1,2 *, Bi-yun Shao1 *, Zheng Xia1 *, Qiong Xie3 , Juan Li1 , Yong-yao Cui1 , Hao Wang1 , Chang-zhi Dong2 , Yan-xing Zhao1 , France Massicot4 , Zhui-bai Qiu3 , Franc¸oise Heymans2 and Hong-zhuan Chen1 1 Department of Pharmacology, Institute of Medical Sciences, Shanghai Jiao Tong University School of Medicine, Shanghai, P.R. China 2 Unite´ de Pharmacochimie Mole´culaire et Syste`mes Membranaires (EA2381), Laboratoire de Pharmacochimie Mole´culaire, Universite´ Paris 7-Denis Diderot, Baˆtiment Lavoisier, Paris, France 3 Department of Medicinal Chemistry, School of Pharmacy, Fudan University, Shanghai, P.R. China 4 Universite´ Paris Descartes, Faculte´ des Sciences Pharmaceutiques et Biologiques, Laboratoire de Toxicologie, Paris, France Abstract Amyloid-b-induced neuroinflammation plays a central role in the extensive loss of cholinergic neurons and cognitive decline in Alzheimer’s disease. The acetylcholinesterase (AChE) inhibitors are the first class of drugs used to enhance surviving cholinergic activities. However, their limited effectiveness following long-term treatment raises a need for new multi-target therapies. We report herein a novel piperazine derivative compound PMS1339 possesses multifunctional properties including anti-platelet-activating factor, AChE inhibition, Ab aggregation inhibition and cognitive improvement. PMS1339 could signifi- cantly inhibit both mice brain AChE (IC50=4.41¡0.63 mM) and sera butyrylcholinesterase (BuChE, IC50=1.09¡0.20 mM). PMS1339 was also found to inhibit neuronal AChE secreted by SH-SY5Y cell line (IC50=17.95¡2.31 mM). Enzyme kinetics experiments performed on electric eel AChE indicated that PMS1339 acts as a mixed type competitive AChE inhibitor. Molecular docking studies using the X-ray crystal structure of AChE from Torpedo californica elucidated the interactions between PMS1339 and AChE: PMS1339 is well buried inside the active-site gorge of AChE interacting with Trp84 at the bottom, Tyr121 halfway down and Trp279 at the peripheral anionic site (PAS). Thioflavin T-based fluorimetric assay revealed the ability of PMS1339 to inhibit AChE-induced Ab aggregation. In-vivo study indicated PMS1339 (1 mg/kg i.p.) reversed scopolamine-induced memory impairment in mice. Overall, these findings indicated that PMS1339 exhibits tri-functional properties in vitro and cognitive improvement in vivo, and revealed the emergence of a multi-target-directed ligand to tackle the determinants of Alzheimer’s disease. Received 17 February 2009; Reviewed 12 March 2009; Revised 30 March 2009; Accepted 9 April 2009; First published online 22 May 2009 Key words : AChE inhibitor, Alzheimer’s disease, amyloid b, PAF antagonist, PMS1339. Introduction Alzheimer’s disease (AD) is a progressive neuro￾degenerative disorder considered to be the leading cause of dementia affecting the elderly population. Clinically, this neuropathology is characterized by an insidious loss of memory and cognitive abilities in close connection with an impairment of brain cholin￾ergic neurotransmission (Walsh & Selkoe, 2004). Despite enormous research efforts devoted to AD so far, an effective treatment seems to be out of reach. The use of acetylcholinesterase (AChE) inhibitors to en￾hance surviving cholinergic activity is still the main therapy for Alzheimer’s patients. These compounds temporally aim to alleviate their disabilities in terms of improving cognition as well as behavioural and func￾tional daily activities but do not stop the disease’s Author for correspondence : Professor H.-Z. Chen, Department of Pharmacology, Institute of Medical Sciences, Shanghai Jiao Tong University School of Medicine, 280 South Chongqing Road, Shanghai 200025, P.R. China. Tel. : +86 21 6467 4721 Fax : +86 21 6467 4721 Email : hongzhuan_chen@hotmail.com (H.-Z. Chen) Email : heymans@ccr.jussieu.fr (Professor F. Heymans). * These authors contributed equally to this work. International Journal of Neuropsychopharmacology (2009), 12, 1409–1419. Copyright f CINP 2009 doi:10.1017/S1461145709000455 ARTICLE

1410 .M. Mizan Ezoulin et al. progression(Landmark Reikvam, 2008; Onor et al 2007; Raina et aL. 2008). As a result, the classic single. target therapeutic approach solely based on the cholinergic hypothesis appears to be questionable and inappropriate to tackle a complex multifactorial drome such as AD, and other avenues are onsequently being exp A new level of understanding about the molecular Fig. 1. Chemical structure of PMS1339 basis underlying the pathology of Ad has highlighted its complexities and showed, along with loss of chol- with improved AChE inhibitory activity and anti-PAF inergic neurons, the presence of additional hallmarks property, in the present study we describe a new can- involving inflammation and oxidative stress which didate molecule(PM$1339, MW 658.9, Fig. 1),one ere reported to play a key role in Ad pathogenesis of the promising piperazine derivatives synthesized, and progression(Eikelenboom et al. 2006; Garcia- which shares some structural similarities with Alloza et al. 2009; Heneka &z OBanion, 2007: Uryu PMS777. This molecule encompasses the chemical et al. 2002; Yao et aL. 2004). Indeed, it is well established properties required for the dual activities, i.e. the that all signs of inflammatory microglial and astroglial presence of a quinolinium moiety as an AChE catalytic activation are present all around B-amyloid(AB) anionic site-binding element, and a piperazine moiety deposits and alongside axons of neurons with intra- as a PAF receptor antagonism element connected to cellular neurofibrillary tangles. Many converging lines each other by a long methylene tether. The aim of this of evidence suggest that platelet-activating factor work was to further characterize the multifunctional (PAF), a potent pro-inflammatory mediator, is greatly properties of PMS1339 including anti-PAF, AChE implicated in those inflammatory events(Engelberts inhibition, AB aggregation inhibition and cognitive et al. 1991; Lo et al. 1997), particularly in brain injury improvement. (Bate et al. 2004; Bazan et aL. 2002), thus promoting neuronal death and dementing disorders(Bate et al Methods 2006; Hershkowitz Adunsky, 1996). As reported by Bulger et aL.(2002), cytokine reduction occurs in a Animals PAF-dependent manne Along with other investigators(Kwon et aL. 2007, New Zealand White rabbits(2.5-3.0 kg) were pro- Rodriguez- Franco et al. 2006; Rosini et aL. 2005, 2008 ided by the Animal Center of Paris 7 University Yao et al. 2004; Zhu et al. 2009), we hypothesized that Kunming mice (20-30 g) were purchased from a more satisfactory clinical approach might be the Shanghai Jiao Tong University School of Medicine and design of compounds able to act simultaneously at were maintained on a 12-h light/dark cycle(lights on 06: 00 hours)with free access to food and water. All different levels of the neurodegenerative process, studies were performed in accordance with the Guide these compounds should meet the innovative ap- proach'one molecule, multiple targets. In this regard, for the Care and Use of Laboratory Animals as adop- our research focused on the discovery of drugs that ted and promulgated the National Institutes of inhibit both AChE and PAF activities with the purpose Health, USA not only of restoring acetylcholine(ACh) levels in the brain of AD patients but also of preventing the Drugs and chemicals inflammatory process PMS1339 was synthesized in our laboratory; tacrine Previously, we have published the synthesis and(a mixed-mode inhibitor of AChE), the peripheral preliminary in-vitro pharmacology of a series of 25- inhibitor propidium iodide and scopolamine were disubstituted tetrahydrofuran derivatives designed as purchased from Sigma(USA)and used as reference dual PAF and AChE inhibitors (Le Texier et al. 1996). compounds. Acetylthiocholine iodide (AcThCh) Among them, the promising compound PMS777 was S-butyrylthiocholine iodide(BTCh), 5,5-dithio-bis- shown to inhibit AChE activity, reverse scopolamine- 2-nitrobenzoic acid (DTNB), electric eel and re- induced dementia in mouse models, prevent PAF- combinant human AChE were obtained from Sigma induced neurotoxicity(Li et al. 2006)and LPS-induced Ap40 peptide was purchased from Biosource(USA) oxidative/inflammatory disturbances in a human Forebrain AChE and sera butyrylcholinesterase neuroblastoma cell lines(Ezoulin et al. 2005). Con-(BuChE)were obtained from Kunming mice. All other tinuing our efforts to design novel multi-acting drugs reagents were obtained from commercial sources

progression (Landmark & Reikvam, 2008; Onor et al. 2007; Raina et al. 2008). As a result, the classic single￾target therapeutic approach solely based on the cholinergic hypothesis appears to be questionable and inappropriate to tackle a complex multifactorial syndrome such as AD, and other avenues are consequently being explored. A new level of understanding about the molecular basis underlying the pathology of AD has highlighted its complexities and showed, along with loss of chol￾inergic neurons, the presence of additional hallmarks involving inflammation and oxidative stress which were reported to play a key role in AD pathogenesis and progression (Eikelenboom et al. 2006; Garcia￾Alloza et al. 2009; Heneka & O’Banion, 2007; Uryu et al. 2002; Yao et al. 2004). Indeed, it is well established that all signs of inflammatory microglial and astroglial activation are present all around b-amyloid (Ab) deposits and alongside axons of neurons with intra￾cellular neurofibrillary tangles. Many converging lines of evidence suggest that platelet-activating factor (PAF), a potent pro-inflammatory mediator, is greatly implicated in those inflammatory events (Engelberts et al. 1991; Lo et al. 1997), particularly in brain injury (Bate et al. 2004; Bazan et al. 2002), thus promoting neuronal death and dementing disorders (Bate et al. 2006; Hershkowitz & Adunsky, 1996). As reported by Bulger et al. (2002), cytokine reduction occurs in a PAF-dependent manner. Along with other investigators (Kwon et al. 2007; Rodriguez-Franco et al. 2006; Rosini et al. 2005, 2008; Yao et al. 2004; Zhu et al. 2009), we hypothesized that a more satisfactory clinical approach might be the design of compounds able to act simultaneously at different levels of the neurodegenerative process; these compounds should meet the innovative ap￾proach ‘one molecule, multiple targets’. In this regard, our research focused on the discovery of drugs that inhibit both AChE and PAF activities with the purpose not only of restoring acetylcholine (ACh) levels in the brain of AD patients but also of preventing the inflammatory process. Previously, we have published the synthesis and preliminary in-vitro pharmacology of a series of 2,5- disubstituted tetrahydrofuran derivatives designed as dual PAF and AChE inhibitors (Le Texier et al. 1996). Among them, the promising compound PMS777 was shown to inhibit AChE activity, reverse scopolamine￾induced dementia in mouse models, prevent PAF￾induced neurotoxicity (Li et al. 2006) and LPS-induced oxidative/inflammatory disturbances in a human neuroblastoma cell lines (Ezoulin et al. 2005). Con￾tinuing our efforts to design novel multi-acting drugs with improved AChE inhibitory activity and anti-PAF property, in the present study we describe a new can￾didate molecule (PMS1339, MW 658.9, Fig. 1), one of the promising piperazine derivatives synthesized, which shares some structural similarities with PMS777. This molecule encompasses the chemical properties required for the dual activities, i.e. the presence of a quinolinium moiety as an AChE catalytic anionic site-binding element, and a piperazine moiety as a PAF receptor antagonism element connected to each other by a long methylene tether. The aim of this work was to further characterize the multifunctional properties of PMS1339 including anti-PAF, AChE inhibition, Ab aggregation inhibition and cognitive improvement. Methods Animals New Zealand White rabbits (2.5–3.0 kg) were pro￾vided by the Animal Center of Paris 7 University. Kunming mice (20–30 g) were purchased from Shanghai Jiao Tong University School of Medicine and were maintained on a 12-h light/dark cycle (lights on 06:00 hours) with free access to food and water. All studies were performed in accordance with the Guide for the Care and Use of Laboratory Animals as adop￾ted and promulgated by the National Institutes of Health, USA. Drugs and chemicals PMS1339 was synthesized in our laboratory; tacrine (a mixed-mode inhibitor of AChE), the peripheral inhibitor propidium iodide and scopolamine were purchased from Sigma (USA) and used as reference compounds. Acetylthiocholine iodide (AcThCh), S-butyrylthiocholine iodide (BTCh), 5,5k-dithio-bis- 2-nitrobenzoic acid (DTNB), electric eel and re￾combinant human AChE were obtained from Sigma. Ab40 peptide was purchased from Biosource (USA). Forebrain AChE and sera butyrylcholinesterase (BuChE) were obtained from Kunming mice. All other reagents were obtained from commercial sources. N N N N H O (CH2)10 Br Fig. 1. Chemical structure of PMS1339. 1410 J. M. Miezan Ezoulin et al

Multi-target effects of PMS1339 1411 Inhibition of PAF-induced platelet aggregation substrate(AcThCh) concentrations ranging from 0.033 lood from the carotid artery of New Zealand White to 0.167 mM in the absence and in the presence of rabbits was collected directly into 5-ml plastic syringes The k and v. values for AChe were calculated containing 3.8%(w/v)sodium citrate as anticoagulant by regression analysis of Lineweaver-Burk plots to achieve a final volume proportion of 9: 1(blood citrate). Platelet-rich plasma(PRP)was obtained from (1/velocity us. 1/[substrateD the supernatant fraction of blood after centrifugation at room temperature for 10 min at 800 rpm, and the Molecular modelling t-poor plasma was obtained by centrifuging To explore the binding mode of PMS1339 in AChE the remaining sedimentation for 10 min at 3000 rpm. active site, a molecular modelling study was per- The final cell count in PRP was adjusted to4×10° platelets/ml with platelet-poor plasma formed employing the docking protocol GOLD (ones o Platelet aggregation was determined by the optical et al. 1997) on the basis of the X-ray crystallographic method of Born Cross(1963) with an aggregometer structure of Torpedo californica AChE (TCAChE) Molecular simulations were performed on an incubated with vehicle or PMs1339(0. 1, 0. 25, 0.5, 0.75, R14000 SGI Fuel workstation with software package 1 uM)for 5 min at 37C. Then platelet aggregation wa induced by the addition of PAF at a final concentration were used unless otherwise indicated. The coordinates of 0.1 uM. Inhibition of aggregation at 5 min was of TcAChE structure were obtained from the donepezil (E2020)/TCAChE complex [Protein Data Bank(PDB the presence of PMS1339 with that in the presence of Hetero-atoms and water molecules were removed vehicle alone. The average inhibition rate for each concentration was calculated in order to determine from the PDB file, and all hydrogen atoms were added the ICso value( the concentration required to inhibit subsequently. The structure of PMS1339 was gener- platelet aggregation by 50%) ated in SYBYL, and Gasteiger-Huckel partial charges were assigned. The resultant input ligand was ob- Cholinesterase inhibitory actiuity in vitro tained after 1000 steps of energy minimization using ipos force field AChE and BuChE activities were determined accord Molecular docking was performed using GOLD 3.0 ing to a modified Ellman's method (Ellman et al. 1961) (Cambridge Crystallographic Data Centre, UK)(ones using mouse forebrain(AChE)and sera(BuChE)of et al. 1997). The active site was defined as residues with Kunming mice. Briefly, mouse forebrain homogenates at least one atom within a radius of 10 A from any [1: 9 w/v in 0.05 M phosphate buffer solution(PBS)] or atom of E2020. Thirty genetic algorithm(GA)runs sera(1: 19 w/v in 0.05 M PBS)were pre-incubated with were performed. For each GA run, the default GA PMS1339 or tacrine for 20 min at 37C in 0.05 M PBS settings were used. All conformations that GOLD (PH7. 2), containing 0.25 mM DTNB. The substrates, generated were evaluated with the scoring function 0.5 mM AcThCh or BTCh, were then quickly added. X-SCORE 1.2. 1(Wang et al. 2002) The reaction was terminated by the addition of 100 ul eserine (1 mM), and the colour production was measured spectrophotometrically at 412 nm. Inhibition of A ChE-induced Ap aggregation The percent inhibition of the enzyme reaction in AB40 peptide was dissolved from the lyophilized he presence of the inhibitors was determined by powder in hexafluoro-2-iso-propanol (HFIP)to make a comparison with control, and the average inhibition solution of 2 mg/ml HFIP was removed by vapor rate for each concentration was calculated in order ation under vacuum. Then AB was dissolved in an- to deduce the ICso value (the inhibitor concentration hydrous dimethyl sulfoxide(DMSO)to make a 2.3 mM required for 50% inhibition of AChE activity) for each stock solution est dru For each experiment, 2 ul of the 2.3 mM AB-DMSO stock solution was added to a final volume of 20 ul of Determination of kinetic parameters samples containing different concentrations of tested To estimate the kinetic parameters Km (Michaelis con- inhibitors and incubated for 48 h at room temperature. stant), apparent Km(Km, app), Vmax(maximum velocity Ap was diluted from the stock solution into 0. 215 of reaction) and apparent Vmax(V max app), the enzyme PBS(pH8.0)(resulting in 10% residual DMSO) of electric eel AChE was determined at make a final concentration of 230 uM AA

Inhibition of PAF-induced platelet aggregation Blood from the carotid artery of New Zealand White rabbits was collected directly into 5-ml plastic syringes containing 3.8% (w/v) sodium citrate as anticoagulant to achieve a final volume proportion of 9: 1 (blood￾citrate). Platelet-rich plasma (PRP) was obtained from the supernatant fraction of blood after centrifugation at room temperature for 10 min at 800 rpm, and the platelet-poor plasma was obtained by centrifuging the remaining sedimentation for 10 min at 3000 rpm. The final cell count in PRP was adjusted to 4r108 platelets/ml with platelet-poor plasma. Platelet aggregation was determined by the optical method of Born & Cross (1963) with an aggregometer at 37 xC under stirring (900 rpm). PRP was pre￾incubated with vehicle or PMS1339 (0.1, 0.25, 0.5, 0.75, 1 mM) for 5 min at 37 xC. Then platelet aggregation was induced by the addition of PAF at a final concentration of 0.1 mM. Inhibition of aggregation at 5 min was calculated by comparing the extent of aggregation in the presence of PMS1339 with that in the presence of vehicle alone. The average inhibition rate for each concentration was calculated in order to determine the IC50 value (the concentration required to inhibit platelet aggregation by 50%). Cholinesterase inhibitory activity in vitro AChE and BuChE activities were determined accord￾ing to a modified Ellman’s method (Ellman et al. 1961) using mouse forebrain (AChE) and sera (BuChE) of Kunming mice. Briefly, mouse forebrain homogenates [1: 9 w/v in 0.05 M phosphate buffer solution (PBS)] or sera (1: 19 w/v in 0.05 M PBS) were pre-incubated with PMS1339 or tacrine for 20 min at 37 xC in 0.05 M PBS (pH 7.2), containing 0.25 mM DTNB. The substrates, 0.5 mM AcThCh or BTCh, were then quickly added. The reaction was terminated by the addition of 100 ml eserine (1 mM), and the colour production was measured spectrophotometrically at 412 nm. The percent inhibition of the enzyme reaction in the presence of the inhibitors was determined by comparison with control, and the average inhibition rate for each concentration was calculated in order to deduce the IC50 value (the inhibitor concentration required for 50% inhibition of AChE activity) for each test drug. Determination of kinetic parameters To estimate the kinetic parameters Km (Michaelis con￾stant), apparent Km (Km,app), Vmax (maximum velocity of reaction) and apparent Vmax (Vmax,app), the enzyme activity of electric eel AChE was determined at substrate (AcThCh) concentrations ranging from 0.033 to 0.167 mM in the absence and in the presence of two concentrations of PMS1339 (0.33 and 1.33 mM). The Km and Vmax values for AChE were calculated by regression analysis of Lineweaver–Burk plots (1/velocity vs. 1/[substrate]). Molecular modelling To explore the binding mode of PMS1339 in AChE active site, a molecular modelling study was per￾formed employing the docking protocol GOLD (Jones et al. 1997) on the basis of the X-ray crystallographic structure of Torpedo californica AChE (TcAChE). Molecular simulations were performed on an R14000 SGI Fuel workstation with software package SYBYL 6.9 (Tripos Inc., USA). Standard parameters were used unless otherwise indicated. The coordinates of TcAChE structure were obtained from the donepezil (E2020)/TcAChE complex [Protein Data Bank (PDB; Berman et al. 2000) code 1EVE] (Kryger et al. 1999). Hetero-atoms and water molecules were removed from the PDB file, and all hydrogen atoms were added subsequently. The structure of PMS1339 was gener￾ated in SYBYL, and Gasteiger–Hu¨ ckel partial charges were assigned. The resultant input ligand was ob￾tained after 1000 steps of energy minimization using Tripos force field. Molecular docking was performed using GOLD 3.0 (Cambridge Crystallographic Data Centre, UK) (Jones et al. 1997). The active site was defined as residues with at least one atom within a radius of 10 A˚ from any atom of E2020. Thirty genetic algorithm (GA) runs were performed. For each GA run, the default GA settings were used. All conformations that GOLD generated were evaluated with the scoring function X-SCORE 1.2.1 (Wang et al. 2002). Inhibition of AChE-induced Ab aggregation Ab40 peptide was dissolved from the lyophilized powder in hexafluoro-2-iso-propanol (HFIP) to make a solution of 2 mg/ml. HFIP was removed by evapor￾ation under vacuum. Then Ab was dissolved in an￾hydrous dimethyl sulfoxide (DMSO) to make a 2.3 mM stock solution. For each experiment, 2 ml of the 2.3 mM Ab-DMSO stock solution was added to a final volume of 20 ml of samples containing different concentrations of tested inhibitors and incubated for 48 h at room temperature. Ab was diluted from the stock solution into 0.215 M PBS (pH 8.0) (resulting in 10% residual DMSO) to make a final concentration of 230 mM Ab. Multi-target effects of PMS1339 1411

1412 J M. Miezan Ezoulin et al. For co-incubation experiments, aliquots(16 ul)of with a 30-s inter-trial interval. Each day, a trial was human recombinant AChE(Sigma-Aldrich)(final initiated by placing each mouse in the water facing the concentration 2.3 uM, AB/rHuAChE molar ratio 100: 1) pool wall in one of the four quadrants. The daily order in the presence of 2 ul of the tested inhibitors at vari- of entry into individual quadrants was randomized so ous concentrations were added. Blanks containing AB, that all four quadrants were used once every day. The rHuAChE, AB plus inhibitors at various concentra- mouse was allowed 90 s to locate the hidden platform tions in 0. 215M PBS(pH8.0)were prepared. The final When successful, it was allowed a 30-s rest period on volume of each vial was also 20 ul. Inhibitor stock the platform, and the time spent in locating the plat solutions(10 mM) were prepared in methanol form was recorded as escape latency. If unsuccessful Aggregation was followed by thioflavin T-binding within the allotted time period, the mouse was as technique. After incubation, the samples containing signed a latency of 90 s and then physically placed on AB, AB plus rHuAChE or AB plus rHuAChE in the platform and also allowed a 30-s rest period. The the presence of inhibitors were diluted with 50 mM escape latency and swim distance as well as swim glycine-NaoH buffer (pH8.5) containing 1.5uM ths were recorded by the video-tracking system. thioflavin T(Sigma-Aldrich) to a final volume of On day 5, each mouse was subjected to a 60-s probe 2.0 ml. The fluorescence signal was monitored by a trial twice in which the hidden platform had been re- PerkinElmer L$45 fluorometer(excitation at 446 nm moved completely. The mouse was placed in the pool and emission at 490 nm). A time scan of fluorescence to swim for 60 s from the opposite and side position of was performed during 300 s, and the intensity values the target quadrant respectively with a 30-s inter-trial were averaged after subtracting the background flu- interval. Time spent in the target quadrant and the orescence from 1.5 uM thioflavin T and rHuAChE. proportions of swim distance in the target quadrant The percentage inhibition of rHuAChE-induced ag- were determined by the analysis system, as well as gregation due to the presence of increasing con- total distance. Time percent and distance percent in centrations of the inhibitor was calculated by the the target quadrant were calculated, which were taken following expression: 100-(IFi/IFo x 100), where IFi as a measure of spatial memory and IFo are the fluorescence intensities obtained for A Scopolamine (0.5 mg/kg i.p. was administered or AB plus AChE in the presence and in the absence 20 min before the test once a day for all 5 d, and of inhibitor, respectively, after subtracting the fluor- PMS1339(0.1 and 1 mg/kg i p. was injected 10 min escence of respective blanks. Inhibition curves and prior to scopolamine administration linear r com- pounds and the ICso was extrapolated, when possible. Statistical analysis Data were expressed as mean+sE M. Statistical sig In-vivo assessment of spatial memory nificance of the observed differences was assessed by The ability of PMS1339 to reverse scopolamine one-way analysis of variance(ANOVA). The statistical induced spatial memory and recall deficits was eval- significance level was set at p<0.05 or p<0.01 uated in the Morris water maze as previously described (Terry, 2001). Briefly, a large black circular pool (140 cm diameter) filled with water to depth of 30 cm(maintained at 22+1.0C). The pool Inhibition of PAF-induced platelet aggregation was divided into four equal quadrants, and a black platform(9 cm diameter)was submerged - cm be- In vitro, 0.1 M PAF markedly induced rabbit platelet low the surface of water in the centre of one quadrant aggregation. PMS1339 significantly inhibited PAF- (target quadrant). The platform was made invisible induced platelet aggregation to the mice and remained in one location for the dependent manner( Fig. 2), with an ICso value of entire test. A high-resolution exview HAD camera 332+44 nM (n=3). Tacrine, at the same concentra- (Shenzhen Hong Tianzhi Electronics Co Ltd, China) tions, failed to exhibit any inhibition against PAF- vas suspended over the centre of the pool, its images induced platelet aggregation( data not shown) being monitored by a video-tracking system(Morris Water Maze Video Analysis System(DigBeh-MM) Cholinesterase inhibitory activities in vitro Shanghai Jiliang Software Technology Co Ltd, China). PMS1339 inhibited forebrain AChE and sera BuChE The male Kunming mice weighing 20-30g were activities in mice in a concentration-dependent given four trials per day for four consecutive days, manner with an ICso value of 4.41+0.63 uM and

For co-incubation experiments, aliquots (16 ml) of human recombinant AChE (Sigma-Aldrich) (final concentration 2.3 mM, Ab/rHuAChE molar ratio 100: 1) in the presence of 2 ml of the tested inhibitors at vari￾ous concentrations were added. Blanks containing Ab, rHuAChE, Ab plus inhibitors at various concentra￾tions in 0.215 M PBS (pH 8.0) were prepared. The final volume of each vial was also 20 ml. Inhibitor stock solutions (10 mM) were prepared in methanol. Aggregation was followed by thioflavin T-binding technique. After incubation, the samples containing Ab, Ab plus rHuAChE or Ab plus rHuAChE in the presence of inhibitors were diluted with 50 mM glycine-NaOH buffer (pH 8.5) containing 1.5 mM thioflavin T (Sigma-Aldrich) to a final volume of 2.0 ml. The fluorescence signal was monitored by a PerkinElmer LS45 fluorometer (excitation at 446 nm and emission at 490 nm). A time scan of fluorescence was performed during 300 s, and the intensity values were averaged after subtracting the background flu￾orescence from 1.5 mM thioflavin T and rHuAChE. The percentage inhibition of rHuAChE-induced ag￾gregation due to the presence of increasing con￾centrations of the inhibitor was calculated by the following expression: 100 – (IFi/IFor100), where IFi and IFo are the fluorescence intensities obtained for Ab or Ab plus AChE in the presence and in the absence of inhibitor, respectively, after subtracting the fluor￾escence of respective blanks. Inhibition curves and linear regression parameters were obtained for com￾pounds and the IC50 was extrapolated, when possible. In-vivo assessment of spatial memory The ability of PMS1339 to reverse scopolamine￾induced spatial memory and recall deficits was eval￾uated in the Morris water maze as previously described (Terry, 2001). Briefly, a large black circular pool (140 cm diameter) was filled with water to a depth of 30 cm (maintained at 22¡1.0 xC). The pool was divided into four equal quadrants, and a black platform (9 cm diameter) was submerged y1 cm be￾low the surface of water in the centre of one quadrant (target quadrant). The platform was made invisible to the mice and remained in one location for the entire test. A high-resolution exview HAD camera (Shenzhen Hong Tianzhi Electronics Co. Ltd, China) was suspended over the centre of the pool, its images being monitored by a video-tracking system (Morris Water Maze Video Analysis System (DigBeh-MM), Shanghai Jiliang Software Technology Co. Ltd, China). The male Kunming mice weighing 20–30 g were given four trials per day for four consecutive days, with a 30-s inter-trial interval. Each day, a trial was initiated by placing each mouse in the water facing the pool wall in one of the four quadrants. The daily order of entry into individual quadrants was randomized so that all four quadrants were used once every day. The mouse was allowed 90 s to locate the hidden platform. When successful, it was allowed a 30-s rest period on the platform, and the time spent in locating the plat￾form was recorded as escape latency. If unsuccessful within the allotted time period, the mouse was as￾signed a latency of 90 s and then physically placed on the platform and also allowed a 30-s rest period. The escape latency and swim distance as well as swim paths were recorded by the video-tracking system. On day 5, each mouse was subjected to a 60-s probe trial twice in which the hidden platform had been re￾moved completely. The mouse was placed in the pool to swim for 60 s from the opposite and side position of the target quadrant respectively with a 30-s inter-trial interval. Time spent in the target quadrant and the proportions of swim distance in the target quadrant were determined by the analysis system, as well as total distance. Time percent and distance percent in the target quadrant were calculated, which were taken as a measure of spatial memory. Scopolamine (0.5 mg/kg i.p.) was administered 20 min before the test once a day for all 5 d, and PMS1339 (0.1 and 1 mg/kg i.p.) was injected 10 min prior to scopolamine administration. Statistical analysis Data were expressed as mean¡S.E.M. Statistical sig￾nificance of the observed differences was assessed by one-way analysis of variance (ANOVA). The statistical significance level was set at p<0.05 or p<0.01. Results Inhibition of PAF-induced platelet aggregation In vitro, 0.1 mM PAF markedly induced rabbit platelet aggregation. PMS1339 significantly inhibited PAF￾induced platelet aggregation in a concentration￾dependent manner (Fig. 2), with an IC50 value of 332¡44 nM (n=3). Tacrine, at the same concentra￾tions, failed to exhibit any inhibition against PAF￾induced platelet aggregation (data not shown). Cholinesterase inhibitory activities in vitro PMS1339 inhibited forebrain AChE and sera BuChE activities in mice in a concentration-dependent manner with an IC50 value of 4.41¡0.63 mM and 1412 J. M. Miezan Ezoulin et al

Multi-target effects of PMS1339 1413 with a proper conformation in the gorge and favour- ably interacted with some hydrophobic residues. A hydrogen bond was observed between the amide nitrogen atom of PMS1339 and the hydroxyl hydrogen of Tyr121 side-chain located at the middle of the gorge. Distance between the hydrogen bond donor was 2.61A Inhibition of AChE-induced Ap aggregation The preparation of the initial solutions of AB peptides 0.00 0.25 0.50 0.75 1.00 is critical for the test and is meant to prevent any [PMS 1339](u/M) aggregation. HFIP alters the biological activity of AB Fig. 2. Effects of PMS1339 on platelet-activating factor(PAF)- by acting as a denaturant of proteins, breaking their induced platelet aggregation in vitro. The platelet-rich plasma secondary and tertiary structures and inducing was incubated with vehicle or PMS1339 and then stimulated a-helices. In 100% DMSO, AB appears to be mono- with PAF (O1 uM). The values shown represent the meric(Snyder et aL. 1994)and has no B-sheet character meantsEM of three independent experiments Once diluted in PBS, the first step in AB self-assembly is the formation of B-sheet dimers from monomers. 1.09+0.20 (M, respectively. PMS1339 also showed a The residual DMSo has no significant effect on fibril low degree of selectivity for BuChE us. AChE (Table 1) In contrast, tacrine was more selective for BuChE than growth(Shen& Murphy, 1995) Our search for AB aggregation inhibition was for AChE prompted by the ability of PMS1339 to interact with Determination of kinetic parameters the AChe peripheral anionic site that is well-known to induce AB aggregation. PMS1339 was found to inhibit Enzymatic analysis of AcThCh hydrolysis by electric rHuAChE catalytic activity with an ICso value of eel AChe yielded the following kinetic values: Km 74.7+6.3 nM(n=6). The effects of PMS1339 have been (28.35+3.00 uM) and Vmax(94.32+2.60 AM/min). At compared to tacrine, a compound known as a mixed concentrations 0.33 and 1.33 uM, PMS1339 increased type AChE inhibitor, and also to propidium iodine, a the Km value by 16% and 152%, respectively, and re- purely non-competitive AChE inhibitor binding to the duced Vmax by 25 % and 66%, respectively (Table 2) peripheral anionic site. The analysis of linear transformation of the We first measured in-vitro rHuAChE-induced AB Michaelis-Menten equation(Lineweaver-Burk)(Fig 3) aggregation and it was shown that rhuAChE showed that as inhibitory concentrations increased, promotes AB aggregation by 155%. These data were PMS1339 increased both the slopes and the intercept results from two independent experiments. Control with the x(1/[AcThChI and y(1/Vmax)axes. measurements showed PMS1339 and propidium exhibit no significant absorption at the wavelength Molecular modelling of thioflavin excitation and have no direct effects on PMS1339 simultaneous ly contacted both the catalytic 48h incubation, PMS1339 significantly inhibited bulky diphenylmethyl and piperazine moieties buried rHuAChE-induced Ap aggregation displaying an apparent ICso value of 45.1 uM. At 200 uM, PMS1339 deep into the catalytic site. One of the phenyl groups showed an inhibition of 68.6%(Fig. 5). Propidium contacted to the side-chain of His440 and Phe331 by edge-to-edge hydrophobic interaction. The other displayed an apparent ICse value of 12.8 uM with an inhibition of 98% formed face-to-face hydrophobic interaction with the the same concentrations, did not show a significant indole ring of Trp84. The nearest hydrophobic contact was 3. 19A inhibitory activity against the rHuAChE-induced AB Both the quinolinium moiety and the long methyl aggregation(data not shown) ene tether were involved in the interactions between PMS1339 and AChE, including cation- interaction Reversal of spatial memory deficits in Morris wvater and T-T stack interaction. Quinolinium group, the polar cationic head, of PMS1339 extended to the outer During the first four consecutive training days, the PAS centred by Trp279. The methylene tether folded escape latency and swim distance for animals to reach

1.09¡0.20 mM, respectively. PMS1339 also showed a low degree of selectivity for BuChE vs. AChE (Table 1). In contrast, tacrine was more selective for BuChE than for AChE. Determination of kinetic parameters Enzymatic analysis of AcThCh hydrolysis by electric eel AChE yielded the following kinetic values: Km (28.35¡3.00 mM) and Vmax (94.32¡2.60 mM/min). At concentrations 0.33 and 1.33 mM, PMS1339 increased the Km value by 16% and 152%, respectively, and re￾duced Vmax by 25% and 66%, respectively (Table 2). The analysis of linear transformation of the Michaelis–Menten equation (Lineweaver–Burk) (Fig. 3) showed that as inhibitory concentrations increased, PMS1339 increased both the slopes and the intercept with the x (1/[AcThCh]) and y (1/Vmax) axes. Molecular modelling PMS1339 simultaneously contacted both the catalytic and peripheral sites of AChE, as shown in Fig. 4. The bulky diphenylmethyl and piperazine moieties buried deep into the catalytic site. One of the phenyl groups contacted to the side-chain of His440 and Phe331 by edge-to-edge hydrophobic interaction. The other formed face-to-face hydrophobic interaction with the indole ring of Trp84. The nearest hydrophobic contact was 3.19 A˚ . Both the quinolinium moiety and the long methyl￾ene tether were involved in the interactions between PMS1339 and AChE, including cation–p interaction and p–p stack interaction. Quinolinium group, the polar cationic head, of PMS1339 extended to the outer PAS centred by Trp279. The methylene tether folded with a proper conformation in the gorge and favour￾ably interacted with some hydrophobic residues. A hydrogen bond was observed between the amide nitrogen atom of PMS1339 and the hydroxyl hydrogen of Tyr121 side-chain located at the middle of the gorge. Distance between the hydrogen bond acceptor and donor was 2.61 A˚ Inhibition of AChE-induced Ab aggregation The preparation of the initial solutions of Ab peptides is critical for the test and is meant to prevent any aggregation. HFIP alters the biological activity of Ab by acting as a denaturant of proteins, breaking their secondary and tertiary structures and inducing a-helices. In 100% DMSO, Ab appears to be mono￾meric (Snyder et al. 1994) and has no b-sheet character. Once diluted in PBS, the first step in Ab self-assembly is the formation of b-sheet dimers from monomers. The residual DMSO has no significant effect on fibril growth (Shen & Murphy, 1995). Our search for Ab aggregation inhibition was prompted by the ability of PMS1339 to interact with the AChE peripheral anionic site that is well-known to induce Ab aggregation. PMS1339 was found to inhibit rHuAChE catalytic activity with an IC50 value of 74.7¡6.3 nM (n=6). The effects of PMS1339 have been compared to tacrine, a compound known as a mixed type AChE inhibitor, and also to propidium iodine, a purely non-competitive AChE inhibitor binding to the peripheral anionic site. We first measured in-vitro rHuAChE-induced Ab aggregation and it was shown that rHuAChE promotes Ab aggregation by 155%. These data were results from two independent experiments. Control measurements showed PMS1339 and propidium exhibit no significant absorption at the wavelength of thioflavin excitation and have no direct effects on the fluorescence of this dye (data not shown). After 48 h incubation, PMS1339 significantly inhibited rHuAChE-induced Ab aggregation displaying an apparent IC50 value of 45.1 mM. At 200 mM, PMS1339 showed an inhibition of 68.6% (Fig. 5). Propidium displayed an apparent IC50 value of 12.8 mM with an inhibition of 98% at 200 mM. Conversely, tacrine, at the same concentrations, did not show a significant inhibitory activity against the rHuAChE-induced Ab aggregation (data not shown). Reversal of spatial memory deficits in Morris water maze During the first four consecutive training days, the escape latency and swim distance for animals to reach 100 80 60 40 20 0 Aggregation (%) 0.00 0.25 0.50 0.75 1.00 [PMS 1339] (µM) Fig. 2. Effects of PMS1339 on platelet-activating factor (PAF)- induced platelet aggregation in vitro. The platelet-rich plasma was incubated with vehicle or PMS1339 and then stimulated with PAF (0.1 mM). The values shown represent the mean¡S.E.M. of three independent experiments. Multi-target effects of PMS1339 1413