In Silico, in Vitro and in Vivo Investigation of a Newly Synthesized Ionic Compound As Anti-Alzheimer Multi-Target Dual AChE/BuChE Inhibitor Possessing Neuro-Protective Effects Against Aβ Induced Neurotoxicity in PC12 Cells and Alzheimer Rat Model

Multi-target anti Alzheimer’s disease (AD) compounds are promising leads for the development of AD modifying agents. Ionic compounds containing quaternary ammonium moiety were synthesized and their multi-targeted anti-AD effects were examined in the current study. Compound 5g possessed suitable aqueous solubility and cell toxicity. It also showed non-competitive dual hAChE/hBuChE inhibition activity. Compound 5g reversed the Aβ-treated PC12 cells’ morphology alteration and reduced PC12 cells’ death. Compound 5g possessed anti-oxidative stress activity through anti-oxidant, anti-ROS production and anti-lipid peroxidation mechanisms. It also reduced the expression of IL-1β and TNF-α genes. Furthermore, compound 5g LDH inhibition, reduction of neuroinammation and prevention of autophagy-apoptosis were approved by the results of in vitro studies. Compound 5g delivery to brain was conrmed by in vivo studies. Administration of compound 5g to Aβ-induced AD rat models improved their cognition function and spatial memory learning behavior. TNF-α and NFkB down-regulated in compound 5g treated AD rats’ hippocamp. Besides, compound 5g reversed the up-regulation of AChE in Aβ treated rats’ hippocamp. Molecular modeling studies conrmed the interaction of compound 5g with both steric and catalytic sites of ChE enzymes. The newly synthesized quaternary ammonium containing derivative (compound 5g) possessed multi-target anti-AD ecacy based on in vitro and in vivo studies and its ecacy in AD rat models were approved by behavioral and molecular investigations. The newly synthesized compound 5g could be a start point for ionic multi-target anti-AD drugs. moiety was investigated in the current study by evaluating its biologic activity toward a number of AD-related targets. Cytotoxicity of the selected inhibitors studied on MCF7 and B16 cell lines using MTT assay [36]. The neuroprotective effect against Aβ induced PC12 cell death was studied. The expression of pro-inammatory cytokines in PC12 cells and in the rat brain were also investigated. Furthermore, in vivo, studies were conducted to evaluate the ecacy of selected compounds on the AD rat model. kit (HITACHI / to do this the proper dilution of the tissue lysates supernatant were subjected to the reagent which contains Cholinesterase enzymes to which 3-to the (Fe(CN) ) . The results could be analyzed using colorimetric assays.


Introduction
Alzheimer's disease (AD) is a chronic neurodegenerative disease known mainly by losing of short-term memory [1]. The rapid growth of AD necessitates the acceleration of therapeutics discovery and development [2][3][4][5]. Neuronal misfunctions and their gradual death that results in memory loss and cognitive decline are the main characteristics of AD. The dynamic loss of synaptic neurons results in the atrophy in the hippocampus and frontal and tempo-parietal cortex [6,7]. The most speci c symptoms of AD are the accumulation of amyloid-beta (Aβ) plaques around neurons, hyperphosphorylation of microtubules associated Tau protein, which in turn leads to the intracellular neuro brillary tangles (NFT), and the reduction of acetylcholine (ACh) level in the synaptic cleft [8][9][10][11][12]. While the pathogenesis of AD has not been understood well yet, but the interacting cascade of elements presents within a network that its core is the accumulation of Aβ and aggregation of hyperphosphorylated Tau protein. Disintegration of axons, degenerative dysfunction of synapsis, neuroin ammation, dysregulation of membranes, and dysfunction of brain metabolic pathways has been reported[6, 7,13]. Oxidative stress and free radical formation, metal dyshomeostasis, excitotoxic processes beyond a large number of AD-associated genes are known as events with the effect on AD progression rate [12].
ACh transmission was studied well in AD. ACh enhancement improves cognitive and behavioral symptoms [14]. The current FDAapproved drugs are anti-cholinesterase (ChE) drugs (donepezil, rivastigmine, and Galantamine) [15] or NMDA antagonists Positive charge centers such as quaternary ammonium facilitates the interaction of AChE/BuChE inhibitors with the binding site, while the lower penetration of the permanent charged quaternary ammonium from the blood-brain barrier [22] limited their application as functional groups in such drugs. Some permanently charged quaternary ammonium containing molecules were designed and investigated mostly as a peripheral inhibitor of ChE, which has the potential application in myasthenia gravis rather than AD [23,24]. Recent investigations suggested an important role for the choline transporter system in the physiologic action of hAChE. It was also introduced as a promising route for the delivery of positively charged drugs such as drugs with quaternary ammonium moieties to the CNS [25,26]. The available evidence for the transporting of donepezil to the CNS by choline transporter [27,28] and the well-known capability of choline transporters as a vector for permanently charged compounds' delivery to the brain [29] are encouraging researchers to take advantage of this drug delivery rout in newly designed AD drugs by incorporating choline transporter binding moieties [30].
Some researchers introduced a number of derivatives with permanently charged quaternary ammonium moieties substituted in para position of the benzene and investigated the probable improvement of dual-site inhibition of hAChE and hBuChE [31], improving water solubility [32] and at the same time to increase the choline transporter binding capability and consequent potential delivery by improving the steric surrounding of the cationic nitrogen [27]. Some piperazenium derivatives and evaluate their biological activities against AD in our recent paper [33], 25 . Two newly designed derivatives, including quaternary ammonium substitution ( gure 2) along with other derivatives from our previous paper [34], [35] were investigated for their dual AChE/BuChE inhibition activity.
The SAR of the synthesized molecules toward hAChE and hBuChE inhibitory activity was investigated using molecular docking and pharmacophore mapping methods. The anti-AD e cacies of the most potent compound with quaternary ammonium moiety was investigated in the current study by evaluating its biologic activity toward a number of AD-related targets. Cytotoxicity of the selected inhibitors studied on MCF7 and B16 cell lines using MTT assay [36]. The neuroprotective effect against Aβ induced PC12 cell death was studied. The expression of pro-in ammatory cytokines in PC12 cells and in the rat brain were also investigated. Furthermore, in vivo, studies were conducted to evaluate the e cacy of selected compounds on the AD rat model. The details of derivatives synthesizing and molecular structure characterizing were reported in our previous papers [34,35].

Extraction of hAChE and hBuChE from human blood
The obtaining of the blood samples from a healthy volunteer was approved by local ethics (IR.TBZMED.REC.868). The blood sample (2.0 mL) was collected in a heparinized tube and mixed with 8.0 mL of sodium phosphate buffer (0.1 M, pH 7.4)). The mixture was centrifuged (5 minutes at 3000 g) and washed with 2-3 volumes of normal saline. After three times of centrifugation and washing, 0.1 ml of the packed RBC was transferred to another tube. To obtain the hemolysate, 6.0 mL of distilled water was added to RBC, and the mixture was centrifuged (1000 g) after 5 minutes of incubation in ice. Aliquots of the erythrocyte membranes were stored at -20˚C until use [64,65]. ChE activity was measured before analysis.

3. hAChE and hBuChE inhibition assay
Modi ed Ellman's method was performed for the measurement of ChE activity [66]. All synthesized compounds, donepezil and galantamine, were weighed and dissolved in DMSO (to prepare 10 -3 M stock solutions). For the preparation of the working standards (10 -4 -10 -9 M), stock solutions diluted in 100 mM phosphate buffer (pH 8.0). The nal concentration of DMSO in the working solutions was less than 1.3 (V/V %).
To investigate the anti-cholinesterase activity of compounds, 550 μl phosphate buffer solution (100 mM, pH 8.0), 150 μl DTNB solution (0.5 mM), 150 μl of the studied compound solution and 150 μl ATCI solution (1.0 mM) (for hAChE) were added to test cuvette (1 mL). The mixture was gently shacked and incubated at 37˚C for 5 min. The reaction was begun by the addition of 50 μl of the hemolysate to the test cuvette (total volume of the mixture was 1050 μL). Absorbance was monitored at the wavelength of 412 nm over 10 min. The same method described was applied for hBuChE Inhibition assay, while instead of ATCI, BuTCI was used. Before docking, co-crystallized inhibitors and unnecessary water molecules were extracted from the crystal structure, and hydrogens were added using the protein preparation module of gold software. The binding mode of donepezil was studied using Ligandscout software, and the binding pocket, as well as pharmacophores, were calculated. The binding cavity was designed according to the binding mode of these compounds to the enzymes and the obtained information from literature about the important amino acids for interaction in the binding gorge. Studied molecules were docked against the desired enzymes using the gold-score method which followed by a chemscore re-docking procedure. The best conformer was visualized using ligandscout software.
To validate the developed docking method, the co-crystallized compounds i.e. donepezil, and buthyrilethiocholine, respectively, for 4EY7 and 4BDS, were removed, and the drawn structures were docked against the designed binding cavity. The resulted conformers were superimposed onto the original co-crystallized compound, and the orientation and mode of interaction were investigated. The obtained RMSD values were regarded as error criteria of the docking procedure. The results showed that the developed docking method was done by RMSD values of less than 2.0 A .

Cell toxicity assay (MTT assay)
Toxicity evaluation of synthetic compounds that showed acceptable inhibitory effects was carried out on cell lines B 16 and MCF 7 using the MTT assay. MCF7 is a breast cancer cell line obtained from the Pasteur Institute of Iran, and the results could be regarded as the ability of the tested compounds cytotoxic activity against cancerous cells. B16 is a murine tumor cell line which used for research as a model for human skin cancer cell line obtained from Pasteur Institute of Iran.
MTT assay is a colorimetric method to evaluate the viability of cells. The reduction of MTT in the mitochondria of living cells leads to the production of a formazan product with purple or violet color, which is not soluble in water [36].
For the MTT method, the cells were passaged in RPMI 1640 with 10% FBS (Gibco, Grand Island, NY, USA). After cell culture, cells were separated with trypsin-EDTA of the ask. (Gibco, Grand Island, NY, USA). 200 microliters of a cell suspension containing 10 4 cells per well of 96-well plates load, and for 24 hours at 37 ° C, the concentration of 5% CO2 and appropriate humidity were pre-incubated. After this time, cells were exposed to different concentrations of compounds and MTT (Sigma-Aldrich) solution with 5mg/ml concentration, for 48 and 4 hours were placed respectively. Then 200 microliters of DMSO (Merck-Germany) and 25 ml Sorensen's buffer as reaction stopper were added to each plate. The plate was located on a plate stirring device after 40 minutes. Absorption at 570 nm was measured by Elisa plate reader. The data were analyzed by Graphpad prism7 software. For validation of this method, positive control (culture medium and cells) and negative control (culture medium, cells, and DMSO) were placed.
Rat pheochromocytoma cells PC12 were used to investigate the neuroprotective effect of compound 5g on neurotoxicity. The culture medium was RPMI 1640 medium that was contained 10% horse serum, 5% fetal calf serum, and gentamicin 50 mg/ml. plastic asks were coated with 0.03% poly-L-ornithine and were used for the incubation of monolayer cultures (density = 0.1 to 0.3× 10 6 cells/cm 2 ) in a 95% air, 5% CO and humidi ed atmosphere at 37ºC. The culture medium was refreshed every 48 h. PC12 cells that were grown on 10-cm dishes were exposed to 5 µM Aβ (1-42) (rat/mouse, ab120959) in the presence or the absence of compound 5g at two concentrations (0.1 and 1.0 µM).

4.3.2.DCFH-DA assay
2′,7′-dichloro uorescein-diacetate (DCFH-DA) was utilized to measure the ROS level. To perform DCFH-DA assay, 10 µM of it was added to the cells of studied groups. To be sure that the DCFH-DA was incorporated into all membrane-bound vesicles the DCFH-DA treated cells were incubated for 15 min at room temperature. This incubation time is enough for esterase to cleave di-acetate group. Then the DAPI counterstained nuclei of the cells were observed under a uorescent microscope. The excitation wavelength was 358 nm and the emission intensity was recorded at 461 nm. The emission of DCF at 530 nm was measured after excitation at 485 nm using a plate reader. Thermoscinti c uorescence Multiscan spectrophotometer (Finland) was used for all ourimetric studies.

Tunnel assay
For performing theTUNEL assay, control and treated cells were transferred to slides and exposed to PBS buffer (containg 4% paraformaldehyde, pH 7.4) for 1 h at room temperature. The obtained slides were rinsed with PBS and incubated in blocking solution (methanol 3% in H 2 O 2 ) for 10 minutes. Then they incubated with 0.1% Triton X-100 in 0.1% sodium Citra for 2 minutes in ice. after washing with PBSthe slides were placed in a wet pan and 50 µl of the reaction mixture was added to each sample and then incubated at 37 ºC for 60 minutes. The nuclei of cells were stained with DAPI and were further studied according to the procedure explained in previous section.

TAOC assay
An Assay Kit for measuring the Total Antioxidant Capacity (T-AOC) (Elabscience Biotechnology Inc.) was used to measure the TAOC of the compound 5g. The applied method is based on the inhibition of ABTS oxidation to the ABTS+ by the studied compound. The concentration of the ABTS+ can be obtained by measuring the absorbance of ABTS+ at 414 nm. The studied cells were collected, and after the addition of cold PBS, homogenized using an ultrasonic bath. The supernatant was transferred to a proper tube, and the reagent was added to the solutions. The absorbance intensities and the total protein concentration were recorded for all samples.
The calculations were performed based on the kit provider guideline.

MDA assay
Biocore Diagnostik (ZellBio) MDA assay Kit was used for the assessment of lipid peroxidation. The method is based on the formation of MDA (Malondialdehyde) -TBA (thiobarbituric acid) adduct from the MDA-TBA reaction at high temperature. Then MDA is measured in hot (90-100°C) acidic media using colorimetric measurements at 532 (530-540 nm). In summary, after the preparation of the cell homogenate, the reagent was added to the samples, and the results were recorded and manipulated, as noted by the kit provider.

LDH assay
LDH release was measured using a CytoSelect LDH Assay Kit (Cell BioLabs, San Diego, CA) based on the standard protocol. PC12 cells were placed in the 96-well plates (1×104 cells/well) and cultivated 24 h prior to experiment performing. The cells were treated with Aβ1-42 (20 µM), with or without various concentrations donepezil and compound 5g for 48 h. Then 90 µl of the supernatant was transferred to a well containing 10 µl LDH reagent and incubated for 30 min. The optical density was recorded at 450 nm using Epoch™ spectrophotometer (BioTek Instrument).

IL-1β and TNF-α genes expression analysis:
SYBR Green method based real-time PCR was used to the measurement of mRNA level of the IL-1β and TNF-α genes in PC12 cells. Total RNA (SinaClon, Tehran, Iran) was isolated using RNX-PLUS solution. Revert Aid Reverse Transcriptase Kit (Thermo Fisher, Waltham, MA, USA) was used to the synthesis of the complementary DNA (cDNA). β-actin was utilized as an endogenous control for mRNA expression. The relative amount of gene expressions was calculated using 2−ΔΔCT method. PC12 cells were treated with Aβ (5µM) and the level of the IL-1β and TNF-α genes were investigated in the presence and absence of the donepezil and compound 5g.
. In vivo studies 6.1. Animals 60 adult male Wistar rat with the age of eight weeks and weight of 250 -280 g were used for animal studied. Standard cages that were exposed to 12/12 h light/dark cycle at 23 ± 2 °C temperature were used to keeping the animals. the access to the food was ad libitum. All experiments were done based on the Guide for the Care and Use of Laboratory Animals of the National Institute of Health (NIH; Publication No. 85-23, revised 1985). This study was ethically approved by Health Ministry of the Islamic Republic of Iran under the grant number of (NO: 97015588) and the obtaining of the blood samples from a healthy volunteer was approved by local ethics (IR.TBZMED.REC.868). An infusion pump was used to infuse Aβ 1-42 (50 pmol per animal) (10.3390/molecules22112007). The same protocol was utilized for Sham-operated animals, except for the injection of vehicle instead of all rats were individually housed following 1 day of surgery then returned to the initial care state.
The rats were divided in 6 different groups (n=10). Animals in sham surgery and AD groups received 1 ml/kg/day normal salin via oral route. Other animals in AD groups received donepezil (2 mg/kg/day) or compound 5g (1, 2 or 4 mg/kg/day) via similar route for 21 days.

Morris water maze (MWM) protocol
To perform MWM studies we used an apparatus in which the black circular pool (120 cm diameter and 60 cm height) of it lled with water (24-25 C, depth 30 cm). the pool was equipped with a 10 cm diameter submerged transparent escape platform that was placed 1 cm above (visible platform) or 1 cm below (hidden platform) the water level. Then, the pool was divided into four hypothetical quadrants for test start positions. Data were recorded using a video camera that was connected to a computerized tracking system (EthoVision XT). The camera was placed above the pool center.

Visible platform
The hidden platform was begun 1 day before the visual version of the water maze. To perform this experiment the platform was marked by beacon and the rat was allowed to nd the platform spontaneously within 60 s and left on the platform for 10 s. in the case that platform was not found by rat, it was rmly directed to the platform by investigator. The animals' performance on the visual platform task con rmed their su cient vision and the motor and mental competence to do the task.

Hidden platform
The working memory could be evaluated using hidden platform. The hidden platform was 1 cm below the water surface. For the spatial learning (acquisition and probe trials) task, in each trial, the time spent by the rat to nd the platform was recorded as an escape latency time.

Probe test
Reference memory was evaluated a probe test was performed. To do this on the next day after the last learning trial, a single probe trial was performed in which animals were allowed to freely swimming without platform for 60 s. The analysis of probe trial shows the index of memory, in which the time spent in the platform quadrant was noted.

Novel object recognition (NOR) Test
A NOR test was done in 3 sessions (habituation, training, and retention) with moderate modi cations. The Plexiglassopen-eld box (50 ×50 ×30 cm) and regular objects with various shape and textures were utilized. The rat nose direction to the object (distance of ≤2 cm) and rearing up against the object was noted as exploration. The total locomotor activity (in habituation session) and the spent time with each object was recorded using video camera and scored using EthoVision XT video tracking software.
After the placing of the rats in the box in the absence of objects for 10 min for habituation. They were trained using 2 identical objects (A and A'). The total time spent to explore objects was recorded. In the retention session, the animals was returned to the same task by replacing of the one of the objects with a new object named B.
Discrimination index (DI) was calculated and utilized as an index of the recognition memory through following equation.
where N stands for the time that was spent to explore the new object and F is the time that was spent to explore the familiar object in retention sessions.

Rat brain tissue sampling
After the nalization of the behavioral test brain tissues were removed for enzyme activity and cytokine level investigations. Prior to the removing of brain tissue, deeply anesthetized animals following by i.p. injection of ketamine/xylazine (90/10 mg/kg) and decapitated. As soon as the brain tissue was removed it transferred on ice the hippocampal tissue was isolated. The isolated hippocampal tissue was frozen in liquid nitrogen and stored at −70 °C for further analysis.
6.8. Cholinesterase activity assay in the rat brain The activity of the cholinesterase enzyme in rats treated with compound 5g was studied using the cholinesterase quantitative assay kit (HITACHI 917 / MODULAR P). to do this the proper dilution of the tissue lysates supernatant were subjected to the reagent which contains Butyrylthiocholin. Cholinesterase enzymes hydrolyze the butyrylthiocholin to thiocholin which reduces the yellow (Fe(CN) 6 ) 3-to the colorless (Fe(CN) 6 ) 4-. The results could be analyzed using colorimetric assays.
6.9 Measurement of the expression of pro-in ammatory cytokines (TNF and NFkB) and AChE in rat brain using western blot analysis The expression of proin ammatory cytokines (TNFa and NFkB), AChE and GADH in rat brain in the presence of compound 5g and donepezil was studied using western blot. The tissue samples were lysed and the lysates were centrifuged at 4 C 12000 rpm for 10 minutes Protein concentration of supernatants were determined by Bradford method. Bovine serum albumin was used as standard protein. After mixing the samples with sample buffer (containing Tris, glycerol, -mercapto ethanol, Bromo phenol blue and SDS), they heated for 10 minutes in boiling water. Proper amount of protein extract were subjected to 12% SDS-PAGE gels and the resulted spots were transferred to a polyvinylidene uoride (PVDF) membrane (Millipore, Billerica, MA, USA). The membrane were shaken 5 times (15 min) with the blocking agent (5% defatted milk powder in TBST (Tris buffered saline-Tween 20)) at room temperature before the incubation with primary antibody. After the blocking the membrane were incubated primary antibodies for 18 hours. The primary antibody was diluted with blocking buffer. The primary antibodies were β-actin (sc-47778, 1:300) as internal standard, ACHE (E-AB-70014, 1:1000), NFκB-p65 (E-AB-22066, 1:1000), TNFa (SC-130349, 1:1000) The membrane were washed three times (15 min) with TBST, and were incubated (75 min) with secondary antibodies (m-IgGκBP-HRP (sc-516102, 1:1000) and mouse anti-rabbit IgG-HRP (sc-2357, 1:1000)). Enhanced chemiluminescence (ECL) (Millipore, USA) was used for visualization. Optical density analysis was performed for semi quantitative measurements.

Statistical analysis:
The obtained data for all cell based and animal studies were statistically analyzed using One-Way ANOVA, Two-Way ANOVA or Posthoc analysis. GraphPad Prism software (GraphPad Software, La Jolla, CA, USA, www.graph pad.com) and Excel software were used for visualization of graphs. p< 0.05 was considered to be statistically signi cant.

Results And Discussion
Synthesize of the quaternary ammonium derivatives The synthetic route of the key intermediate is shown in Figure 2. 4-ourobenzaldehyde (1a) reacted with various secondary amines (2a-f) by nucleophilic substitution reaction [35] and compounds 3a-f were produced. Then 1b-d, 3a-f and 3h compounds reacted with 5, 6-dimethoxy-2, 3-dihydro-1H-inden-1-one. More details of the synthesis procedure optimization and characteristic of synthesized compounds could be found in our previous papers [34,35].
We synthesized the designed compounds in the current study with a minor revision comparing to our previous papers [34,35], to avoid superfluous byproducts produced. We conducted the reaction at room temperature, using EtOH as the solvent and aqueous NaOH (10 %) as the base. The quaternary ammonium compounds (5g-i) were synthesized from (5f-h) reacting with methyl iodide. The details of the chemical characterization of two newly synthesized compounds (5g and 5i) using 13 C NMR are presented in gure 3 and detailed characterization and IR and 1 H NMR results are presented and methods section.

hAChE and hBuChE inhibitory activity of synthetized molecules
The anti-cholinesterase activity of the synthesized compounds was studied against hAChE and hBuChE. Compounds 5h, 5j, 5k, and 5l were practically insoluble in the applied solvents, and consequently, their inhibitory activity was not further evaluated.
The IC 50 values (Table 1)  hBuChE inhibitory evaluations showed that compound 5g inhibitory activity (IC 50 =4 μm) is higher than donepezil (IC 50 =5.25 μM) and galantamine (IC 50 =31.6 μM). The selectivity index for hAChE to hBuChE inhibition was calculated for studied compounds, and the results ( Table 1) showed that compound 5i possessed much higher selectivity toward hAChE in comparison with donepezil and other studied compounds. Compound 5g is almost inactive against hBuChE. Also, compound 5f was more selective toward hAChE in comparison with 5g.
Kinetic studies results (Table 1) showed that the synthesized compounds possessed mixed non-competitive inhibition, except for compound 5d that showed competitive inhibition. According to the results, inhibition constants (k i ) of compounds 5f and 5d were smaller than their non-competitive inhibition constant (ki′) values. This nding showed that the tendency of these compounds to bind to the active site is more than the steric site. In addition, k i values of compounds 5f and 5g were smaller than ki value of compound 5d, which shows the higher a nity of compounds 5f and 5g to the steric site. In conclusion, the better inhibitory activity of compounds 5f and 5g in comparison with compound 5d can be a result of their tendency to bind to both steric and active sites of enzymes. Double reciprocal plots and secondary plots of compound 5g inhibitory activity are shown in Figure 4.

SAR analysis of the studied compounds' cholinesterase inhibitory activity
Incorporating the quaternary ammonium to the para position of phenyl ring (compounds 5g and 5i) resulted in increased ChE inhibitory activity signi cantly. Comparing compounds 5i and 5g shows that when this quaternary ammonium is a part of piperazine ring (5g) the inhibitory effect against both enzymes increased signi cantly. Corresponding non-quaternary amine substituted compounds (i.e. 5f and 5h), showed much less ChE inhibitory activity ( Table 1). Comparison of compounds 5f and 5h con rmed the importance of piperazine ring. Substitution of piperazine with morpholine (compound 5c) leads to a signi cant reduction in inhibition e cacy in both enzymes, which shows the importance of a positively charged moiety rather than a functional group with hydrogen bonding capability.
A similar range of activities for compounds 5a-e against both enzymes revealed that the presence of positive charge in the exact distance from the aromatic linker is important for ChE inhibitory activity in which replacing piperazine with piperidine moiety (compound 5a) leads to signi cant activity reduction (Table 1). This nding con rms the importance of the presence of a positively charged moiety in this position [37]. This parameter also is much more important than hydrophobic moieties or the size of the group, as it's obvious from compound 5i, in which the ring is substituted with quaternary ammonium, while hAChE inhibitory activity is about 10 times more than compounds 5a-e which have a different ring in this position. Simultaneously, a signi cant reduction in hBuChE inhibitory activity of compound 5i shows a signi cant impact of hydrophobic interactions on the interaction of studied compounds with hBuChE binding pocket. The same nding was approved by a signi cant reduction of compound 5f activity against hBuChE compared with 5g, which has one more methyl group. Also, loss of action of compounds 5h-5l against hBuChE con rms the ndings above of the importance of positive charge and hydrophobic groups for both enzymes.
Toxicity pro ling of compounds 5g and 5f The cytotoxicity of the compounds with higher hAChE inhibitory activity (5f and 5g) was studied using MTT assay on MCF7, B16 cell lines, and the resulted LD 50 values are represented in Table 2. It is obvious that compounds 5g and 5f possessed lower cell growth inhibitory activity against studied cells (MCF7, B16). Comparing the LD 50 values of these compounds with donepezil revealed that they show less cell toxicity. Galantamine LD 50 values toward both cell-lines are higher than the studied compounds. The obtained LD 50 values for both studied compounds were signi cantly higher than their IC 50 values toward hAChE and hBuChE.

Inhibition of Aβ -induced death of PC12 cells by compound 5g
The cytotoxic effect of Aβ aggregation in neuronal cells is one of the main hallmarks of AD. Attenuation of Aβ aggregates is an understudy strategy for developing anti-AD drugs [38,39]. One of the frequently applied in vitro models for studying the neuroprotective effect of drug candidates is PC12 cell death induced by Aβ 1-40 [40]. The underlying mechanism of cell death due to the accumulation of Aβ aggregates is not clearly known yet, but some studies suggested the crosstalk between autophagy, apoptosis, ROS production, and in ammation as probable mechanisms [41,39]. Autophagy modulation has been suggested as a target for AD drugs. The autophagy activation mediates neuroprotection against induced cytotoxicity by Aβ, oxidative stress and apoptosis [42,43,40]. ROS role in the neurological disorders development have been described in different studies [44]. Rising number of studies are indicating the transaction between ROS and autophagy as a determinant factor in the neuronal homeostasis modulation [44]. Drugs that modulate autophagy and/or apoptosis attracted much attention in recent years [45,39].
To examine the multi-targeted capability of the compound 5g, we studied its neuroprotective effect against Aβ-induced PC12 cell  Figure 5b shows the decreased viability of PC12 cells' after incubation with Aβ. Apoptotic bodies were seen easily; cells grew more slowly than normal cells. The morphology of the Aβ treated cells changed in a way that they are less clear, and they kept their spindle shape [47]. The addition of compound 5g at both concentrations (0.1 and 1.0 µM) was reversed the Aβ induced cell death ( gure 5c and 5d). The protection was increased at a higher concentration of compound 5g.
The prevention of Aβ-induced neurotoxicity may occur via different mechanisms such as acting as an anti-oxidant agent, attenuating neuroin ammation, reduction of Aβ bril formation, inhibition of cell membrane damage, and some other mechanisms [39]. The probable mechanism of the Aβ-induced cell death reduction by compound 5g was studied, and the results are reported in the following sections.

Compound 5g effect on oxidative stress
Oxidative stress is one of the AD hallmarks, and drugs with antioxidant activity are candidates for MTD anti-AD agents. The antioxidant activity of drugs attributed to their capacity for biological system protection against the ROS and reactive nitrogen species (RNS) damage. Total antioxidant activity (TAC) is the general ROS and/or RNS scavenging capacity. Acute oxidative stress can lead to the rapid up-regulation of autophagy via post-translational modi cations of key autophagy regulators. ROS can induce single-strand DNA breaks, speci cally in telomeric regions, resulting in collapsed replication forks, un-replicated single-stranded DNA, and telomeric loss. Also, ROS can target lipid double bonds that will lead to the production of reactive aldehydes like malondialdehyde (MDA) via lipid-peroxidation procedure [48]. Reactive aldehydes can disrupt the cell membrane structure and function by interaction with membrane proteins, which consequently leads to the neuro-toxicity in the AD brain [49,50]. To examine the effect of compound 5g on Aβ induced death in PC12 cells via oxidative stress-related mechanisms, we studied compound 5g effect on TAC, ROS production, and lipid peroxidation (MDA level).

Total anti-oxidant, anti-ROS production and autophagy modulation capacity of compound 5g
We used 20, 70-dichloro uorescein di-acetate (DCFH-DA) assay to the measurement of the ROS level. Cellular esterase's can converted nonpolar DCFH-DA (an oxidant-sensing uorescent probe) to its polar derivative (DCFH). DCFH is a non-uorescent moiety which can be oxidized by intracellular ROS to highly uorescent DCF. Fig 6 A shows the images obtained using the uorescence microscope for Aβ treated cells in the presence and absence of the compound 5g.
The counterstaining of the nuclei of the cells was done using 4′,6′-diamidino-2-phenylindole (DAPI). Using the DAPI staining, the cell nuclei along with the apoptotic bodies that include condensed and fragmented nuclei) can be illustrated ( Figure 6A). The nuclei of living or dead cells can emit blue uorescence upon binding of DAPI to adenine-thymine regions of DNA. An increased number of the cells with small, condensed nuclei after incubation with Aβ shows the enhancement of the apoptotic cells.
The morphological evaluation of DAPI stained cells ( Figure 6A) shows the capability of the compound 5g on the prevention of Aβ induced apoptosis. The apoptotic cells tend to lose cell structure. They also showed altered nuclear condensation. The control cells keep their intact and evenly shaped structures along with nuclear condensation.
Fluorescence intensities ( Figure 6B) indicate the ROS level in the studied samples. Much intense green uorescence of Aβ-treated cells in comparison with normal cells ( gure 6A) shows enhanced production of ROS. The presence of the compound 5g leads to a reduction in ROS production. The DCF uorescent intensity of the normal cells (20%) increased in Aβ treated cells ( Figure 6B). In contrast, compound 5g leads to less enhancement of the DCF uorescent intensity. ROS reduction in the presence of the compound 5g could occur both via radical scavenging property of compound 5g and autophagy modulation.
The total antioxidant activity of compound 5g was investigated using the uorescein based assay for oxygen radical absorbance capacity. As seen in Fig. 6C, the lowest absorbance of uorescein observed in Aβ-treated cells (69%), while in the presence of compound 5g (0.1 and 1 µM) the uorescein absorbance was increased from to 74% and 87%, respectively. These results indicate the capability of the compound 5g for the cell protection against Aβ-induced cell death is partially via anti-oxidant activity.
Lipid peroxidation inhibitory effect of compound 5g and Preventive effect of compound 5g on membrane damage The capability of compound 5g for the inhibition of lipid peroxidation was studied by the measurement of the MDA level in Aβ treated PC12 cells. Figure 7A shows the MDA concentration in the studied cells. MDA levels was increased signi cantly in the Aβ 1-42 treated PC12 cells in comparison with control cells. The increased level of the MDA due to the Aβ 1-42 , was reduced by the compound 5g. The results showed that the inhibition of lipid peroxidation plays a role in compound 5g prevention against Aβ induced cell death.
Plasma membrane damage is one of the key features of different forms of cellular damage, such as apoptosis and necrosis. Lactate dehydrogenase (LDH) is rapidly released into the cell culture supernatant when the plasma membrane is damaged [51]. Also, the intracellular LDH level is a measure of cell viability. LDH activity is signi cantly increased from about 20% in normal cells, to73.6% in Aβ treated cells (Fig. 7B). This nding suggests that the oxidative modi cation by Aβ  contributes to the impaired function of the PC12 cells. The compound 5g resulted in a reduction in LDH levels at both studied concentrations (68% for 0.1 µM and 55% for 1.0 µM), respectively. It could be concluded that the compound 5g contributes to the prevention of cellular membrane damage mechanisms. As mentioned, cellular membrane damage could occur via different mechanisms due to different pathways activation such as oxidative stress, apoptosis, autophagy, and other mechanisms.
The effect of compound 5g on DNA fragmentation and morphology of the apoptotic cells The degree of the DNA fragmentation and morphology of the apoptotic cells were studied using a terminal deoxynucleotidyl transferase nickend labeling (TUNEL) assay. Apoptosis index for the normal cells and Aβ treated cells in the presence, and the absence of the compound 5g are shown in gure 8A. The counterstained nuclei by DAPI and merged images are presented as well.
The much intense green color in the Aβ treated cells compared with control cells shows that the Aβ cell toxicity happens in part via an apoptotic mechanism. The apoptotic index (percentage of the apoptotic cells, Figure 8B) increased from about 18% in control cells to 60% in the Aβ treated cells. The addition of the compound 5g to the incubation medium resulted in the reduction of the apoptotic index from 60% to 35% and 25% for the compound 5g concentration of 0.1 and 1.0 µM, respectively.
The TUNEL assay also revealed apoptotic characteristics in Aβ treated cells. The untreated cells showed blurred coloration, while apoptotic cells were stained by TUNEL reagent ( Figure 8A). These results indicate the apparent DNA fragmentation and apoptotic nuclei. Also, the morphological variation and losing cell integrity of apoptotic cells are obvious compared to the normal cells.
Compound 5g could prevent Aβ induced apoptosis by preventing both loss of cell integrity and DNA fragmentation. Cell membrane integrity loss is a hallmark of necrotic cell death, and necrotic cell death has been suggested as a player in neurodegenerative disease. According to the results, it could be concluded that compound 5g prevents PC12 cell death by contributing to oxidative stress, autophagy, apoptosis, and maybe necrosis. It also reduces radicals, prevents lipid peroxidation, and inhibits cell membrane damage.
Effect of compound 5g on neuro-in ammation Activated microglia mediates the Aβ peptide based neuroin ammation. interleukin (IL)-1β, IL-6, and tumor necrosis factor-α (TNF-α) are pro-in ammatory cytokines which can be increased following the microglia activation and could in uences the brain tissue and play a role in neurodegeneration as well as in neuroprotection [52]. Anti-in ammatory compounds that could reduce the expression of TNF-α and IL-1β have been investigated for their anti-AD potencies 7 . We studied the expression of mRNA levels of TNF-α and IL-1β in the presence and absence of compound 5g and donepezil. The mRNA level of IL-1β and TNF-α in Aβ treated PC12 cells increased in comparison to the normal cells ( Figure 9). Also, compound 5g prevented the enhancement of the IL-1β and TNF-α mRNA levels due to the Aβ induced cell toxicity at concentrations of 0.1 µM (p=0.0272 for IL-1β and p=0.01 for TNF-α) and 1 µM( p=0.0002 for IL-1β and p=0.0002 for TNF-α). Furthermore, decreased expression levels of these genes were observed in compound 5g (1 µM) treated cells compared to control cells (p=0.0397 for IL-1β and p=0.0478 for TNF-α). The results were illustrated in gure 9 and Table 3. According to the results compound 5g is capable for decreasing of in ammatory genes expression in which its potency increases at higher doses. These ndings suggested an anti-in ammatory mechanism for the compound 5g in addition to other pharmacological effects.
In vivo anti-Alzheimer e cacy of compound 5g Effect of the compound 5g on the recognition function of Aβ induced AD rats The capability of the compound 5g on the recognition function of rats was studied using the novel object recognition (NOR) task. NOR measures a speci c form of recognition memory, and it has been increasingly used as an experimental tool in assessing drug effect on memory. The results of NOR test ( Figure 10) revealed that there is no signi cant difference among studied groups in the locomotor activity (p = 0.39) and total observation time (p = 0.77) (Fig 10A and 10B). This approves the reliability of the task results for episodic memory. It also shows that the task results are independent from locomotor or observation biases. Discrimination index (DI) factor is an episodic-like memory index that was calculated for the studied groups and the results indicated that the DI between groups are signi cantly different (p < 0.001). DI reduced in the normal saline group (p < 0.001), which indicates the impairment of episodic-like memory. Furthermore, a signi cant increase of DI in donepezil (p < 0.01) and compound 5g (1 µM) treated groups (p < 0.001) (Fig 10 C) was observed, which could be an indicator of protective effect of donepezil and compound 5g against episodic-like memory impairment.
According to the results, it could be concluded that although compound 5g includes a permanently charged amin group, it reaches the brain of AD rats. As mentioned in the introduction choline transporter has been suggested as a root for transportation of the permanently charged molecules to the brain. As can be seen from the results ( Figure 10C), by increasing the compound 5g dose from 0.5 to 1 µM, an obvious improvement was observed in the recognition function of rats. This could be a result of the role of a transporter in compound 5g transportation to the brain. As we used the Aβ model of AD, and Aβ has been suggested as an inhibitor of choline transporter, the signi cantly lower e cacy at lower concentrations of compound 5g could be a result of their disability to reach the brain due to the inhibition of choline transporter. Moreover, the e cacy of compound 5g in the improvement of cognition function of AD rats is comparable with donepezil (0.5 µM) as a positive control. Due to the probable dependence of compound 5g transportation to the brain to choline transporter and its inhibition by Aβ, its application of albumin may lead to different results.
Effect of the compound 5g on the spatial learning and memory function of Aβ induced AD rats In addition, we conducted a Moris water maze(MWM) test to investigate the spatial learning and memory function of Aβ-induced AD in rats. The escape latency time was studied during the trainingdays in control and treated groups. The comparison of the results showed that there is a signi canteffects of day and group (p < 0.001), as well as group×day (p = 0.04). Such results indicated that despite the fact that the task was learned by animals across training days,different learning outcomes were observed among groups. Furthermore, posthoc analysis of the obtained results revealed a longer escape latency time for the normal saline-treated group in the2-4 days of the training session (p < 0.001, p < 0.01, and p < 0.05, respectively), which shows the lower target-platform nding abilityof animals in this group. Furthermore, escape latency time on the 3rd and 4th days were signi cantly decreased in the presence of donepezil and compound 5g (p < 0.05 or p < 0.01) (Fig. 11A). Also the time that spent to nd the target quadrant was signi cantly different between groups in the probe session (p < 0.001).The time spentin the target quadrant was different between sham and normal-saline-treatedgroups (p < 0.01), that con rms the impairment of spatial memory. Also, a signi cant difference was observed between the normal-saline-treatedgroup with donepezil and compound 5g (1 µM) treated groups (p < 0.05)groups (Fig. 11B).
Effect of compound 5g on the expression of pro-in ammatory cytokines (TNFa and NFκB) in Aβ-induced AD rat brain We used the western blot analysis of the hippocamp tissue biopsy obtained from AD rats' brain to investigate the effect of compound 5g on the pro-in ammatory cytokines (TNFa and NFκB) level. Minor downregulation of TNFα in hippocamp tissue was observed in compound 5g treated AD rats' brain ( Figure 12). We showed that a signi cant reduction ouccures in TNFα mRNA expression in Aβ+ compound 5g treated PC12 cells. NFkB downregulated due to the compound 5g administration signi cantly in a dose-dependent manner ( Figure 12).. Activated NFkB regulates the expression of COX-2 enzyme, TNF, IL-1, IL-6, IL-8 cytokines, adhesion, and cell cycle regulatory molecules [39]. The compound 5g mediated reduction of NFkB may results in neuroin ammation modulation of neuronal cells. Downregulation of NFkB and TNFα was observed in donepezil treated AD rats' hippocamp ( Figure 12).

Effect of compound 5g on total ChE activity and AChE level in rat brain
It is well known that the ChE activity and expression show a diverse correlation with inhibitors and physiologic conditions such as variations in Aβ precursor's level. It is reported previously that the Aβ treated neuronal cells show downregulation of AChE expression. Also some studies reported that Aβ could enhance the ChE hydrolyzing activity, which could be a result of the formation of soluble ultra-reactive acetylcholine-hydrolyzing Aβ-ChE complexes [53]. Our results showed that the Aβ treatment leads to reducing the ChE activity and downregulation of AChE in rat brain. The reduction of ChE activity can be a result of its downregulation or complex formation with Aβ.
For the evaluation of the compound 5g effects on the ChE activity in the Aβ-treated rat's brain, the butyrile thiocholine concentration was measured using a colorimetric method. Moreover, the AChE level was quantitatively measured using the western blot method. The results are shown in gure 13.
According to the results compound 5g (0.5 µM) prevented the downregulation of AChE in Aβ-received rats' brains. Compound 5g also decreased ChE activity in comparison with the control group ( gure 13B) while the activity reduction was less than Aβ treated group and more than donepezil treated group.
Donepezil could not prevent the Aβ induced downregulation of AChE while it inhibited the ChE activity reduction due to Aβ. Compound 5g (0.1) almost completely reversed the regulation pro le of AChE, while its effect on ChE activity was not signi cant comparing with Aβ treated group at all studied doses. ChE activity reduction due to compound 5g could be a result of ChE inhibition by it. While the prevention of downregulation of AChE could be a result of non ChE inhibitory anti-AD effects of compound 5g which was investigated and reported in this study.
These results could be due to the differences in ChE inhibition mechanism of donepezil and compound 5g in a way that donepezil possess competitive inhibition while compound 5g possess mixed non-competitive inhibition based on the kinetic studies. Furthermore, compound 5g is inactive against BuChE while donepezil inhibits both enzymes that could be a reason for their different e cacy on the reversion of Aβ induced ChE activity and AChE expression variation in part.

Molecular docking analysis and pharmacophore mapping
To optimize and validate the docking procedure, donepezil was docked to the binding cavity and the results compared with cocrystalized donepezil with hAChE (pdb code 4EY7). Figure 5 shows the superimposition of docked donepezil on co-crystalized donepezil with hAChE. The developed docking procedure was able to predict the binding mode well and resulted in high geometric similarity to the ligand conformation observed in the crystal structure.
Active compounds were docked in the binding pocket of hAChE (pdb code 4EY7) and hBuChE (pdb code: 4BDS). The best conformer (highest score and best orientation) of each compound was further studied using MOE (add reference: Molecular operation environment v2019.0102, Chemical Computing Group, Montreal, Canada) and LigandScout [54] software to investigate ligandprotein interactions and pharmacophore mapping of them. Figure 14 shows the 3D illustration of the positioning of the compounds 5f and 5g in the binding cavity of hAChE and hBuChE. Three main common structural characteristics of the binding pocket of both hAChE and hBuChE (Figure1) are the availability of a deep narrow gorge (20 A), the catalytic active (anionic) subsite (CAS) in conjunction with a choline-binding pocket (esteratic subsite) makes the catalytic machinery located in the bottom of the gorge beyond a peripheral acyl pocket [55] and an anionic site in the entrance of the binding gorge which is known as peripheral anionic site (PAS). The results show that the availability of a positive charge in an appropriate position facilitates interaction with Glu197 in hBuChE, and a p-cation interaction with Trp86 and Trp82 of hAChE and hBuChE, respectively. Also, a hydrophobic interaction of one of the methyl groups of compound 5g piperazine ring with Trp86 and Trp82 of hAChE and hBuChE have been observed (Figure 14), which could be a partial replacement of Donepezil phenyl ring interaction with those amino acids. In comparison with donepezil, compound 5g lacks one of the main interactions with Trp86 (i.e. p-p stacking with phenyl ring of donepezil) in hAChE, which could be a reason for lower potency. At the same time, a strong contact (which could be a hydrogen bond rather than an ionic interaction) with Glu197 of hBuChE leads to a bit higher e cacy of compound 5g in comparison with donepezil. Recent studies suggested that Glu197, which is a conserved amino acid within cholinesterase enzymes, has a signi cant role in the binding of substrate/inhibitors with hBuChE binding sites [62]. Molecular docking studies of the interaction between hBuChE and the studied compounds compared with donepezil revealed that the compound 5g possess similar interactions with the binding pocket while compound 5f lacks interaction with Trp82, which could be resulting from one methyl group removing. Quaternary ammonium itself does not lead to this variation in interaction, as the results showed that the consequent amino moiety of piperazine ring converted to the quaternary amino moiety due to ionization in physiologic pH.
The carbonyl moiety of the indanone ring in compound 5f interacts with catalytic triad amino acids of hAChE via hydrogen bonds [63]. Also, the indanone carbonyl group establishes a hydrogen bond with Phe295 backbone NH. The additional interaction with the acyl binding pocket of hBuChE is obvious for both compounds. Benzilidine group of compounds 5g and 5f possess interaction with Tyr341 and Tyr332 of hAChE and hBuChE, respectively.
The exibility of CAS interacting fragment of the molecule is important to enable its movement through the bottleneck of the gorge, and the introduction of one double bond to the linker leads to the reduction of the possibility of compounds to reach the CAS. This rigidity leads to the obvious con rmation orientation variation in comparison with donepezil for compounds 5g and 5f, which is presented in gure 15.
This miss-orientation can be observed in both enzymes. Molecular dynamics simulation of donepezil interaction with hAChE revealed that placement of the donepezil in the active site is facilitated by a conformational change (90° bending), which leads to suitable orientation. At the same time, this miss-orientation leads to an extra interaction with both enzymes' acyl binding pocket, which is not available in donepezil.
Exploring the water molecules in the active site of both enzymes showed that there are some conserved water molecules in hAChE binding gorge, which are necessary for the interaction. The results of molecular docking without removing those water molecules showed that there is the possibility of hydrogen bonding between methoxy moieties and at least one water molecule in PAS. Investigation of the hAChE-donepezil complex shows one water molecule in the suitable distance for such an interaction in PAS. The results are consistent with the previous studies [37], which suggested conserved water molecules in similar situations. Also, according to the reported B-factors for those water molecules, they could be regarded as conserved molecules. Our studies suggested no similar hydrogen bonding in PAS of hBuChE.
Interaction with the peripheral site inhibits the interaction of the hAChE with Aβ, and accordingly, the developed compounds can be regarded both as dual inhibitors [37] and hAChE-hBuChE inhibitors.

Conclusion
Newly synthesized indanone ionic analogs' structures were characterized using spectroscopic methods. The synthesized compounds cholinesterase inhibitory activity were examined against hAChE and hBuChE. The studied compounds possessed IC 50 values in the range 6 ->75 μM and 4 ->1000 μM respectively against hAChE and hBuChE, except for compounds 5f and 5g, which exhibited IC 50 values of 0.9 and 0.1 μM for hAChE and 126 and 4 μM for hBuChE.
Kinetic studies revealed that all compounds have mixed non-competitive inhibition mechanisms except compound 5d, which showed a competitive mechanism. A molecular modeling study of the compounds showed that compound 5g has dual binding site interaction. In addition, the availability of permanent quaternary ammonium facilitates interaction with the oxyanion hole, which enables the inhibition of hAChE driven Aβ aggregation by the developed compounds. In addition to the inhibition of hAChE compound 5g possessed better hBuChE inhibitory activity comparing with donepezil. Furthermore, compound 5g had neuroprotection against Aβ-induced neurotoxicity. Its anti-oxidant, anti-ROS production, anti-lipid peroxidation, LDH inhibition, antiapoptotic, and anti-neuroin ammation activity was studied and con rmed using Aβ treated PC12 cells. In addition, in vivo administration of compound 5g to Aβ-induced AD rats improved their cognition function and spatial memory, and learning function. Western blot analysis of Aβ-induced AD rats' brain biopsies con rmed NfKB reduction, AChE elevation, and CHE activity reduction in studied rats. Compound 5g is a promising lead for the development of hAChE and hBuChE dual inhibitors with the capability of multi target anti-AD activity, while its delivery to the brain has been con rmed by in vivo studies.    Figure 1 Schemateic overview of binding gorge of hAChE (left) and hBuChe (right). Catalytic site of enzymes including the anionic site (simillar triad aminoacids), acylpocket (larger in hBuChe), oxianionic hole (simillar) and preferal binding pocket (including aromatic amino acids in hAChE and non aromatic amino acids in hBuChE) are obvious in schematic view. (residue numbers are for 4EY7 and 4BDS pdb codes).

Figure 2
Synthesis of compounds 5a-l (compounds 5g and 5i desighned and synthesized in current study other derivatives were reported in our previous publications21-22).

Figure 6
Inhibiton of ROS production by compound 5g in Aβ-induced neurotoxicity on PC12 cells. (A) The image of DCF and DAPI treated normal, Aβ and Aβ+compound 5g treated cells, (B) DCF ourecent intensity of studied cells and (C) TAC% of studied cells.

Figure 12
Effect of compound 5g on the expression of pro-in ammatory cytokines (TNFα and NFκB), and GADH in AD rats' hippocamp.

Figure 13
Effect of compound 5g on the AChE relative amount and ChE relative activity in AD rats' hippocamp.