The inhibitory effects of bile acids on catalytic and non‐catalytic functions of acetylcholinesterase as a therapeutic target in Alzheimer’s disease

Acetylcholine is a fast‐acting neurotransmitter in synapses and neuromuscular junctions that is decreased in Alzheimer’s disease (AD) by hyper‐activation of acetylcholinesterase (AChE), which leads to progressive loss of memory and neurobehavioral abnormalities. Therefore, AChE inhibitors have therapeutic potential in AD that could include natural compounds such as bile acids. Bile acids, as potent molecules, could improve some types of neurodegenerative diseases via antioxidant effects and other unknown mechanisms. The aim of this study was to investigate beneficial effects of bile acids on AChE catalytic and non‐catalytic functions, amyloid plaque deposit and memory in a rat model of AD. The effects of sodium deoxycholate and cholic acid on AChE activity were assessed by in vitro assay. Then, the bile acids’ potential therapeutic effects were investigated on nucleus basalis of Meynert lesioned rats using behavioral evaluation, biochemical tests and histological methods. Molecular docking simulation was also implemented to investigate the possible interaction between bile acids and AChE. According to the in vitro and in vivo results, sodium deoxycholate could efficiently inhibit the catalytic function of the enzyme by interacting with the catalytic site, while cholic acid interacted with the peripheral anionic site and inhibited chaperone activity of the enzyme that led to the reduced amyloid plaque deposition. The co‐administration of cholic acid and sodium deoxycholate showed these compounds are able to simultaneously inhibit the catalytic and non‐catalytic functions of the enzyme. This study clarifies the roles of natural bile acids in the nervous system and in AChE function through multiple experimental and simulation methods.


INTRODUCTION
Alzheimer's disease (AD) is a progressive dementia, hallmarked by amyloid plaque deposition and characterized by memory loss and cognitive dysfunction (Dong et al., 2012). Cumulative evidence strongly supports that decreased acetylcholine (ACh) neurotransmitter, as a result of acetylcholinesterase (AChE) hyperactivation, plays an important role in the occurrence and development of neurological disorders related to AD in patients and animal models (Anand and Singh, 2013). Therefore, AChE inhibition by various drugs, such as donepezil, rivastigmine and galantamine, are used to improve symptoms of AD pathophysiology (Anand and Singh, 2013).
AChE which belongs to the α/β hydrolase fold family is an ACh-hydrolyzing enzyme in the synapse and neuromuscular junction that terminates cholinergic impulses in nerves (Holmquist, 2000). Despite the catalytic activity, AChE participates in multiple biological functions, such as neuronal cell differentiation, dendrite and axon formation, cell adhesion and neurodegeneration in Alzheimer's disease (Dong et al., 2004). Two sites of ligand interaction contribute to the catalytic and non-catalytic functions of AChE: a catalytic active site (CAS) at the bottom of a deep and narrow gorge and a peripheral anionic site (PAS) that is surrounded by negatively charged and hydrophobic residues (Johnson and Moore, 2006). Previous studies confirmed that AChE plays an effective role in Aβ plaque deposition by mediating PAS chaperone activity (García-Ayllón, 2011). Therefore, natural and synthetic compounds that effectively bind to the active sites, CAS and PAS, could potentially inhibit ACh hydrolysis and amyloid plaque formation (Holmquist, 2000;Liu et al., 2013).
It was also previously reported that endogenous bile acids are potent anti-apoptotic agents in neuronal cells that are exposed to Aβ (Viana et al., 2010). They are further known as powerful neuroprotective agents in various neurodegenerative diseases. Bile acids are structurally known as a group of antioxidants that are derived from cholesterol. This group can pass through the blood brain barrier (BBB) and was also shown to protect neuronal cells against oxidative damage (Lo et al., 2013). The neuroprotective mechanisms of bile acids are not completely understood but prior studies suggest that bile acid derivatives could potentially block the apoptosis process through stabilization of the mitochondrial membrane potential (Δψ m ) and regulation of gene expression (Viana et al., 2010;Ackerman and Gerhard, 2016). Therefore, the main goal of this study was to evaluate the effects of cholic acid (CA) and sodium deoxycholate (SDC), as main bile acids, on the catalytic and non-catalytic functions of AChE. The principal binding site of bile acids on AChE were investigated by molecular docking analysis. Their potential therapeutic effects were investigated on nucleus basalis of Meynert lesioned (NBML) rats, a widely accepted model of AD that manifests pathophysiological and behavioral signs of the disease (Toledano and Alvarez, 2004;Gratwicke et al., 2013). Previous results demonstrated that impaired memory related to AD is dependent on hippocampal function (Moodley et al., 2015;Sadeghi et al., 2018). Therefore, we investigated CD and/ or SDC for beneficial effects on the NBML rats' hippocampus through biochemical analysis, immunological evaluation and histological studies.

Animals
In vivo study was conducted on experimental animals by using 4.5-6 months adult male Wistar rats, weighing 250-300 g obtained from the animal house. All tests were carried out at 20-25°C (12 hours light and 12 hours dark). Municipal tap water was used as drinking water and compressed food as nutrition (consisting of 23% protein, 3.0% crude fat, 7.0% crude fiber: 8% acid insoluble ash, 1-2.5% calcium, 0.9% phosphorus, 0.5-1% sodium, 12% moisture). All the experimental works were approved by the Ethical Committee of Tabriz Uni-versity of Medical Science (Tabriz, Iran) and conform to the European Communities Council Directive of 24 November 1986 (86/609/EEC).

AChE activity assessment in hippocampal primary culture homogenate
Primary neuronal cultures of rat hippocampus were prepared according to a previous study with some modifications (Seibenhener and Wooten, 2012). Dissected hippocampus (isolated from neonatal rat) was triturated by trypsin and mechanical disruption. The cells were differentiated in neurobasal medium supplemented with B27 and FGF2 for 7 days. The resulting cells were homogenized in cold buffer phosphate, 28 mM (pH 7.8), containing Triton X100 (Sadeghi et al., 2018). The homogenized samples were then centrifuged at 13,000 g for 20 min and subsequent supernatants were considered for AChE assay. AChE activity was determined based on spectroscopic monitoring of 5-thio-2-nitrobenzoate anion production at 430 nm during the reaction between 5,5′-Dithiobis(2-nitrobenzoic acid) (DTNB) and thiocholine. In this reaction, acetylcholine iodide as a substrate was converted to thiocholine that reacts with DTNB and produces 5-thio-2-nitrobenzoate. The final concentrations of DTNB and substrate were 0.33 mM and 1.56 mM, respectively, in all assay medium. The activity of the enzyme was expressed as μmole of ACh hydrolyzed per minute per milligram protein in the absence and presence of bile acids (10 μM), individually. The homogenate's protein concentration was evaluated according to the Bradford method (Bradford, 1976). All assays were done in five replicates.

Molecular docking analysis
In order to precisely assess the affinity of CA and SDC to AChE, Auto Dock software (version 4.2) was used to simulate the binding process of bile acid molecules to AChE enzyme (Morris et al., 2009). For this aim, firstly, the crystal structure of AChE (PDB ID: 1DX6) was taken from the Protein Data Bank (https://www.rcsb.org/). Since different subunits of the enzyme are the same, the subunit A was selected to perform docking analysis precisely. The 3D molecular structures of CA and SDC were obtained from the PubChem website (https://pubchem.ncbi.nlm.nih.gov). Their structural conformations were minimized by Gaussian 03 programs based on a theoretical level of B3LYP with 6-31G basis set. Polar hydrogen atoms were added into AChE as a target and Kollman charges were also calculated. In addition, the rotation of bile acid molecules was defined and rotatable bonds were detected. To identify the ideal binding site of each bile acid on AChE protein, a docking calculation using the Lamarckian genetic algorithm (LGA) search method was performed. All other calculation parameters were default settings. Chimera 1.10.1 program was used to visualize the obtained results (Rizvi et al., 2013).

In vivo experimental design
NBML rats were used as an AD model as follows: rats were anesthetized with ketamine (125 mg/kg, i.p.) and xylazine (10 mg/kg, i.p) and then placed in a Stoelting stereotaxic apparatus. The midline of clean scalp was notched and a burr hole was drilled into the skull. Following that, ibotenic acid was injected at the NBM nucleus with coordination as follows: AP=1.2, ML=±3.2, and DV=7.5 mm from the surface of skull (Paxinos and Watson, 2005). Each side of the NBM was injected with 1 μl of 5 μg/μl ibotenic acid solution with a microinjection pump at the speed of 120 μ lit/h. After 7 days of recovery, the animals were divided randomly into three groups as follows: Control: rats were submitted to surgery and injected with 1 μl saline (as equivalent shock obtained by surgery and local injection) and were orally administered saline after recovery for 25 days.
Rat model of AD: NBML rats in which both NBM nuclei were bilaterally destroyed by ibotenic acid during surgery. These rats received saline orally after recovery for 25 days.
NBML rats treated with bile acids: this group consisted of 24 rats whose NBM nuclei were bilaterally destroyed by ibotenic acid. After recovery, the AD-like rats were randomly divided into three groups (8 rats in each group) and two groups were treated orally with 30 mg/kg of SDC or CA for 25 days. The third group received 15 mg/kg SDC+15 mg/kg CA to study synergic effects of both bile acids. The dose of bile acids and the experimental time duration were selected according to a previous study (Lo et al., 2013).
To confirm the increase of bile acid levels in the brain tissue of rats after 25 days of treatment (15 mg/kg SDC+15 mg/kg CA), the concentration of CA and SDC were evaluated in brain tissue by using CA and DCA ELISA Kit (Cell BIOLABS and MYBIOSOURCE). According to the obtained results, CA and DCA concentrations in the control group were calculated at 0.1±0.02 nmol/g and 0.06±0.01 nmol/g, respectively, which increased to 0.24±0.08 nmol/g and 0.18±0.03 nmol/g, respectively, in rats that received combined bile acids during the 25 days. Therefore, oral administration of bile acids enhanced these molecules' concentration more than two-fold in brain tissue.

Western blotting analysis
Western blotting was carried out according to a standard protocol (Sadeghi et al., 2017). After SDS-PAGE, the proteins were transferred onto PVDF membrane under 140 V for 1.5-2 h in transfer buffer. The membrane was probed with the primary [Anti-Glial Fibrillary Acidic Protein (GFAP) antibody (G3893, Sigma-Aldrich), Anti-beta Amyloid 1-42 antibody (ab10148, Abcam), rat specific anti-acetylcholinesterase antibody (ab31276, Abcam), rat specific anti-amyloid precursor protein (APP) antibody (ab2072, Abcam)] and secondary specific antibodies after blocking and washed four times in TBST (50 mM Tris, pH 7.5, 150 mM NaCl, 0.05 % Tween 20). An ECL detection kit was used to visualize bands containing specific proteins according to the manual. Anti β-actin (1:1000) (Cell Signaling Technology) was used as a housekeeping protein to control protein concentration.

Histological studies
For histological analysis, the rats from each experimental group were sacrificed under deep ketamine and xylazine anesthesia and before fixation with 4% formaldehyde and were subsequently perfused with 0.3% sodium sulphide in 0.1 M phosphate buffer. The separated brains were stored in 30% sucrose for cryoprotection and sectioned on a freezing microtome at 8 μm. The tissue sections were stained with Thioflavin T, a general marker of amyloid deposits, and studied by a fluorescent microscope. To this purpose, the tissue sections were incubated with freshly filtered 0.05 % Thioflavin T solution in PBS for 8 min at room temperature (Batelli et al., 2008).

Statistical analysis
SPSS10 software was used to analyze the data. Analysis and identification of the differences between experimental groups and control were performed using the t-test and p<0.05 was considered as a significant difference level.

Bile acids bind to the active site gorge and inhibit AChE catalytic activity
AChE belongs to the serine hydrolase family that contains the α/β hydrolase fold as a common structural element (Holmquist, 2000). A hallmark of AChE is its catalytic activity that can be assessed by the Ellman method, a sensitive, rapid and reliable method for detecting very low amounts of AChE catalytic function (Sadeghi et al., 2018). In the present study, primary hippocampal neuron cultures were used as a source of AChE to investigate the possible effects of bile acids (CA, SDC and CA+SDC) on catalytic activity of AChE. The results showed that bile acids could significantly inhibit hydrolyzing activity of the enzyme. The catalytic function of AChE was reduced from 6.37±0.52 to 5.56±0.31 μmol/min.mg (protein) in the presence of CA (10 μM), while SDC reduced AChE activity to 3.64±0.37 μmol/min.mg (protein). Further, it was shown that 10 μM of combined bile acids (5 μM CA+ 5 μM SDC) could reduce the enzyme activity to 2.07±0.23 μmol/ min.mg (protein). The maximum rates of AChE inhibition occurred in the presence of SDC+CA and decreased more than three-fold (Table I).

Molecular docking simulation
Molecular docking simulation revealed that both of the evaluated bile acids were capable of binding to the active site gorge of AChE at different affinities. The binding energy was calculated as -7.76 and -8.94, respectively, for the binding of CA and SDC to the active site gorge. It was shown that the active site of AChE consists of catalytic residues (Ser 203, His 447 and Glu 334), which are located within the gorge, similar to other AChEs in mammals (Johnson and Moore, 2006;García-Ayllón et al., 2011;Liu et al., 2017). Also, some anionic and hydrophobic amino acids surround the entrance of the active site known as the PAS site (Tyr 286, Trp 286, Phe 338, Val 294 and Leu 289) (Johnson and Moore, 2006;García-Ayllón et al., 2011). Fig. 1 shows the binding sites of bile acids on the active site of AChE. It was indicated that CA due to the bulky structure could bind to PAS and is not able to reach to the catalytic site at deep of the gorge so could limit substrate access to the catalytic site. While SDC could pass through the PAS tunnel and reach to CAS which limits substrate binding to the active site. All docking results are listed in Table I.

Bile acids treatment attenuated catalytic activity of hippocampal AChE that was increased by NBM lesion
NBM lesion causes impaired spatial navigation according to Fig. S1 (supplementary data) in experimental animals. The pathogenesis of AD has been linked to AChE activation that is also accompanied by β-amyloid deposition (Dong et al., 2004;Johnson and Moore, 2006). Therefore, to investigate whether NBM lesion and oral administration of bile acids could affect AChE activity in the rat hippocampus, the enzyme activity in homogenate tissue was evaluated for all groups. In agreement with our previous study (Sadeghi et al., 2013), the results showed that AChE activity was significantly increased in NBML rats up to 2982.71±152.26 nmol/min.mg protein, while the enzyme activity in the control hippocampus was calculated to be 525.34±34.52 nmol/min.mg protein (Fig. 2). NBML rats that treated with 30 mg/kg of SDC and/or CA for 25 days, showed reduced AChE activity compared with the NBML rats that treated with saline. Fig. 2 shows that the activity of AChE was remarkably reduced (3-fold) in NBM lesioned rats were administered 30 mg/kg of SDC. NBML rats received 30 mg/kg of CA revealed a two-fold reduction in AChE activity in hippocampal tissue. The enzyme activity was, however, remarkably inhibited in the rats that received both of the bile acids (520.13±50.25 nmol/min.mg protein), indicating synergistic effects.

Bile acids improved abnormal expression of GFAP, Aβ and AChE
Western blot analysis confirmed the increased expression of GFAP, a well-known biomarker for neuronal  (Lei et al., 2015). In this study, it was revealed that the GFAP content of hippocampal tissue was reduced in the rats treated with bile acids, especially the rats that received CA+SDC daily. Our results also revealed that AChE expression was enhanced in hippocampal tissue of NBML rats, while bile acids significantly reduced its expression. Fig. 3 reveals that the effects of CA and the combination of both bile acids are more significant than SDC treatment. According to the results, the level of Aβ content was six-fold in the NBML rats compared with the control. Interestingly, CA caused a greater decrease in the level of Aβ in hippocampal cells (more than five-fold) than SDC (near to two-fold). Our results also confirmed nonlinear cumulative effects of CA and SDC in the reduction of AChE expression.

Bile acids improved histopathological abnormalities induced by NBM lesion
In this investigation, histological studies were done as confirmation of biochemical results. Thioflavin staining demonstrated a clear abundance of amyloid plaques in the hippocampal tissue of NBML rats; however, deposits were absent in the control rats. In agree-ment with a previous study (Lo et al., 2013), thioflavin plaque numbers were reduced in the hippocampus of bile acids-supplemented-NBML rat compared to the NBML rats that treated with saline. Thioflavin plaque  numbers significantly decreased in the groups treated with SDC and CA. Fig. 4 also shows cumulative effects of both bile acids in AD rats that received SDC and CA simultaneously and resulted to decreased number and size of plaques.

DISCUSSION
We investigated the effects of bile acids (SDC and CA) on hippocampal AChE activity as an important target in AD therapy. The in vitro investigation of AChE was carried out using Ellman's method with crude enzyme that was extracted from primary neuronal cultures approximating neonatal rat hippocampus (Sadeghi et al., 2018). The results showed that both bile acids could significantly inhibit AChE hydrolytic activity (Table I) possibly due to substrate accessibility restriction. AChE protein possesses two sites for ligand interactions: an active (hydrolytic) site within a deep narrow gorge (CAS), including Ser203, Glu334 and His447, and a peripheral anionic site (PAS) that consists of a group of negatively charged and hydrophobic residues surrounding the entrance site (Prasher, 2004;Liu et al., 2017). The CAS is located within the inner site of the enzyme and substrate molecules must pass through the PAS tunnel to reach to the CAS, the distance between the two position estimated to be approximately 12 Å. Molecular docking studies revealed that SDC, upon binding to the CAS, could inhibit esterase activity more significantly; however, since CA could not reach to the catalytic site within the gorge, it could not effectively inhibit hydrolytic activity but could possibly limit substrate access by binding to the PAS site. Structural and docking analysis of bile acids revealed that the presence of hydroxyl group at C 7 of the sterol rings limited bile acid from penetrating to the active site gorge possibly due to the increasing molecular dimensions that lead to decreased flexibility. Interestingly, AChE activity decreased more significantly in the presence of an equivalent combination of CA and SDC, which suggests synergistic inhibitory effects. Regarding the different physiological roles of PAS and CAS in biological systems, the effects of CA (binds to PAS), SDC (binds to CAS) and CA+SDC (bind to PAS and CAS) on biochemical and molecular aspects of AD were investigated by using NBML rats, a recognized animal model of AD.
Local injection of ibotenic acid damaged acetylcholine manufacturing cells in the NBM and caused AChE hyper-activation in hippocampus that ultimately led to AD pathophysiological signs (Wenk, 1990;Anand and Singh, 2013). According to the results, oral administration of bile acids for 25 days enhanced CA and DCA concentrations in the brain tissue more than two-fold. Therefore, similar to the in vitro assay, oral administration of SDC and CA effectively decreased the activity of hippocampal AChE. NBML rats that received both of bile acids in similar doses (15 mg/kg CA+15 mg/kg SDC) exhibited minimal hippocampal AChE hydrolytic activity (near control values), which indicates non-linear cumulative effects similar to the in vitro results. Immunoblotting analysis also confirmed enhanced AChE expression, due to NBM lesion, decreased significantly with bile acids treatment, especially in rats that received CA and CA+SDC. NBM lesion caused an approximately 4-fold increase in AChE expression but its activity increased more than 5.6-fold. Therefore, activation and increased expression of AChE were taking place simultaneously. Previous experiments indicate that bile acids could regulate gene expression through binding to the G-protein-coupled bile acid membrane receptor 1 (TGR5) and bile acid nuclear Farnesoid X receptor (FXR) at the membrane and inside the nucleus, respectively (Mertens et al., 2017). SDC treatment reduced AChE expression slightly but decreased its catalytic activity more than 4-fold. The reduction of AChE expression and catalytic activity in the presence of CA seems to be equal; therefore, it appears that CA reduces AChE's hydrolyzing activity primarily through regulation of gene expression. Therefore, these results reveal that the expression and activity of AChE enzyme could be regulated by different pathways. AChE is colocalized with Aβ in pre-amyloid deposits, mature senile plaques and cerebral blood vessels in the brain of AD patients, which suggests its role in promotion and development of Aβ plaques (Inestrosa et al., 1996). In vitro and in silico analyses confirmed that Aβ has a marked tendency to interact with the PAS site mainly via hydrophobic interactions (García-Ayllón et al., 2011). The increased activity and expression of AChE in NBML rats suggests that the chaperone role of AChE also improved, which is associated with the PAS site and is responsible for Aβ polymerization. Interestingly, it was observed that CA binds to the PAS site and reduces amyloid plaques more significantly than SDC, which mainly interacts with CAS. It is possible that CA could have bound to PAS, inhibited the chaperone function of AChE and attenuated Aβ plaques deposition, leading to the decrease of plaque number and size. NBM lesion by ibotenic acid caused cholinergic system impairment and led to the increased GFAP content in hippocampal cells that confirmed AD is a neurodegenerative disease (Moodley et al., 2015;Lei et al., 2015). The treatment of rats with CA or CA+SDC resulted in a strong decrease in neuronal damage compared to rats that received SDC. Simultaneous AChE upregulation and increased expression of Aβ and GFAP suggests a possible role for AChE in the regulation of the genes involved in apoptosis as a non-catalytic function (Dong et al., 2004). The effect of bile acids at the molecular level resulted in the refinement of behavioral abnormalities caused by NBM lesion in rats. Fig. S1 reveals that NBML rats treated with CA+SDC for 25 days took less time to reach the platform than rats that received normal saline, SDC or CA.
Overall, this is an original report of the bile acids' therapeutic effects on NBM lesion-induced pathology through in vivo, in vitro and in silico experiments, which results suggests a novel approach in AD therapy. Our results confirmed that SDC more strongly inhibited hydrolytic functions, possibly by binding to the catalytic site, while CA binds to the PAS site and inhibited non-catalytic functions of AChE such as plaque formation and gene regulation. Therefore, in the presence of both bile acids, CAS and PAS were inhibited simultaneously and both of the catalytic and non-catalytic functions were attenuated, resulting in the inhibition of ACh hydrolysis and reducing size and number of plaques. Thus, this study suggests AChE regulation is an essential role for bile acids as liver-derived molecules in the CNS. Considering the similarity between the NBML model and AD at molecular and physiological levels, CA and SDC may be potential therapeutic adjuvants for patients that are suffering from AD or other related cognitive disorders. This study additionally provides new structural information for the rational design of new inhibitors against AChE.

SUPPLEMENTAL MATERIAL
Fig. S1. Spatial learning evaluation by the Morris water maze test. The rats' latency to reach the platform increased in NBML rats in comparison with control on 3 consecutive test days. 30 mg/kg of bile acids decreased latency time for AD rats to reach the platform. Treatment with CA+SDC improved spatial navigation in NBML rats more significantly in comparison with other groups. Each data point indicates the mean ± S.E.M. The asterisks (*) indicate significant differences (p<0.05) according to Duncan's multiple range test.