Memory-Enhancing Effects of Mangosteen Pericarp Water Extract through Antioxidative Neuroprotection and Anti-Apoptotic Action

Mangosteen has long been utilized as a traditional medicine in Southeast Asia. Diverse extracts of mangosteen pericarp and its bioactive xanthones exhibit various bioactivities. However, the pharmacological potential of mangosteen pericarp water extract (MPW) has not been reported yet. This study used primary cultured rat cortical cells to investigate the effect of MPW on neurotoxicity. We found that MPW inhibited neurotoxicity and production of reactive oxygen species triggered by Aβ(25–35) or excitatory amino acids. MPW inhibited caspase 3 activation and DNA fragmentation in Aβ(25–35)- or N-methyl-D-aspartate-treated cells, suggesting an anti-apoptotic action. Additionally, MPW reduced lipid peroxidation and scavenged 1,1-diphenyl-2-picrylhydrazyl radicals, assuring its antioxidant property. Furthermore, MPW suppressed β-secretase and acetylcholinesterase activities. These findings prompted us to evaluate its effect on memory dysfunction in scopolamine-treated mice using Morris water maze test. Oral administration of MPW at the dosage of 50, 100, or 300 mg/kg for four days significantly decreased the latency time to find the platform and markedly increased the swimming time in the target quadrant. Taken together, our results suggest that MPW exerts memory-enhancing effect through antioxidative neuroprotection and anti-apoptotic action. Accordingly, MPW may have a potential to prevent or treat memory impairment associated with Alzheimer’s disease.


Introduction
Alzheimer's disease (AD) is a chronic neurodegenerative disorder accompanied by progressive memory and cognitive impairment. The pathological features of AD include the extracellular accumulation of senile plaques and the intracellular aggregation of neurofibrillary tangles. Senile plaques are mainly composed of aggregated β-amyloid (Aβ) generated from amyloid precursor protein (APP) upon enzymatic cleavage by βand γ-secretases [1,2]. Aβ is thought to play an important role in neurotoxicity both in vitro and in vivo. Aβ peptide causes destabilization of calcium homeostasis and induction of apoptosis, ultimately leading to neurotoxicity [3]. Accordingly, the Aβ pathway has been one of the promising targets for the development of putative disease-modifying drugs to treat AD. Various approaches to combat the Aβ pathway, including inhibitors of βor γ-secretase, or anti-Aβ monoclonal antibodies, have been challenged. However, no drug has been introduced into the clinic yet [4]. Furthermore, Aβ peptide is considered to induce oxidative damage in neuronal cells by generating reactive oxygen species (ROS) [5]. Thus, oxidative stress either triggered by Aβ peptide or any other stimuli may be one of the critical causative factors in AD. Based on these findings, antioxidants that alleviate ROS and mitigate oxidative stress-induced neuronal cell death have also been proposed as an intriguing approach to prevent or treat AD [6]. Unfortunately, however, the results from clinical studies have been disappointing to date. Considering the multifactorial and heterogeneous characteristics of the pathophysiology of AD, various approaches have been attempted to explore the favorable efficacies of antioxidant therapies in AD [7,8].
Glutamate is the most abundant excitatory amino acid neurotransmitter in the central nervous system (CNS). Glutamate transmission is mediated by ionotropic receptors, such as N-methyl-D-aspartic acid (NMDA), α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid, and kainate receptors, as well as by metabotropic (G protein-coupled) receptors [9]. Although it is well documented that glutamate transmission is involved in synaptic plasticity and cognitive functions in the brain [10], an excessive quantity of glutamate release is also known to be involved in many acute and chronic neuropathological conditions, including cerebral ischemia, epilepsy, and AD [11]. Accumulating evidence indicates that excess glutamate overactivates NMDA receptors and causes a redundant influx of calcium ions into cells and excessively generates ROS, which can eventually trigger apoptotic or necrotic neuronal cell death, resulting in excitotoxicity [12,13]. Thus, targeting glutamate transmission to prevent excitotoxicity, specifically by modulating NMDA receptor function, has been an attractive approach for the development of pharmacological interventions to treat neurological disorders such as AD. Memantine, a non-competitive NMDA receptor antagonist, is a well-known drug clinically used for the treatment of AD [14].
Mangosteen (Garcinia mangostana L.) is a tropical tree cultivated in Southeast Asia. The seeds and pericarps of the fruit have been traditionally used for centuries to treat skin infections, wounds, and dysentery [15][16][17]. Additionally, mangosteen products containing pulp and pericarps are sold worldwide, including in the U.S. market as nutritional supplements in the form of beverages or tablets [15,18,19]. The mangosteen pericarp has been reported to contain at least 50 different bioactive compounds, such as xanthones, polyphenols, and catechins [16,20]. Compared to the edible aril parts of the fruit, the pericarp consists of 10 times more phenolic compounds and 20 times more antioxidant activities [19].
Diverse extracts prepared from mangosteen pericarp through different methods have been reported to exhibit a wide variety of biological activities, including antioxidant, antiinflammatory, analgesic, anti-cancer, and neuroprotective effects [15,20]. The methanol extract and chloroform fraction of mangosteen pericarp inhibit the proliferation of human hepatocellular, breast, and colorectal cancer cells and exhibit antioxidant activities [21,22]. Another study has reported that a 50% ethanol mangosteen pericarp extract (MPE) shows antioxidative and neuroprotective activities against H 2 O 2 -induced oxidative stress in NG108-15 neuroblastoma cells [23]. Various xanthone derivatives, including αand γmangostins have been found in mangosteen pericarp and exhibit many of these biological activities [16]. We have recently demonstrated that γ-mangostin, not α-mangostin, isolated from mangosteen pericarp inhibits H 2 O 2 -induced oxidative neurotoxicity in the primary culture of rat cortical cells and reverses scopolamine-induced memory impairment in mice [24]. However, no reports are currently available on the effects of the mangosteen pericarp water extract (MPW) prepared by simplified and easily applicable processes without using organic solvents. Therefore, to elucidate the pharmacological potential of MPW, the present study explored the antioxidative and neuroprotective effects and memory-enhancing properties of MPW using primary cultures of rat cortical cells and scopolamine-treated mice, respectively.
We first investigated the effects of MPW on neuronal cell damage, ROS production, and apoptotic processes triggered by Aβ or excitatory amino acids in the cultured cortical cells. The antioxidative property of MPW was further evaluated by cell-free bioassays, measuring the levels of lipid peroxidation (LPO) and 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical formation in the presence of MPW. Moreover, we examined the impact of MPW on the activities of β-secretase as well as acetylcholinesterase. Finally, we used the Morris water maze (MWM) test to assess the effect of orally administered MPW on the scopolamineinduced memory dysfunction in mice.

Animals
Timed-pregnant Sprague-Dawley (SD) rats and ICR mice were purchased from Daehan Biolink (Chungbuk, Korea). Animals were maintained in the animal facility with controlled temperature (22 ± 2 • C), relative humidity (40-60%), and a 12-h light/ 12-h dark cycle. Animals were provided with a standard chow diet and water ad libitum. All the experimental steps, including the care, handling, and use of animals, were performed following the international guidelines (Guide for the Care and Use of Laboratory Animals, Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council; National Academy Press: Washington DC, 1996). Prior to the study, the rationale, design, and protocols of the animal experiments were approved by the Institutional Animal Ethical Committee of Dongguk University (Approval Number: IACUC-2016-035-2).

Preparation and Analysis of MPW
The dried pericarps of Garcinia mangostana L. (7.9 kg) were obtained from Indonesia and refluxed with distilled water (7.9 L) for 2 h at 100°C, yielding the crude extract of MPW (298.7 g). Chromatographic separation of the analytes was performed using the Acquity UPLC system with a BEHC18 column (2.1 × 50 mm, 1.7 µm, Waters, Milford, MA, USA). An aliquot of the sample (2 µL) was injected into the UPLC system for analysis. The mobile phase consisted of 0.1% formic acid in water for solution A and in acetonitrile for solution B. The mobile phase consisting of (A) and (B) was delivered at a flow rate of 0.4 mL/min by the following programmed gradient elution: 30% isocratic for 3 min, 30→70% (B, v/v) in 4 min, 70% isocratic for 2 min, 70→100% (B) in 2 min, 100% (B) isocratic for 2 min, 100→30% (B) in 0.5 min, and 5% (B) isocratic for 0.5 min for reconditioning post-run. The column temperature was maintained at 40°C.
Mass spectrometry was performed on the Xevo G2 Q-TOF MS (Waters, Milford, MA, USA) equipped with an electronic spray ion source under the positive-ion mode. The capillary and cone voltages were 2.7 kV and 15 V, respectively. Nitrogen, used as the cone, and desolvation gas were set at 0 and 600 L/h, respectively. The temperatures of the source and desolvation were 100 and 250°C, respectively. Mass spectra were recorded in the range of m/z 50-1200. Spectroscopic data, including UPLC-UV and mass chromatogram, for MPW are depicted in the Supplementary Materials.

Primary Cultures of Rat Cerebrocortical Cells
Primary cultures of rat cerebrocortical cells were carried out as previously described [25,26]. In brief, cerebral cortices were dissected from timed-pregnant SD rat embryos on the 17th day of gestation and mechanically dissociated into single cells by trituration with fire- polished Pasteur pipettes. The cerebrocortical cells containing neuronal and non-neuronal cells were seeded on either 35-mm culture dishes or 24-well plates, pre-coated with a mixture of laminin and poly-L-lysine. The respective seeding density was 6 × 10 6 cells/dish or 6 × 10 5 cells/well in MEM supplemented with 25 mM glucose, 2 mM glutamine, 5% HS, 5% FBS, and 1% antibiotic-antimycotic agent. The cells were maintained at 37 • C in an incubator with a humidified atmosphere of 95% air and 5% CO 2 . On the 7th day of seeding, the cells were treated with 10 µM cytosine arabinoside to arrest the proliferation of non-neuronal cells. The experiments were performed 10-11 days after plating.
The stock solution of MPW was prepared in sterilized distilled water at the concentration of 30 mg/mL and appropriately diluted prior to the treatment. Following the desired treatments, cell viability was assessed using the MTT reduction assay, as previously described [26,28]. Briefly, MTT was added to the culture of the treated cells at a final concentration of 1 mg/mL in phosphate-buffered saline (PBS), and then the cells were incubated for 3 h at 37 • C. After removing the culture media, 500 µl dimethylsulfoxide (DMSO) was added to each well, and the formazan crystal products were dissolved by incubating at 37 • C for 15 min. The absorbance was measured at 550 nm by a microplate reader (SpectraMax M2 e , Molecular Devices, Sunnyvale, CA, USA). The cell viability was expressed as a percentage of control-treated cells.

Measurement of Intracellular ROS Levels
Intracellular ROS levels in the absence or presence of MPW were spectrofluorometrically evaluated using the fluorogenic dye DCFH-DA as a probe [26]. Briefly, the cultured cells were washed with HCSS and then treated with DCFH-DA at a final concentration of 10 µM in MEMG for 30 min at 37 • C. Subsequently, the cells were washed with HCSS and then subjected to the corresponding insult (100 µM glutamate, 300 µM NMDA, or 40 µM Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) ) in MEMG for 2 h in the absence or presence of various concentrations of MPW. Intracellular ROS levels were assessed by measuring the level of 2 ,7 -dichlorofluoresceinderived fluorescence in a microplate reader (SpectraMax M2 e , Molecular Devices) with the excitation and emission wavelengths at 490 nm and 520 nm, respectively. Intracellular ROS levels were expressed as a percentage of the level determined in the control-treated cells.

TUNEL Assay
The effect of MPW on the NMDA-or Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) -induced apoptosis was assessed by the detection of fragmented DNA using a TUNEL assay kit (DeadEnd™ Colorimetric TUNEL System, Promega, Madison, WI, USA), as previously described [26,30]. In brief, the cells were treated with 100 µM NMDA or 40 µM Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) in MEMG for 2 h in the absence or presence of MPW (30 µg/mL), washed with PBS, and then fixed in 4% paraformaldehyde for 25 min at 4 • C. The cells were washed with PBS and then incubated in 0.2% Triton X-100 for 5 min for permeabilization. Following another round of washing, the cells were incubated in the equilibration buffer for 10 min at room temperature and then treated with the terminal deoxynucleotidyl transferase reaction mixture, consisting of the biotinylated nucleotide mix, for 60 min to permit the nick-end-labeling reaction. After the termination of the reaction, the cells were immersed in the saline sodium citrate solution and then incubated with streptavidin-conjugated HRP. The cells were washed twice with PBS and stained with diaminobenzidine. Finally, after washing the cells with PBS again, the TUNEL-positive cells stained in dark brown color were detected using a TS-100 inverted microscope (Nikon, Tokyo, Japan). The TUNEL-positive cells were manually counted from four randomly selected fields and expressed as percentages of total cell numbers, as described [26,30]. The counting was performed blind.

Assessment of the LPO Levels in Rat Brain Homogenates
The effect of MPW on the LPO levels in rat forebrain homogenates was examined as previously described [32]. Briefly, an aliquot of SD rat forebrain homogenate and various concentrations of MPW were added to a reaction mixture containing 10 µM Fe 2+ and 100 µM L-ascorbic acid and incubated at 37 • C for 1 h to allow the lipid peroxidation reaction to proceed. After stopping the reaction by adding 28% w/v trichloroacetic acid and 1% w/v TBA, the mixture was heated at 100 • C for 15 min and centrifuged at 3000 rpm for 10 min at 4 • C to remove the precipitates. The absorbance of the supernatant was measured at 532 nm on a microplate reader (SpectraMax M2 e , Molecular Devices). The percent inhibition of the LPO was quantitated using the following equation: where Abs control and Abs sample denote the absorbance of the control (without MPW) and experimental sample (with MPW), respectively.

DPPH Radical Scavenging Activity Assay
DPPH radical scavenging activity in the absence or presence of MPW was measured as previously described [32]. Briefly, the reaction mixture containing various concentrations of MPW and methanolic solution of DPPH (150 µM) was incubated at 37 • C for 30 min. The absorbance was then measured at 520 nm on a microplate reader (SpectraMax M2 e , Molecular Devices). The radical scavenging activity was determined using the following equation: Radical scavenging activity (%) = 100 × (Abs control − Abs sample )/Abs control where Abs control and Abs sample denote the absorbance of the control (without MPW) and experimental sample (with MPW), respectively.
2.11. In Vitro Assessment of β-Secretase Activity β-Secretase activity in the absence or presence of MPW was measured using a βsecretase FRET assay kit (Invitrogen, Carlsbad, CA, USA) according to the instructions of the manufacturer with some modifications [26]. In brief, 10 µL of the assay buffer containing various concentrations of MPW was added to 20 µL of the substrate (750 nM) into a 96-well plate. Then, 10 µL of β-secretase (1.0 U/mL) was added to the reaction mixture and incubated for 2 h at room temperature. The fluorescence was measured using a microplate reader (SpectraMax M2 e , Molecular Devices) with the excitation and emission wavelengths at 545 and 585 nm, respectively.

In Vitro Assessment of Acetylcholinesterase (AChE) Activity
AChE activity in the absence or presence of MPW was measured according to the method of Ellman, with some modifications [33]. In brief, a reaction mixture consisting of 100 µL of 10 µM 5,5 -dithiobis (2-nitrobenzoic acid), 10 µL of 75 mM acetylthiocholine iodide, and 10 µL of various concentrations of MPW in phosphate buffer (pH 8.9) was incubated at 37 • C for 10 min. Subsequently, 100 µL of 0.2 U/mL AChE enzyme was added, and the mixture was incubated for 15 min at 37 • C. The absorbance of the mixture was read at 410 nm by a microplate reader (SpectraMax M2e, Molecular Devices). The percent inhibition of the AChE activity was quantitated using the following equation: where Abs control and Abs sample denote the absorbance of the control (without MPW) and the experimental sample (with MPW), respectively.

The Morris Water Maze (MWM) Test
The MWM test was performed using a circular pool (120 cm in diameter and 45 cm in height), according to the procedures previously described with some modifications [34]. The pool was filled with water containing black food dye to the depth of 30 cm and placed under dim light. The water temperature was maintained at 22-23 • C throughout the experiment. The pool was conceptually divided into four quadrants, to which four different shapes of visual cues were attached. A white platform (10 cm in diameter and 29 cm high) was subsequently placed in one of the quadrants and submerged 1 cm below the water surface.
On the first day of the experiment, the mice were acclimated to the pool for 120 s in the absence of the platform. During the next four consecutive days (days 1-4) of training trials, the mice were given two trial sessions of 60 s each per day with the platform in place. Once the mice reached the platform within 60 s, they were permitted to stay on it for 10 s. If the mice could not find the platform within 60 s, they were placed on it for 10 s. After each training trial, the animals were delivered to their cages and dried under an infrared lamp. The interval time between each trial session was 30 min. During each training trial session, the time to find the hidden platform (latency time) was recorded using a video camera-based EthoVision Maze Test System (Noldus Information Technology, Wageningen, Netherlands). On day 5 of the experiment (probe trial), the platform was removed, and the mice were allowed to swim for 60 s to search for it. The swimming time the mice spent in the target quadrant where the platform had previously been placed and the total swimming distance were recorded during the probe trial sessions.
To test the effects of MPW on memory impairment, mice were randomly divided into seven groups of 8 mice, and MPW prepared in normal saline was orally administered at the dosages of 10, 50, 100, and 300 mg/kg from day 1 to day 4, respectively. Donepezil was also administered at 10 mg/kg to serve as a positive control. After 30 min of MPW or donepezil administration, intraperitoneal injection of scopolamine (2 mg/kg) was used to induce memory impairment. After 30 min of scopolamine administration, training trials were performed each day from day 1 to day 4, as described above. On day 5, the probe trial was performed after the hidden platform was removed. The control group received vehicle (normal saline) only.

Statistical Analysis
All the experiments were carried out at least three times independently. Quantitative data are expressed as the mean ± S.E.M. Statistical analyses were performed using one-way ANOVA followed by Tukey's post hoc test with SigmaPlot 12.5 software (Systat Software, San Jose, CA, USA). p < 0.05 was considered statistically significant.

Preparation and Analysis of MPW
MPW was prepared from the dried pericarps of Garcinia mangostana L. and analyzed via UPLC-UV and Mass spectrometry, as described above. The yield of MPW was 3.8% (w/w). The spectroscopic chromatograms of MPW from UPLC-UV and Mass analyses exhibited several prominent peaks as well as many minor peaks (see Supplementary Materials). Among the peaks detected, both αand γ-mangostins were found to be the major constituents of MPW. In addition, the peaks representing mangostanin and mangostanol were also observed along with several unidentified minor peaks. In this study, we evaluated the protective effects of MPW against the oxidative neurotoxicity triggered by various insults, such as Aβ or excitotoxic amino acids, in primary rat cortical cells. Furthermore, we investigated the memory-improving effect of MPW on scopolamine-induced impairment of spatial learning and memory in mice.

Effects of MPW on A(25-35)-Induced Caspase 3 Activation and DNA Fragmentation in Primary Rat Cortical Cells
To elucidate the probable mechanism(s) by which MPW suppressed A induced neurotoxicity, we examined the effect of MPW on the A(25-35)-in The cultured cells were exposed to 40 µM Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) in MEMG for 24 h in the absence or presence of MPW at the indicated concentrations. The cell viability was determined by the MTT reduction assay at 22-24 h of the exposure, as described in the Materials and Methods. The viability of each treated group is expressed as a percentage of that of the vehicle-treated control cells. (B) The cultured cells were pre-incubated with 10 µM DCFH-DA for 30 min at 37°C in the dark and then treated with 40 µM Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) in MEMG in the absence or presence of MPW at the indicated concentrations. The intracellular ROS levels were measured by the fluorescence detection of 2 ,7 -dichlorofluorescein, as described in the Materials and Methods. The ROS levels in each treated group are expressed as percentages of those in the vehicle-treated control cells. Each data point represents the mean ± S.E.M. from at least three independent experiments conducted in duplicate (#, p < 0.05 vs. the vehicle-treated control cells without MPW; *, p < 0.05 vs. the Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) -treated cells without MPW).

Effects of MPW on Glutamate-or NMDA-Induced Neurotoxicity and ROS Generation in Primary Rat Cortical Cells
We next evaluated the effect of MPW on excitotoxic neurotoxicity induced by glutamate or NMDA in primary rat cortical cells. The cells were treated with 100 μM glutamate or 300 μM NMDA for 15 min in the absence or presence of MPW at various concentrations, and the cell viability was assessed by the MTT reduction assay at 22-24 h of the exposure. As illustrated in Figure 3A, treatment of the cells with glutamate or NMDA reduced the cell viability by approximately 40%, compared to the viability of the vehicle-treated control cells. The reduced viability of the glutamate-or NMDA-treated cells was reversed by MPW treatment (Figure 3A). While the glutamate-induced toxicity

Effects of MPW on Glutamate-or NMDA-Induced Neurotoxicity and ROS Generation in Primary Rat Cortical Cells
We next evaluated the effect of MPW on excitotoxic neurotoxicity induced by glutamate or NMDA in primary rat cortical cells. The cells were treated with 100 µM glutamate or 300 µM NMDA for 15 min in the absence or presence of MPW at various concentrations, and the cell viability was assessed by the MTT reduction assay at 22-24 h of the exposure. As illustrated in Figure 3A, treatment of the cells with glutamate or NMDA reduced the cell viability by approximately 40%, compared to the viability of the vehicle-treated control cells. The reduced viability of the glutamate-or NMDA-treated cells was reversed by MPW treatment ( Figure 3A). While the glutamate-induced toxicity was significantly suppressed by MPW at the concentrations of 30 µg/mL and above, the NMDA toxicity was significantly suppressed by MPW at as low as 1 µg/mL and completely reversed at 30 µg/mL and above. 021, 10, 34 11 of 23 was significantly suppressed by MPW at the concentrations of 30 μg/mL and above, the NMDA toxicity was significantly suppressed by MPW at as low as 1 μg/mL and completely reversed at 30 μg/mL and above. Additionally, treatment of cells with glutamate or NMDA increased the intracellular ROS levels to approximately 140% or 170%, respectively, compared to the vehicle-treated control cells ( Figure 3B). The glutamate-or NMDA-induced increases in ROS production were inhibited by MPW in concentration-dependent manners ( Figure  3B).  Additionally, treatment of cells with glutamate or NMDA increased the intracellular ROS levels to approximately 140% or 170%, respectively, compared to the vehicle-treated control cells ( Figure 3B). The glutamate-or NMDA-induced increases in ROS production were inhibited by MPW in concentration-dependent manners ( Figure 3B).

Effects of MPW on NMDA-Induced Caspase 3 Activation and DNA Fragmentation in Primary Rat Cortical Cells
We then examined the effect of MPW on the NMDA-induced apoptosis. In agreement with the previous report [26], the exposure of cultured cells to NMDA markedly increased the level of cleaved caspase 3, and this effect was suppressed by MPW treatment ( Figure 4A). Similarly, DNA fragmentation triggered by NMDA was remarkably reduced by MPW at 30 µg/mL ( Figure 4B).

Effects of MPW on NMDA-Induced Caspase 3 Activation and DNA Fragmentation in Primary Rat Cortical Cells
We then examined the effect of MPW on the NMDA-induced apoptosis. In agreement with the previous report [26], the exposure of cultured cells to NMDA markedly increased the level of cleaved caspase 3, and this effect was suppressed by MPW treatment (Figure 4A). Similarly, DNA fragmentation triggered by NMDA was remarkably reduced by MPW at 30 g/mL ( Figure 4B).

Effects of MPW on LPO and DPPH Radical Formation
The antioxidant properties of MPW were further investigated by evaluating its ability to inhibit LPO initiated by Fe 2+ and L-ascorbic acid in rat brain homogenates. In addition, the radical scavenging activity of MPW was assessed using stable free radical DPPH as a probe. Our results demonstrated that MPW considerably inhibited lipid peroxide formation in rat brain homogenates ( Figure 5A) and effectively scavenged DPPH radicals ( Figure 5B) in concentration-dependent manners. The calculated IC 50 values were 58.9 and 37.1 µg/mL, respectively. 2021, 10, 34 13 of 23

Effects of MPW on LPO and DPPH Radical Formation
The antioxidant properties of MPW were further investigated by evaluating its ability to inhibit LPO initiated by Fe 2+ and L-ascorbic acid in rat brain homogenates. In addition, the radical scavenging activity of MPW was assessed using stable free radical DPPH as a probe. Our results demonstrated that MPW considerably inhibited lipid peroxide formation in rat brain homogenates ( Figure 5A) and effectively scavenged DPPH radicals ( Figure 5B) in concentration-dependent manners. The calculated IC50 values were 58.9 and 37.1 g/mL, respectively.

Effect of MPW on β-Secretase Activity
Aβ peptide is derived from APP upon enzymatic cleavage by β-and -secretases [1,2]. To evaluate the impact of MPW on the synthesis of Aβ peptide, we measured βsecretase activity in the presence of various concentrations of MPW using an in vitro βsecretase FRET assay kit. We found that MPW inhibited β-secretase activity, with an IC50 value of 31.0 g/mL ( Figure 6).

Effect of MPW on β-Secretase Activity
Aβ peptide is derived from APP upon enzymatic cleavage by βand γ-secretases [1,2]. To evaluate the impact of MPW on the synthesis of Aβ peptide, we measured β-secretase activity in the presence of various concentrations of MPW using an in vitro β-secretase FRET assay kit. We found that MPW inhibited β-secretase activity, with an IC 50 value of 31.0 µg/mL ( Figure 6). 2021, 10, 34 14 of 23 Concentration of MPW ( g/mL)

Effect of MPW on AChE Activity
We also evaluated the effect of MPW on AChE activity, determined using Ellman's colorimetric method [33]. As shown in Figure 7, MPW significantly and concentrationdependently inhibited AChE activity at the concentrations tested. The maximal inhibition achieved at 100 g/mL of MPW was approximately 50% of the enzyme activity measured in the absence of MPW.

Effect of MPW on AChE Activity
We also evaluated the effect of MPW on AChE activity, determined using Ellman's colorimetric method [33]. As shown in Figure 7, MPW significantly and concentrationdependently inhibited AChE activity at the concentrations tested. The maximal inhibition achieved at 100 µg/mL of MPW was approximately 50% of the enzyme activity measured in the absence of MPW.

Effect of MPW on AChE Activity
We also evaluated the effect of MPW on AChE activity, determined using Ellman's colorimetric method [33]. As shown in Figure 7, MPW significantly and concentrationdependently inhibited AChE activity at the concentrations tested. The maximal inhibition achieved at 100 g/mL of MPW was approximately 50% of the enzyme activity measured in the absence of MPW.

Effect of MPW on Scopolamine-Induced Spatial Learning and Memory Impairment in Mice
Based on our findings from cell-based and cell-free in vitro assays in this study, we finally carried out the MWM test to examine whether MPW could improve spatial learning and memory impairment induced by scopolamine in mice. MPW was orally administered once daily at the dosages of 10, 50, 100, and 300 mg/kg for 4 consecutive days, as illustrated in Figure 8A. The latency time to reach the platform was measured on each day of the training trial at 1 h after MPW administration ( Figure 8B). The latency time in the vehicletreated control group showed a dramatic decrease from day 1 to day 4. In contrast, the latency time in the scopolamine group was not considerably changed during 4 days of the training trial, indicating that spatial learning capability of mice was impaired by scopolamine injection. The latency time in groups administered with MPW or donepezil, a positive reference drug, showed a tendency to be decreased during the training trial, as compared to that of the scopolamine group. Particularly, the MPW groups (50 and 300 mg/kg) and the donepezil group (10 mg/kg) significantly shortened the latency time on day 4 (*, p < 0.05 vs. the scopolamine group; Figure 8B).
On the following day (day 5), the probe trial was performed after removing the platform. The representative images tracking the swimming paths are provided ( Figure 8C). During the probe trial, the swimming time each mouse spent in the target quadrant where the platform had previously been located is shown ( Figure 8D). The swimming time the scopolamine-treated mice spent in the target quadrant was markedly decreased relative to that of the control group (#, p < 0.05 vs. the control group), indicating that spatial memory function was also impaired in this group. However, the time spent in the target quadrant was significantly restored in the groups treated with MPW at the dosages of 50 mg/kg and above. The reduced swimming time of the scopolamine group was also recovered by donepezil administration. Interestingly, MPW administration (50-300 mg/kg of dosages) recovered scopolamine-induced spatial memory impairment to the same extent as, or even better than, that of the donepezil group. MPW at 10 mg/kg did not show a significant effect ( Figure 8D). The total swimming distances measured on day 5 in all the test groups were not significantly affected by MPW or donepezil treatment (data not shown).

Effect of MPW on Scopolamine-Induced Spatial Learning and Memory Impairment in Mice
Based on our findings from cell-based and cell-free in vitro assays in this study, we finally carried out the MWM test to examine whether MPW could improve spatial learning and memory impairment induced by scopolamine in mice. MPW was orally administered once daily at the dosages of 10, 50, 100, and 300 mg/kg for 4 consecutive days, as illustrated in Figure 8A. The latency time to reach the platform was measured on each day of the training trial at 1 h after MPW administration ( Figure 8B). The latency time in the vehicle-treated control group showed a dramatic decrease from day 1 to day 4. In contrast, the latency time in the scopolamine group was not considerably changed during 4 days of the training trial, indicating that spatial learning capability of mice was impaired by scopolamine injection. The latency time in groups administered with MPW or donepezil, a positive reference drug, showed a tendency to be decreased during the training trial, as compared to that of the scopolamine group. Particularly, the MPW groups (50 and 300 mg/kg) and the donepezil group (10 mg/kg) significantly shortened the latency time on day 4 (*, p < 0.05 vs. the scopolamine group; Figure 8B).
On the following day (day 5), the probe trial was performed after removing the platform. The representative images tracking the swimming paths are provided ( Figure  8C). During the probe trial, the swimming time each mouse spent in the target quadrant where the platform had previously been located is shown ( Figure 8D). The swimming time the scopolamine-treated mice spent in the target quadrant was markedly decreased relative to that of the control group (#, p < 0.05 vs. the control group), indicating that spatial memory function was also impaired in this group. However, the time spent in the target quadrant was significantly restored in the groups treated with MPW at the dosages of 50 mg/kg and above. The reduced swimming time of the scopolamine group was also recovered by donepezil administration. Interestingly, MPW administration (50-300 mg/kg of dosages) recovered scopolamine-induced spatial memory impairment to the same extent as, or even better than, that of the donepezil group. MPW at 10 mg/kg did not show a significant effect ( Figure 8D). The total swimming distances measured on (A)

Discussion
Mangosteen has been utilized as a traditional medicine to treat various health problems, such as wounds, infections, inflammation, and diarrhea, in Southeast Asia [35].
Mangosteen pericarp and its bioactive xanthones have been documented to elicit biological actions in the brain, and thus, may exhibit beneficial effects to prevent or treat various brain disorders, including AD, Parkinson's disease, and depression [36]. Diverse extracts, such as methanol, chloroform, and 50% ethanol extracts, prepared from mangosteen pericarp were reported to exhibit a wide variety of biological activities, including antioxidant, anti-inflammatory, analgesic, anti-cancer, and neuroprotective effects [15,20]. However, the antioxidant and neuroprotective effects of the water extract (MPW) have not been reported yet. Since MPW has been prepared without using any organic solvents, it is expected to have advantages with regard to safety concerns when administered to animals and humans. In this study, we prepared MPW from dried mangosteen pericarps and used primary cultured rat cortical cells to investigate its effects on the neurotoxicity triggered by various insults. Furthermore, we also evaluated the effect of MPW on scopolamine-induced memory dysfunction in mice by the MWM test.
We revealed the protective effects of MPW against the neurotoxicity, oxidative stress, and apoptotic processes induced by numerous insults, such as Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) , glutamate, and NMDA. In addition, MPW possessed antioxidant properties, inhibiting LPO and scavenging DPPH radicals. Moreover, MPW was demonstrated to inhibit β-secretase and AChE activities, which are associated with the typical pathological features of AD. Finally, this study showed that the impairment of the spatial learning and memory function in the scopolamine-treated mice was markedly alleviated by orally administered MPW. Our findings suggest that MPW may exert memory-enhancing effects, presumably through antioxidative neuroprotection as well as anti-apoptotic action.
Glutamate is the crucial excitatory neurotransmitter in the CNS. The interaction between glutamate and its excitatory amino acid receptors is necessary for normal synaptic function. However, for any reason, if the receptor activation becomes prolonged or excessive, the target neurons will be damaged and ultimately die [9]. The excitotoxicity of glutamate is mediated, in most cases, by the interaction with NMDA receptors, which leads to uncontrollable rises of intracellular Ca 2+ concentrations, resulting in cell lysis and death [46]. In the present study, the treatment of the cultured cells with glutamate or NMDA considerably suppressed the cell survival, compared to that of the vehicle-treated control cells ( Figure 3A). The decreased viability of the glutamate-or NMDA-treated cells was reversed by MPW, suggesting the neuroprotective properties of MPW against excitotoxicity.
Excessive NMDA receptor activity also triggers a cascade of apoptotic-like cell death. NMDA receptor hyperactivation activates the p38 mitogen-activated kinase and then myocyte enhancer factor 2C (MEF2C). MEF2C is a transcription factor that influences neuronal injury and apoptosis. Additionally, overactivation of NMDA receptors accentuates the toxic effects of free radicals, including ROS and nitric oxide, and activates caspases and apoptosis-inducing factors [47]. Further evidence has demonstrated that the activation of NMDA receptors is required for Aβ-induced over-production of ROS [48], which eventually causes the neurodegeneration and memory dysfunction seen in the brains of AD patients [36,49,50]. Therefore, dysregulation of NMDA receptor activity and oxidative stress may play dual deleterious roles in AD [51]. Notably, our results showed that glutamate or NMDA treatment of the cells caused a significant increase in ROS production, which was considerably attenuated by MPW ( Figure 3B). Moreover, it was revealed that MPW treatment reduced the apoptotic processes, including NMDA-induced caspase-3 activation and DNA fragmentation (Figure 4). These results indicate the protective effects of MPW against the neurotoxicity, oxidative stress, and apoptotic processes in glutamateor NMDA-treated cells.
The antioxidant properties of MPW were further illustrated in cell-free bioassays, with the inhibition of LPO induced by Fe 2+ and L-ascorbic acid in rat brain homogenates ( Figure 5A) and scavenging activity of DPPH radicals ( Figure 5B). These antioxidant effects of MPW may be correlated with its inhibitory effects on the ROS generation triggered by Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) , glutamate or NMDA (Figures 1B and 3B). Our findings are consistent with previous reports involving other extracts prepared from mangosteen pericarps. The 50% ethanol extract exhibited potent antioxidant effects with high free-radical scavenging and neuroprotective actions in H 2 O 2 -treated NG108-15 cells [23]. Interestingly, the water-soluble partition of methanol or ethanol extracts successfully protected SK-N-SH neuroblastoma cells from cytotoxicity and inhibited the increases in ROS levels and caspase 3 activity induced by Aβ  or H 2 O 2 and polychlorinated biphenyls treatments, respectively [52]. Moreover, the antioxidant and anti-apoptotic activities of the water-soluble partition of the ethanol extract were observed in the brains of scopolamine-treated mice [52]. It would be interesting if we could compare the in vitro and in vivo efficacies of these extracts with those of MPW under the same experimental conditions and identify bioactive components in these preparations.
Spectroscopic chromatograms of MPW revealed that it contained αand γ-mangostins and considerable amounts of mangostanin and mangostanol, along with several unidentified minor peaks (Supplementary Materials). The αand γ-mangostins were shown to have neuroprotective activities against glutamate-induced HT22 hippocampal neuronal cell death by upregulating HO-1 protein level [53]. In another study, however, only γmangostin, not α-mangostin, was reported to protect the primary rat cortical cells against both H 2 O 2 -and xanthine/xanthine oxidase-induced oxidative neuronal death as well as ROS production [24]. This xanthone also showed its significant anti-apoptotic activities, suppressing H 2 O 2 -induced DNA fragmentation and activations of caspases-3 and -9. In contrast to α-mangostin, γ-mangostin effectively reduced DPPH radical formation and LPO in rat brain homogenates [24]. Although data related to the effects of mangostanin and mangostanol on neuronal cells are still limited, mangostanin was suggested to exhibit antioxidant effects through induction of quinone reductases in the Hepa 1c1c7 mouse hepatoma cells [54]. Therefore, it can be speculated that γ-mangostin and possibly α-mangostin and mangostanin may contribute, at least in part, to the antioxidative neuroprotection and memory-enhancing effects of MPW. However, further investigation is needed to identify the bioactive constituent(s) of MPW. Collectively, the MPW-induced antioxidative neuroprotection observed in our study is consistent with the effects of various mangosteen pericarp extracts and bioactive xanthones.
β-Secretase, along with γ-secretase, produces Aβ by sequentially cleaving APP. Inhibition of β-secretase is considered to be a potential strategy to lower cerebral Aβ levels to treat and prevent AD [55]. In this study, MPW potently inhibited β-secretase activity with an IC 50 value of 31.0 µg/mL (Figure 6), suggesting a promising MPW-mediated suppression of Aβ production. Intriguingly, β-secretase activity was also abolished by xanthones from mangosteen pericarp, including αand γ-mangostins [24,53]. Particularly, α-mangostin diminished Aβ  and Aβ  production through the inhibitions of βsecretase and possibly γ-secretase activities in the amyloidogenic pathway in primary rat cortical cells [56]. The inhibition of β-secretase activity by MPW observed in this study may also be due to its constituents including αand γ-mangostins.
Moreover, the intellectual deficiency in AD patients has been linked to a selective loss of cholinergic function in specific parts of the brain, such as the cortex and hippocampus [57]. Therefore, the primary therapeutic strategy in the treatment of AD is based on the restoration of the cholinergic function [58]. The inhibitors of cholinesterase, especially AChE, are designed to attenuate the breakdown of acetylcholine and retain its activity at cholinergic synapses. Clinical applications of the currently approved cholinesterase inhibitors are generally considered symptomatic treatments to improve cognitive, daily, and global functions, and some behavioral manifestations of AD [59]. As shown in Figure 7, MPW significantly attenuated AChE activity at the concentrations tested in this study. From our preliminary experiments, we observed some, but significant toxicity in the cells treated with MPW for 24 h at the concentration of 300 µg/mL, while no toxicity was detected up to 100 µg/mL of MPW (data not shown). Accordingly, the effects of MPW were examined at the concentrations up to 100 µg/mL in this study. The inhibition of AChE activity by MPW at 100 µg/mL was approximately 50% of the enzyme activity measured in the absence of MPW (Figure 7). Growing evidence has demonstrated the inhibition of AChE activities by MPE or MPE components. The water-soluble partition of ethanol MPE dramatically diminished AChE activity of SK-N-SH cells [60]. Moreover, pretreatment with 50% ethanol MPE in streptozotocin-injected male SA mice induced a dose-dependent decrease in AChE activity in the brain [61]. Additionally, α-mangostin and mangostanol were reported to exhibit selective inhibition of AChE activity assessed by Ellman's colorimetric method [62]. Since both α-mangostin and mangostanol have been found in MPW (Supplemental Materials), these components may be responsible for the inhibition of AChE activity. Taken together, our study demonstrated using various cell-based and cell-free in vitro models that MPW exhibited antioxidative and anti-apoptotic neuroprotection as well as inhibition of β-secretase and AChE activities.
Injection of scopolamine, an anti-muscarinic agent, is known to cause progressive impairment of learning and memory in animals [63,64]. Moreover, the scopolamine-induced memory deficit is associated with the enhancement of oxidative stress and induction of apoptotic processes in the brain [65,66]. Thus, scopolamine-induced amnesia is widely used as an animal model of memory dysfunction to evaluate potential effects of drugs for AD [66,67]. In this study, we also employed this scopolamine-induced amnesia model to evaluate the effect of orally administered MPW by using the MWM test ( Figure 8A). We found that the administration of MPW or donepezil, a well-known AChE inhibitor clinically used to treat AD patients, significantly decreased the latency time during the training trial and markedly increased the time spent in the target quadrant ( Figure 8B-D). Interestingly, the memory-enhancing effect of MPW at the dosages of 50-300 mg/kg was nearly comparable to or even better than that of the donepezil-treated group. Notably, similar results were previously reported for the water-soluble partition of ethanol extracts, which was also revealed to improve scopolamine-induced memory dysfunction in mice, as assessed via the MWM and passive avoidance tests [60]. Moreover, pretreatment of streptozotocin-injected mice with 50% ethanol extract increased the habituation memory and cognitive function, as measured by the open field and Y-maze tests [61]. Furthermore, mangosteen pericarp diet exhibited neuroprotective and antioxidant activities in triple transgenic AD mice. Hippocampal Aβ deposition was also attenuated, which might further diminish the impairment of spatial memory retrieval in the MWM test [68]. Considerably, the oral administration of γ-mangostin successfully reversed the scopolamine-induced memory deficits in mice [24]. Taken together, our in vitro and in vivo findings in the current study imply that MPW exerts memory-enhancing effects in scopolamine-treated mice, probably through its antioxidative neuroprotection and anti-apoptotic action.

Conclusions
In this study, we used primary cultured rat cortical cells to investigate the effect of MPW on neurotoxicity induced by various insults. We found that MPW inhibited neurotoxicity and production of intracellular ROS triggered by Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) or excitatory amino acids. MPW inhibited caspase 3 activation and DNA fragmentation in Aβ (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) or NMDA-treated cells, suggesting that MPW has an anti-apoptotic action. Additionally, MPW inhibited LPO and scavenged DPPH radicals, thus assuring its antioxidant properties. Moreover, MPW suppressed β-secretase and AChE activities, which might also influence its effect on memory function. Finally, the oral administration of MPW significantly reversed the memory impairment in scopolamine-treated mice, with decreased latency time to find the platform and markedly increased swimming time in the target quadrant. These findings demonstrated the antioxidative neuroprotection and anti-apoptotic action of MPW, and thereby the spatial memory function of scopolamine-treated mice was improved. Accordingly, our study provides in vitro and in vivo preclinical evidence to support therapeutic potential of mangosteen pericarp, particularly its water extract, for prevention or treatment of cognitive impairment associated with AD.