Brain mitochondria effectively supply the adenosine triphosphate (ATP) necessary for maintaining neuronal activity and function, which, in turn, are responsible for learning and memory [1
]. Intact and healthy mitochondria are indispensable for synapse formation and dendritic remodeling [2
]. Because amyloid β (Aβ) is taken up and subsequently imported into mitochondria, mitochondrial dysfunction is prominent in Alzheimer’s disease (AD) model mice, which show a progressive impairment in cognition [3
]. In rat hippocampal slices, blockade of mitochondrial ATP production leads to deficits in long-term potentiation (LTP)—a critical mechanism for learning and memory.
Prevention and treatment of cognitive impairment is the goal of dementia and AD treatment. Natural compounds, such as resveratrol and green tea, are considered AD preventive therapeutics that reduce Aβ40 levels in cerebrospinal fluid and plasma [4
] or enhance cognition, as confirmed by a mini-mental state examination (MMSE); however, only a few such agents are in clinical trials [5
]. FDA-approved drugs, including Aricept (donepezil) and Razadyne (galantamine), both of which are cholinesterase inhibitors, prevent memory deficits, but cause side effects such as nausea, vomiting, and muscle cramps [7
]. Other agents include minocycline, a tetracycline antibiotic that targets neuroinflammation in AD [8
], and memantine, an N-methyl-D-aspartate (NMDA) receptor antagonist that attenuates damage caused by reactive oxygen species (ROS) in the brain [9
]. Although mitochondrial dysfunction is a causative factor in memory deficits in AD [3
], how mitochondrial modulation by natural compounds alters the expression of key synaptic proteins is not known. Thus, we explored the effect of mitochondrial-modulating natural compounds on hippocampal learning and memory.
extract (SCE) has been reported to exert neuroprotective effects [10
], including amelioration of cognitive deficits in a mouse model of chronic, unpredictable mild stress [11
], and improvement of synaptic morphology and plasticity in ovariectomized mice [12
]; it also protects against oxidative stress in hepatocytes [13
]. Supplementation with ascorbic acid (AA), a water-soluble vitamin C found in plants, has been reported to prevent impairment of synaptic plasticity and hippocampal LTP attributable to oxidative damage induced by lead, a neurotoxic metal [14
]. High doses of AA have also been shown to reduce amyloid plaque accumulation in AD model mice. However, whether SCE and AA directly regulate mitochondrial activity is still unknown. Mixtures of bioactive natural plant extract or nutrients are known to produce synergistic effects [15
]. Because SCE and its metabolites can induce ROS in cells [10
] and AA is known as an antioxidant [18
], combining SCE with AA would have the benefit of reducing the oxidative stress of ROS which is increased by enhancement of mitochondrial respiration. In the present study, we demonstrated that a mixture of the natural compounds SCE and AA acted through modulation of mitochondria—the main contributors to intracellular ATP production—to synergistically enhance memory storage and recognition memory in association with an increase in the expression of synaptic plasticity-regulating proteins.
2. Materials and Methods
2.1. Cell Culture
mHippoE-14 mouse embryonic hippocampal cell line was cultured in Dulbecco’s Modified Eagle’s medium (DMEM, Sigma-Aldrich, MO, USA), 10% FBS (Hyclone, MA, USA), 1% penicillin and streptomycin (Hyclone, MA, USA) at 37 °C under 5% CO2 and 21% O2 condition.
2.2. Plant Material and General Procedures of Natural Products
The Schisandra chinensis was purchased from TCM market in Seoul, Korea in September 2017. Voucher specimens (GL0680) were authenticated by us and were deposited by Dr. Chun Whan Choi at the herbarium of Bio-center, Gyeonggi Institute of Science & Technology Promotion, Suwon, South Korea. 1H and 13C NMR experiments were performed on a Bruker Ascend 700 MHz spectrometer with tetramethylsilane (TMS). LC-ESI-MS were obtained on a Triple TOF 5600+ instrument (AB SCIEX, MA, USA) and HRESI-MS on a LTQ Orbitrap XL instrument (Thermo Scientific, MA, USA). Thin Layer Chromatography (TLC) was conducted on Silica gel 60 F254 (Merck, Darmstadt, Germany) and Silica gel 60 RP-18 F254S (Merck, Darmstadt, Germany) plates. Column chromatography (CC) was performed using Silica gel 60 (70~230 mesh, Merck, Germany), ODS-A (12 nm S-7 μm, YMC GEL, Tokyo, Japan), and preparative high performance liquid chromatography (HPLC) was performed on LC-8A (Shimadzu, Japan).
2.3. Isolation and Determination of Schisandrin from Schisandra Chinensis Extract
The dried Schisandra chinensis (600 g) were ground and powdered. The powder was extracted with 70% EtOH two times at room temperature (each time for 2 days) and the combined extracts were concentrated under vacuum at 40 °C to yield 75.7 g (92 brix) of extract. The extract (70 g) was separated by Diaion HP-20 chromatography using gradient mixtures as eluents (water: MeOH; 100:0, 70:30, 30:70, 0:100), (F001-004). Compounds 1 (63.2 mg) were isolated from F003 by preparative HPLC (column: YMC-Pack ODS-A, 5 μm, 250 × 20 mm I.D., Japan, 8 mL/min, 10–35% MeCN, 40 min). Structures of Compound 1 (schisandrin) were elucidated by chemical evidence on the basis of NMR spectroscopic and MS data, and as well as by comparison with those reported.
2.4. Oxygen Consumption Rate (OCR) Measurement
mHippoE-14 cells were plated 2 × 104 cells at each well and incubated in media containing SCE, ascorbic acid (AA, Sigma-Aldrich, MO, USA) or SCE and AA mixture (10 ug/mL) for 24 h. After measurement of basal OCR, ATPase inhibitor oligomycin A (20 µg/mL, Sigma-Aldrich, MO, USA), uncoupler carbonyl cyanide 3-chlorophenylhydrazone (CCCP, 50 µM, Sigma-Aldrich, MO, USA) and mitochondrial complex I inhibitor rotenone (20 μM, Sigma-Aldrich, MO, USA) were sequentially added to each well and OCR was measured at 37 °C using XF24 analyzer (Seahorse, MA, USA).
2.5. Animal Experiments
Eight-week-old male C57BL/6 mice were used for the experiments. These mice were maintained at 22 °C under a 12-h light-dark cycle. Animal experiments were performed from 9 a.m. to 6 p.m., which is light phage of the cycle. Animal experiments were approved by the Institutional Animal Care and Use Committee of Chungnam National University (Ethical approval number, 201903A-CNU-46). Eight-week-old male mice were intraperitoneally injected with SCE, AA, or a mixture of SCE and AA (10 mg/kg per injection) three times at 24-h intervals. Control (CN) mice were injected with saline and an equal amount of N-methyl-2-pyrrolidone (NMP; Sigma-Aldrich, MO, USA) as the SCE-injection group. Behavioral testing was performed after injection. We used five mice per group for behavior tests. We performed Western blotting three times independently with different samples.
2.6. Behavioral Tests
Open-field test: Mice injected with saline (CN), SCE, AA, or an SCE and AA mixture were placed in a 40 × 40 × 40 cm box and their movements were recorded for 60 min. Distance moved and heat-map images were obtained using EthoVision XT 11.5 software.
Fear memory test: Fear tests consisting of fear conditioning, contextual memory, and cued-fear memory tests were performed using a fear chamber (Coulbourn Instruments, MA, USA). Mice were placed in the fear chamber and fear-conditioning tests were performed as three trials of stimulus, consisting of a 20-s tone (3 kHz, 80 dB), and 1 s of 0.4-mA electric shock after a 5-min habituation period. The duration of each trial was 1 min. Contextual-fear memory was analyzed 24 h after fear conditioning by placing mice in the same chamber for 5 min without a stimulus. Cued-fear memory tests were performed 24 h after contextual-fear memory tests by placing mice in a novel chamber and applying the same tone used in the fear-conditioning test for 3 min after a 5-min habituation period. The threshold for freezing time was set to 0.75 s, and freezing time and testing time, expressed as a percentage, were calculated using FreezeFrame software.
Novel object-recognition (NOR) test: The NOR test consisted of two video-recorded sessions. In session 1, two identical objects were placed at an equal distance from the center of the base of a cylinder (diameter 25 cm, height 40 cm). Mice were placed in the cylinder and allowed to freely explore the objects and space for 10 min. After this, mice were returned to their cage and the cylinder was cleaned with 70% ethanol. In session 2, one object was replaced with a novel object with a different texture and shape, and mice were allowed to explore the objects for 10 min. Time spent touching the objects with the forepaw and sniffing was considered as exploration and was measured.
2.7. Protein Extraction and Western Blot
Proteins of mice hippocampus and frontal cortex was prepared after 3 injections of SCE, AA, SCE and AA mixture (10 mg/kg) using an RIPA buffer consisting of 50 mM Tris–HCl pH 7.5, 150 mM NaCl, 0.1% SDS, 0.5% deoxycholate and 1% Nonidet P-40. A 10% phosphatase inhibitor and 10% protease inhibitor cocktail (Roche, Basel, Switzerland) was added before using the RIPA buffer. Next, 12 μg of protein was mixed with a 5X sample buffer and loaded on SDS-PAGE gel. The Protein sample was run on gel by electrophoresis and transferred by 200 mA to polyvinylidene fluoride (PVDF) membrane (Millipore, MA, USA). The membrane was blocked by 5% skim milk and incubated with primary antibodies including anti-PSD95 (Thermo Scientific, MA, USA), GluR1 (Abcam, Cambridge, UK), GAD65 (Abcam, Cambridge, UK), Gephyrin (Synaptic Systems, Goettingen, Germany), and α-Tubulin (Sigma-Aldrich, MO, USA) antibodies at 4 °C overnight. The membrane was incubated with a secondary anti-mouse or rabbit IgG horseradish peroxidase antibody (HRP, Pierce Biotechnology, MA, USA), which corresponds to the primary antibody host, for 1 h at room temperature and each protein band was visualized by the ECL system (Thermo Scientific, MA, USA). For ECL detection, we used medical X-ray film blue (AGFA CP-BU NEW, Belgium), developer solution, and fixer solution for the ECL detection.
2.8. Statistical Analysis
Statistical analysis of data was performed by Prizm5 software (Graphpad, CA, USA) and data are presented as mean ± SD. One-way ANOVAs and two-way ANOVAs were used for analyzing significance of difference between groups and p < 0.05 was considered statistically significant.
Cognitive defects with dendritic loss and impairment of synaptic plasticity are associated with aging and neurodegenerative diseases, including AD [28
]. Mitochondrial dysfunction and reduced expression of oxidative phosphorylation (OxPhos) complex proteins are also found in these conditions, resulting in a decrease in the ATP pool. In the present study, we analyzed the beneficial effects of SCE and AA on mitochondrial respiration and behavior. SCE is enriched for a number of bioactive ingredients, including gomisin and schisandrin, as confirmed by HPLC (Figure S1
), and is known to enhance cognitive performance and provide neuroprotection [10
]. AA is present in various plants and is used as a dietary supplement. An AA deficiency in the brain impairs cognition and increases amyloid plaque deposition [30
]. AA is known to scavenge mitochondria-generated ROS [31
], but the effects of SCE or AA on mitochondrial respiration in the brain has not previously been tested. Consistent with previous reports that combinations of plant extracts exhibit augmented biological activity compared with individual plant extracts [32
], we found that a mixture of SCE and AA at a 4:1 ratio (10 μg/mL) exerted synergistic effects on mitochondrial respiration. However, this effect was absent at lower ratios of SCE in the mixture, implying that not only do SCE and AA in the mixture additively increase mitochondrial activity, the inclusion of AA may contribute to the scavenging of ROS, produced as a result of enhanced mitochondrial respiration. However, which component of SCE is the primary contributor to the efficacy of the SCE-AA mixture remains unknown.
Neurons exploit mitochondrial-derived ATP for the regulation of synaptic plasticity and dendritic remodeling involved in learning and memory function. Hippocampal synaptic plasticity has been implicated in learning and memory, and the loss of mitochondrial ATP production has been reported to impair LTP in the hippocampus [1
]. Because neurons exclusively depend on glucose metabolisms, deletion of estrogen-related receptor gamma (ERRγ) in the cerebral cortex and hippocampus, which promotes mitochondrial oxidative phosphorylation in the brain, decreases metabolic capacity and impairs LTP and memory formation [34
]. Although synthetic agents and chemicals, including vildagliptin and low-dose USP methylene blue are known to increase mitochondrial respiration resulting in amelioration of cognitive impairment in AD [35
], there are no FDA-approved plant extracts or natural compounds for improving memory through modulation of synaptic plasticity and mitochondrial activity. We demonstrated that the combination of SCE and AA enhanced mitochondrial respiration of hippocampal neurons and increased expression of key synaptic plasticity-related proteins in the hippocampus upon injection as a mixture (4:1 ratio) in mice.
PSD95 is a major synaptic element that binds to postsynaptic NMDA receptors [36
]. Glutamatergic neurotransmission mediated by the AMPA receptor is indispensable for synaptic plasticity. In aging and AD model mice, PSD95 expression is diminished, leading to postsynaptic alterations in the cortex and cognitive decline [38
]. The glutamatergic AMPA receptor is critical for learning and memory and interacts with the PSDs molecule [39
]. The GluR1 subunit of the AMPA receptor is downregulated in postmortem brains of AD patients, and LTD decays more rapidly in hippocampal slices from mice with a genetic mutation in GluR1 [40
]. Unlike PSD95, GluR1 levels in the hippocampus were highest in mice treated with a mixture of SCE and AA, but they were also increased in mice injected with SCE or AA only. We demonstrated that the SCE-AA mixture increased both PSD95 and GluR1—excitatory synaptic transmission regulatory proteins—in the hippocampus to improve cognition in mice through an enhancement of mitochondrial respiration. It has been reported that BDNF induces mitochondrial accumulation at presynaptic sites of hippocampal neurons, and that responses to synaptic stimuli in the form of synaptic plasticity are highly dependent on ATP provided by mitochondria [42
]. In addition, BDNF treatment increases PSD95 levels and transport to dendritic spines, and enhances synapse formation by postsynaptic neurons [27
]. Viewed in this context, our findings suggest that induction of BDNF by the SCE-AA mixture enhances PSD95 protein expression in the hippocampus and frontal cortex, as well as recruitment of mitochondria during learning and memory storage.
Additionally, it is reported that a decrease in synaptic proteins such as synapsin-I, SAP97, and PSD95 were observed in a traumatic brain injury mouse model within 24 h by increasing oxidants and decreasing antioxidants [44
]. This implies that short stimulation can cause the change of synaptic protein expression. Furthermore, a decrease in PSD 95 expression also observed at 12 h, 24 h and 72 h after exposure of sevoflurane for 2 h in 2–3 months old mice accompanying impaired short-term memory [45
]. From these reports and our results, it is suggestive that the SCE-AA mixture can reduce the expression of PSD95 by short-term treatment. However, the limitation is that we only observed the short-term effects of the SCE-AA mixture and a long-term effect of the SCE-AA mixture in synaptic plasticity needs to be investigated in further studies.
The hippocampus is activated by contextual exposure, and the amygdala reacts to environments recognized as dangerous, whereas the medial prefrontal cortex associates information from these two brain regions and formulates a response [22
]. Defects in the hippocampus impair freezing responses to context as well as cued fear conditioning and recognition memory. As we observed in behavioral tests, administration of the SCE-AA mixture enhanced both recognition memory and memory storage (Figure 2
and Figure 3
) and was correlated with an induction of PSD95 and BDNF in the hippocampus rather than the frontal cortex (Figure 4
). These observations imply that the SCE-AA mixture mainly affects hippocampal memory storage, accompanied by modest effects on frontal cortex-associated memory formation. Whether the combinatorial effects of an SCE-AA mixture on learning and memory observed in healthy mice translates to mitigation of disease-associated cognitive dysfunction in scopolamine-induced memory impairment or AD models will require further investigation.