Chronic stress induces Alzheimer’s disease-like pathologies through DNA damage-Chk1-CIP2A signaling

Acute and chronic stress induce DNA damage in animal models

To explore whether DNA damage occurs under different stress conditions, we established three stress animal models. Chronic unpredictable mild stress (CUMS) model is a reliable chronic stress model, rats are randomly stimulated with different stressors to induce anxiety- and depression-like behaviors [25]. Another common chronic stress model is chronic restraint stress (CRS) model. In this model, mice are restrained movements for 3 hours per day for 21 consecutive days. It represents a predictable, mild and effective stress model. At last, another group of mice are limited movements for 3 hours just once, which is used to induce acute restraint stress (ARS) with physical and emotional dysfunction temporarily [26]. We observed significant DNA damage in the brain of CUMS rat model (Figure 1A–1D), which was manifested by increased γH2A.X level in hippocampus (Figure 1A, 1B). The DNA damage was prominent in hippocampal CA1, DG and brain cortex, and mainly occurred in neurons, manifested by increased immunostaining of γH2A.X in NeuN-positive cells, not in GFAP-positive or Iba1-positive cells (Figure 1C, 1D and Supplementary Figure 1A–1C). Similarly, DNA damage was observed in the hippocampus of the CRS model (Figure 1A, 1B). In addition to chronic stress, the acute restraint stress also caused DNA damage. Compared to control, increased DNA damage occurred in hippocampal CA1 region and cortex of ARS model, not in hippocampal CA3 and DG regions (Figure 1E, 1F). These data collectively indicate that stress induces DNA damage, and neurons in CA1 region and brain cortex may be more susceptible to DNA damage than those in other regions.

Figure 1. Acute and chronic stress induce DNA damage in animal models. (A) Representative immunoblots of γH2A.X and β-actin in hippocampal tissues from chronic unpredictable mild stress (CUMS), chronic restraint stress (CRS) model and control animals. (B) Quantification of the relative protein levels of γH2A.X, which were normalized to the β-actin levels. n = 3 per group. (C) Representative fluorescence images of γH2A.X (green), NeuN (red) and DAPI (blue) in hippocampus (CA1 and DG region) and cortex in CUMS rats. Scale bar: 50 μm. (D) Quantitative analysis of the fluorescence intensity of γH2A.X in C. n = 5 per group. (E) Representative fluorescence images of γH2A.X (green), NeuN(red) and DAPI (blue) in hippocampal CA1 region and cortex in acute restraint stress (ARS) mice. Scale bar: 50 μm. (F) Quantitative analysis of the fluorescence intensity of γH2A.X in E. n = 4 per group. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Stress hormone causes DNA damage, activates Chk1-CIP2A pathway, results in tau hyperphosphorylation, Aβ overproduction and synaptic impairments in primary neurons

To mimic chronic stress in primary neurons, and evaluate whether AD-related pathologies appear in stress condition and its potential mechanisms, we used two stress hormones, corticosterone (CORT) and corticotropin-releasing factor (CRF) to treat primary neurons. The results showed that the protein level of γH2A.X was increased in dose-dependent manner in primary neurons treated with CORT and CRF for 24 hours (Supplementary Figure 2B, 2D). Meanwhile, LDH assay showed 300 or 400 μM CORT which induced DNA damage also caused significant cytotoxicity, whereas 50 nM or 100 nM CRF did not induce significant cell death (Supplementary Figure 2A, 2C). In addition, we attempted to treat primary neurons with lower dose of CORT (10 or 50 μM) for 48 hours. However, 50 μM CORT induced cytotoxicity but no DNA damage (Supplementary Figure 2A, 2B). Thus, we used CRF to treat primary neurons to mimic the chronic damage induced by stress. While 50 nM CRF did not cause tau hyperphosphorylation at Thr231 and Ser396, it was less stable than 100 nM CRF in causing AD-like pathologies. So, we finally chose primary neurons treated with 100 nM CRF for 24 hours as the chronic stress cell model. We found that besides causing DNA damage, 100 nM CRF incubation induced tau hyperphosphorylation (at Thr231, Ser396 and Ser404) and phosphorylation of APP at Thr668 in primary neurons (Figure 2A, 2B). Moreover, Aβ₄₀ and Aβ₄₂ levels were increased in culture media of neurons after CRF treatment, suggesting that CRF could induce Aβ overproduction (Figure 2C). We further detected whether CRF impaired synaptic function, and found that the presynaptic protein Synapsin I and postsynaptic proteins including PSD95 and GluA1 were reduced (Figure 2D, 2E). These results indicate that CRF promotes AD-like pathologies and synaptic impairments in primary neurons.

Figure 2. Stress hormone activates DNA damage-Chk1-CIP2A pathway, results in tau hyperphosphorylation, Aβ overproduction and synaptic impairments in primary neurons. (A–E) Primary neurons were treated with 100 nM corticotropin-releasing factor (CRF) for 24 hours. (A) Representative immunoblots of γH2A.X, Chk1-pS317, Chk1-pS345, Chk1, CIP2A, Tau-pT231, Tau-pS396, Tau-pS404, Tau-5 (total tau), APP-pT668, APP and β-actin. (B) Quantification of the relative protein levels; non-phosphorylated proteins such as γH2A.X and CIP2A were normalized to the β-actin levels; phosphorylated Chk1-pS317, Chk1-pS345, Tau-pT231, Tau-pS396, Tau-pS404 and APP-pT668 were normalized to corresponding total Chk1, tau (Tau-5) and APP respectively. n = 3 or n = 6 per group. (C) The relative concentration of Aβ₄₀ and Aβ₄₂ in supernatant and cell lysate of primary neurons detected by ELISA. n = 3 or n = 4 per group. (D) Representative immunoblots of Synapsin I, PSD95, GluA1 and β-actin. (E) Quantification of the relative protein levels, which were normalized to the β-actin levels. n = 3 per group. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

We further explored the role of Chk1-CIP2A signaling in inducing AD-like pathologies under stress condition and found CRF treatment resulted in a marked increase in active forms of Chk1 (phosphorylated at Ser317 and Ser345) as well as CIP2A in neurons (Figure 2A, 2B). Based on the previous findings that upregulated ChK1-CIP2A signaling promoted AD pathologies, we demonstrate that CRF could induce DNA damage and activate Chk1-CIP2A signaling pathway, eventually resulting in AD related pathologies and synaptic impairments.

Chronic stress induces cognitive deficits in mice, which could be prevented by vitamin C and Chk1 inhibitor

Next, we established chronic restraint stress mouse model to further investigate the impact of chronic stress on cognitive function and the role of DNA damage-Chk1-CIP2A signaling pathway in this process. Meanwhile, on the basis of in vitro results, we also tried to interfere the possible key pathogenic mechanism through targeting DNA damage and Chk1 activation by vitamin C and Chk1 inhibitor respectively in CRS models. Only male C57BL/6 mice were used because this study is not to examine possible sex differences in the response to chronic restraint stress, but rather to identify the possible mechanism that cause cognitive deficits under this stress. In addition, better molding, less variability [27] and better therapeutic effect [28] may be achieved with male rodents than females. As shown in Figure 3, CRS models were successfully established after 21 consecutive days of restraint. In open field test (OFT), the center time (s) was decreased in CRS group compared with control group, suggesting that CRS group showed anxiety phenotype. The total moving time (s) and distance (cm) showed no difference, indicating neither group had impaired locomotion activity (Figure 3B). Elevated plus-maze test (EPM) was also used to detect anxiety-like behaviors. CRS mice showed decreased tendency of time in open arm (%) (Figure 3C). In sucrose preference test, CRS mice exhibited reduced preference for sucrose compared with control mice (Figure 3D). And immobility time (s) was increased both in tail suspension test (TST) and forced swimming test (FST) (Figure 3E, 3F). These data indicated that CRS mice exhibited depression-like behaviors. In the detection of cognitive function, we found that CRS showed no preference exploring the new subject in novel objective recognition test (NOR) (Figure 3G, 3H), and impaired spatial learning (Figure 3J) and memory (Figure 3I) in Barnes maze test (BMT). Collectively, chronic stress causes both anxiety/depression and cognitive dysfunction in mice.

Figure 3. Chronic stress induces cognitive deficits in mice, which could be prevented by vitamin C and Chk1 inhibitor. (A) Schematic diagram for establishing chronic restraint stress model and treatment with vitamin C or Chk1 inhibitor. (B) The total time (s, left), distance (cm, middle) and time in central area (s, right) in open field test. (C) The time spending in open arm (%) in elevated plus maze test. (D) The sucrose preference ratio (%) in sucrose preference test. (E) The immobility time (s) in tail suspension test. (F) The immobility time (s) in forced swimming test. (G) The experimental design of novel object recognition test. Above is the training trial, below is the testing trial (after 24 hours). (H) Left: the recognition index to object A and object B in the training trial; Right: the recognition index to object A and object C in the testing trial. (I) The latency of the mice to find the target holes on different training day in Barnes maze test. (J) Errors (explore incorrect holes) (% in total) in the probe trial of Barnes maze test. All data represent mean ± SEM, n=10 per group, *P<0.05, **P<0.01, ***P < 0.001, ****P< 0.0001.

To prevent the activation of DNA damage-Chk1-CIP2A signaling pathway, we used vitamin C as an antioxidant to protect DNA and Chk1 inhibitor GDC-0575, the latter has been identified to effectively prevent AD-like pathologies and cognitive deficit in APP/PS1 mice through targeting Chk1-CIP2A signaling pathway [23]. Our data showed that both interventions could effectively rescue the cognitive impairment induced by stress in CRS mice (Figure 3H–3J). However, Chk1 inhibitor showed no therapeutic effect on the anxiety and depression of CRS mice; and vitamin C only showed the tendency of rescuing the depression phenotype in CRS mice (Figure 3B–3F). These data suggest that stress might induce anxiety, depression and cognitive deficit through different mechanisms, whereas DNA damage-Chk1 signaling pathway might be more specifically involved in promoting cognitive dysfunction.

Vitamin C and Chk1 inhibitor reduce DNA damage, Chk1 activation and CIP2A expression, decrease tau phosphorylation and Aβ levels in mice exposed to chronic stress

To further identify the role of DNA damage-Chk1-CIP2A signaling pathway in cognitive dysfunction caused by stress, we further detected key proteins in this signaling pathway and AD related pathologies. It was found that DNA damage was increased, Chk1 was activated and the expression of CIP2A was elevated in hippocampus and cortex of the brains of CRS mice. These changes were rescued by vitamin C and Chk1 inhibitor (Figure 4A, 4B and Supplementary Figure 3A, 3B). Moreover, vitamin C and Chk1 inhibitor reversed APP and tau hyperphosphorylation, the events downstream of CIP2A upregulation both in hippocampus and cortex in CRS mice (Figure 4C, 4D and Supplementary Figure 3C, 3D). In line with these changes, both treatments decreased Aβ levels in soluble and insoluble fractions of hippocampal tissues from CRS mice (Figure 4E, 4F and Supplementary Figure 3E, 3F). Taken together, these results suggest that chronic stress contributes to AD-like pathologies in which DNA damage-Chk1-CIP2A signaling axis may play an important role.

Figure 4. Vitamin C and Chk1 inhibitor reduce DNA damage, Chk1 activation and CIP2A expression, decrease tau phosphorylation and Aβ levels in hippocampus of mice exposed to chronic stress. (A) Representative immunoblots of γH2A.X, Chk1-pS317, Chk1-pS345, Chk1, CIP2A, β-actin in hippocampal tissues of mice in different groups. (B) Quantification of the relative protein levels; non-phosphorylated proteins such as γH2A.X and CIP2A were normalized to the β-actin levels; phosphorylated Chk1-pS317, Chk1-pS345 were normalized to total Chk1. n = 3 per group. (C) Representative immunoblots of Tau-pT231, Tau-pS396, Tau-pS404, Tau-5, APP-pT668, APP, β-actin in hippocampus of mice in different groups. (D) Quantification of the relative protein expression levels; phosphorylated Tau-pT231, Tau-pS396, Tau-pS404 and APP-pT668 were normalized to Tau-5 and total APP respectively. n = 3 per group. (E) The Aβ₄₀ and Aβ₄₂ in soluble fraction of hippocampal tissues in different groups were detected by ELISA kit. n = 3 per group. (F) The Aβ₄₀ and Aβ₄₂ in insoluble fraction of hippocampal tissues in different groups were detected by ELISA kit. n = 3 per group. All data represent mean ± SEM *P<0.05, **P<0.01, ***P < 0.001, ****P < 0.0001.

Vitamin C and Chk1 inhibitor rescue neuron loss and synaptic impairments in mice exposed to chronic stress

Impaired synaptic plasticity and neuronal damage underlies the cognitive impairment, and we have identified that Chk1 inhibitor has the ability to protect synapses and neurons from the injury of AD pathologies in APP/PS1 mice through reducing Aβ production and tau hyperphosphorylation via downregulating CIP2A [23]. Thus, we evaluated the damage of synapses and neurons in CRS mouse brains and tried to explore whether vitamin C and Chk1 inhibitor exert the same protective effect in this stress model. We detected the levels of pre- and post-synaptic proteins in hippocampal and cortex tissues of the mice in different groups. It was found that the level of presynaptic protein Synapsin I as well as postsynaptic proteins including PSD95 and GluA2 were reduced or showed a tendency of downregulation in CRS mice compared with control mice, and recovered to or even higher than normal levels after supplemented with vitamin C and Chk1 inhibitor (Figure 5A, 5B). Besides, Nissl staining showed neuron loss in CA1 region, but not in CA3 or DG region in hippocampus of CRS mice, was restored by vitamin C and Chk1 inhibitor (Figure 5C, 5D). As mentioned above (Figure 1), neurons in the CA1 region were most vulnerable to DNA damage in hippocampus. This data supports that DNA damage might be an upstream factor of neuronal loss. Taking together, vitamin C and Chk1 inhibitor which inhibit DNA damage-Chk1-CIP2A signaling rescue synaptic impairments and neuronal loss in CRS mice, reinforcing the conclusion that chronic stress causes AD-like cognitive impairment through DNA damage-Chk1-CIP2A signaling.

Figure 5. Vitamin C and Chk1 inhibitor rescue neuron loss and synaptic impairments in mice exposed to chronic stress. (A) Representative immunoblots of Synapsin I, PSD95, GluA1, GluA2, β-actin in hippocampus (left) and cortex (right) of mice in different groups. (B) Quantification of the relative protein expression levels, which were normalized to the β-actin levels. n = 3 per group. (C) Representative images of Nissl staining in the CA1, CA3 and DG region of mice hippocampus. Scale bar: 50 μm (CA1 and CA3) or 100 μm (DG). (D) The quantitative analysis of the area occupied by neurons in the CA1, CA3 and DG region. n = 8-11 per group. All data represent mean ± SEM, *P<0.05, **P<0.01, ****P< 0.0001.

Antibodies and reagents

The primary antibodies used in this research were as follows: anti-γH2A.X (ab26350, Abcam; 80312S, Cell Signaling Technology), anti-NeuN (26975-1-AP, Proteintech), anti-GFAP (12389S, Cell Signaling Technology), anti-Iba1 (ab178846, Abcam), anti-Chk1-pS317 (12302S, Cell Signaling Technology), anti-Chk1-pS345 (39233, Gene Tex; A21009, Abclonal), anti-Chk1 (A7653, Abclonal), anti-CIP2A (A12267, Abclonal), anti-Tau-pT231 (R381181, Zen-Bioscience), anti-Tau-pS396 (381213, Zen-Bioscience), anti-Tau-pS404 (11112, SAB), Tau-5 (ab80579, Abcam), anti-APP-pT668 (6986S, Cell Signaling Technology), anti-APP (60,342–1-Ig, Proteintech), anti-PSD95 (36233S, Cell Signaling Technology; 30255-1-AP, Proteintech), anti-Synapsin- I (5297S, Cell Signaling Technology), anti-GluA1 (13185S, Cell Signaling Technology), anti-GluA2 (13607S, Cell Signaling Technology), anti-β-actin (AC026, Abclonal). The second antibodies for Western blotting were purchased from LICOR-Biosciences (Cat#926-32210, IRDye-800CW Goat anti-Mouse IgG Secondary Antibody; Cat#926-32211, IRDye-800CW Goat anti-Rabbit IgG Secondary Antibody). The second antibodies for immunofluorescence were obtained from Proteintech (Cat#SA00013-1, CoraLite488-conjugated Goat Anti-Mouse IgG; Cat#SA00013-4, CoraLite594-conjugated Goat Anti-Rabbit IgG). Corticosterone (Cat# HY-B1618), Corticotropin-releasing factor (human) (Cat# HY-P0086) and GDC-0575 (Cat#HY-112167) were from MedChemExpress, USA. Vitamin C (Cat# 255564) was purchased from Sigma-Aldrich, USA.

Animals and treatment

C57B6/L mice (male, 8 weeks old, 22 ± 0.5 g) were supplied by Liaoning Changsheng biotechnology. Sprague-Dalley rats (female, 8 weeks old, 250 ± 25 g) were obtained from the Experimental Animal Central of Disease Control and Prevention Center of Hubei Province. All animals were housed under a regular 12/12 hours light/dark cycle, appropriate temperature (22 ± 2° C) and humidity (55 ± 15%) with unrestricted access to food and water.

Mice were randomly divided into four groups: the control group (Con), the chronic restraint stress group (CRS), the CRS treated with Chk1 inhibitor (GDC-0575) group (CRS+GDC) and the CRS treated with vitamin C group (CRS+Vc). GDC-0575 was dissolved in 10% DMSO (25 mg/ml), 40% PEG300, 5% Tween-80, and 45% saline and mixed to obtain a final concentration of 3.75 mg/ml before every gavage. The CRS+GDC group were orally administered with GDC-0575 (25 mg/kg) in 6 days of 3 weeks. Other mice were treated by the corresponding solvent. The vitamin C were solubilized in distilled water and the CRS+Vc group were provided with 0.75 g/ml vitamin C in drinking water one week prior to chronic restraint stress for 4 weeks, while others received normal drinking water. Because vitamin C is easily to be degraded, the solutions were prepared fresh daily.

Mice exposed to chronic restraint stress were put into a well-ventilated 50 mL centrifuge tube to restrain movements for 3 hours (10:00 to 13:00) per day for 21 consecutive days. The acute restraint stress group were limited in a same tube for three hours and sacrificed immediately following stress.

The rats exposed to chronic unpredictable mild stress were given the following stressors: (1) electrical shock to the feet (2 seconds, 0.5 mA, 5 times in 10 minutes); (2) cold water bath (20 minutes); (3) water deprivation (12 hours); (4) food deprivation (24 hours); (5) reversed day/night light cycle (24 hours); (6) physical restraint (2 hours); (7) low temperature (4° C, 2 hours); (8) high temperature (45° C, 20 minutes); (9) Cage tilting (45° angle, 24 hours); (10) tail pinch (90 seconds); (11) crowed living condition (12 hours); (12) moist bedding (24 hours); (13) horizontal vibration (100 rpm, 20 minutes). A combination of 0-3 kinds of stressors were randomly applied each day and the same stressor was not used continuously within 2 days during the period of 28 days.

Behavioral tests

Open field test (OFT)

The test was performed in an empty area (50 cm × 50 cm × 50 cm plastic container, Techman Software Co, Ltd, Chengdu, China). The area was divided into 25 equal squares (5 × 5 sectors), the middle 3 × 3 sectors were denoted as central area, and others were peripheral areas. Mice were placed in the center at the beginning of test and allowed to explore freely for 5 minutes. The time and distance in total and in central area were recorded. The total time and distance were used to assess to locomotor activity. The center duration (%) were calculated to measure anxiety-like behavior.

Elevated plus-maze test (EPM)

Mice were placed in the center of an apparatus with two open arms and two closed arms for 5 minutes (25 cm long, 6 cm wide). The device was elevated 50 cm. Software measured the time spent on the open arms and the number of entries into open and closed arms. The time in open arms (%) were calculated to assess anxiety-like behavior.

Sucrose preference test (SPT)

Sucrose preference test was used to assess depression-like behavior. Mice were placed individually into a cage and given with two identical bottles containing 2% sucrose solution or pure water, respectively for 12 hours. On the second period, the two bottles were changed. Next, mice were given water deprivation for 12 hours. Finally, each mouse was provided with the two bottles mentioned above. After 24 hours, we recorded the consumption of solution in each bottle. The sucrose preference was calculated by the formula: Sucrose preference (%) = [(sucrose consumption) / (water consumption + sucrose consumption)] × 100%.

Tail suspension test (TST)

The tail of each mouse was fixed with adhesive plaster and suspended for 5 minutes. The head was 15 cm above the ground. The immobility and struggling time were recorded during the testing period. More immobility time was considered as depression-like behavior.

Forced swimming test (FST)

Mice were arranged in an open cylinder container (25 cm high×18 cm diameter) filled with fresh water (25 ± 2° C) to 15 cm high and kept the head above the water to breath. The duration of immobility was scored during the 5-min session to assess depression-like behavior.

Novel object recognition test (NOR)

The device was a 50 cm × 50 cm × 50 cm plastic box. Before the test, mice allowed to move freely in this container to adapt experiment. On the first day of training trial, object A and object B were placed on two corners in this box to allow mice to explore the two objects freely for 5 minutes. The exploring time on object A and B were recorded, denoted as TA and TB. On the second day of testing trial, the object B was replaced with another new object C, and the three objects have different color and shape. Mice were returned to the box for 5 minutes to move freely. The time spent on exploring object A and C were measured, denoted as TA and TC. The recognition index (%) was expressed as follows: The first day is TA/ (TA + TB), TB/ (TA + TB), the second day is TA/ (TA + TC), TC/ (TA + TC).

Barnes maze test (BMT)

The test was conducted in a 90 cm-diameter circular plate with 20 round holes (5 cm in diameter) on the edge. The test comprised two steps: training (4 days) and probe (1 day).

During the training trial, a hole was set as escape box (a square box with 10 cm × 10 cm × 6 cm under this hole), where mice could stay away from light. Mice were limited in a cylindrical cage at the center of the device for 5 seconds at the beginning of training. Then, mice were released to explore the area, find and enter the escape box within 3 minutes. If mice did not enter the box, the experimenter guided them into this box at the end of this session. And mice remained for 30 seconds after reaching the escape box. Each mouse was trained for three trails per day, and an interval of less than two hours was guaranteed between adjacent training sessions. In the probe trial, the escape box was removed and mice were allowed to explore the area for 90 seconds. The escape latency (the time until reaching the escape box) during training session and the number of errors (explore incorrect holes) in the probe test were recorded.

Primary neuron culture and treatment

Primary cortical neurons were isolated from E16-E18 C57B6/L mice. Cortical tissues were dissociated, and digested by trypsin for 15 minutes, followed by adding medium containing DMEM/F12 with 10% fetal bovine serum to terminate the digestion. Then cells were suspended again and plated onto 6-well plates coated with poly-D-lysine after suspension and cultured in an incubator (5% CO₂, 37° C). After 6 hours, medium was replaced with neurobasal medium supplemented with 2% B-27, 1% GlutaMAX (2 mM), 1% penicillin (50 U/mL), and streptomycin (50 μg/ mL). The medium was half-changed every 3 days during the culture. After 9 days, cells were treated with CORT and CRF at different concentrations for 24 hours. Then the neurons and culture media were harvested for following tests.

Western blotting

Cultured neurons and brain tissue were lysed with RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China) containing PMSF (1:100) and proteinase inhibitor cocktail (200 mM AEBSF, 30 μM aprotinin, 13 mM bestatin, 1.4 mM E64, and 1 mM leupeptin in DMSO, 1:100) (Yeasen Biotech, Shanghai, China Cat#20124ES03). Then cell lysates or tissue homogenates were added with the loading buffer (final concentration: 50 mM Tris-HCl, pH 7.6, 2% SDS, 10% glycerol, 10 mM DTT, and 0.2% bromphenol blue) and boiled at 100° C for 10 minutes. Next, samples were centrifuged at 12,000 g at 4° C for 10 minutes and sonicated. Finally, the concentration of protein was measured by BCA kit (Beyotime Biotechnology, Shanghai, China).

Protein was electrophoresed in 10% or 12% SDS-polyacrylamide gel and then transferred onto 0.45 μm nitrocellulose membranes (Cytiva, USA). After blocking with 5% non-fat milk for 1 hour, the membranes were incubated overnight at 4° C with primary antibodies. Then they were washed 3 times for 10 minutes each by TBST and incubated with secondary antibodies at room temperature for 1 hour. Finally, these blots were washed again as above and visualized using Odyssey Infrared Imaging System (LICOR Biosciences, USA). The intensity of the protein bands was quantitatively analyzed by Image J software (Rawak Software, Inc.).

LDH cytotoxicity assay

The LDH cytotoxicity assay was performed according to the manufacturer’s procedure (Cat# C0017, Beyotime Biotechnology, Shanghai, China).

ELISA assay

The culture media of primary neurons was collected for ELISA assay directly. For detecting the Aβ levels in cell lysates and brain tissues, cell and brain samples were lyzed in PBS (containing 1:100 PMSF and 1:100 protease inhibitor cocktail) and centrifuged for 10 minutes at 12,000 g at 4° C. The supernatant was RIPA-soluble fraction. The pellets obtained from brain tissues were further dissolved in 70% formic acid as the RIPA-insoluble fraction. The levels of Aβ₄₀ and Aβ₄₂ were detected according to the manufacturer’s instructions (Elabscience Biotechnology, Wuhan, China).

Fluorescence imaging and confocal microscopy

The mice were anesthetized by isoflurane, perfused with 0.9% normal saline and 4% paraformaldehyde (PFA) intracardially. Then the brain was removed and further fixed with 4% PFA at 4° C. After 48 hours, these brains were immersed in 30% sucrose in PB until they plunged to the bottom. Then the brains were embedded in OCT (optimum cutting temperature compound, Sakura, USA), frozen, and sectioned at 30 μm using a cryotome (CM1950, Leica, Germany). Brain slices were washed by PBS three times and permeabilized in 0.5% Triton X-100 for 20 minutes, then incubated with 5% bovine serum albumin. After blocking, primary antibodies, secondary antibodies and DAPI (Beyotime Biotechnology, Shanghai, China) were used in succession. All the images were collected by confocal microscope (LSM780, Zeiss, Germany).

Nissl staining

Brain slices prepared as above were stained according to the manufacturer’s procedure (Leagene Biotechnology, Beijing, China) and images were collected under the microscope (VS120, Olympus, Japan). The area occupied by neurons in the hippocampal CA1, CA3 or DG regions were measured by Image J software.

Statistical analyses

Data were analyzed as the mean ± SEM and analyzed using GraphPad Prism 9 software (GraphPad Software, Inc., La Jolla, CA). The one-way analysis of variance (ANOVA) with Tukey’s post hoc test was used to compare the differences among groups. And Student’s test was used to determine between two groups. The significance was set at P < 0.05, and the test was two-tailed.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.