Mitochondrial dysfunction following repeated administration of alprazolam causes attenuation of hippocampus-dependent memory consolidation in mice

Repeated administration of Alp causes differential expression of mitochondrial functional proteins

Since the hippocampus is inextricably linked to memory acquisition, consolidation, and extraction, a proteomic method was utilized to identify protein changes that occur in the hippocampus of mice with repeated administration of Alp.

The principal component analysis (PCA) can adequately distinguish between the control and Alp groups (Figure 1A). In the DIA proteomics study, > 1869 proteins were identified and quantified with acceptable confidence. Among these, the expression of 439 proteins changed significantly, including 173 upregulated and 266 downregulated proteins (Figure 1B). The information for these 439 DEPs is shown in Supplementary Table 1.

As shown in Figure 1C, cluster heat map was generated for the 439 DEPs with significant changes in relative abundance. The distinction between the control and Alp groups showed that repeated Alp exposure significantly affected the biological processes in the hippocampus.

Among these, we have focused on subcellular localization of DEPs mainly in the cytoplasm, nucleus, and mitochondria (Figure 1D). Moreover, combined with the DEPs Supplementary Table 1, 7 upregulated mitochondrial proteins were identified: cytochrome c oxidase subunit IV (COX 4), cytochrome c oxidase subunit VIa (COX 6A), complex III subunit 5 (RISP), NADH dehydrogenase subunit 5 (ND5), acyl carrier protein (Ndufab1), NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 13 (Ndufa13), and optic atrophy protein 1 (OPA1). In addition, 14 downregulated mitochondrial proteins were identified, including cytochrome c oxidase subunit 6B (COX 6B), cytochrome c oxidase subunit 6C (COX 6C), succinate dehydrogenase complex subunit (CybS), fumarase, complex III subunit 3 (Cob), ATP synthase protein 8 (ATP8), mitochondrial pyruvate carrier 2 (MPC2), NADH dehydrogenase subunit 4 (ND4), NADH dehydrogenase [ubiquinone] iron-sulfur protein 6 (Ndufs6), NADH dehydrogenase [ubiquinone] iron-sulfur protein 5 (Ndufs5), NADH dehydrogenase [ubiquinone] 1 subunit C2 (Ndufc2), NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 7 (Ndufb7), NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 10 (Ndufb10), and mitochondrial fission 1 protein (FIS1) (Table 1).

Further, GO enrichment analysis (Figure 1E) displayed significant changes in subcellular localization of DEPs in mitochondria. It mainly affects cellular components including mitochondrial membrane, mitochondrial protein complex, respiratory electron transport chain, NADH dehydrogenase complex and oxidoreductase complex.

The KEGG pathway analysis was shown in Figure 1F. Totally, 15 of the main pathways were presented, and the top one was metabolic pathways, such as oxidative phosphorylation and other metabolic process including folate, glycan and degradation of valine, leucine, and isoleucine. In addition, many DEPs were enriched in the neurodegenerative diseases’ pathway, including Alzheimer’s, Parkinson’s, and Huntington’s diseases.

The protein network analysis of significantly differentiated proteins (Figure 2A) revealed that these are mitochondrial complex I-related proteins [28–30], mitochondrial complex II-related proteins [31], mitochondrial complex IV-related proteins [32, 33], and mitochondrial dynamics-related proteins [34], which interfere with energy metabolism processes primarily by influencing the mitochondrial electron transport chain (mETC) [35]. Most importantly, the 6 DEPS in above proteomic analysis were selected for further Western blotting validation, and the data (Figure 2B) showed that COX6A1 (p < 0.05), Ndufab1 (p < 0.05) and OPA1 (p < 0.01) in the Alp group were significantly up-regulated compared to the control group, conversely, Ndufs6 (p < 0.05), FIS1 (p < 0.05) and ND4 (p < 0.01) were significantly down-regulated. In summary, repeated administration of Alp leads to a disruption of mitochondrial functional proteins in the mouse hippocampus.

Repeated administration of Alp causes differential expression of related proteins including cycle, mitophagy and cognition

According to the results of KEGG pathway enrichment (Table 2), key proteins in the apoptosis and cell cycle [36–38], mitophagy [39, 40], and cognitive dysfunction-related [41, 42] pathways were identified (Figure 3). Among these DEPs, the levels of B cell lymphoma 2 (BCL2), catenin beta-1 (CTNNB1), microtubule-associated protein tau (MAPT), reticulon 4 (RTN4), reticulon 3 (RTN3), reticulon 1 (RTN1), and a disintegrin and metalloproteinase domain-containing protein 10 (ADAM 10) were significantly downregulated, while the levels of autophagy-related protein 3 (ATG3), NAD-dependent protein deacetylase sirtuin-2 (SIRT2), microtubule-associated protein 1 light chain 3a (Map1lc3a), cyclin-dependent kinase 5 activator 1 (CDK5R1), cyclin-dependent kinase 5 (CDK5), and glial fibrillary acidic protein (GFAP), were significantly upregulated.

Alp promotes HT22 cell apoptosis by influencing the cell cycle

Alp showed a decline in HT22 cell viability in a dose-dependent manner. At 48 h, HT22 cells showed 68.74 ± 5.32% and 59.85 ± 5.23% viability for 50 and 100 μM, respectively (Figure 4A). The lowest concentration of Alp (50 μM) that causes cytotoxicity was subsequently used for further experiments.

To understand the neurotoxicity of Alp-induced HT22 cells, the effects of 50 μM Alp on apoptosis and cell cycle analysis were investigated using flow cytometry. The results showed that after treatment with 50 μM Alp, the percentage of apoptosis was significantly higher than in the control group (p < 0.01) (Figure 4B, 4C).

In addition, Alp can induce HT22 cellular neurotoxicity by interfering with cell cycle distribution, leading to cell cycle arrest. Compared to the control group, Alp significantly stagnated cells in the G1 phase and reduced cell distribution in the S phase and G2/M phase (p < 0.01) (Figure 4D–4F).

The above results suggested that high concentrations of Alp inhibit the proliferation of HT22 cells and promote apoptosis by inducing G1 phase arrest in cells.

Alp induces mitochondrial dysfunction of HT22 and hippocampal neurons

When the MMP was high, JC-1 polymer in the mitochondrial matrix presented red fluorescence, and at low MMP, JC-1 polymer transformed into monomers, which was green fluorescence. The results of mitochondrial fluorescence probe detection showed that the red fluorescence was evenly distributed in the control group of cells, while the red fluorescence was only little in the cells of the Alp group (Figure 5A), indicating that the MMP of the Alp group was significantly reduced compared to the control group (Figure 5B). A decrease in MMP is a significant event of apoptosis [43]. The results of this experiment suggested that Alp-induced apoptosis of HT22 cells may be related to Alp-mediated reduction of MMP.

To further explore how Alp affects mitochondrial function, its effect on cellular energy metabolism in the primary hippocampal neurons was investigated. The results of mitochondrial stress tests are shown in Figure 5C. The basal OCR was the cellular OCR before the addition of 1.5 μM oligomycin at 15 min. FCCP (2 μM) and the oxidative phosphorylation coupling agent (2 μM) were added at 35 min to detect the maximum cell OCR. The respiratory reserve function was the difference between the maximum cell OCR and the basal cell OCR. Finally, 0.5 μM rotenone and 0.5 μM antimycin A were added at 50 min to detect the non-mitochondrial respiration of cells. OCR data (Table 3) showed that compared to the control group, 50 μM Alp significantly reduced the maximum respiratory capacity (p < 0.05) of neurons and increased their proton drain (Figure 5D, 5E) (p < 0.01), which resulted in mitochondrial function damage.

Importantly, the ultrastructure of the mouse hippocampal tissue revealed a whole structure, intact outer membrane, clear inner crest, and uniform internal matrix of the mitochondria in the control group. On the other hand, the number of mitochondria was decreased, the shape became irregular, the inner crest dissolution was missing, most mitochondria appeared as vacuoles, and the membrane was dissolved and connected to each other in the Alp group (Figure 5F).

All these in vitro and in vivo experimental results suggested that Alp induced mitochondrial damage by disrupting mitochondrial ultrastructure and energy metabolism.

Repeated administration of Alp induces memory impairment in vivo

The complete behavioral experiment schematic is shown in Figure 6A. The ICS is a fully automated system that assess home cage behavior and cognitive abilities of mice living in a social environment, respectively. This unique experimental setting promotes natural social behavior in a biologically relevant, rich but highly standardized home cage setting. According to our experimental design, all mice learned how to properly drink before the start of the test, although the learning of this protocol was temporary. According to the ICS program, after the experiment begins, each active operation of the mouse, such as visiting a corner, or each passive behavior, such as jet punishment, promotes the mouse to repeatedly learn the program to reduce the attempt to drink water correctly. Therefore, it is easy to explain that mice obtain memory consolidation through continuous reinforcement operations to reduce the drinking error rate. On the contrary, after the memory consolidation process is damaged, the mice will not be able to quickly and effectively reduce the water drinking error rate.

In this ICS test, the drinking error of mice in the Alp group increased significantly from day 5 compared to the control group (p < 0.05), although it narrowed gradually (Figure 6B). For the NSI group and the Alp + NSI group, there was no significant difference in the drinking error compared to the control group, but it was significantly reduced from day 7 compared to the Alp group (p < 0.01), further indicating that the memory consolidation process in mice was barriered after repeated administration of Alp. Since memory consolidation was impaired, whether repeated administration of Alp affects their learning and acquisition processes needs to be elucidated. Therefore, the learning ability was tested after the re-administration of Alp by NOR test. Interestingly, the results of both the 7^th and 24^th experiments did not show any significant difference in the NOR index in the Alp group, the NSI group and Alp + NSI group compared to the control group (Figure 6C). Hence, it was speculated that the learning ability of mice is not affected during the 24 days of repeated administration of Alp.

To further demonstrate the effect of Alp administration repeatedly on the process of memory consolidation in mice, pathological changes in the hippocampal neurons were observed, and vertebral neurons were counted. As shown in Figure 6D, 6E, HE staining showed that neurons in the control group were arranged regularly with abundant cytoplasm with round and clear nucleoli. Furthermore, distinct pathological changes, such as neuron necrosis, disordered arrangement, and significantly decreased pyramidal neurons in CA2, CA3, and DG regions were observed in the Alp group (p < 0.01). There were no obvious pathological changes in the NSI group, while only minor pathological changes were seen in the CA3 area in the Alp+NSI group. Thus, the mitochondrial protector NSI-189 exerted a protective effect against the above neuronal damage.

Moreover, BDNF is essential for memory consolidation [44], and its expression was evaluated in the hippocampus. The protein expression was significantly downregulated in the hippocampus of mice in the Alp group compared to the control group, further illustrating that repeated administration of Alp impairs the memory consolidation process by reducing the expression of BDNF protein. Moreover, NSI-189 significantly weakens the decline in BDNF expression, thereby protecting the memory consolidation process (Figure 6F, 6G).

Based on these findings, repeated administration of Alp affected the memory consolidation process of learning and memory in mice, while repeated administration of mitochondrial protector NSI along with that of Alp could significantly offset the adverse effects of repeated Alp administration on memory consolidation.

Cell culture and treatment

HT22 cell line (Procell, Wuhan, China) was cultured in DMEM (Gibco, USA) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Life Technologies, USA). All cell cultures were incubated at 37° C under 5% CO₂ in a humidified atmosphere.

Animals and treatments

Adult C57BL/6J male mice (6-week-old, 18 ± 1 g) and neonatal C57BL/6J mice (within 24 h) were obtained from HFK Biotechnology Co., Ltd (Beijing, China).

Before any behavioral and experimental procedures, mice were housed (3 animals per cage) with food and water available ad libitum at 24° C on a 12 h light/dark cycle (lights on at 7 a.m.). The animal welfare and experimentation program were reviewed and approved by the Animal Ethics Committee of State Key Laboratory of NBC Protection for Civilian (LAE-2022-03-001).

In behavioral tests, mice (4-week-old) were divided into the following groups with different treatments (12 mice per group): (1) the control group with 1% sodium carboxymethylcellulose CMC Na (Sigma-Aldrich, USA); (2) the Alp group (Alp) treated with 1.84 mg/kg (ED₅₀, Supplementary Figure 1) Alp (Yimin Pharmaceutical Co., Ltd, China) suspended with 1% CMC Na; (3) the NSI group (NSI) treated with 30 mg/kg NSI-189 (Selleck, USA) suspended with 1 % CMC Na; (4) the Alp and NSI-189 group (Alp + NSI) treated with 1.84 mg/kg Alp and 30 mg/kg NSI-189 suspended in 1% CMC Na. All mice were intragastrically administered once a day from 9:00 a.m. to 10:00 a.m. for 24 consecutive days. After the last administration, all mice were anesthetized with isoflurane and rapidly decapitated, and all tissue were collected.

Proteomics analysis

In proteomic analysis experiments, mice were (6-week-old) divided into two groups (6 mice per group, and all mice were treated by intragastric administration): (1) the control group with 1% CMC Na; (2) the Alp group (Alp) treated with 1.84 mg/kg Alp suspended with 1% CMC Na. All mice were continuously administered for 24 days. After the last administration, all mice were anesthetized with isoflurane and rapidly decapitated on ice, and the fresh hippocampus tissues were collected immediately and stored at −80° C.

Subsequently, above fresh hippocampus samples were placed on ice. The protein lysis buffer containing 1% protease inhibitors and 1% phosphatase inhibitors was added to the tissues and homogenized three times for 1 min each with a tissue grinder. The protein concentration in the sample supernatant was determined by the BCA protein assay kit (Thermo Fisher Scientific, USA).

An equivalent of 100 μg protein was resuspended in 100 mM triethylammonium bicarbonate buffer (TEAB). Trypsin was added to the protein at a trypsin-to-protein mass ratio of 1:50 and incubated at 37° C overnight. An equal amount of trypsin-digested peptide was taken from each sample and resuspended in UPLC loading buffer after vacuum drying (Phase A: 2% acetonitrile, pH 10; Phase B: 80% acetonitrile, pH 10). The mixed peptides were fractionated by Vanquish Flex binary UHPLC chromatography (Thermo Fisher Scientific, USA) on a Acquity UPLC BEH C₁₈ Column (1.7 μm, 2.1 mm × 150 mm, Waters, USA) to increase the proteomic depth. The peptides were redissolved in a loading buffer (2% ACN with 0.1% formic acid) and analyzed by online nanoflow liquid chromatography-tandem mass spectrometry on an EASY-nLC system (Thermo Fisher Scientific, USA) connected to a Q Exactive HF-X quadrupole orbitrap mass spectrometer (Thermo Fisher Scientific, USA).

The Q Exactive HF-X instrument was operated in the data-independent acquisition mode (DIA) to automatically switch between full scan MS and MS/MS acquisition. The survey of full scan MS spectra (m/z 300–1500) was acquired; then, the all-precursor ions were selected into collision cell for fragmentation by high-energy collision dissociation (HCD). DIA was performed with a variable isolation window; each window overlapped by 1 m/z for a total of 40 windows. The DIA data files were analyzed using Spectronaut (Biognosys AG, Version 14, Switzerland) with the default settings, and the retention time prediction type was set to dynamic iRT.

The proteomics statistical analysis was performed as described previously [19]. Proteomic data was analyzed using the one-way analysis of variance (ANOVA), followed Tukey’s post-hoc test. Statistically significant parameters were selected with fold-change (FC) > 1.2 (or < 0.83) and p < 0.05 for the proteomic analysis. DEPs were used for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. Protein-protein interaction analysis was carried out using String v13.5.

Cell viability

HT22 cells were seeded in the 96-wells at a density of 1 × 10⁴ cells/well and incubated overnight at 37° C with 5% CO₂. Then, they were treated with various concentrations (0, 1, 10, 50, and 100 μM, dissolved with 0.5% DMSO, 100 μL) of Alp for 48 h. The Cell Counting Kit-8 (CCK-8, Solarbio Science and Technology Co., Ltd, Beijing, China) reagent was added to each well and incubated at 37° C for 1 h with 5% CO₂, and the absorbance of the samples was measured in a microplate reader (Spark, Tecan, Switzerland) at 450 nm.

Flow cytometry

HT22 cells were plated in 6-well plates at a density of 1.5 × 10⁶ cells per well and treated with drugs for 48 h. Then, cells were trypsinzed and harvested by centrifugation, followed by staining with Annexin V-FITC/PI kit (556547, BD Biosciences, USA); the cells were incubated with Annexin V-FITC and PI at room temperature for 15 min in the dark. The apoptosis rate was examined by flow cytometry (NovoCyte, Agilent, USA) in 30 min. The experiment was repeated three times with four duplicate wells for each group.

Each group of well-grown cells was washed twice with phosphate-buffered saline (PBS), resuspended with pre-chilled in 70% ethanol, fixed at 4° C for 12 h, permeabilized with 0.1% Triton X-100, stained with PI/RNase (550825, BD Biosciences, USA) in PBS, and incubated at 37° C in the dark for 30 min. Subsequently, the cell cycle distribution of each group was detected 3 times by flow cytometry, and the experiment was repeated 3 times.

Mitochondrial activity assay

The JC-1 fluorescent probe (Beyotime, China) was used to detect mitochondrial membrane potential (MMP). HT22 cells were plated in 6-well plates at a density of 1.5 × 10⁶ cells per well and treated with drugs for 48 h. After treatment, the culture medium was discarded, the cells were washed with PBS once, and 1 mL of medium and 1 mL of JC-1 staining solution was added to each well, and incubated for 20 min. Subsequently, the supernatant was discarded, the cells were washed two times with JC-1 staining buffer, and then 2 mL of medium was added. Finally, the images from each well were acquired using a fully automated cell imaging multifunctional microplate detector (Cytation 1, Bio-Tek, USA).

Hippocampal primary neurons culture in vitro

The extraction and in vitro culture of primary cells are based on previous studies [76]. The hippocampal tissue was digested by papaya enzyme (P123425, Aladdin, USA) and filtered through a 40-μm cell sieve to prepare a single-cell suspension. The neurobasal-A culture medium (2429791, Gibco, USA) containing 2% B27 supplements (2415388, Gibco, USA) and 25 μM glutamine (3505061, Gibco, USA) was added to single-cell suspension. The cell culture medium was changed every 3 days, the old medium was aspirated and the neurons were cleaned 3 times by PBS, and finally the fresh medium was added. After 7 days of culture in a humidified atmosphere of 5% CO₂/95% air at 37° C, cell experiments were performed subsequently.

Mitochondrial bioenergetics analysis

Primary hippocampal neurons were seeded in 96-well plates at a density of 1.2 × 10⁴. After 24 h of culture with 5% CO₂ at 37° C, the neurons were randomly divided into two groups, control and Alp, 6 wells each. The neurons in the Alp group were treated with 100 μL medium containing the 50 μM Alp, while the control group were treated with normal medium. After 48 h, neurons were washed with assay solution (culture medium containing 1 mM sodium pyruvate, 2 mM glutamine, and 10 mM glucose) and cultured at 37° C for 45 min without CO₂. The oxygen consumption rate (OCR) was measured by cell energy metabolism analyzer (Seahorse XFe, Agilent Technologies, USA). 20 μL Oligomycin (1.5 μM), 20 μL carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP, 2 μM), and 25 μL rotenone/antimycin A (R/A, 0.5 μM) were dissolved with assay solution and injected sequentially during the assay. Each sample was tested three times. Finally, basal breathing, ATP-related breathing, and maximal and spare respiratory capacity were determined, respectively.

Automated Western blotting

Proteins levels in issues or cells were measured using the Jess western blotting system (ProteinSimple, USA), which is an automated capillary-based size separation and nano-immunoassay system. Jess-specific-reagents (PS-ST05EZ, ProteinSimple), including their rabbit secondary antibodies, were used according to the manufacturer’s protocols. The rabbit anti-GAPDH (ab181602, Abcam), anti-COX6A1 (11460-1-AP, Proteintech), anti-NDUFS6 (ab230464, Abcam), anti-FIS1 (BPS04232S, Abmart), anti-NDUFAB1 (BT55373S, Abmart), anti-ND4 (26736-1-AP, Proteintech) and anti-OPA1 (BPS06076S, Abmart) were diluted at 1:50 in antibody diluent. Digital image of chemiluminescence of the capillary was captured with Compass Simple Western software (version 6.2.0, ProteinSimple) that calculated automatically area (chemiluminescence intensity).

Intelligent cage system (ICS) test

The cognitive function of mice was evaluated by the ICS (New Behavior, TSE Systems, Germany) as described previously. The experimental protocol was further improved by optimizing the program design. The whole 43-day protocol includes intelligent programming, mouse drinking training, and mouse behavior testing, and the details of the experiment are provided in the Supplementary Figure 2.

Novel objects recognition (NOR)

Before starting the tests, each mouse was allowed to explore freely in an open field for 10 min. After 30 min, the mice were placed again in the field to explore freely for 10 min, at which point two identical wooden cylinders were placed symmetrically in the field. After 1 h, one of the wooden cylinders was replaced with a plastic cone with an equal base area (defined as a new object), and the mouse was re-entered into the field to explore freely for 10 min. The time spent exploring each object during the training and testing periods was recorded (VisuTrack, XinRuan, China) separately, and the novel object recognition index was calculated. The novel object recognition index is the percentage of time spent exploring a new object with respect to the time spent exploring two objects.

Hematoxylin-eosin (HE) staining

Hippocampus slices in the ischemic brains were used to observe the changes in neuronal cell morphology by HE staining. All information on tissue sections was scanned and imaged using the slide scanner (Pannoramic, 3DHISTECH, Hungary). The scanning software CaseViewer2.4 (3D-HISTECH, Hungary) was applied to select the tissue target area for 400× imaging, followed by Image ProPlus 6.0 (version 6.0, Media Cybernetics, USA) to count the number of pyramidal neurons in each image, and the number of pyramidal neurons was calculated per unit area.

Immunohistochemistry

The paraffin slices of the hippocampal brain were blocked with 3% bovine serum albumin (BSA) at 37° C for 30 min and then incubated overnight with 1% brain-derived neurotrophic factor (BDNF) antibody (ab108319, Abcam) at 4° C. The standard protocols were followed as described previously [77]. The different regions of the hippocampus were selected to acquire images under a 400× field of view (Eclipse E100, Nikon, Japan). Three non-overlapping regions were randomly selected for each slice, and the Image ProPlus 6.0 image analysis system was used to determine the average optical density values of BDNF-positive cells in each region of the hippocampus under different fields of view.

Transmission electron microscopy (TEM)

The hippocampus tissue was fixed in 1% osmium acid prepared in 0.1 M phosphoric acid buffer PB (pH 7.4) at room temperature for 2 h. Then, the tissue was stained with 2% uranyl acetate in the dark, dehydrated in an acetone gradient solution, and embedded in an epoxy resin. The resin block was used to prepare 60-nm ultrathin sections on an ultramicrotome (UC7, Leica, Germany) and stained with 2.6% lead citrate and 2% uranium acetate. Finally, images of mitochondrial ultrastructure were acquired by transmission electron microscopy (HT7800, Hitachi, Japan).

Statistical analysis

GraphPad Prism 9.0 (GraphPad Software, USA) was used for data analysis. The data conformed to a normal distribution, which was expressed as the mean ± standard error of the mean (SEM). An unpaired t-test was used to compare two groups, and one-way or two-way ANOVA was used to compare multiple groups. p < 0.05 was considered statistically significant difference.

Data availability statement

The detailed data used to support the findings of this study are available from the corresponding author upon written request.