BANF1 promotes glutamate-induced apoptosis of HT-22 hippocampal neurons

Glutamate exposure was fatal to HT-22 neuronal cells that derived from mouse hippocampus. This is often used as a model for hippocampus neurodegeneration in vitro. The targets relevant to glutamate-induced neuronal toxicity is not fully understood. In this study, we aimed to identify crucial factors associated with glutamate-induced cytotoxicity in HT-22 cells. HT-22 cells were treated with 7.5 mM glutamate for 24 h and isobaric tags for relative and absolute quantitation (iTRAQ) proteomic analysis conducted to identify the differentially expressed proteins. Differential proteins were subjected to Gene Ontology analyses. Upregulation of barrier to autointegration factor (BANF1/BANF1) protein was confirmed by RT-qPCR and western blotting. Cell viability was measured by CKK-8 and MTT assays. Cell apoptosis rates and intracellular reactive oxygen species (ROS) levels were detected using flow cytometry. A total of 5811 proteins were quantified by iTRAQ, 50 of which were recognized as significantly differential proteins (fold change ≥ 1.5 and P ≤ 0.05); 26 proteins were up-regulated and 24 were down-regulated after exposure to glutamate. GO enrichment analysis showed that the apoptotic signaling pathway was involved in cell death induced by glutamate. BANF1 expression level was markedly increased in HT-22 cells after glutamate treatment. Further, knockdown of BANF1 alleviated glutamate-mediated cell death with lower ROS levels. In conclusion, we successfully filtered out differential proteins relevant to glutamate-mediated cytotoxicity. BANF1 upregulation promoted glutamate-induced apoptosis of HT-22 cells by enhancing ROS generation.


Introduction
Alzheimer's disease (AD) is a neurodegenerative condition characterized by progressive cognitive and behavioral impairment, and the most common cause of dementia in older adults [1].Approximately 50 million individuals worldwide had AD in 2018, and the number is estimated to triple by 2050 [2].Various factors are reported to be associated with AD development, including hereditary elements, amyloid-beta (Aβ) aggregation, Tau hyperphosphorylation, acetylcholine deficiency, and glutamate toxicity, among others [2,3].Regrettably, many Xinyu Yao and Xiaoyi Xu have contributed equally to this work.drugs targeting Aβ, such as BACEI inhibitors, solanezumab, and bapineuzumab, have failed to efficiently improve cognitive performance in patients with AD [4,5]; therefore, it is imperative to identify molecules with key roles in AD pathogenesis.As an excitatory neurotransmitter, glutamate has important roles in synaptic plasticity and memory [6,7]; however, excessive glutamate triggers excitotoxicity and oxidative stress in neuronal cells, which are closely correlated with AD progression [8].The Glu-HT-22 cell model is an established classical method for studying nerve cell death mediated by oxidative stress [9].In this study, we conducted isobaric tags for relative and absolute quantitation (iTRAQ) analysis using the Glu-HT-22 cell model, with the aim of identifying proteins differentially expressed during glutamate-induced neurotoxicity.
BANF1 is an abundant DNA-binding protein, the subcellular location of which can alter at different phases of the cell cycle [10].BANF1 can also bind directly to histone proteins, LEM-domain proteins, and transcription factors, and participates in many biological processes, including nuclear envelope remodeling, DNA damage response, transcription regulation, and epigenetics [11][12][13].BANF1 is reported to regulate the DNA damage response to oxidative stress by interacting with the NAD + -binding domain of poly [ADP-ribose] polymerase 1 (PARP1), thereby hindering PARP1-mediated DNA repair [14].An A12T mutation in BANF1 is the main cause of Néstor-Guillermo progeria syndrome and contributes to impaired PARP1 activity in patients [15,16], while in the L-HaCaT psoriasis model, BANF1 upregulation alleviates skin inflammation [17]; however, research on the role of BANF1 in AD is lacking.In this study, BANF1 upregulation was detected in the Glu-HT-22 model using iTRAQ analysis.We also investigated the impact of BANF1 on glutamate-mediated hippocampal neuron cell death, providing fresh insights into AD pathogenesis.

Cell culture and cell viability analysis
The HT-22 cell line, which comprise a neuronal line derived from mouse hippocampus, was obtained from Procell (Wuhan, China) and cultured in DMEM containing 10% FBS and 1% penicillin-streptomycin antibiotics.Glutamate (Sigma) was diluted using cell culture medium.Cell viability was measured by cell counting kit-8 (CCK-8; HY-K0301, MedChemExpress).HT-22 cells were seeded in 96-well plates (5 × 10 3 /well) and were cultured in a medium containing various concentrations of Glu (1.8, 3.75, 7.5, 15 mM).After 24 h incubation, 10% CCK-8 was added and incubated at 37 °C for 3h.Absorbance was measured at 450 nm using a microplate reader (Multiskan-GO,thermofisher).For MTT assay, cells were incubated with 20 µL MTT (5 mg/mL) for 2 h, then the culture medium was discarded, 150 µL DMSO added to each well, and optical density values at 490 nm recorded using a microplate reader (Multiskan-GO,thermofisher).

Apoptosis and reactive oxygen species (ROS) detection
HT-22 cells were seeded in 24-well plates (5 × 10 4 /well) and treated with 7.5mM glutamate (Sigma, G0355000) for 24 h.Subsequently, cells were stained using an apoptosis kit (KeyGEN BioTECH, KGA106) and a reactive oxygen species kit (Beyotime, No. S0033S).Flow cytometry (BD FACS Aria II) was used to detect fluorescence signals to analyze apoptosis rates and intracellular ROS levels.

Sample pretreatment for iTRAQ
Cells were fully lysed in RIPA Lysis Buffer and incubated with acetone at − 20 °C overnight, then samples were centrifuged at 13,000×g at 4 ℃ for 15min and supernatants removed.Proteins in the precipitate were resuspended in RIPA Lysis Buffer for BCA quantification.Next, 100 µg protein was treated with 2 µl Reducing Reagent at 37 ℃ for 1 h, then 1 µl Cysteine-Blocking Reagent added.After incubating the mixture at room temperature for 10 min, 100 µl 1M TEAB was added, and samples centrifuged using a 10 kDa ultrafiltration tube.The centrifugation process was repeated three times.Finally, samples were incubated with 2 µl trypsin (1 µg/µl) at 37 °C for 10 h, then centrifuged to collect digested peptides.

iTRAQ labeling and high-performance liquid chromatography (HPLC) component separation
Samples were labeled with iTRAQ tags, as follows: control groups were marked as 113 and 118, and the Glu groups were marked as 114, 115, and 119.After 150 µl of isopropanol was mixed with iTRAQ reagent at room temperature for 2 h, 100 µl ultrapure water was added to terminate the reaction.Then, labeled samples were mixed and freezedried; 40 µl 0.1% formic acid was used to dissolve samples.Fraction separation was conducted by HPLC (Dionex Ultimate3000).Phenomenex chromatographic columns (Gemini-NX 3u C18 110A, 150 × 2.00 mm) were used with Phase A: 20 mM HCOONH 4 , pH 10, and Phase B: 20 mM HCOONH 4 , 98% CAN, pH 10, at flow rate, 100 µL/min.

Bioinformatics enrichment analysis
The database, http://www.geneontology.org,was used for Gene Ontology (GO) enrichment analysis, which included three main modules: biological processes (BP), cellular components (CC), and molecular functions (MF).GO terms with P < 0.05 were considered significantly enriched.In addition, a protein-protein interaction (PPI) network of differentially expressed protein was established based on the STRING database (http:// string-db.org).

Statistical analysis
All data are expressed as mean ± standard deviation, or as a percentage of the control.Differences between two groups were analyzed by unpaired two tailed t-test using GraphPad Prism 8 (GraphPad Software, USA).When p < 0.05, differences were considered statistically significant.P values are presented as *p < 0.05, **p < 0.01, and ***p < 0.001.

Glutamate enhances HT-22 cells ROS production and apoptosis
HT-22 cells were treated with 7.5 mM glutamate for 24 h.As shown in Fig. 1B, exposure to glutamate significantly decreased HT-22 cell viability (p < 0.001).PI/Annexin-FITC staining followed by flow cytometry analysis verified that the rate of apoptosis was elevated (Fig. 1D).Further, ROS levels rose in HT-22 cells incubated with glutamate (Fig. 1C).These results indicate that glutamate may trigger apoptosis in neuronal cells by enhancing ROS production.

Differentially expressed proteins in glutamate-treated HT22 cells
To identify critical molecules associated with glutamateinduced neurotoxicity, iTRAQ quantitative analysis was carried out in two groups of HT-22 cells with and without glutamate treatments.A total of 5811 proteins were identified, of which 50 were considered significantly differentially expressed (fold-change ≥ 1.5 and p ≤ 0.05).Overall, 26 proteins were up-regulated and 24 were down-regulated after exposure to glutamate, as listed in Table 1.

Bioinformatics analysis
To further study the function of the differentially expressed proteins, we conducted GO enrichment analysis.GO terms were divided into three areas: BP, CC, and MF.The BP domain was mainly enriched in translation, negative regulation of transcription by RNA polymerase II, and negative regulation of apoptosis.The CC domain primarily involved the nucleus, extracellular exosome, and cytoplasm.MF domains included structural constituent of ribosome, ATP  binding, and RNA binding (Fig. 2A).In the constructed PPI network of differentially expressed proteins, the top 10 hub proteins were: RPL5, RPL8, PRL27, RPS3A1, RPS4X, RPS20, RPS28, EIF4A2, EEF2, and HSP90B1 (Fig. 2B).

Glutamate induces upregulation of BANF1
The results of iTRAQ analysis (Fig. 3A) demonstrated that BANF1 expression levels were significantly increased after treatment with glutamate (fold-change = 1.9, p = 0.0018).BANF1 upregulation in response to glutamate was also verified by RT-qPCR and western blotting (Fig. 3B, C).These results suggest that upregulation of BANF1 may be associated with glutamate-induced cytotoxicity.

BANF1 promotes glutamate-induced apoptosis in HT-22 cells
To elucidate the role of BANF1 in glutamate-induced cytotoxicity, HT-22 cells were transfected with Banf1 siRNA or negative control siRNA.Transfection with Banf1 siRNA successfully suppressed BANF1 expression (Fig. 4A, B).Compared to the transfection with negative control siRNA, the knockdown of BANF1 ameliorated glutamate-mediated decline in cell viability and lowered the rate of apoptosis (Fig. 4C, D).There was also a decrease in intracellular ROS levels (Fig. 4E).These data suggest that BANF1 promotes glutamate-induced apoptosis of HT-22 cells by boosting ROS generation.

Discussion
Progressive loss of hippocampal neurons is a distinctive pathological feature of AD, which may contribute to memory impairment in patients [18,19].Cerebrospinal fluid glutamate levels are reported to be significantly elevated in patients with AD and researchers have speculated that glutamate results in progressive neuronal loss in AD [20]; however, the underlying mechanism is not fully understood.In this study, we conducted iTRAQ quantification analysis to identify differentially expressed proteins in HT-22 hippocampal neuronal cells treated with glutamate.GO enrichment and PPI network analysis were also carried out to study the function of differentially expressed proteins and explore the potential mechanism underlying glutamate-induced cytotoxicity.GO analysis indicated that translation, negative regulation of transcription by RNA polymerase II, and negative regulation of apoptosis biological processes were enriched among differential proteins.Glutamate can regulate transcription by inducing splicing of the RNA polymerase II-associated cyclin, Ania-6 [21].Further, glutamate can control gene expression by enhancing binding of the nuclear transcription factor, activator protein-1 (AP1), to DNA [22].Apoptosis is closely associated with glutamate-induced cytotoxicity and neurodegenerative disease pathogenesis, and excitotoxicity is mediated by continuous stimulation of the ionotropic glutamate receptor (iGluR), which leads to elevated intracellular calcium levels and triggers a chain of events relevant to apoptosis [23].Excessive glutamate also inhibits activity of the cystine/glutamate antiporter system, System Xc-, and promotes ROS accumulation, a process known as oxidative toxicity [24,25].BID protein-mediated translocation of AIF participates in neuronal apoptosis after treatment with glutamate [26,27].The cellular components mainly enriched for differentially expressed proteins included extracellular exosome, nucleus, and cytoplasm.Exosome release in oligodendrocytes and cortical neurons is triggered by glutamate-mediated Ca 2+ influx [28,29], and massive Na + influx into the cytoplasm via N-methyl-Daspartic acid receptor results in neuronal cell swelling [30].Enriched molecular functions mainly included structural constituents of ribosomes, ATP binding, and RNA binding.Abnormal ribosome function is associated with AD development, and Tau protein can interact with ribosomal protein S6 (RPS6) and suppress protein synthesis regulated by RPS6 [31].
In our PPI analysis, the top 10 hub proteins included RPL5, RPL8, PRL27, RPS3A1, RPS4X, RPS20, RPS28, Eif4a2, Eef2, and HSP90b1, indicating that ribosomes play a vital role in glutamate-induced neurotoxicity.Ribosomes, composed of rRNA and proteins, are responsible for protein synthesis, and aberrant ribosome function contributes to cognitive deficits in patients with AD [32].Further, RPL21, RPL23A, RPL27, RPS5, RPS6, RPS10, and RPS13 mRNA expression levels are significantly Fig. 2 GO and PPI analyses of differentially expressed proteins.GO analysis included biological process, cellular component, and molecular function (A).Protein-protein interaction analysis of differentially expressed proteins was performed using the STRING database (B) down-regulated in the dentate gyrus (DG) of the hippocampus in patients with AD [33].Previous quantitative proteomics studies revealed that 28 ribosomal proteins are considerably up-regulated in brain capillaries, rather than the parenchyma, in patients with AD, including RPL8, RPL27, RPS3A, RPS4X, and RPS20 [34].Eukaryotic initiation factors (eIFs) control translation initiation, while eukaryotic elongation factor 2 (eEF2) can catalyze tRNA migration from site A to site P in ribosomes [35].Hyperphosphorylation of the eEF2 α subunit has been detected in brain tissues from patients with AD and eEF2 kinase inhibition can attenuate pathological damage in mouse models of AD [36,37].HSP90 helps other proteins to fold correctly and maintains protein stability [38], and activation of the AXL/HSP90/PPAR pathway can ameliorate cognitive impairment in patients with AD [39].
Our bioinformatic analysis of differentially expressed proteins reveals that the molecular mechanisms underlying glutamate-mediated HT22 cell death are diverse, and include apoptosis, aberrant energy metabolism, and ribosomal dysfunction.As the immortalized hippocampal cell line, HT22, lacks iGluR, the Glu-HT22 cell model is a classical tool for study of oxidative toxicity caused by System Xc-inhibition [9].This transporter system takes in extracellular cystine and transports glutamate out of the cell at a 1:1 ratio, in favor of glutathione synthesis [40].High concentrations of extracellular glutamate suppress System Xc-activity, resulting in depletion of glutathione and elevated ROS production [24,25].Furthermore, our iTRAQ quantification analysis found for the first time that BANF1 is significantly up-regulated in HT22 cells after exposure to glutamate.Total BANF1 expression levels are elevated, along with its translocation from the nuclear envelope to chromatin, in U2OS cells treated with 200 M H 2 O 2 , indicating that BANF1 can respond to oxidative stress [14].Evidence from our study suggests that BANF1 can promote glutamate-induced apoptosis in HT22 cells via enhancing ROS production, which may provide a potential target for AD treatment.
This study has limitations.Further research, including animal experiments, is warranted, to verify the upregulation and biological function of BANF1.Additionally, the specific mechanism by which BANF1 induces ROS generation was not established in this study.Montes de Oca et al. identified four transcription factors that can directly bind to BANF1, including PC4, NonO, Requiem, and LEDGF [41], among which PC4 can bind directly to the N-terminal transactivation domain (TAD) of p53 [42,43]; p53 both upregulates the expression of apoptosis-related proteins and inhibits expression of the SLC7A11 subunit of System Xc, leading to intracellular ROS accumulation [44,45].Therefore, further experiments to determine whether BANF1 triggers apoptosis via activation of PC4 and P53 will be among the logical next steps to build on our findings.

Conclusion
A total of 50 differentially expressed proteins (26 upregulated and 24 downregulated) relevant to glutamate-induced cytotoxicity in a neuronal cell line were filtered out using iTRAQ technology.We also demonstrate for the first time that BANF1 upregulation by glutamate can trigger HT22 cell apoptosis by increasing ROS production.

Fig. 1
Fig. 1 Glutamate enhanced ROS production and apoptosis in HT-22 cells.Glutamate with gradient concentrations were applied to HT-22 cells, then cell counting kit-8 (CKK-8) assay was performed to evaluate cell viability (A).HT-22 cells were treated with 7.5 mM gluta-

Fig. 3 Fig. 4
Fig. 3 Glutamate induced the upregulation of BANF1.HT-22 cells were treated with 7.5 mM glutamate for 24 h.Mass spectrometry data analysis revealed significant changes of iTRAQ labeling intensity in a Table1 differentially expressed proteins after exposure to glutamate detected by ITRAQ quantitative analysis