CENPF interaction with PLA2G4A promotes glioma growth by modulating mTORC1 and NF-κB pathways

Abstract

Background

Glioma is the most common primary malignant tumor of the central nervous system, and due to the limited effectiveness of traditional single-target therapies, there is an urgent need for new therapeutic targets. Centromere protein F (CENPF) belongs to the centromere protein family and is mainly involved in the regulation of the cell cycle. CENPF has recently been found to play a key role in tumorigenesis and tumor progression, but its role in gliomas has not been well studied.

Methods

The expression level and clinical information of CENPF were obtained by analyzing the TCGA, CGGA and GEO databases. Immunohistochemistry and western blot analysis were used to quantitatively detect the expression of CENPF in glioma tissues and cell lines. Gene set enrichment analysis (GSEA) of TCGA and GSE16011 datasets was used to explore the molecular mechanism of the CENPF. CENPF-interacting proteins were detected by molecular docking and co-immunoprecipitation (Co-IP). After silencing CENPF, CCK-8 assay, Transwell assay and flow cytometry were used to detect changes in cell proliferation, invasion, cell cycle and apoptosis, and Western blot was used to detect changes in signaling pathway protein levels.

Results

Bioinformatics analysis showed that CENPF was generally highly expressed in gliomas and was associated with poor prognosis. This result was confirmed in glioma samples from our hospital. Multivariate Cox regression analysis showed that CENPF was an independent prognostic marker for gliomas. Western blot analysis in vitro showed that CENPF was overexpressed in the U251 and LN229 cell lines; therefore, these two cell lines were selected for subsequent experiments. GSEA analysis showed that CENPF was mainly involved in the G2/M phase-mediated cell cycle and P53 signaling pathway. Flow cytometry analysis confirmed that silencing CENPF induced G2/M phase arrest and increased apoptosis in glioma cells. Subsequent experiments confirmed that CENPF influences the epithelial-mesenchymal transition (EMT) process through the mTORC1 signaling pathway. Molecular docking and Co-IP assay revealed that CENPF exerts its effects by interacting with PLA2G4A promoting the downstream signaling pathway. Finally, we found that silencing CENPF combined with a PLA2G4A inhibitor (AACOCF3) induced glioma cell apoptosis and exhibited anti-glioma effects.

Conclusions

This study found that CENPF plays a key role in promoting tumorigenesis through its interaction with PLA2G4A. This study provides a theoretical foundation for advancing multi-targeted therapies in glioma and for developing strategies to overcome tumor drug resistance.

Introduction

Gliomas are the most common and most aggressive primary tumors of the central nervous system [1, 2]. According to the 2021 WHO Classification of Central Nervous System Tumors, glioblastoma (GBM), as a high-grade glioma, is highly invasive and heterogeneous, with a high recurrence rate and poor prognosis [3]. The median survival time is only 15 months and the 5-year survival rate is < 10% [4]. Currently, the standard treatment for glioma patients includes surgical resection combined with radiotherapy and temozolomide chemotherapy [5]. However, the prognosis of glioma remains poor. Therefore, new glioma treatments are urgently needed to improve patient survival.

Molecular biomarkers combined with targeted therapy is one of the hot topics in current tumor treatment research. The 2021 World Health Organization Classification of Central Nervous System Tumors further strengthens the application of molecular pathology in gliomas and more clearly relies on molecular markers for diagnosis. It emphasizes the role of molecular markers such as TERT promoter mutation, MGMT promoter methylation, EGFR amplification, and IDH mutation [3, 6, 7]. The new classification specifically emphasizes that glioblastomas are exclusively designated for IDH wild-type tumors, thereby establishing a clear demarcation from previous classifications where both IDH-mutant and wild-type tumors could be categorized as such. Although larotrectinib and entrectinib for the treatment of neurotrophic tyrosine receptor kinase (NTRK) fusion tumors, everolimus (mTOR inhibitor) for the treatment of tuberous sclerosis-related subependymal giant cell astrocytoma, and erlotinib (EGFR inhibitor) for the treatment of adult high-grade gliomas have good efficacy, they have not yet reached the stage of precise individualized treatment [8,9,10,11,12]. Therefore, it is urgent to discover new key molecules in the occurrence and development of gliomas and to provide valuable biomarkers and therapeutic targets for glioma treatment.

Centromere protein F (CENPF) is a mammalian mitogen protein which encodes a protein involved in the formation of the centromere-kinetochore complex and is an important component involved in the G2/M phase of the cell cycle [13]. In recent years, high expression of CENPF has been shown to be associated with the prognosis of various cancers, especially tumor metastasis and poor prognosis in patients with gastric cancer or breast cancer [14, 15]. In addition, CENPF was found to be significantly upregulated in prostate cancer, hepatocellular carcinoma, and papillary thyroid cancer, and is often associated with shorter survival time [16,17,18]. Additionally, our previous investigation revealed a significant association between CENPF and unfavorable prognosis in glioma through comprehensive transcriptome analysis [19]. However, the specific mechanism of CENPF in glioma remains unclear and needs to be studied.

In this study, we through bioinformatics analysis, found that CENPF is upregulated in glioma and significantly correlated with the prognosis of gliomas. In vitro, we used well-established glioblastoma cell lines (U251 and LN229) to investigate the effects of CENPF on gliomas [20, 21]. The results showed that high expression of CENPF contributed to the proliferation and invasion of glioma cells via the mTORC1 signaling pathway. Further experiments demonstrated that CENPF interacts with PLA2G4A and jointly participates in regulating the proliferation of gliomas. Silencing CENPF combined with the PLA2G4A inhibitor arachidonyl trifluoromethyl ketone (AACOCF3) can significantly inhibit the malignant phenotype of glioma, indicating that CENPF may be a potential target for the treatment of glioma.

Materials and methods

Datasets

The glioma datasets utilized in the bioinformatics analysis originated from the TCGA, CGGA and GEO databases. Four glioma microarray data (GSE54004, GSE16011, and GSE4290) were obtained from the GEO database. The TCGA sequencing data were downloaded from GDC (https://portal.gdc.cancer.gov/). The RNA sequencing data for glioma patients (CGGA693 and CGGA325 datasets) were obtained from the CGGA (http://www.CGGA.org.cn/) database.

Clinical tissue samples

The 60 glioma tissues (33 males and 27 females, with a median age of 49 years) and 10 adjacent normal tissues in this study were obtained from patients who underwent radical glioma resection in Linyi People’s Hospital (Linyi, Shandong Province) from January 2014 to December 2020. All glioma patients did not receive chemotherapy or radiotherapy before surgery. All samples underwent confirmed pathological diagnosis by two neuropathologists in accordance with the 2016 WHO classification. Fresh glioma tissues were stored at -80°C, while a portion of the tissues was fixed in 4% paraformaldehyde (PFA). The present study was approved by the Ethics Committee of Linyi People’s Hospital and conducted in accordance with the principles outlined in the Declaration of Helsinki.

Somatic alteration data

The mutation data for glioma patients were obtained from the TCGA cohort (https://portal.gdc.cancer.gov/). Somatic copy number alterations and mutation levels were compared between the high and low CENPF expression groups based on their expression levels.

Gene set variation analysis (GSVA)

Using the R package “GSVA”, these features were used to quantify the enrichment level of relevant genes for each sample. The analysis was performed with the “GSVA” R package, which calculates enrichment scores for each gene set in each sample, providing a non-parametric, unsupervised method to assess gene set enrichment across the dataset. The genesets used for enrichment analysis were sourced from the Molecular Signatures Database (https://www.gsea-msigdb.org/gsea/). Visualization and further statistical analysis were performed using “limma” and “ggplot2” R packages.

Gene set enrichment analysis (GSEA)

Pathway enrichment analysis of target genes was performed using Gene Set Enrichment Analysis (GSEA) based on the TCGA and GSE16011 datasets. Both datasets were stratified into high- and low- expression groups according to the median expression level of CENPF. Annotation files were obtained from the GSEA database (https://www.gsea-msigdb.org/gsea/). The R packages “ggplot2”, “clusterProfiler,” and “enrichplot” were employed to visualize the bioinformatics analysis results.

Immunohistochemistry (IHC) and staining evaluation

Immunostaining was performed using an ultrasensitive SP IHC kit (Maxin Bio-technology, Beijing, China) according to the manufacturer’s protocol. The sections were incubated with the primary antibody targeting CENPF (1:100, Proteintech, China) at 4 °C overnight, washed three times with PBS, and then incubated with HRP-conjugated polymer for 1 h. After adding 3,3’-diaminobenzidine tetrahydrochloride (DAB) and counterstaining with hematoxylin, the staining was analyzed under a bright field microscope (Nikon, Tokyo, Japan). Brown-yellow nuclei were observed under the microscope and considered positive. Ten fields of view were randomly selected from each section, and each section was scored according to the number of positive cells in each field of view and the staining intensity. The expression of CENPF in glioma was evaluated by two independent pathologists who were blinded to the pathology of glioma. IHC was used to semiquantitatively evaluate the positivity of CENPF as follows: the degree of staining was scored according to the percentage of positive tumor cells (< 5%, 0; 5–25%, 1; 26–50%, 2; 51–75%, 3; >75%, 4). The staining intensity was scored into four levels (negative, 0; weak, 1; moderate, 2; strong, 3). The staining intensity was multiplied by the percentage of positive cells to obtain the scoring result. Pathological sections with a score of ≥ 4 were divided into the CENPF high expression group, and pathological sections with a score of < 4 were divided into the CENPF low expression group. Additionally, the presence of Ki-67 and P53 was determined by the percentage of immunopositive cells relative to the total number of cells in five fields, regardless of staining intensity. For Ki-67, immunoexpression was considered negative if immunostaining was present in < 20% of tumor cells. It was considered positive if immunostaining was present in > 20% of tumor cells. For P53, immunoexpression was considered negative if immunostaining was observed in < 10% of tumor cells (wild type) and positive if > 10% of the examined tumor cells were immunopositive (mutant type) [22].

Cell culture

The human glioma cell lines LN229, U251, SF295, A172, TJ905 and the normal astrocyte cell line HA1800 used in this study were obtained from the Central Laboratory of Linyi People’s Hospital (Linyi, Shandong Province, China). All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; C111995500 BT, Gibco, USA) containing 1% double antibiotics (penicillin-streptomycin solution, P1400, Solarbio, China) and 10% fetal bovine serum (FBS; 10270106, Gibco, USA) in a cell culture incubator at 37 °C with 5% CO₂. In the PLA2G4A inhibition assay, upon reaching approximately 60% cell confluence, the culture medium was replaced with complete medium supplemented with AACOCF3 (25 µM, HY-108611, MCE) and incubated for 24 h. The cell lines were treated with 5µM MHY1485 (5mM, HY-B0795, MCE) and cultured for 24 h before western blot assay.

Western blotting

After washing with PBS, the cells were collected by centrifugation, lysed with RIPA extraction reagent, and the protein concentration was determined using a bicinchoninic acid (BCA) protein assay kit. The concentration was calculated according to the bicinchoninic acid protein assay kit (Beyotime). The proteins in the samples were separated on SDS-PAGE gels, and then the membranes were incubated with primary and secondary antibodies. Finally, the samples were washed with PBST and developed by chemiluminescence. The primary antibodies used were as follows: anti-CENPF (28568-1-AP, Proteintech), anti-PLA2G4A (68133-1-Ig, Proteintech), anti-p-PLA2G4A (28925-1-AP, Proteintech), anti-mTOR (28273-1-AP, Proteintech), anti-p-mTOR (67778-1-Ig, Proteintech), anti-S6K (14485-1-AP, Proteintech), anti-p-S6K (28735-1-AP, Proteintech), anti-4E-BP1 (#9644, CST), anti-p-4E-BP1 (#2855, CST), anti-NF-κB (66535-1-Ig, Proteintech), anti-p-NF-κB (#3033, CST), anti-CCNB1 (28603-1-AP, Proteintech), anti-CCNB2 (21644-1-AP, Proteintech), anti-CDK1 (AF6108, Affinity), anti-E-Cadherin (AF0131, Affinity), anti-N-Cadherin (22018-1-AP, Proteintech), anti-Vimentin (10366-1-AP, Proteintech), anti-Bcl-2 (AF6139, Affinity), and anti-Bax (AF0120, Affinity).

Silencing of CENPF by SiRNA transfection

Small interfering RNA (siRNA) to CENPF (GenePharma, Shanghai, China) was transfected with Lipofectamine™ 2000 reagent (Thermo Fisher Scientific, USA) according to the manufacturer’s protocol. Silencing efficiency was assessed by western blotting 48 h after transfection. The negative control (NC) sequences were as follows: Forward: 5’-UUCUCCCGAACGUGUCACGUTT-3’, reverse: 5’-ACGUGACACGUUCGGAGAATT-3’.The siRNA sequences that generated effective knockdown were as follows: siCENPF#1 forward: 5’-CAAGCUUCAGUUACUGUCAAAUGAA-3’, reverse: 5’-UUCAUUUGACAGUAACUGAAGCUUG-3’; CENPF#2 forward: 5’-CAGGGCUCUUCAGAAUGCAUUUCUG-3’, reverse: 5’-CAGAAAUGCAUUCUGAAGAGCCCUG-3’ and CENPF#3 forward: 5’-CCGAGAGAAAUUGACUUCUAAAGAA-3’, reverse: 5’-UUCUUUAGAAGUCAAUUUCUCUCGG-3’.

Overexpression of PLA2G4A by lentiviral transduction

For overexpression of PLA2G4A, ORF lentiviral expression vector EX-L5215-Lv105 (PLA2G4A) and the corresponding empty vector EX-EGFP-Lv105 were purchased from Genecopoeia (GenePharma, Shanghai, China). The overexpression sequence of PLA2G4A was: forward: 5’-GCGGTAGGCGTGTACGGT-3’, reverse: 5’-ATTGTGGATGAATACTGCC-3’. Viruses were produced in LN229 and U251 cell lines according to the manufacturer’s instructions, and EndoFectin transfection reagent was used to transfect with the overexpression vector EX-L5215-Lv105 (PLA2G4A) and the corresponding empty vector EX-EGFP-Lv105 in proportion. After transfection to construct a stable overexpression cell line, western blot was used for verification. Finally, the transfected cells were selected with puromycin and used for subsequent experiments.

Co-lmmunoprecipitation (Co-lP) and mass spectrometry

Cells were lysed in IP buffer (R0100, Solarbio, China) containing a cocktail of protease inhibitors and phosphatase inhibitors. The lysate was incubated with the appropriate primary antibody or an equal amount of IgG to the primary antibody at 4 °C with rotation overnight. Then 50ul of Protein A/G agarose beads (SC-2003, Thermo Fisher Scientific, China) were added and incubated at 4 °C with rotation for 4 h. The immunoprecipitated complexes were then washed three to four times with lysis buffer and eluted with Co-IP antibody elution buffer (T10007M, Abmart, China) according to the instructions. The eluted complexes were separated by SDS-PAGE for mass spectrometry analysis or western blotting. For liquid chromatography tandem mass spectrometry analysis, samples obtained by the above method were subjected to SDS-PAGE, and the gel was stained with Coomassie Brilliant Blue Fast Stain G-250. Protein bands were cut out from the gel and then analyzed on LC-MS.

Flow cytometry

To evaluate the effect of CENPF on glioma cell apoptosis and cell cycle, BD FACSCanto flow cytometer was used for detection. Annexin V-FITC/PI double staining apoptosis detection kit (BestBio) was used to detect cell apoptosis according to the instructions; cell cycle was detected using a cell cycle detection kit (BestBio) according to the instructions. FlowJo software was used to analyze glioma cell apoptosis data, and Modfit software was used to analyze glioma cell cycle data.

Cell proliferation assays

Cell viability was measured using the Cell Counting Kit-8 (CCK-8) assay according to the manufacturer’s instructions (Beyotime, China). After transfection, cells were seeded in 96-well plates at a density of 2000 cells per well; absorbance values ​​were measured from 0 to 3 days after transfection. 10 µL of CCK-8 solution was added to each well of the 96-well plate and incubated for another 2 h. Then, the absorbance at 450 nm (OD 450) was measured using a microplate reader (Bio-Rad, USA).

Transwell invasion assay

After trypsin digestion, LN229 and U251 cells were placed in Transwell chambers (24-well, 8.0 mm pore size, Corning) at a density of 8 × 10^4/well and cultured at 37 °C for 48 h. 100 µL of cell suspension without FBS was added to the inner chamber, and 500 µL of DMEM medium containing 10% FBS was added to the outer chamber. The non-invaded cells in the upper chamber were removed, and the cells attached to the Matrigel-coated Transwell filter were fixed with pre-cooled 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet at room temperature for 15 min. Finally, the cells that migrated to the lower side of the membrane in each upper chamber were photographed under a Nikon Eclipse Ti microscope at a field of view of 200 times.

Molecular docking

To further explore the protein-protein interaction between CENPF and PLA2G4A, the three-dimensional structure was downloaded from the Uniprot (https://www.uniprot.org). The water molecules and organic matter in the target protein were removed. The processed protein was then imported into AutoDockTools for hydrogenation, charge assignment, and the addition of atomic types. The docking simulations were carried out using AutoDockVina, and the resulting conformations were visualized and analyzed in PyMOL to assess the potential interaction between CENPF and PLA2G4A.

Statistical analysis

All experiments were conducted at least three times and are presented as mean ± standard deviation (SD). Comparisons between two independent groups were performed using the chi-square test and two-tailed Student’s t-test. One-way ANOVA was utilized to assess statistical significance among three or more groups. The Kaplan-Meier curve was used to analyze the prognosis of gliomas, and the log-Rank test was used to detect the survival difference between the groups. The χ² test was employed to analyze the relationship between CENPF expression and clinicopathological features. Univariate and multivariate Cox regression analyses were conducted to identify independent risk factors affecting the clinical prognosis of glioma patients. All data were statistically analyzed using R software (version 4.4.0), Prism 8, and SPSS (version 28.0). The statistical significance was determined when the P value was less than 0.05.

Results

CENPF is highly expressed in glioma and is associated with poor prognosis in the database

To explore whether CENPF is differentially expressed in glioma, we analyzed the mRNA expression levels of CENPF in TCGA, CGGA, and GEO datasets. Our results demonstrated that CENPF expression in glioma tissues was significantly upregulated compared to normal tissues, with higher expression levels correlating with increased WHO grade (Fig. 1A-F). Survival analysis of the database showed that the expression of CENPF was associated with poor prognosis of patients (P < 0.001) (Fig. 1G-I). Subsequent investigations revealed elevated CENPF mRNA levels in IDH-1 wild-type, 1p/19q non-codel status, as well as in response to chemotherapy (Fig. S1A-F). These results suggest that CENPF expression contributes to the malignant progression of gliomas. These results suggest that CENPF expression contributes to the malignant progression of gliomas.

Fig. 1

CENPF is increased and correlates with poor prognosis of glioma in database. (A) Differential CENPF mRNA expression in the TCGA cohort. (B, C) The mRNA level of CENPF increased with the increase of grade in the CGGA (CGGA 325 and CGGA 693) database. (D-F) CENPF mRNA expression in the TCGA cohort. Differential CENPF mRNA expression in microarray (GSE54004, GSE16011 and GSE4290). (G-I) Kaplan-Meier curves were drawn based on the CENPF mRNA levels in the TCGA, CGGA 693, and GSE16011 datasets. *P < 0.05; **P < 0.01; ***P < 0.001

Somatic variations between high- and low- CENPF group

We further investigated the differential changes in CENPF expression associated with somatic mutations in the TCGA cohort. The top 20 driver genes with the highest mutation frequencies in the high- and low- CENPF groups were further analyzed. The results of somatic variations indicated statistically significant differences in the mutation frequencies of IDH1, EGFR, CIC, PTEN, ATRX and TTN between the CENPF high expression group and the low expression group (Fig. 2A, B). Figure 2C shows the TMB difference between the high- and low- CENPF group (P < 0.001). Survival analysis showed that patients with high TMB and high expression of CENPF had the shortest survival, while patients with low TMB and low expression of CENPF had a significantly prolonged survival (P < 0.001) (Fig. 2D). Next, we explored the signaling pathways involved in the high CENPF group. The KEGG analysis showed that the high expression CENPF group showed higher levels of P53 signaling pathway, RNA degradation, cell cycle and DNA replication (Fig. 2E).

Fig. 2

Correlation Between CENPF expression and somatic variants. (A-B) Somatic variants were analyzed, with low CENPF expression (blue) on the left and high CENPF expression (red) on the right. (C) TMB score in high and low CENPF expression. (D) Kaplan-Meier curves for glioma patients in the TCGA cohort, stratified by both TMB and CENPF expression levels. (E) Heatmap showing differential enrichment of signaling pathway in high and low CENPF expression

Expression of CENPF in glioma tissues and cell lines

To verify the prognostic value of CENPF in gliomas, we used western blot to detect the expression of CENPF in paracancerous tissues (n = 4) and glioma tissues (n = 12) from Linyi People’s Hospital. The results showed that the expression of CENPF was positively correlated with the WHO grade of gliomas (P < 0.001) (Fig. 3A, B). IHC results demonstrated that CENPF was primarily localized in the nucleus, with some staining in the cytoplasm. CENPF was expressed in all 60 glioma tissue specimens, and expression levels increased with ascending WHO grade (Fig. 3C). Clinicopathological characteristics analysis showed that CENPF expression was significantly associated with WHO grade (χ² = 9.479, P = 0.002), P53 mut (χ² = 5.725, P = 0.016), and Ki-67 expression (χ² = 16.194, P = 0.001) (Fig. 3D; Table 1). In univariate and multivariate Cox regression analyses, WHO grade (HR = 16.304, P < 0.001), CENPF expression (HR = 3.939, P = 0.008), Ki-67 expression (HR = 4.650, P = 0.006) and P53 mutation (HR = 4.653, P = 0.003) were identified as independent prognostic factors (Table 2). Kaplan-Meier analysis showed that patients with high CENPF expression had a significantly worse prognosis compared with patients with low CENPF expression (P < 0.001) (Fig. 3E).

Fig. 3

The expression of CENPF was correlated with clinicopathological parameters of glioma in our hospital. (A, B) Western blot detection of CENPF levels in paracancerous tissues and glioma tissues. (C) IHC analysis was performed to detect the expression levels of CENPF in glioma tissues compared to control. (D) The correlation between CENPF expression and proliferation biomarkers (Ki-67, P53mut) using IHC. (E) Kaplan-Meier curves of glioma patients with high and low CENPF expression. The bars indicate the means ± SDs; *P < 0.05; **P < 0.01; ***P < 0.001

Table 1 Correlation between CENPF expression and clinicopathological features in gliomas

Table 2 Univariate and multivariate Cox analysis of CENPF in gliomas from our hospital

Silencing CENPF inhibits glioma cell proliferation and invasion

In vitro experiment, we detected the protein expression levels of CENPF in human glioma cell lines and found that CENPF was overexpressed in U251 and LN229 cell lines (Fig. 4A, B). Therefore, U251 and LN229 cell lines were selected for subsequent experiments. We silence endogenous CENPF by transfecting specific siRNA, constructed three groups of U251 and LN229 cell lines, and detected the silencing efficiency by Western blot. After transfection of siRNA, the expression of CENPF in U251 and LN229 cell lines was significantly lower than that in the control group (Fig. 4C-E). We selected siCENPF#1 and siCENPF#3 with the best silencing efficiency for subsequent experiments. The CCK-8 experiment showed that in U251 and LN229 cells, the cell survival rate in the siCENPF group was lower than that in the control group (Fig. 4F-G). The transwell experiment showed that the cell invasion ability in the siCENPF group was lower than that in the control group in U251 and LN229 cells (Fig. 4H). The above results suggested that silencing CENPF can significantly inhibit the growth of glioma cells.

Fig. 4

Silencing CENPF has an anti-glioma effect. (A, B) Western blot detection and quantitative analysis of the relative expression levels of CENPF protein in glioma cell lines. (C-E) Western blot detection of the interference effect of silencing CENPF in U251 and LN229 cells. (F, G) CCK-8 assay was used to evaluate the proliferation ability of U251 and LN229 cells with silencing CENPF. (H) Transwell assay to detect the invasion ability of U251 and LN229 cells after silencing CENPF for 48 h. The bars indicate the means ± SDs; *P < 0.05; **P < 0.01; ***P < 0.001

Silencing of CENPF induces G2/M arrest and apoptosis in glioma cells

The GSEA results on GO terms showed that CENPF was involved in the glioma cell cycle and apoptosis involving the P53 signaling pathway (Fig. 5A, B). Flow cytometry was used to analyze the effects of siCENPF on apoptosis and the cell cycle of glioma cell lines. Compared with the control group, the apoptosis rate of glioma cells (U251 and LN229) increased after silencing of CENPF (Fig. 5C). Quantitative data for apoptosis were presented in Fig. 5D-E (P < 0.05). Additionally, the cell cycle results showed that silencing CENPF caused glioma cells to mainly arrest in the G2/M phase (Fig. 5F). The cell cycle statistics were shown in Fig. 5G-H (P < 0.05). We then detected the changes in apoptosis and cell cycle-related proteins. The results showed that after silencing CENPF, the protein level of pro-apoptotic factor BAX increased, and the protein level of anti-apoptotic factor Bcl-2 decreased (Fig. 5I-K). After silencing CENPF, the expression of G2/M-related molecules such as CCNB1, CCNB2 and CDK1 decreased significantly (Fig. 5I-K). Taken together, these observations suggested that silencing CENPF promotes apoptosis and inhibits cell cycle progression in glioma cells.

Fig. 5

Silencing of CENPF leads to cell cycle arrest and apoptosis in glioma cells. (A, B) GO analysis showed that the expression of CENPF in glioma was related to cell cycle and apoptosis in TCGA and GSE16011 cohort. (C-E) Flow cytometry to detect the apoptosis of U251 and LN229 cells after silencing CENPF for 48 h. (F-H) Flow cytometry to detect the cell cycle of U251 and LN229 cells after silencing CENPF for 48 h. (I-K) After silencing CENPF for 48 h, Western blotting was used to detect the relative expression levels of cell cycle-related proteins and apoptosis-related proteins. The bars indicate the means ± SDs; *P < 0.05; **P < 0.01; ***P < 0.001

CENPF induces EMT by activating the mTORC1 signaling pathway

To further explore the relevant signaling pathways mediated by CENPF, we performed a GSEA enrichment analysis based on TCGA and GSE16011 datasets. The GSEA results showed that CENPF was significantly correlated with the G2/M checkpoint, mTORC1 signaling pathway, DNA repair and Epithelial-mesenchymal transition (EMT) (Fig. 6A, B). Western blotting results showed that the expression levels of p-mTOR, p-S6K and p-4EBP1 proteins in glioma cells (U251 and LN229) were down-regulated after silencing CENPF (Fig. 6C) and the protein quantification was shown in Fig. 6D-E. At the same time, western blotting was used to detect the expression of EMT-related proteins. The results showed that compared with the control group, the expression level of N-Cadherin in the siCENPF transfection group was downregulated, and the expression level of E-Cadherin protein was significantly upregulated, indicating that the cell invasion ability was reduced (Fig. 6C-E).

Fig. 6

CENPF induced EMT and activated mTORC1 pathway. (A, B) Results based on TCGA and GSE16011 cohort showed that the expression of CENPF in glioma was associated with the mTORC1 signaling pathway and EMT process. (C-E) After silencing CENPF treatment of U251 and LN229 cells for 48 h, western blot was used to detect the levels of related proteins in the mTORC1 signaling pathway and EMT process. The bars indicate the means ± SDs; *P < 0.05; **P < 0.01; ***P < 0.001

We further treated CENPF-silenced glioma cells with MHY1485 (an mTORC1 signaling activator) to confirm the role of CENPF in the mTORC1 pathway. The results showed that the expression levels of p-mTOR, p-S6K, and p-4EBP1 were reduced in U251 and LN229 cells with siCENPF compared with the control group. Treatment with MHY1485 was able to reverse the reduction in the expression of these proteins (Fig. S2A-C). These results confirmed that CENPF regulated the growth and EMT process of glioma cells through the mTORC1 signaling pathway.

CENPF interacts with PLA2G4A and modulates NF-κB and mTORC1 signaling pathways

To better understand the mechanism of CENPF, particularly given its ability to directly bind proteins and exert regulatory functions, we employed mass spectrometry and Co-IP to identify CENPF-binding proteins. KEGG analysis of differentially expressed proteins by mass spectrometry showed that they were mainly involved in the immune system, antineoplastic drug resistance, replication/repair and cell growth/death (Fig. S3A). Subsequently, we identified PLA2G4A as an interacting protein (Fig. S3B). GSEA analysis of PLA2G4A showed that it was mainly involved in immune response, DNA replication, cell cycle checkpoint signaling and NF-κB pathway (Fig. S3C). Correlation analysis showed that the expression of CENPF and PLA2G4A was positively correlated in the CGGA database (R = 0.525, P < 0.001) (Fig. 7A). Molecular docking analysis showed that the binding energy between CENPF and PLA2G4A was − 20.8 kcal/mol, indicating an exceptionally strong interaction between the two molecules. The docking result was displayed in a 3D graph (Fig. 7B), and there were four hydrogen bonds between CENPF and PLA2G4A. Furthermore, Co-IP assay was used to verify the interaction between CENPF and PLA2G4A (Fig. 7C). Several studies have reported that PLA2G4A can activate the NF-κB signaling pathway, which is consistent with GSEA analysis [23]. Western blot analysis confirmed that in U251 and LN229 cells, overexpression of PLA2G4A increased its phosphorylation level and activated the NF-κB pathway (Fig. 7D).

Fig. 7

CENPF interacts with PLA2G4A and regulates both the mTORC1 and NF-κB signaling pathways. (A) CENPF was positively correlated with the expression of PLA2G4A in the CGGA database. (B) The 3D docking conformation of CENPF with PLA2G4A. (C) Co-IP experiments demonstrated that CENPF interacts with PLA2G4A in LN229. (D) After stably overexpressing PLA2G4A using lentivirus, western blot was used to detect the expression levels of NF-κB pathway proteins in U251 and LN229 cells. (E, F) CCK-8 assay detected the effects of CENPF silencing, PLA2G4A overexpression and the combination on the proliferation of U251 and LN229 cells. (G) Transwell assay detected the effects of CENPF silencing, PLA2G4A overexpression and the combination on the invasion ability of U251 and LN229 cells. (H-J) The expression of mTORC1 pathway and EMT process proteins in U251 and LN229 cells treated with CENPF silencing, PLA2G4A overexpression, or both. The bars indicate the means ± SDs; *P < 0.05; **P < 0.01; ***P < 0.001

Next, we explored the interaction between CENPF and PLA2G4A and its effects on tumor growth and downstream pathways. We combined silencing CENPF with overexpressed PLA2G4A to explore their effects in cell function experiments. CCK-8 assay showed that the viability of glioma cells in the CENPF silencing group decreased, while overexpression of PLA2G4A could partially restore the cell viability induced by CENPF silencing (Fig. 7E-F). Transwell assay showed that compared with the control group, the invasion ability of cells in the CENPF silencing group decreased, while overexpression of PLA2G4A could restore the invasion ability of cells in the silencing group (Fig. 7G). Western blot was used to explore the effects of silencing CENPF and overexpressing PLA2G4A on signaling pathways. Western blot analysis showed that the expression levels of p-4EBP1 and N-Cadherin in U251 and LN229 cells with CENPF silencing were lower compared with the control group, but the expression of p-PLA2G4A, p-P65 and E-Cadherin was increased (Fig. 7H-J). When PLA2G4A was overexpressed, the expression of mTORC1 signaling pathway and NF-κB signaling pathway proteins were increased (Fig. 7H-J). These results further demonstrate that targeting CENPF inhibits tumor growth by modulating the mTORC1 signaling pathway, while simultaneously activating the NF-κB signaling pathway to exert a compensatory effect.

Silencing CENPF in combination with AACOCF3 further inhibits glioma growth

AACOCF3 is a widely used and selective slow-binding inhibitor of PLA2G4A [24, 25]. To investigate the potential synergistic effect of silencing CENPF in combination with AACOCF3 on glioma growth, we analyzed the expression of EMT and apoptosis-related proteins. Western blot results showed that silencing CENPF in conjunction with AACOCF3 significantly inhibited the expression of Bcl-2, N-Cadherin, and Vimentin, while the expression of Bax and E-Cadherin was upregulated, indicating that glioma cells underwent epithelial transformation accompanied by increased apoptosis (Fig. 8A-C). Furthermore, CCK-8 and transwell assays showed that silencing CENPF and AACOCF3 alone could inhibit the proliferation and migration of gliomas (Fig. 8D-F). Notably, the combination of CENPF silencing and AACOCF3 treatment produced significantly greater inhibition of proliferation and migration compared to either treatment alone (Fig. 8D-F). This finding underscores the synergistic inhibitory effect on glioma growth resulting from the combination of CENPF silencing and AACOCF3 treatment.

Fig. 8

Silencing CENPF combined with AACOCF3 further inhibits glioma growth. (A-C) Western blot was used to detect the effects of CENPF silencing, AACOCF3 treated and the combined on the expression levels of apoptosis-related and EMT-related proteins in U251 and LN229 cells. (D, E) CCK-8 assay was used to detect the effects of CENPF silencing, AACOCF3 treated and the combined on the proliferation of U251 and LN229 cells. (F) Transwell assay was used to detect the effects of CENPF silencing, AACOCF3 treated and the combined treatment on the invasion ability of U251 and LN229 cells. The bars indicate the means ± SDs; *P < 0.05; **P < 0.01; ***P < 0.001

Discussion

Recent studies have shown that the alterations in the genome and immune microenvironment play a critical role in the initiation and progression of tumors [26]. With the updated classification of gliomas in the 2021 WHO classification of central nervous system tumors, precise molecular targeting based on genomic alterations offers new hope for genetic precision therapy [3]. Among central nervous system malignancies, glioblastoma (GBM) is the most lethal, marked by rapid progression, a high recurrence rate, and a propensity for drug resistance. Glioma recurrence is a common occurrence due to the unique invasiveness of glioma cells, which allows them to penetrate brain parenchyma and infiltrate normal cells. Therefore, molecular targeted therapy has emerged as a promising treatment approach for gliomas, as conventional treatments (surgery combined with radiation and chemotherapy) have shown limited efficacy against gliomas [27].

The CENPF is located on chromosome 1q41, with a full-length molecular weight of approximately 350 kDa and containing 3210 amino acids [28]. It is a centromere-associated protein that regulates the cell cycle and is involved in recruiting spindle checkpoint proteins to maintain checkpoint responses. CENPF is involved in the segregation of sister chromatids during the cell cycle and is expressed in a cell cycle-dependent pattern, with low expression levels in G0/G1 cells, accumulation in the nuclear matrix during the S phase, highest expression levels in G2/M cells, and rapid degradation upon completion of mitosis [29]. In previous studies, the role of CENPF was primarily associated with the regulation of mitosis. However, recent investigations have provided evidence that CENPF functions as an oncogene, with its overexpression linked to the progression of various cancers [15, 30, 31]. Sun et al. found that CENPF increased the secretion of parathyroid hormone-related protein (PTHrP) and promoted breast cancer bone metastasis by activating the PI3K-AKT-mTORC1 signaling pathway [15]. Further experiments showed that knocking down CENPF can significantly reduce the resistance of triple-negative breast cancer to doxorubicin chemotherapy through the Rb-E2F1 axis [32]. Recent evidence suggests that the expression of CENPF was associated with the clinical characteristics of gastric cancer patients, and CENPF modified by increased m6A promoted gastric cancer metastasis and angiogenesis via the CENPF/FAK/MAPK and EMT axes [14]. However, the underlying mechanisms of CENPF expression in gliomas remain unclear.

In our previous studies, we systematically integrated multiple microarray expression profiles and found that CENPF is a core molecule through a series of bioinformatics analyses [19]. In this study, we observed a significant upregulation of CENPF across multiple glioma datasets, with its expression positively correlating with the increasing grade of gliomas. Database survival analysis also showed a strong correlation between CENPF expression and overall survival of glioma patients. Analysis of somatic variations showed that CENPF expression was significantly correlated with the mutation frequencies of IDH1, EGFR, CIC, PTEN, ATRX and TTN. These bioinformatics analyses suggest that CENPF may serve as a new prognostic indicator and potential molecular therapeutic target for glioma patients.

The results of bioinformatics analysis were verified in glioma samples from our hospital. Analysis of clinical pathological parameters showed that CENPF expression was an independent prognostic factor for glioma. Kaplan-Meier analysis showed that the overall survival of glioma patients with high CENPF expression was significantly lower than that of glioma patients with low CENPF expression. Subsequently, we conducted in vitro experiments to assess the expression of CENPF in glioma cell lines and identified U251 and LN229 as the cell lines with the highest expression for subsequent investigations. Functional experiments found that silencing CENPF could significantly inhibit cell proliferation and invasion levels, thereby inhibiting cell growth.

To explore the molecular mechanism of CENPF inhibiting glioma growth, we performed a GO enrichment analysis. The GO enrichment analysis showed that CENPF was mainly involved in the glioma cell cycle (G2/M phase) and apoptosis involving the P53 signaling pathway. In cell cycle regulation, activated CCNB1-dependent CDK1 triggers the transition of cells from the G2 phase to the mitotic phase [33]. P53 is a tumor suppressor gene that is involved in regulating apoptosis, cell cycle arrest, senescence, and metabolism [34]. It has been reported that P53 inhibits CDK1 and CCNB1 at the G2 checkpoint, preventing cells from entering the mitotic phase [35]. CENPF is mainly involved in sister chromosome separation during the cell cycle, and its dysregulated expression will lead to abnormal cell growth [36]. Based on the bioinformatics results, we used flow cytometric analysis to explore the effects of silencing CENPF on cell cycle and apoptosis. The results indicated that silencing CENPF led to an accumulation of cells in the G2/M phase, along with an increased level of apoptosis, resulting in a significant inhibition of glioma cell viability. We also detected cell cycle and apoptosis-related proteins by western blot experiments, and their expression patterns supported the above results. We then used GSEA enrichment analysis to explore the signaling pathways in which CENPF was involved. GSEA results showed that CENPF was involved in the mTORC1 signaling pathway and EMT process. The mTOR is a highly conserved serine/threonine kinase that primarily exists as mTORC1 and mTORC2 complexes within cells. It facilitates intracellular protein synthesis, supplying the necessary resources for tumor cell growth [37]. The activation of mTORC1 accelerates the transcription of specific intracellular mRNAs and increases the ability of intracellular protein synthesis [38, 39]. The mTORC1/S6K/4E-BP1 signaling directly regulates cell growth and plays a key role in regulating protein translation efficiency, cell proliferation, and oncogenic transformation [40]. When CENPF was silenced in the glioma cell line, the phosphorylation levels of the mTORC1/S6K/4E-BP1 axis decreased, indicating that CENPF functions as an upstream regulator of mTORC1. EMT is a process in which epithelial cells lose connectivity and polarity while acquiring mesenchymal properties and invasive ability, which gives cancer cells higher metastatic potential and drug resistance and is now considered a key process in tumor progression [41]. The EMT process is also one of the key behaviors driving invasion and drug resistance in glioma cells and deserves further study [42]. Accumulating evidence indicates that the mTORC1 signaling pathway plays a crucial role in the invasion and migration of various cancer cells [43]. Numerous studies have demonstrated that mTORC1 regulates the EMT process, thereby influencing the invasive ability of tumors [44, 45]. Cai et al. found that fatostatin induces ferroptosis and EMT through the mTORC1 signaling pathway in glioblastoma [46]. We also employed western blot to assess the changes in the expression levels of EMT-related proteins. Our results showed that silencing CENPF resulted in downregulation of the mesenchymal marker N-cadherin and upregulation of the epithelial marker E-cadherin, indicating that glioma cells switched from a mesenchymal to an epithelial phenotype, thereby reducing invasiveness. Rescue experiments showed that after treatment with MHY1485, silencing of CENPF partially reversed the inhibitory effect on the mTORC1 signaling pathway in glioma cells. Taken together, the above results suggest that CENPF regulates glioma cell growth and the EMT process through the mTORC1 signaling pathway.

To further investigate the mechanism of CENPF in glioma, mass spectrometry analysis suggested its potential interaction with PLA2G4A, which was subsequently validated through molecular docking and Co-IP analyses. PLA2G4A is the most abundant isoform of cytosolic phospholipase A2 (cPLA2), which cleaves fatty acyl bonds at the sn-2 position of glycerophospholipids and is an enzyme that regulates membrane phospholipid homeostasis and arachidonic acid release in inflammatory responses [47]. PLA2G4A is one of the major phospholipases present in the brain and has the function of surrounding lysosomes, making them susceptible to phospholipase activation [23]. Studies have found that the activation of PLA2G4A is associated with lysosomal membrane permeabilization in traumatic brain injury and may play a role in Alzheimer’s disease and other potential age-related neurodegenerative diseases [47]. Several studies have shown that PLA2G4A exerts its effects by activating the NF-κB signaling pathway [23, 48, 49]. We experimentally confirmed that overexpression of PLA2G4A increased its phosphorylation level and activated the NF-κB pathway in gliomas. Next, we explored the interaction mechanism between CENPF and PLA2G4A in glioma cells and its impact on signaling pathways. We combined siCENPF with overexpressed PLA2G4A to explore its effect on the malignancy of glioma. Our findings indicate that PLA2G4A exhibits a higher affinity for CENPF, preferentially interacting with it to activate the mTORC1 pathway. Only when CENPF expression decreases and PLA2G4A is abundant does PLA2G4A undergo phosphorylation to activate the NF-κB signaling pathway. Silencing CENPF in glioma may activate the NF-κB signaling pathway as a form of self-rescue for the tumor, which underlies mechanisms of drug resistance and immune escape [50].

The activation of the NF-κB pathway is associated with tumorigenesis and drug resistance. Studies have suggested that inhibiting the NF-κB signaling pathway may serve as a potential therapeutic strategy for suppressing glioma cell proliferation and survival [51]. Kurupe et al. found that ADAR3 enhances the resistance of GBM cells to temozolomide by activating NF-κB signaling [52]. Other studies have demonstrated that Calpain-2 can downregulate DNA damage signaling proteins to block DNA damage recognition and active NF-κB signaling to inhibit the sensitivity of GBM cells to temozolomide [53]. Based on this, we further explored whether silencing CENPF and blocking the NF-κB signaling pathway could have an enhanced inhibitory effect on glioma growth. AACOCF3, an effective and selective slow-binding inhibitor, inhibits PLA2G4A, which has been demonstrated in previous studies to be involved in the NF-κB signaling pathway [25, 54]. Our results showed that silencing CENPF combined with AACOCF3 could significantly inhibit the EMT process and increase apoptosis, which could more effectively inhibit the growth of gliomas.

Conclusion

In summary, our study showed that CENPF was significantly upregulated in glioma and correlated with poor prognosis. We found that CENPF, as an oncogene, promoted the malignant phenotype of glioma by activating the mTORC1 signaling pathway and inducing EMT process. Subsequent experiments revealed that CENPF interacted with PLA2G4A and modulated the mTORC1 and NF-κB signaling pathways. Based on our mechanistic findings, a dual-targeting approach is proposed, which provides a promising avenue for developing more effective glioma treatments. We believe these advancements significantly enhance the novelty and impact of our study, providing valuable insights into the molecular underpinnings of glioma progression and potential therapeutic interventions.