https://orcid.org/0000-0001-5712-3624
Figures
Abstract
Glioblastoma (GBM) is a fast-growing primary brain tumor with high mortality and recurrence rates. Dysregulation of the cell cycle is a hallmark of GBM, and cyclin B1 (CCNB1) is a key regulator of the cell cycle. However, the role of CCNB1 in GBM remains unclear. In this study, we found that CCNB1 mRNA and protein expression levels were significantly higher in GBM tissues than normal tissues. High CCNB1 mRNA expression was associated with poorer prognosis in GBM patients. Single-cell and spatial transcriptomics data revealed that CCNB1+ cells represent a proliferative subcluster in GBM, annotated as proliferative cells, and characterized by the upregulation of cell cycle-related pathways. CCNB1 inhibition decreased the proliferation of GBM cells and impaired cell cycle progression from S phase to G2/M. Additionally, resveratrol could inhibit the expression of CCNB1 and its interacting gene polo-like kinase 1 (PLK1). Importantly, through in vitro and in vivo experiments, we found that resveratrol suppressed GBM cell growth with low toxicity. CCNB1 silencing combined with resveratrol treatment further inhibited the proliferation of GBM cells. Collectively, these data suggest that CCNB1 is highly expressed in GBM and may promote GBM progression. Inhibition of CCNB1 may represent a potential therapeutic strategy for GBM.
Citation: Liu B, Peng D, Chen Y, Li Q, Hu Y, Liu S, et al. (2026) Multi-omics analysis identifies CCNB1 as a cell cycle factor driving glioblastoma progression and its inhibition by resveratrol. PLoS One 21(3): e0344872. https://doi.org/10.1371/journal.pone.0344872
Editor: Amir Faisal, Lahore University of Management Sciences, Syed Babar Ali School of Science and Engineering, PAKISTAN
Received: August 1, 2025; Accepted: February 26, 2026; Published: March 16, 2026
Copyright: © 2026 Liu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The RNA-seq data of U87 and U251 cells treated with resveratrol or DMSO are provided within the Supplementary material. Other datasets are downloaded from the Cancer Genome Atlas (TCGA) database (https://www.cancer.gov/about-nci/organization/ccg/research/structural-genomics/tcga), Genotype-Tissue Expression (GTEx) database (https://www.gtexportal.org/home/), Chinese Glioma Genome Atlas (CGGA) database (http://www.cgga.org.cn/), UALCAN (https://ualcan.path.uab.edu/) database and the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/).
Funding: This work was supported by grants from Beijing Municipal Health Commission (JYY2023-2), National Natural Science Foundation of China (82372782, 82341048), and Beijing Municipal Science & Technology Commission, Administrative Commission of Zhongguancun Science Park Z241100009024044.
Competing interests: The authors have declared that no competing interests exist.
Introduction
GBM is the most aggressive and lethal primary brain tumor in adults, characterized by rapid malignant progression [1,2]. Although radiotherapy, chemotherapy and comprehensive treatments have benefited GBM patients to some extent, the overall prognosis remains poor, with a median survival of less than 2 years [3,4]. The rapid malignant progression of GBM is attributed to the vigorous proliferation of tumor cells.
Sustaining proliferative signaling and enabling replicative immortality are two important hallmarks acquired during the development of human tumors [5], including GBM. Dysregulation of the cell cycle is an important manifestation of these two biological capacities in GBM, which can lead to rapid proliferation of GBM cells. As a key cell cycle regulator, CCNB1 can regulate cell mitosis at the G2/M phase through the interaction with cyclin-dependent kinase 1 (CDK1) [6]. CCNB1 is frequently altered in tumors, leading to uncontrolled cell division and cell cycle of tumor cells [7,8], thus promoting rapid proliferation of tumors. High expression of CCNB1 is also related to invasion, resistance and recurrence of various tumors [9]. For instance, CCNB1 has been reported as an independent risk factor for the recurrence of non-muscle-invasive bladder cancer and prostate cancer [10,11]. Additionally, CCNB1 can mediated SIRT3 activation and enhance tumor radioresistance of multiple tumor cells [12]. Although, several studies have reported that CCNB1 is highly expressed in GBM or other tumors [13,14], and silencing the expression of CCNB1 or other cell cycle-related genes promote the anticancer activities of temozolomide [15], the role of CCNB1 in GBM still lacks sufficient evidence. Thus, we aimed to further understand CCNB1-induced cell cycle dysregulation and to develop a novel therapeutic strategy in GBM.
Resveratrol is a natural polyphenolic compound found in grapes, berries, and peanuts. It exerts antiproliferative, pro-apoptotic, and anti-inflammatory effects in various types of cancer, including breast, colorectal, multiple myeloma, and prostate cancers [16–18], and has shown the clinically beneficial to patients [19]. Ahmad et al. first reported that resveratrol can inhibit cyclin D1/D2-CDK6, cyclin D1/D2-CDK4, and cyclin E-CDK2 complexes, thus causing arrest of the cell cycle and promoting apoptosis of cancer cells [20]. Baek et al. revealed that resveratrol can inhibit STAT3 activation, which suppresses the proliferation and invasion of head and neck tumor cells [21]. Moreover, in GBM cells, resveratrol inhibits STAT3 signaling pathway [22], and induces the production of reactive oxygen species, activation of caspases 3/7 and arrest of the cell cycle [23], thereby attenuating proliferation of GBM cells and increasing their sensitivity to TMZ. However, there is still insufficient evidence to show whether resveratrol can down-regulate the expression of CCNB1.
In this study, we demonstrated that the expressions of CCNB1 mRNA and protein were higher in GBM tissues compared within normal brain tissues. The expression of CCNB1 mRNA increased with the progression of pathological grade. High CCNB1 mRNA expression indicated poor prognosis of GBM patients. Prognostic model showed that CCNB1 and its interacting genes may serve as important predictors of overall survival (OS) in GBM patients. Furthermore, single-cell and spatial transcriptomic datasets identified a subpopulation of CCNB1+ proliferative cells in the mitotically active phase of the cell cycle. Notably, we revealed that CCNB1 silencing led to significant accumulation of cells in S phase and decreased the growth of GBM cells. Resveratrol suppressed GBM cell proliferation and inhibited the expression of CCNB1 and its interacting gene PLK1. CCNB1 inhibition combined with resveratrol treatment further suppressed the proliferation of GBM cells. The inhibition of CCNB1 expression and cell cycle may be a new strategy for the treatment of GBM.
Materials and methods
Cell culture
The human glioma cell lines U87 and U251 and human astrocyte cell (HA) were purchased from the Cell Bank of the Chinese Academy of Sciences. GBM cells were stably cultured in DMEM (Dulbecco's modified Eagle's medium, basic 1×) containing 10% fetal bovine serum (Gibco) and in an incubator (37 °C, 5% CO2). Human astrocyte cells were cultured in Astrocyte medium (1801, Sciencell, USA). Mycoplasma surveillance and STR sequencing were routinely performed.
Data collection and processing
RNA-seq data for glioma tissues (CGGA-693, CGGA-325 and CGGA-301) and normal brain tissues were downloaded from the CGGA database [24]. Using the GEPIA2 database [25], we performed a combined analysis of The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) data to examine the differential expression of CCNB1 between glioma and normal tissues. CCNB1 expression levels in various cancer tissues were obtained from the TCGA database. Protein expression data for CCNB1 in GBM and normal brain tissues were retrieved from the UALCAN database [26]. Based on the Human Protein Atlas (HPA) database (https://www.proteinatlas.org/), we analyzed the expression of CCNB1 mRNA and protein across various normal tissues. Immunohistochemical staining images of GBM histological sections were obtained from the “pathology atlas” module in the HPA database.
Analysis of CCNB1 expression and prognosis of GBM patients
GBM patients were divided into high and low CCNB1 expression groups (cutoff = the median CCNB1 mRNA expression levels). Kaplan-Meier survival analysis was then performed to investigate the association between CCNB1 mRNA expression and patient outcomes in CGGA-693, CGGA-325, CGGA-301 and TCGA cohorts. The R package pROC (version 1.17.0.1) was used to perform receiver operating characteristic (ROC) analysis and to obtain Area Under the Curve (AUC) values. Specifically, OS data and CCNB1 mRNA expression levels from CGGA and TCGA GBM patients were used, and ROC analysis at 1-, 3-, and 5-year time points were conducted using the roc function in pROC. Finally, the final AUC results were obtained by calculating AUC and confidence intervals (The ci function in pROC).
Single-cell RNA-seq (scRNA-seq) analysis
The scRNA-seq data was obtained from CGGA database and GSE214966 dataset [27,28]. The expression profile was read using the R package (Seurat, version 5.1.0). Exclusion of low-quality scRNA-seq data: Cells with fewer than 3 counts, fewer than 200 detected genes, and those that meet the criteria of nFeature_RNA > 200, and percent.mt < 20 were excluded from the analysis. Then, the data were standardized, normalized and PCA dimensionality reduction. Subsequently, the cell cluster identification and differentiation were performed by using FindNeighbors and FindClusters functions. Cell subsets were described by unsupervised cell placement using t-distributed random neighbor embedding (t-SNE, the resolution parameter was 0.5). The FindAllMarkers function identified the marker genes of each cluster, and we manually annotated each cell population by specific marker genes (with default parameters). The FeaturePlot function showed the distribution of expression patterns of different genes in tSNE space, and the DotPlot function was used to visualize gene expression across various cell types. Finally, we performed the Kyoto Encyclopedia of Genes and Genomes (KEGG) and the Gene Ontology (GO) enrichment analysis of differentially expressed genes (DEGs) in the proliferative cell subpopulation.
Spatial transcriptome data acquirement and processing
The spatial transcriptomics data of glioma tissues were obtained from the Heiland dataset [29]. The primary tool used for processing the spatial transcriptomics data was the R package SPATA2, developed by Jan Kueckelhaus, Dieter-Henrik Heiland, and Simon Frerich. The processing included all data filtering, normalization, dimensional reduction, and visualization. The spatial distribution of GO enrichment pathway was analyzed using “getGeneSets” function.
PPI network design and the establishment of prognostic model
By utilizing the STRING website (https://string-db.org/), we queried “CCNB1” in the protein name module and “Homo sapiens” in the organism module. We observed the available CCNB1 interacting partners (the specific criteria for selecting interacting genes: network type = full STRING network, meaning of network edges = evidence, active interaction sources = textmining, experiments and co‑expression, medium confidence = 0.4, maximum number of interactors = no more than 10 interactors).
Univariate analysis with cox-regression of OS was used to evaluate the independent prognostic value of CCNB1 and its 10 interacting genes in GBM patients. Forest plots were generated using R packages. After excluding interacting genes without prognostic value, LASSO regression analysis was employed to construct a prognostic model based on genes within the CCNB1 interaction network. The risk score formula = (−0.1498 × ESPL1) + (0.3847 × CDC20) + (0.2241 × CKS2) + (0.0842 × PLK1) + (0.6745 × ANAPC4) + (−0.2054 × CCNB1). Kaplan-Meier curves were used to assess the survival of GBM patients (high risk score vs low risk score groups, cutoff = median score). Based on the “pheatmap” package survival state diagrams, risk curves, and heatmaps were created. Based on the “timeROC” R package, ROC curves for 1-, 3-, and 5-year OS were plotted. Finally, a nomogram model was established to predict the survival of GBM by using “rms” R package.
Cell counting kit-8 (CCK8)
The U87 and U251 cells (1.0 × 104 cells/well) were seeded into 96-well plates. Cell proliferation was detected by CCK8 reagent (K1018, APExBIO, USA) for 6, 24, 48, and 72 h of resveratrol (S1396, Selleck, China) treatment. The assay was performed according to the manufacturer’s instructions. The concentration of resveratrol was determined based on previous studies and preliminary experiments, and 50 μM was selected as a stable treatment concentration [30,31].
Intracranial tumor model construction
All animal handling, surveillance and experimentation were performed in accordance with the approval from the Beijing Neurosurgical Institute, Laboratory Animal Care (approval number: 2023054). All personnel involved in animal handling received formal training in laboratory animal care and handling. Experiments were performed in accordance with relevant guidelines and regulations. All authors complied with the ARRIVE guidelines. All surgery was performed under anesthesia, and all efforts were made to minimize suffering.
The five-week-old NOD-Prkdcscid IL2rgnull (NPG) female mice were obtained from Beijing Vitalstar Biotechnology Co., Ltd. (Beijing, China). The GFP-luciferase lentiviruses were purchased from Genechem (Shanghai, China) and used to construct U87 cell lines labeled with GFP/luciferase. Mice were anesthetized with 1.25% tribromoethanol at a dose of 0.2 mL per 10 g of body weight. Using stereotactic instrument (Stoelting Co., USA) and microinfusion pump, U87 cells (5 × 105), labeled with GFP/luciferase, were implemented into the brain of NPG mice (5 mice/group, a total of 10 mice). Ten-day post-tumor injection, resveratrol (50 mg/kg, 1 time/day) was administered orally for 28 days. The control group was given the same dose of PBS solution. The body weight of the mice was monitored at 0, 10, 17, 24 and 31 days. The oral dosage of resveratrol was determined based on previous studies, and the experiments were conducted using doses within established safety ranges [32].
Mice anesthetized with 2% isoflurane (R510-22–16, RWD Life Science, China). The in vivo imaging system (PerkinElmer, USA) was used to capture the bioluminescent images (intraperitoneal injection of D-fluorescein, 150 mg/kg) (122799, PerkinElmer, USA). The animals are monitored daily by well-trained experimenters, and tumor growth was tracked using bioluminescence imaging. Mice were monitored and maintained until predefined humane endpoints are reached, including the onset of neurological symptoms, a body weight loss of ≥15–20%, sustained behavioral abnormalities, any other observable signs of significant distress, or until 60 days had elapsed. No mice were found dead during the experiment. Once the mice reached endpoint criteria, they were immediately euthanized. Mice were humanely euthanized using a carbon dioxide chamber. CO₂ was introduced at a flow rate of 20–30% of the chamber volume per minute to induce unconsciousness. Death was confirmed by the absence of corneal reflex, respiration, and heartbeat for at least 5 minutes. Finally, 9 mice reached endpoint criteria and were euthanized during experiment, and one mouse was lived until 60 days. The main organs (heart, liver, spleen, lungs, and kidneys) were collected for hematoxylin-eosin (H&E) staining to evaluate drug toxicity.
RNA sequencing (RNA-seq)
The U87 and U251 cells were treated with resveratrol (50μM) or DMSO for 48hs. RNA sequencing was performed by Hangzhou Lianchuan Biological Information Co., Ltd. The DEGs were identified (|fold change| > 2 and p < 0.05). Finally, the Volcano map showed the DEGs and GO enrichment analysis of DEGs were performed by using OmicStudio [33].
H&E staining
The main organs (heart, liver, spleen, lung, and kidney) of the mice were embedded in paraffin and sectioned. Paraffin sections were then deparaffinized and hydrated. H&E staining was performed using a kit (G1120, Solarbio, China) according to the manufacturer’s instruction.
RT-qPCR
Total RNA from cells was extracted using TRIzol reagent (Invitrogen). The RNA was then reverse transcribed into cDNA using a reverse transcription kit (A5001, Promega, USA). The qPCR reaction mixture was prepared, typically consisting of 10 μL of PowerUp™ SYBR™ Green Mix (A25741, Applied Biosystems™), 1 μL of each primer (10 μM), 2 μL of cDNA template, and 6 μL of ddH2O. The total reaction volume was 20 μL. The specific primers for each gene were as follows:
- CCNB1: (F) 5’-AACTTTCGCCTGAGCCTATTTT-3’
- (R) 5’-TTGGTCTGACTGCTTGCTCTT-3’
- PLK1: (F) 5’-AAAGAGATCCCGGAGGTCCTA-3’
- (R) 5’-GGCTGCGGTGAATGGATATTTC-3’
- CKS2: (F) 5’-TTCGACGAACACTACGAGTACC-3’
- (R) 5’-GGACACCAAGTCTCCTCCAC-3’
- CDC20: (F) 5’-GACCACTCCTAGCAAACCTGG-3’
- (R) 5’-GGGCGTCTGGCTGTTTTCA-3’
- ANAPC4: (F) 5’-TGAACCTCTTGGACTAGATGCT-3’
- (R) 5’-CATTGCCACATAAAGCCACCG-3’
- ESPL1: (F) 5’-TAAATCTACCGTTGTCTGGATGC-3’
- (R) 5’-CTCGGGATAAGTGCCTGGC-3’
- GAPDH: (F) 5’-GGTGGTCTCCTCTGACTTCAACA-3’
- (R) 5’-GTTGCTGTAGCCAAATTCGTTGT-3’
Western blot (WB) analysis
Cells were lysed on ice for 60 minutes using RIPA buffer supplemented with PMSF (1:100 dilution, Sigma). After centrifugation, the supernatant was collected as a protein extract. The protein concentration of the samples was determined using the BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein samples were mixed with loading buffer and heated at 95°C for 5 minutes, then loaded onto a 10% SDS-PAGE gel. Subsequently, electrophoresis and transfer to the membrane were performed. The membrane was blocked with a 5% non-fat milk solution. The primary antibodies were incubated overnight at 4°C. The membrane was washed three times with TBST buffer, each for 8 minutes. The membrane was then incubated with secondary antibodies for 1 hour. Chemiluminescence detection was performed using an ECL kit (Thermo Fisher Scientific), and the chemiluminescent signals were recorded and quantitatively analyzed. For simultaneous detection of CCNB1 and GAPDH, the membrane was first incubated with anti-CCNB1 antibody and detected by chemiluminescent, followed by incubating with the anti-GAPDH antibody on the same membrane. The antibodies were as follows: CCNB1 (28603–1-AP, Proteintech, China); GAPDH (60004–1-Ig, Proteintech, China).
Small interfering RNA (siRNA) production and infection
Three specific siRNA sequences targeting the CCNB1 were designed, and the sequences were as follows:
- siRNA1: 5’GCUGAAUUCUGCACUAGUUdTdT’3;
- siRNA2: 5’GGUAAAUCAAGGACUUACAdTdT’3;
- siRNA3: 5’CUGACAACACUUAUACUAAdTdT’3.
SiRNA transfection was performed using Lipofectamine™ 3000 transfection reagent (Thermo Fisher Scientific). 48 hours post-transfection, the mRNA or protein expression levels of the CCNB1 and its interacting genes were detected.
5-Ethynyl-2’-deoxyuridine (EdU) assay
Cells were treated using the Cell-Light EDU Apollo567 In Vitro Kit (Ribobio Co., Ltd, China). EdU solution was added to the culture medium to a final concentration of 10 μM, and cells were incubated for 2 hours. After treatment, cells were washed twice with PBS to remove residual culture medium. Then, cells were fixed with 4% paraformaldehyde (solarbio, P1110) at room temperature for 20 minutes. After fixation, cells were washed three times with PBS. Cells were incubated with EdU staining solution for 30 minutes. The cell nuclei were stained with Hoechst solution for 20 minutes. EdU-stained cells were observed and imaged using a fluorescence microscope. The percentage of EdU-positive cells in each field of view was calculated based on the fluorescence microscope images.
Cell cycle assay
The cell cycle distribution of U251 and U87 cells was analyzed using the Cell Cycle and Apoptosis Analysis Kit (C1052, Beyotime) according to the manufacturer’s instructions. Briefly, cells were harvested by trypsinization, fixed in 70% ethanol, and stained with propidium iodide (PI) solution. Cell cycle profiles were acquired using an Accuri™ C6 Plus flow cytometer (BD Biosciences, San Jose, CA, USA) based on red fluorescence following excitation at 488 nm. Data were analyzed using FlowJo software (v10.9.0).
Statistical analysis
Statistical analyses were primarily conducted using GraphPad Prism 8 software. To compare differences between two groups, an unpaired t-test was performed. For comparisons among multiple groups, a one-way analysis of variance (ANOVA) was applied. Survival data were evaluated using the log-rank test. p < 0.05 was regarded as statistically significant.
Results
CCNB1 promotes the progression of GBM
GBM has features of rapid progression. CCNB1 has been widely regarded as a clinical prognostic biomarker in various cancers, such as liver and breast cancer. However, its role in glioma has not been systematically explored. Therefore, we aimed to further investigate the expression of CCNB1 and its function in GBM. First, we compared the expression levels of CCNB1 in various normal tissues. According to the data from the Human Protein Atlas (HPA) database, the expression of CCNB1 mRNA was highest in thymus tissue, followed by expression in the tonsil. CCNB1 mRNA expression level was low in brain tissues compared to other organs (S1A Fig). Moreover, CCNB1 protein was undetectable in brain tissue (S1B Fig). However, distinct from normal brain, GBM displayed considerable CCNB1 expression level in TCGA (S1C Fig). GBM had much higher level of CCNB1 expression than low grade glioma (LGG). We further investigated CCNB1 expression in glioma tissues.
As shown in Fig 1A, CCNB1 mRNA expression was higher in GBM tissues compared to normal tissues in the CGGA-693 dataset (p < 0.0001). In contrast, CCNB1 mRNA expression in LGG was not significantly different from that in normal tissues (p = 0.1980). Additionally, CCNB1 mRNA expression was higher in GBM than in LGG (p < 0.0001). Similar results were observed in the CGGA-325 and CGGA-301 datasets (Fig 1B, 1C). Likewise, combined analysis of TCGA and GTEx datasets also showed that CCNB1 mRNA expression was elevated in GBM but not in LGG compared to normal tissues (Fig 1D). Moreover, CCNB1 mRNA expression in GBM was higher than that in LGG in TCGA (p < 0.0001, Fig 1E). According to the UALCAN database, CCNB1 protein levels were also higher in GBM tissues compared to normal tissues (p = 1.06E-04, Fig 1F). Similarly, immunohistochemistry analysis revealed no detectable CCNB1 protein in normal tissues, while CCNB1 protein was observed in GBM tissues (Fig 1G).
(A-D) The expression level of CCNB1 mRNA in normal brain tissues, LGG and GBM tissues in CGGA-693 (A), CGGA-325 (B), CGGA-301 (C), and TCGA + GTEx (D) database. *p < 0.05. (E) The expression level of CCNB1 mRNA in LGG and GBM tissues in TCGA database. (F) CCNB1 protein level in UALCAN database. (G) Immunohistochemistry analysis of CCNB1 protein in normal brain tissues and GBM tissues obtained from HPA database. Scale bar = 200μm. (H-I) The relationship between CCNB1 mRNA expression and OS in GBM patients in CGGA (H), and TCGA (I) clinical cohorts. (J-K) ROC curve analysis of CCNB1 expression for 1-, 3-, and 5-year OS in CGGA (J), and TCGA (K) GBM clinical cohorts.
We further explored the relationship between CCNB1 mRNA expression and prognosis in GBM patients. Patients were divided into high and low expression groups based on the median CCNB1 mRNA expression levels. Results indicated that high CCNB1 expression was associated with poorer prognosis in GBM patients in the CGGA-693 (p = 0.0302), CGGA-325 (p = 0.0105), and CGGA-301 (p = 0.0544) datasets (Fig 1H). However, in the TCGA dataset, CCNB1 expression did not correlate with prognosis in GBM patients (p = 0.5085, Fig 1I). We further assessed the predictive power of CCNB1 expression for 1-, 3-, and 5-year OS in the CGGA-693, CGGA-325, and CGGA-301 cohorts (Fig 1J). Only the AUC values for 5-year survival were greater than 0.7. In contrast, the AUC values for 1- and 3-year survival were below 0.6 in the TCGA cohort (Fig 1K). These findings suggest that CCNB1 mRNA and protein are both highly expressed in GBM, and high CCNB1 mRNA expression is associated with poor prognosis in GBM patients.
Investigation of CCNB1 in single-cell sequencing data
To further investigate the potential mechanisms by which CCNB1 influences GBM and its corresponding cell subpopulations, we analyzed scRNA-seq data obtained from the CGGA database. After rigorous quality control and data filtering, 6,148 cells and 24,990 genes were retained for downstream analysis. Using the FindClusters function, cells were grouped into 15 distinct clusters and manually annotated into 12 cell types based on common marker genes (Fig 2A, 2B). Identified cell types include oligodendrocytes, regulatory T cells, microglia cells, T cells, macrophages, astrocytes, neurons, and oligodendrocyte precursor cells (Fig 2A). Additionally, based on the marker genes MKI67 and CDK1, we identified a subcluster of cells that proliferated actively, which we labeled as proliferating cells (Fig 2A, 2C). The results showed that cell cycle-related CCNB1, PLK1, CCNA2, CDC20 and CKS1B marked the same group of cells, indicating that CCNB1+ cells are tumor cells in the proliferative phase (Fig 2C, S2A Fig). We further performed differential gene expression analysis between proliferative cells and other cell subclusters. KEGG enrichment analysis indicated that highly expressed genes were not only associated with cell cycle and cellular senescence but also with pathways related to carbon metabolism, RNA degradation, and longevity regulation (Fig 2D). GO enrichment analysis showed that these genes were involved in cell cycle processes such as chromosome segregation, RNA splicing, proteasome-mediated ubiquitin-dependent protein catabolic process (Fig 2E).
(A) T-SNE plot displaying the annotated 12 cell types in the scRNA-seq data from GBM tissue in CGGA database. Clusters are color-coded and labeled to highlight cellular heterogeneity. (B) Dot plot showing the expression patterns of selected marker genes across different cell types. Dot size indicates the percentage of cells expressing each gene, while color intensity represents the average expression level. (C) T-SNE of MKI67, CDK1, CCNB1 and PLK1 expression by proliferative cells. The KEGG (D) and GO (E) enrichment analysis of upregulated genes in proliferative cells subset.
We further validated our findings using the scRNA-seq dataset GSE214966, which yielded results consistent with those obtained from the CGGA scRNA-seq database. Through clustering and GO enrichment analysis, we similarly identified proliferative cells (S2B-S2C Fig). Notably, the marker genes MKI67 and CDK1 showed co-localized expression with CCNB1, PLK1, CCNA2, CDC20 and CKS1B (S2D Fig). Furthermore, using Neftel’s gene set [34], we further subdivided the proliferative tumor cell population into two subgroups: proliferative cells in the G1/S phase and those in the G2/M phase (S3A-S3B Fig). To elucidate the specific cell cycle stage in which CCNB1 is involved, we performed a correlation analysis between CCNB1 expression levels and the gene set scores for the G1/S and G2/M phases proliferative tumor cells. The analysis revealed that CCNB1 exhibited a significantly higher correlation coefficient with G2/M phase cells than G1/S phase cells, suggesting that CCNB1 primarily exerts its function during the G2/M phase in GBM (S3C-S3D Fig). These results support that CCNB1+ cells constitute a proliferative tumor cell subpopulation, and CCNB1 may be involved in not only cell cycle processes but also in pathways related to metabolism, and protein ubiquitination in GBM cells.
Investigation of CCNB1 in spatial transcriptomics
To characterize the spatial landscape of CCNB1+ cells, we conducted spatial transcriptomics on a public dataset acquired from GBM patients [29]. We obtained the 269-T sample containing both tumor and adjacent non-tumor tissues, as well as tumor tissues from the 243-T, 265-T, and 275-T samples (Fig 3A). These 4 spatial transcriptomic datasets were each derived from 4 independent GBM patients. In the spatial transcriptomic analysis, established cell cycle-associated molecular markers including MKI67, CDK1, CCNB1, PLK1, CDC20, CCNA2, and CKS1B were employed to identify and spatially localize proliferating tumor cells. Analysis using the SPATA2 algorithm revealed that MKI67+ and CDK1+ proliferative cells were predominantly distributed in the tumor region of the 269-T sample, while being nearly absent in the adjacent non-tumor tissue (Fig 3B). Similarly, CCNB1+ and PLK1+ cells were primarily located within the tumor tissue and scarcely found in the adjacent non-tumor region (Fig 3C), which aligned with the results from bulk RNA-seq analysis (Fig 1A-1D). Furthermore, spatial transcriptomic analysis of tumor tissues showed co-localization of CCNA2+, CDC20+, CKS1B+, CCNB1+ cells with MKI67+ and CDK1+ proliferative cells, concentrated in specific regions (Fig 3A-3C, S3E Fig), indicating the specificity of spatial distribution of this proliferating cell subpopulation in GBM tissues. This finding was consistent with the results from scRNA-seq analysis (Fig 2C), suggesting that this subpopulation may play a critical role in GBM progression. Further pathway analysis revealed that this cell subpopulation was characterized by chromosome segregation, RNA splicing, regulation of cell cycle phase transition, DNA replication, and mitotic nuclear division (Fig 3D).
(A) H&E staining of four tumor samples (269-T, 243-T, 265-T, 275-T) with zoomed-in regions. (B-C) Spatial expression patterns of key cell cycle-related genes (MKI67, CDK1, CCNB1, PLK1) across the tumor samples. Deeper blue color indicates higher expression of the gene at the corresponding spatial location. The differences in color intensity across different regions present the difference in gene expression levels among various regions in the spatial transcriptome. (D) Spatial functional spatial enrichment maps showing biological processes, including chromosome segregation, RNA splicing, regulation of cell cycle phase transition, DNA replication, and mitotic nuclear division. Color intensity indicates relative expression or activity levels, with darker shades representing higher expression/activity.
Interaction gene set analysis and the establishment of prognosis model
Using the STRING database, we explored the CCNB1 interacting genes, including CDK1, ESPL1, CDK2, CDC20, CKS2, PLK1, CKS1B, CDC27, ANAPC10, and ANAPC4 (Fig 4A). In both the TCGA and CGGA datasets, CCNB1 showed a positive correlation with these interacting genes (S4, S5 Figs). Based on the CGGA cohort, univariate analysis with Cox regression of OS was performed to investigate the relationship between CCNB1 interacting genes and GBM patient prognosis. The results revealed that high expression of ESPL1 (p = 0.032, hazard ratio = 1.284), CDC20 (p = 0.008, hazard ratio = 1.441), CKS2 (p = 0.012, hazard ratio = 1.744), PLK1 (p = 0.021, hazard ratio = 1.466), ANAPC4 (p = 0.011, hazard ratio = 2.206), and CCNB1 (p = 0.029, hazard ratio = 1.486) was associated with poor prognosis in GBM patients (Fig 4B). We constructed a prognostic risk score model using the LASSO regression model: RiskScore = (−0.1498 × ESPL1) + (0.3847 × CDC20) + (0.2241 × CKS2) + (0.0842 × PLK1) + (0.6745 × ANAPC4) + (−0.2054 × CCNB1). To assess the accuracy of the prognostic model, patients were divided into two groups based on the median riskScore. The Kaplan-Meier curve showed that GBM patients in the high riskScore group had a worse prognosis (p = 0.0062, Fig 4C). The risk curve and scatter plot indicated that the risk of death increased with the riskScore (Fig 4D, 3E). Heatmaps showed that all six genes were high-risk genes associated with poor prognosis in GBM patients (Fig 4F). Univariate analysis with cox-regression of OS was used to further evaluate the independent predictive ability of the riskScore for OS. The results indicated that radiotherapy (p = 0.038, Hazard ratio = 0.716), chemotherapy (p < 0.001, Hazard ratio = 0.463), and IDH mutation (p = 0.027, Hazard ratio = 0.711) were protective factors significantly reducing the risk of death in GBM patients, while a high riskScore (p < 0.001, Hazard ratio = 2.718) significantly increased the risk of death (Fig 4G). To visually predict the 1-, 3-, and 5-year OS probabilities of GBM patients, we developed a nomogram based on several clinical factors, including PRS type, gender, age, radiotherapy, chemotherapy, IDH mutation, 1p/19q codeletion, MGMT methylation, and the riskScore (Fig 4H). ROC curves and AUC suggested that the predictive accuracy of riskScore (0.691) was much higher than that of gender (0.516), grade (0.555), radiotherapy (0.512), chemotherapy (0.421), IDH mutation (0.423), 1p/19q codeletion (0.507) and MGMT methylation (0.468). From the overall prediction, riskScore + clinical (0.774) performed the best (S6 Fig).
(A) PPI network of CCNB1 and its interacting genes (CDK1, ESPL1, CDK2, CDC20, CKS2, PLK1, CKS1B, CDC27, ANAPC10, ANAPC4). (B) Forest plot of univariate Cox regression analysis of CCNB1 interacting genes in GBM patients. (C) RiskScore is obtained based on LASSO regression model (RiskScore = (−0.1498 × ESPL1) + (0.3847 × CDC20) + (0.2241 × CKS2) + (0.0842 × PLK1) + (0.6745 × ANAPC4) + (−0.2054 × CCNB1). The Kaplan-Meier curve showed that GBM patients in the high riskScore group had a worse prognosis (cutoff = the median riskScore). (D-E) The risk curve(D), and scatter plot (E) showed the relationship between risk of death and the riskScore. (F) Heat maps showed the relationship between CCNB1 and its interacting genes expression and prognosis in GBM patients. (G) Forest plot showing the hazard ratios and p-values for various clinical factors and molecular markers. Factors include gender, age, radiotherapy (Radio), chemotherapy (Chemo), IDH mutation, 1p/19q co-deletion, MGMT methylation, and risk score. (H) Nomogram predicting 1-, 3-, and 5-year survival probabilities, based on risk score, PRS type, gender, age, Radio, Chemo, IDH mutation, 1p/19q co-deletion, and MGMT methylation status. Each variable contributes points to calculate overall survival probability.
CCNB1 promotes GBM cell proliferation and cell cycle
We first detected the expression of CCNB1 in certain glioma cell lines and human astrocyte (HA) cells. Compared to HA cells, the expression level of CCNB1 mRNA was higher in glioma cells (S7A Fig). Additionally, among various glioma cells, the expression level of CCNB1 was higher in U251 and U87 cells (S7A-S7B Fig). Therefore, U251 and U87 cells were selected for the further research. Three siRNAs targeting CCNB1 were used to specifically knock down CCNB1. The efficiency and specificity of CCNB1 knockdown were verified by WB assay (Fig 5A). Only siRNA1 and siRNA2 specifically knocked down the CCNB1 expression, and these were used for subsequent research. EdU and CCK8 assays confirmed that CCNB1 silencing inhibited the proliferation of U251 and U87 cells (Fig 5B-5D) (p < 0.0001). Cell cycle assay confirmed that CCNB1 knockdown led to a significant accumulation of cells in S phase with a concomitant reduction in the G2/M and G1 population of U251 and U87 cells (Fig 5E-5F). Moreover, CCNB1 knockdown decreased the expression of its interacting partners PLK1 and CDC20 in both U251 and U87 cells (S7C-S7D Fig) (p < 0.001), which may lead to the inhibition of the proliferation and cell cycle in GBM cells. Collectively, these results verified that CCNB1 promotes GBM cell proliferation and cell cycle progression.
(A) WB assay was used to evaluate the efficiency and specificity of three siRNAs targeting the CCNB1. (B) EdU assay was used to assess the changes in DNA synthesis capacity of U251 and U87 cells following CCNB1 knockdown. (C) The quantification of the EDU positive rate (n = 3). (D) CCK8 assay detected the proliferation of GBM cells after CCNB1 knockdown (n = 3). Flow cytometric analysis of cell cycle distribution in U251 (E) and U87 (F) cells after CCNB1 knockdown. Left: representative DNA content histograms. Right: quantification of the percentages of cells in G1, S, and G2/M phases (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Resveratrol inhibits the growth of GBM cells and the expression of CCNB1
It is reported that resveratrol has anti-cancer properties in initiation and progression of various cancers by regulating the cell cycle [35,36]. Its ability to penetrate the blood-brain barrier and its minimal toxicity to normal brain cells has garnered significant interest in the field of central nervous system diseases. CCNB1 and its interacting genes are key factors in cell cycle regulation, which play an important role in the proliferation of gliomas [37]. Thus, we further investigated whether resveratrol inhibited the expression of CCNB1 and its interacting genes in GBM cells.
First, the CCK8 assay confirmed that resveratrol significantly inhibited the proliferation of U251 cells (p = 0.0416) and U87 (p = 0.0143) (Fig 6A). the EdU assay also confirmed that resveratrol suppressed the proliferation of U251 and U87 cells (Fig 6B-6C). To further elucidate the mechanism by which resveratrol inhibited GBM cell proliferation, RNA-seq analysis was performed on resveratrol-treated and untreated U251 and U87 cells. A total of 4,253 and 3,307 DEGs (|fold change| > 2, p < 0.05) were identified in resveratrol-treated U251 and U87 cells, respectively (Fig 6D, S7E Fig). Among these, 2,231 genes were upregulated and 2,022 were downregulated in resveratrol-treated U251 cells, while 1,648 genes were upregulated and 1,659 were downregulated in resveratrol-treated U87 cells. Notably, resveratrol treatment reduced the expression of CCNB1 in U251 and U87 cells (Fig 6D, S7E Fig). GO enrichment analysis of the downregulated genes in resveratrol-treated U251 and U87 cells showed that these genes were primarily enriched in cell cycle-related biological processes, such as spindle organization and mitotic sister chromatid separation (Fig 6E, S7F Fig). Moreover, RT-qPCR confirmed that resveratrol could reduce the expression level of CCNB1 and its interacting gene PLK1 (Fig 6F). Finally, the CCK8 assay confirmed that CCNB1 knockdown combined with resveratrol treatment could further inhibit the proliferation of U251 (p = 0.0013) and U87 (p = 0.0214) cells (Fig 6G).
(A) CCK8 assay detected the proliferation of U251 and U87 cells after resveratrol treatment (n = 3). (B) EdU assay was used to assess the changes in DNA synthesis capacity of glioma cells after resveratrol treatment. (C) The quantification of the EDU positive rate (n = 3). (D) Volcano plot showed the DEGs in resveratrol and PBS treated U251 cells. (E) The GO enrichment analysis of downregulated genes in U251 cells treated with resveratrol. (F) RT-qPCR detected the expression of CCNB1 and its interacting genes after resveratrol treatment in U251 and U87 cells (n = 3). (G) CCK-8 assay showing cell proliferation of U251 and U87 cells under siNC, siCCNB1, Res, and siCCNB1 combined with resveratrol treatments over time (n = 3). (H) Representative bioluminescent images (at 10, 17, 24, and 31 days) of the NPG mice were captured after U87 cells were intracranially injected into NPG mice. CON = control, RES = resveratrol. Ten-day post-tumor injection, resveratrol (50 mg/kg, 1 time/day) was administered orally for 28 days. (I) Upper panel: bioluminescent intensity of glioma-bearing mice on day 31 (n = 5), Lower panel: Kaplan‒Meier survival curves of glioma-bearing mice (resveratrol treatment group or PBS treatment group, n = 5). RES: resveratrol. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Excitingly, in vivo experiments demonstrated that resveratrol suppressed intracranial tumor growth (p = 0.0360) and prolonged the survival of tumor-bearing mice (p = 0.0018) (Fig 5H-5I). H&E staining analysis revealed that resveratrol treatment did not induce toxic effects on major organs, such as the heart, liver, spleen, lungs, and kidneys, compared to the control group (S7G Fig). There was no statistically significant difference in body weight between the resveratrol-treated group and the control group (S7H Fig).
Collectively, these results indicate that resveratrol inhibits GBM cell growth and reduces the expression of CCNB1 and PLK1. CCNB1 silencing combined with resveratrol treatment could further inhibit GBM cell proliferation.
Discussion
The most prevalent primary malignant tumor of the central nervous system is GBM, which exhibits high mortality and recurrence rates [3,38]. Despite advancements in treatment, the prognosis for patients with GBM remains unsatisfactory, with a median survival rate of only 14.6 months [39]. The dysregulation of the cell cycle is a hallmark of GBM and plays a crucial role in promoting tumor proliferation and resistance to treatment [40]. GBM cells frequently exhibit alterations in key cell cycle regulators, such as cyclins, CDKs, and checkpoint proteins, which drive uncontrolled cell division and tumor growth [41,42]. Targeting the cell cycle in GBM has emerged as a potential therapeutic strategy, with several studies exploring inhibitors of CDKs and other cell cycle-related proteins as treatments for GBM [43,44]. However, it remains unclear whether inhibiting CCNB1 expression can be used as a therapeutic strategy for GBM.
In this study, we found that both CCNB1 mRNA and protein levels were elevated in GBM tissues compared to normal tissues. High CCNB1 mRNA expression was associated with poor prognosis of GBM patients in CGGA datasets. However, in the TCGA cohort, although CCNB1 mRNA expression was elevated in GBM compared to normal and LGG tissues, it was not correlated with patient prognosis. Furthermore, analysis through the GlioVis database (https://gliovis.bioinfo.cnio.es/#tab-6146-4) showed that in certain datasets (such as Rembrandt and Gravendeel microarray data), CCNB1 expression was not associated with the prognosis of GBM patients (data not shown). The discrepancy in CCNB1’s prognostic value between TCGA and CGGA cohorts may be attributable to differences in patient populations.
Notably, single-cell and spatial transcriptomic data revealed a subcluster of CCNB1+ proliferative cells. This cell subcluster exhibited active pathways including the cell cycle, chromosome segregation, RNA splicing, and DNA replication, indicating that these cells were in a highly proliferative state. Particularly in the spatial transcriptomics data of the 269-T sample, we observed that CCNB1 and its associated cell cycle pathways were predominantly enriched in tumor cell area but not in adjacent non-tumor tissue. These findings highlight that the cell cycle signaling pathway is a significant hallmark of GBM, with CCNB1 playing a key role in disrupting the cell cycle processes and promoting GBM progression. Liu et al. revealed the downregulation of cell cycle pathways in recurrent GBM [45]. These findings suggest that disturbance in the cell cycle may play a critical role in the evolution of GBM (initiation to recrudesce). This highlights the necessity for further investigation into the molecular mechanisms underlying the relationship between cell cycle dysregulation and GBM tumor progression.
In the CGGA dataset, ROC curve analysis showed that the predictive value of CCNB1 mRNA expression for OS increased over time, suggesting its potential as a long-term survival predictor. However, in the TCGA dataset, the predictive value of CCNB1 mRNA expression for OS remained below 0.6, indicating that differences in CCNB1 expression among different ethnic groups should be considered when analyzing its prognostic significance. The cell cycle is regulated by multiple factors, including CCNB1, CDKs, and checkpoint proteins [40]. We attempted to construct a prognostic model using CCNB1 interacting genes. The prognostic model indicated that a high riskScore correlated with poor prognosis in GBM patients, and the riskScore was an independent prognostic factor for predicting OS in GBM patients. The cell cycle plays a critical role in cancer development and progression [40]. Numerous studies have demonstrated that tumor prognostic models based on cell cycle-related genes perform well in assessing patient OS [46,47]. Therefore, our prognostic model constructed with CCNB1, and its interacting genes offers a novel strategy for clinically evaluating OS in GBM patients.
Given the importance of CCNB1 and its interacting genes in regulating the cell cycle in GBM, we aimed to inhibit this biological process to suppress GBM progression. Resveratrol has demonstrated anticancer effects in various cancers by modulating the cell cycle [20]. However, the molecular mechanisms through which resveratrol inhibits GBM proliferation remain underexplored. Limited studies suggest that resveratrol has good blood-brain barrier penetration, low toxicity, and can reduce GBM cell proliferation and enhance sensitivity to TMZ by increasing G2/M phase arrest [22,23,48]. In our study, both in vivo and in vitro experiments have shown that resveratrol can inhibit GBM cell proliferation with low toxicity. In GBM cells, resveratrol can suppress the expression of CCNB1 and its interacting gene PLK1. CCNB1 knockdown also inhibited PLK1 expression. PLK1 is activated in the G2 phase and drives cells into mitosis [49]. CCNB1 activates CDK1, which mediates the initiation of mitosis (G2/M transition) [50]. The compound resveratrol exhibited inhibitory effects on GBM cell proliferation and acted as a CCNB1 and cell cycle inhibitor. Additionally, Ahmad et.al reported that resveratrol treatment led to a dose- and time-dependent decrease in CDK1/CDK2 kinase activity [20]. Jeong et.al reported that resveratrol could decrease the expression of CDK1 expression [51]. Therefore, by means of a CDK1 inhibitor or in combination with CDK1/2 inhibitors, resveratrol may represent a new strategy for GBM therapy.
However, this study has some limitations. First, we identified a subpopulation of CCNB1+, PLK1+, MKI67+and CDK1+ cells, referred to as proliferative cells. The role of these cells in the GBM tumor microenvironment requires further elucidation. Second, while we revealed potential mechanisms by which resveratrol inhibits GBM cell proliferation at the transcriptional level, additional experimental validation is necessary. Finally, although our findings found that resveratrol could inhibit the expression of CCNB1 and PLK1, it is still unclear whether the inhibition of GBM cells proliferation by resveratrol is achieved by directly suppressing CCNB1 expression or through other mechanisms. Whether resveratrol inhibits the expression of CCNB1 and PLK1 through direct transcriptional regulation or via other mechanisms requires further investigation.
Conclusion
In summary, the expression of CCNB1 is higher in GBM tissues compared to normal brain tissues and LGG tissues. CCNB1 is associated with poor prognosis in GBM patients. CCNB1 can promote the proliferation and cell cycle progression of GBM cells. Resveratrol could inhibit the expression of CCNB1 and PLK1 and suppress GBM cell growth. CCNB1 silencing combined with resveratrol treatment could further inhibit the proliferation of GBM cells. Arresting cell cycle procession by inhibiting CCNB1 expression provides a novel therapeutic strategy for the treatment of GBM patients.
Supporting information
S1 Fig. CCNB1 expression in normal organs and various tumor tissues.
(A) CCNB1 mRNA expression of normal organs in HPA database. (B) CCNB1 protein level of normal organs in HPA database. (C) CCNB1 mRNA expression across tumor tissues in TCGA database. A log2(x + 1) transformation was applied to each expression value to determine the relative levels of CCNB1 expression. CESC: Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma, HNSC: Head and Neck Squamous Cell Carcinoma, COAD: Colorectal Adenocarcinoma, READ: Rectal Adenocarcinoma, UCEC: Uterine Corpus Endometrial Carcinoma, LUSC: Lung Squamous Cell Carcinoma, BLCA: Bladder Urothelial Carcinoma, ESCA: Esophageal Carcinoma, GBM: Glioblastoma, BRCA: Breast Cancer, STES, Stomach and Esophageal carcinoma, STAD: Stomach Adenocarcinoma, LUAD: Lung Adenocarcinoma, CHOL: Cholangiocarcinoma, PAAD: Pancreatic Adenocarcinoma, LIHC: Liver Hepatocellular Carcinoma, THCA: Thyroid Cancer, KIRC, Kidney Renal Clear Cell Carcinoma, PRAD: Prostate Adenocarcinoma, LGG: Low Grade Glioma, KIPAN, Pan-kidney cohort (KICH+KIRC+KIRP), KIRP: Kidney Renal Papillary Cell Carcinoma, PCPG: Pheochromocytoma and Paraganglioma.
https://doi.org/10.1371/journal.pone.0344872.s001
(TIF)
S2 Fig. scRNA-seq data analysis of cell cycle regulator.
(A) T-SNE of CCNA2, CDC20 and CKS1B expression by proliferative cells in CGGA database. (B) T-SNE plot displaying the annotated 8 cell types in the scRNA-seq data from GSE214966. (C) GO enrichment analysis of upregulated genes in proliferative cells subset. (D) T-SNE of MKI67, CDK1, CCNB1, PLK1, CCNA2, CDC20 and CKS1B expression by proliferative cells in GSE214966.
https://doi.org/10.1371/journal.pone.0344872.s002
(TIF)
S3 Fig. scRNA-seq and spatial transcriptomics data analysis of cell cycle regulator.
T-SNE of G1/S gene set score and G2/M gene set score level in CGGA database (A) and GSE214966 (B). Scatter plots depict the correlation between CCNB1 expression levels and G1/S gene set score (left) and G2/M gene set score (right) in CGGA database (C) and GSE214966 (D). (E) Spatial expression patterns of cell cycle-related genes (CCNA2, CDC20 and CKS1B) across the tumor samples. Deeper blue color indicates higher expression of the gene at the corresponding spatial location. The differences in color intensity across different regions present the difference in gene expression levels among various regions in the spatial transcriptome.
https://doi.org/10.1371/journal.pone.0344872.s003
(TIF)
S4 Fig. The relationship between CCNB1 and its interacting genes.
The relationship between CCNB1 and its interacting genes were analyzed by using CGGA dataset download from GlioVis database(http://gliovis.bioinfo.cnio.es/). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
https://doi.org/10.1371/journal.pone.0344872.s004
(TIF)
S5 Fig. The relationship between CCNB1 and its interacting genes.
The relationship between CCNB1 and its interacting genes were analyzed by using TCGA dataset download from GlioVis database. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
https://doi.org/10.1371/journal.pone.0344872.s005
(TIF)
S6 Fig. ROC curves for various factors predicting survival.
The AUC values for riskScore, gender, age, Radio, Chemo, IDH mutation, 1p/19q co-deletion, MGMT methylation, and a combined model (riskScore + Clinical, AUC = 0.774) was shown. Higher AUC values indicated better predictive accuracy.
https://doi.org/10.1371/journal.pone.0344872.s006
(TIF)
S7 Fig. CCNB1 expression in GBM cells and the effects of resveratrol on NPG mice.
(A) The expression levels of CCNB1 in various glioma cell lines were assessed using RT-qPCR. (B) WB analysis was used to detect the protein levels of CCNB1 in various glioma cell lines. Right: Relative quantification for the WB analysis. RT-qPCR detected the expression of CCNB1 and its interacting genes after resveratrol treatment in U251 (C) and U87 (D) cells (n = 3). (E) Volcano plot showed the DEGs in resveratrol and PBS treated U87 cells. (F) The GO enrichment analysis of downregulated genes in U87 cells treated with resveratrol. (G) H&E staining of heart, liver, spleen, lungs, and kidneys in resveratrol treated group and control group. Con = control, Res = resveratrol, Scale bar = 100μm. (H) Body weight changes in mice treated with resveratrol or PBS solution. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
https://doi.org/10.1371/journal.pone.0344872.s007
(TIF)
References
- 1. Schaff LR, Mellinghoff IK. Glioblastoma and other primary brain malignancies in adults: A review. JAMA. 2023;329(7):574–87.
- 2. Lapointe S, Perry A, Butowski NA. Primary brain tumours in adults. Lancet. 2018;392(10145):432–46. pmid:30060998
- 3. Tan AC, Ashley DM, López GY, Malinzak M, Friedman HS, Khasraw M. Management of glioblastoma: State of the art and future directions. CA Cancer J Clin. 2020;70(4):299–312. pmid:32478924
- 4. Xue C, Wu X, Zhang X, Liu Z. The role of magnetic resonance imaging-guide surgery in cytokines of immune cells and effect in glioma patients. Curr Probl Surg. 2025;70:101819. pmid:40812987
- 5. Hanahan D. Hallmarks of Cancer: New Dimensions. Cancer Discov. 2022;12(1):31–46. pmid:35022204
- 6. Malumbres M, Barbacid M. Mammalian cyclin-dependent kinases. Trends Biochem Sci. 2005;30(11):630–41. pmid:16236519
- 7. Janss AJ, Maity A, Tang CB, Muschel RJ, McKenna WG, Sutton L, et al. Decreased cyclin B1 expression contributes to G2 delay in human brain tumor cells after treatment with camptothecin. Neuro Oncol. 2001;3(1):11–21. pmid:11305412
- 8. Cortese K, Daga A, Monticone M, Tavella S, Stefanelli A, Aiello C, et al. Carnosic acid induces proteasomal degradation of Cyclin B1, RB and SOX2 along with cell growth arrest and apoptosis in GBM cells. Phytomedicine. 2016;23(7):679–85. pmid:27235706
- 9. Li B, Cheng J, Wang H, Zhao S, Zhu H, Li C, et al. CCNB1 affects cavernous sinus invasion in pituitary adenomas through the epithelial-mesenchymal transition. J Transl Med. 2019;17(1):336. pmid:31585531
- 10. Kim S-K, Roh Y-G, Park K, Kang T-H, Kim W-J, Lee J-S, et al. Expression signature defined by FOXM1-CCNB1 activation predicts disease recurrence in non-muscle-invasive bladder cancer. Clin Cancer Res. 2014;20(12):3233–43. pmid:24714775
- 11. Ersvær E, Kildal W, Vlatkovic L, Cyll K, Pradhan M, Kleppe A, et al. Prognostic value of mitotic checkpoint protein BUB3, cyclin B1, and pituitary tumor-transforming 1 expression in prostate cancer. Modern pathology: an official journal of the United States and Canadian Academy of Pathology, Inc. 2020;33(5):905–15.
- 12. Liu R, Fan M, Candas D, Qin L, Zhang X, Eldridge A. CDK1-Mediated SIRT3 Activation Enhances Mitochondrial Function and Tumor Radioresistance. Molecular Cancer Therapeutics. 2015;14(9):2090–102.
- 13. Chen H, Huang Q, Dong J, Zhai D-Z, Wang A-D, Lan Q. Overexpression of CDC2/CyclinB1 in gliomas, and CDC2 depletion inhibits proliferation of human glioma cells in vitro and in vivo. BMC Cancer. 2008;8:29. pmid:18230152
- 14. Chen Q, Ouyang L, Liu Q. Cyclin B1: A potential prognostic and immunological biomarker in pan-cancer. Biomol Biomed. 2024;24(5):1150–69. pmid:38581717
- 15. Mokgautsi N, Kuo YC, Tang SL, Liu FC, Chen SJ, Wu ATH. Anticancer activities of 9-chloro-6-(piperazin-1-yl)-11H-indeno[1,2-c] quinolin-11-one (SJ10) in glioblastoma multiforme (GBM) chemoradioresistant cell cycle-related oncogenic signatures. Cancers. 2022;14(1).
- 16. Rauf A, Imran M, Butt MS, Nadeem M, Peters DG, Mubarak MS. Resveratrol as an anti-cancer agent: A review. Critical Reviews in Food Science and Nutrition. 2018;58(9):1428–47.
- 17. Ren B, Kwah MX-Y, Liu C, Ma Z, Shanmugam MK, Ding L, et al. Resveratrol for cancer therapy: Challenges and future perspectives. Cancer Lett. 2021;515:63–72. pmid:34052324
- 18. Malaguarnera L. Influence of Resveratrol on the Immune Response. Nutrients. 2019;11(5):946. pmid:31035454
- 19. Singh AP, Singh R, Verma SS, Rai V, Kaschula CH, Maiti P, et al. Health benefits of resveratrol: Evidence from clinical studies. Med Res Rev. 2019;39(5):1851–91. pmid:30741437
- 20. Ahmad N, Adhami VM, Afaq F, Feyes DK, Mukhtar H. Resveratrol causes WAF-1/p21-mediated G(1)-phase arrest of cell cycle and induction of apoptosis in human epidermoid carcinoma A431 cells. Clinical cancer research: an official journal of the American Association for Cancer Research. 2001;7(5):1466–73.
- 21. Baek SH, Ko J-H, Lee H, Jung J, Kong M, Lee J, et al. Resveratrol inhibits STAT3 signaling pathway through the induction of SOCS-1: Role in apoptosis induction and radiosensitization in head and neck tumor cells. Phytomedicine. 2016;23(5):566–77. pmid:27064016
- 22. Wu M, Song D, Li H, Ahmad N, Xu H, Yang X, et al. Resveratrol Enhances Temozolomide Efficacy in Glioblastoma Cells through Downregulated MGMT and Negative Regulators-Related STAT3 Inactivation. Int J Mol Sci. 2023;24(11):9453. pmid:37298405
- 23. Jhaveri A, Deshpande P, Pattni B, Torchilin V. Transferrin-targeted, resveratrol-loaded liposomes for the treatment of glioblastoma. J Control Release. 2018;277:89–101. pmid:29522834
- 24. Zhao Z, Zhang K-N, Wang Q, Li G, Zeng F, Zhang Y, et al. Chinese Glioma Genome Atlas (CGGA): A Comprehensive Resource with Functional Genomic Data from Chinese Glioma Patients. Genomics Proteomics Bioinformatics. 2021;19(1):1–12. pmid:33662628
- 25. Tang Z, Kang B, Li C, Chen T, Zhang Z. GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res. 2019;47(W1):W556–60. pmid:31114875
- 26. Chandrashekar DS, Karthikeyan SK, Korla PK, Patel H, Shovon AR, Athar M, et al. UALCAN: An update to the integrated cancer data analysis platform. Neoplasia. 2022;25:18–27. pmid:35078134
- 27. Yu K, Hu Y, Wu F, Guo Q, Qian Z, Hu W, et al. Surveying brain tumor heterogeneity by single-cell RNA-sequencing of multi-sector biopsies. Natl Sci Rev. 2020;7(8):1306–18. pmid:34692159
- 28. Lee S, Weiss T, Bühler M, Mena J, Lottenbach Z, Wegmann R, et al. High-throughput identification of repurposable neuroactive drugs with potent anti-glioblastoma activity. Nat Med. 2024;30(11):3196–208. pmid:39304781
- 29. Ravi VM, Will P, Kueckelhaus J, Sun N, Joseph K, Salié H, et al. Spatially resolved multi-omics deciphers bidirectional tumor-host interdependence in glioblastoma. Cancer Cell. 2022;40(6):639-655.e13. pmid:35700707
- 30. Cilibrasi C, Riva G, Romano G, Cadamuro M, Bazzoni R, Butta V, et al. Resveratrol Impairs Glioma Stem Cells Proliferation and Motility by Modulating the Wnt Signaling Pathway. PLoS One. 2017;12(1):e0169854. pmid:28081224
- 31. Wang G, Dai F, Yu K, Jia Z, Zhang A, Huang Q, et al. Resveratrol inhibits glioma cell growth via targeting oncogenic microRNAs and multiple signaling pathways. Int J Oncol. 2015;46(4):1739–47. pmid:25646654
- 32. Chen Q, Xie L, Wang J, Su X, Ye X, Lin X. Resveratrol inhibits lipopolysaccharide‑induced MUC5AC expression and airway inflammation via MAPK and Nrf2 pathways in human bronchial epithelial cells and an acute inflammatory mouse model. Mol Med Rep. 2025;31(6):157. pmid:40211706
- 33. Lyu F, Han F, Ge C, Mao W, Chen L, Hu H, et al. OmicStudio: A composable bioinformatics cloud platform with real-time feedback that can generate high-quality graphs for publication. Imeta. 2023;2(1):e85. pmid:38868333
- 34. Neftel C, Laffy J, Filbin MG, Hara T, Shore ME, Rahme GJ, et al. An Integrative Model of Cellular States, Plasticity, and Genetics for Glioblastoma. Cell. 2019;178(4):835-849.e21. pmid:31327527
- 35. Kursvietiene L, Kopustinskiene DM, Staneviciene I, Mongirdiene A, Kubová K, Masteikova R, et al. Anti-Cancer Properties of Resveratrol: A Focus on Its Impact on Mitochondrial Functions. Antioxidants (Basel). 2023;12(12):2056. pmid:38136176
- 36. Song B, Wang W, Tang X, Goh RMW-J, Thuya WL, Ho PCL, et al. Inhibitory Potential of Resveratrol in Cancer Metastasis: From Biology to Therapy. Cancers (Basel). 2023;15(10):2758. pmid:37345095
- 37. Wu Y, Song Y, Wang R, Wang T. Molecular mechanisms of tumor resistance to radiotherapy. Mol Cancer. 2023;22(1):96. pmid:37322433
- 38. Qian X, Yang L, Shi Y. Investigation of the impact of magnetic resonance imaging-assisted surgery on immune cell cytokine levels and efficacy in patients with gliomas. Curr Probl Surg. 2025;63:101640. pmid:39922633
- 39. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJB, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352(10):987–96. pmid:15758009
- 40. Matthews HK, Bertoli C, de Bruin RAM. Cell cycle control in cancer. Nat Rev Mol Cell Biol. 2022;23(1):74–88. pmid:34508254
- 41. Furnari FB, Fenton T, Bachoo RM, Mukasa A, Stommel JM, Stegh A, et al. Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes Dev. 2007;21(21):2683–710. pmid:17974913
- 42. Le Rhun E, Preusser M, Roth P, Reardon DA, van den Bent M, Wen P, et al. Molecular targeted therapy of glioblastoma. Cancer Treat Rev. 2019;80:101896. pmid:31541850
- 43. Liu H, Qiu W, Sun T, Wang L, Du C, Hu Y, et al. Therapeutic strategies of glioblastoma (GBM): The current advances in the molecular targets and bioactive small molecule compounds. Acta Pharm Sin B. 2022;12(4):1781–804. pmid:35847506
- 44. Huang S, Xiao J, Wu J, Liu J, Feng X, Yang C, et al. Tizoxanide Promotes Apoptosis in Glioblastoma by Inhibiting CDK1 Activity. Front Pharmacol. 2022;13:895573. pmid:35694267
- 45. Liu J, Cao S, Imbach KJ, Gritsenko MA, Lih T-SM, Kyle JE, et al. Multi-scale signaling and tumor evolution in high-grade gliomas. Cancer Cell. 2024;42(7):1217-1238.e19. pmid:38981438
- 46. Xiang M, Gao Y, Zhou Y, Wang M, Yao X. A novel nomogram based on cell cycle-related genes for predicting overall survival in early-onset colorectal cancer. BMC Cancer. 2023;23(1):595. pmid:37370046
- 47. Zhao L, Jiang L, He L, Wei Q, Bi J, Wang Y, et al. Identification of a novel cell cycle-related gene signature predicting survival in patients with gastric cancer. J Cell Physiol. 2019;234(5):6350–60. pmid:30238991
- 48. Luís Â, Marcelino H, Domingues F, Pereira L, Cascalheira JF. Therapeutic Potential of Resveratrol for Glioma: A Systematic Review and Meta-Analysis of Animal Model Studies. Int J Mol Sci. 2023;24(23):16597. pmid:38068922
- 49. Raab CA, Raab M, Becker S, Strebhardt K. Non-mitotic functions of polo-like kinases in cancer cells. Biochim Biophys Acta Rev Cancer. 2021;1875(1):188467. pmid:33171265
- 50. Xie B, Wang S, Jiang N, Li JJ. Cyclin B1/CDK1-regulated mitochondrial bioenergetics in cell cycle progression and tumor resistance. Cancer Lett. 2019;443:56–66. pmid:30481564
- 51. Jeong JK, Kang MH, Gurunathan S, Cho SG, Park C, Park JK, et al. Pifithrin-α ameliorates resveratrol-induced two-cell block in mouse preimplantation embryos in vitro. Theriogenology. 2015;83(5):862–73.