Ferroptosis resistance-related TEP1 as a novel prognostic biomarker involved in immune cell infiltration and tumour progression in glioblastoma

Abstract

Background

TEP1 is a ribonucleoprotein complex component that binds to telomerase, and its role in GBM development is unclear. This study was designed to investigate the regulatory mechanism of TEP1 in GBM development.

Methods

TEP1 expression was verified using the TCGA and CGGA databases. Kaplan–Meier survival curve, univariate Cox regression, and multivariate Cox regression analyses were employed to assess the prognostic and predictive value of TEP1. KEGG and GO analyses were conducted to identify signalling pathways and functional gene sets enriched in TEP1-related genes. Immune infiltration was evaluated using the CIBERSORT and ssGSEA algorithms. Single-cell sequencing data were utilized to pinpoint TEP1-enriched cell populations. Drug susceptibility data from the CCLE, GDSC, and CTRP databases were integrated to explore potential therapeutic implications. Further validation was performed using immunofluorescence, quantitative real-time PCR, Western blotting, and in vitro and in vivo ferroptosis assays.

Results

TEP1 expression was significantly elevated in GBM tissues compared to normal brain tissues, correlating with copy number alterations and poor overall survival. Elevated TEP1 expression correlated with increased tumour malignancy, IDH wild-type status, and radiotherapy resistance. Functional analyses revealed significant associations between TEP1 expression and pathways involved in lipid metabolism, apoptosis, immune response, cell cycle regulation, and ferroptosis. Immune profiling highlighted enrichment of TEP1 in immune cells, particularly plasmacytoid dendritic cells and natural killer T cells. Elevated TEP1 expression was associated with altered sensitivity to various anticancer drugs, notably ferroptosis modulators. In vitro and in vivo experiments confirmed that TEP1 significantly inhibited ferroptosis induced by RSL3, thereby promoting GBM cell survival and tumour progression.

Conclusion

Overall, we found that TEP1 is a promising biomarker for GBM and reduces the susceptibility of GBM cells to ferroptosis, providing a novel target for GBM treatment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-025-04114-0.

Introduction

Glioblastoma (GBM) is one of the most lethal and treatment-resistant malignant tumors. GBM accounts for 14.2% of all tumors and 50.9% of all malignancies of the central nervous system (CNS) [1]. After standard treatment (surgical resection followed by chemoradiotherapy), the median progression-free survival is 6.9 months, the median overall survival (OS) is 14.6 months, and the 5-year survival rate is less than 5% [2]. Moreover, the available drugs for the clinical treatment of GBM are limited; temozolomide (TMZ) is the only first-line agent still in clinical use. However, prolonged use of TMZ may lead to reduced sensitivity or chemotherapy resistance [3]. Thus, a deeper understanding of the molecular mechanisms underlying GBM development and progression and the identification of new therapeutic targets are needed to improve GBM treatment strategies and outcomes.

Telomere-associated protein 1 (TEP1), located on chromosome 14 (14q11.2), encodes a protein consisting of 2627 amino acids. Poderycki et al. demonstrated that the p80 homology domain of TEP1 is essential for its interactions with telomerase RNA and vault RNA (vRNA), as well as for its localization to ribosomal particles [4]. While the loss of p80 does not affect telomerase activity in cell extracts, changes in telomere length indicate that TEP1 may play a role in telomerase–telomere interactions or in maintaining telomere structure [5]. Previous studies have revealed that elevated levels of TEP1 are associated with increased risks of bladder cancer [6] and breast cancer [7]. Additionally, TEP1 is widely expressed in neuronal cells, particularly in astrocytes. TEP1 plays a critical role in cellular responses to hypoxia and is significantly involved in regulating neuronal migration, which is closely associated with conditions such as neonatal asphyxia and cerebral palsy [8]. However, the specific functions of TEP1 in GBM remain unclear and need to be further investigated.

Ferroptosis is a recently discovered form of necrotic, nonapoptotic cell death that is regulated by iron-dependent lipid peroxidation and modulated by various cellular metabolic and disease-related signalling pathways [9, 10]. The accumulation of reactive oxygen species (ROS) within cells, regulated by the iron-dependent Fenton reaction and glutathione (GSH) depletion, is the direct cause of ferroptosis [9]. Decreased activity of glutathione peroxidase 4 (GPX4) and the accumulation of redox-active iron are key hallmarks of ferroptosis [11]. Another study revealed that telomere-protective proteins, such as telomeric repeat binding factor 2 (TRF2), maintain telomere stability by preventing telomeres from being recognized as double-strand breaks (DSBs) through the promotion of T-loop formation at telomere ends [12, 13]. In contrast, the inhibition of telomere-protective proteins induces telomere instability and dysfunction, leading to ferroptosis [14]. Additionally, in mTep1 knockout mice, while the ribosomal structure remains intact, stable binding of vRNA to ribosomal particles is disrupted, reducing vRNA levels and stability. This finding underscores the critical role of TEP1 in stabilizing and recruiting vRNA to ribosomal particles [15]. Ribosome stalling, collisions, and premature translation termination lead to a rapid decrease in GPX4 levels, thereby sensitizing cells to ferroptosis [16]. However, it is unclear whether TEP1 influences cellular activities via the regulation of ferroptosis.

In this study, we comprehensively analysed the expression levels of TEP1, the functional enrichment of related coexpressed genes, and the correlation of TEP1 expression with drug sensitivity in GBM using publicly available datasets. These findings demonstrate that TEP1 has the potential to serve as a biomarker significantly correlated with prognosis and a promising therapeutic target for GBM. Additionally, a series of in vitro and in vivo experiments were conducted and confirmed that TEP1 overexpression decreases GBM cell sensitivity to ferroptosis. In summary, our results reveal novel mechanisms of action of TEP1, thereby providing alternative therapeutic targets for GBM treatment.

Methods

Data acquisition and processing

Gene expression data, associated mutation data, and clinical data for GBM patients were retrieved from The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/) and from the Chinese Glioma Genome Atlas (CGGA) database via Gliovis (http://gliovis.bioinfo.cnio.es/). Additionally, the GSE4290 dataset was obtained from the Gene Expression Omnibus database (GEO, https://www.ncbi.nlm.nih.gov/geo/). Protein abundance and gene expression data for the same samples were sourced from the Clinical Proteomic Tumour Analysis Consortium (CPTAC, https://proteomics.cancer.gov/programs/cptac). The data were subsequently analysed to assess differential expression between normal and tumour samples using the appropriate R package for normalization, preprocessing, and log transformation.

Survival and Cox risk regression analyses

Following the collection of TEP1 gene expression data, along with patient clinical and prognostic information, Kaplan-Meier survival curves were further analyzed. Univariate and multivariate Cox regression analyses were performed to identify independent predictors of overall survival (OS). The variables investigated included pathological grade, age, gender, recurrence, IDH status, co-deletion of 1p19q, and TEP1 expression.

Identification and functional enrichment of differentially expressed genes

Differentially expressed genes (DEGs) between high and low TEP1 samples utilizing 388 GBM samples from the CGGA database or 170 GBM samples from the TCGA database, were identified using a |log2 fold change| >0.3 and an adjusted p-value < 0.05 as the threshold. The results were visualized using the “limma” R package [17]. Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed on DEGs using the “clusterProfiler” R package. Additionally, Gene set enrichment analysis (GSEA) was conducted on all gene expression data [18].

Tumor immune microenvironment analysis

In this study, based on the TCGA cohort of 170 GBM samples, the Immuno-Oncology Biological Research (IOBR) tool was used to perform immuno-tumor biology calculations [19]. The relative infiltrate abundance and immunological signatures of all GBM samples were assessed using Cibersort. Additionally, the ssGSEA algorithm was employed to calculate the infiltration levels of 28 types of immune cells and their correlation with TEP1 expression.

Single-cell RNA sequencing analysis

Single-cell RNA sequencing data from GSE182109 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE182109), which includes 22 primary GBM samples sequenced using Smart-seq3 [20], were analysed. Low-quality cells were filtered on the basis of criteria including nFeature-RNA (< 200 or >10,000), mitochondrial content (>15%), and a dual-cell ratio of more than 7.5%. The dataset was subsequently debatched, integrated using Harmony, and annotated for cell types on the basis of reference materials and the “SingleR” algorithm [21]. TEP1 expression was visualized using a weighted kernel density algorithm, and its expression across different cell types and T-cell subtypes was assessed using the Seurat v4.4.0 package. Finally, the pseudotemporal dynamics of TEP1 expression were analysed using Monocle2.

Drug sensitivity analysis

The “OncoPredict” R package was utilized to predict in vivo drug responses in cancer patients [22]. OncoPredict fits tissue gene expression profiles to half-maximal inhibitory concentration (IC50) or area-under-the-curve (AUC) values and corresponding gene expression data from cancer cell lines. The tissue gene expression data were obtained from GBM samples in the CGGA (n = 388) and TCGA (n = 170) databases. These IC50 or AUC values were sourced from the Genomics of Drug Sensitivity in Cancer (GDSC; https://www.cancerrxgene.org/), the Cancer Therapeutics Response Portal (CTRP; http://portals.broadinstitute.org/ctrp/), or the Cancer Cell Line Encyclopedia (CCLE; https://sites.broadinstitute.org/ccle/). This approach yields drug response estimates for patient tissues. Drug sensitivity was subsequently evaluated using an unpaired t test, and a P value < 0.05 was considered to indicate statistical significance.

Cell culture

A set of human GBM cell lines, including U87MG (RRID: CVCL_0022), T98G (RRID: CVCL_0556), LN229 (RRID: CVCL_0393), A172 (RRID: CVCL_0131), and U251 (RRID: CVCL_0021), was sourced from the American Type Culture Collection (ATCC). The human glial cell line HEB (RRID: CVCL_S549) was obtained from Sun Yat-Sen University (Guangdong, China). Prior to experimentation, these cell lines were authenticated by short tandem repeat (STR) profiling.

Before starting the assay, early-passage stocks were retrieved from cryostorage, thawed, and cultured for two to three passages. The cells were maintained in DMEM (4.5 g/L glucose, Gibco 11995065) supplemented with 10% FBS (Gibco, 16140071) and 1% penicillin/streptomycin (Gibco, 10378016). All the cultures were maintained at 37 °C in a 5% CO₂ humidified incubator. Furthermore, Mycoplasma screening was conducted for each cell line (Mycoplasma Detection Kit, Yeasen Biotechnology) prior to starting the experiments.

Primary cell culture

The patient-derived primary GBM cell line (P2#GBM) was established from surgical resection specimens of human glioblastoma. Fresh tumor fragments were immediately placed in ice-cold sterile PBS (Gibco, 10010023) and processed within 2 h. Necrotic areas, non-neoplastic parenchyma, residual blood clots, and connective tissue were meticulously excised. The remaining viable tumor was washed three times with cold PBS and mechanically minced into 1–2 mm³ fragments.

Minced fragments were enzymatically dissociated in 0.25% trypsin-EDTA (Gibco, 25200072) at 37 °C for 20–30 min with gentle agitation (100 rpm). Digestion was stopped by adding an equal volume of complete medium (DMEM, 4.5 g/L glucose, supplemented with 10% FBS, and 1% penicillin/streptomycin). The suspension was passed through a 70 μm cell strainer (Falcon, 352350) and centrifuged at 1000 × g for 3 min. The cell pellet was resuspended in complete medium and cultured at 37 °C in a humidified 5% CO₂ incubator.

RNA isolation and qRT-PCRRNA

Total RNA was extracted from cultured tumour cells or tumour samples using RNAiso Plus (TaKaRa, 9109; Shiga, Japan) following the manufacturer’s protocol. One microgram of total RNA was reverse transcribed in a 20 µL reaction mixture using the PrimeScript RT Reagent Kit with gDNA Eraser (TaKaRa, RR047A). Then, 1 µL of cDNA was added to a 20 µL PCR mixture containing SYBR Premix Ex Taq II Tli RNaseH Plus (TaKaRa, RR820A). Threshold cycle (Ct) values were determined using a Quant Studio 5 Real-Time PCR System (Thermo Fisher Scientific). The sequences of primers used were as follows: human TEP1, forward: 5’-GACGCCCCAGGCATATACAA-3’, reverse: 5’-TGCTGCCTGGCATACAAAGA-3’; and human GAPDH, forward: 5’-GGAGCGAGATCCCTCCAAAAT-3’, reverse: 5’-GGCTGTTGTCATACTTCTCATGG-3’. The relative expression of each was quantified using the 2-ΔΔCt method with GAPDH as the internal reference gene.

Western blotting and antibodies

Protein extracts were prepared by lysing cells and tumour specimens in RIPA buffer (Sigma, R0278), followed by protein concentration determination using a BCA protein assay kit (Solarbio Life Sciences, PC0020; Beijing, China). After quantification, 40 µg of total protein was separated on 10–15% SDS‒PAGE gels and transferred to PVDF membranes (Millipore, IPVH00010; Billerica, MA). The membranes were subsequently blocked in 5% skim milk (BD Biosciences, 232100; San Jose, CA) or 5% BSA (Solarbio Life Sciences, A8020) in TBS containing 0.1% Tween 20 (Sigma, P9416). The sections were incubated with primary antibodies overnight at 4 °C, followed by a one-hour incubation with an HRP-conjugated secondary antibody (CST) at room temperature. The protein bands were visualized using a Tanon 5500 Chemiluminescence Imaging System (Tanon Science & Technology; Shanghai, China) with Immobilon ECL Ultra Western HRP Substrate (Millipore, WBULS0500). The antibodies used for Western blotting included anti-TEP1 antibody (A9844, ABclonal) and anti-GAPDH antibody (AC001, ABclonal).

Immunohistochemistry

To process the samples for immunohistochemistry (IHC) analysis, they were fixed in 4% PFA for 24 h, followed by paraffin embedding and sectioning at a thickness of 3 μm. After deparaffinization and rehydration, 3% hydrogen peroxide was used to block endogenous peroxidase activity, and antigen retrieval was performed prior to standard serum blocking. The sections were then incubated overnight at 4 °C with specific primary antibodies and exposed to a secondary antibody. Peroxidase activity was visualized using DAB, and the sections were counterstained with Mayer’s haematoxylin. Microscopy images were obtained using an Olympus IX71 system (RRID: SCR_022185). The antibodies used for IHC were as follows: anti-Ki67 (ZM-0166, ZSGB-BIO); anti-TEP1 (YT4704, Immunoway); anti-4-HNE (ab48506, Abcam); and GPX4 (DF6701, Affinity).

Semiquantitative analysis was based on the immunoreactivity score. The percentage of positively stained cells was scored as follows: 0 (none), 1 (1–10%), 2 (10–50%), 3 (51–80%), and 4 (81–100%). The staining intensity was scored as 0 (none), 1 (weak), 2 (moderate), or 3 (strong). The final immunoreactivity score was calculated by multiplying the score for staining intensity by the score for the percentage of positive cells.

Cell viability assay

Cell viability was assessed by plating 5,000 cells per well in 96-well plates one day prior to the initiation of drug treatment. After exposure to the specified agents, the medium was replaced with fresh medium containing Cell Counting Kit-8 (CCK-8) reagent (B34304; Bimake). The plates were incubated at 37 °C for 1 h, and the absorbance at 450 nm was measured using a CLARIOstar microplate reader (BMG Labtech, RRID: SCR_019751).

Cell colony formation assay

Cell colony formation was assessed by plating 200 cells per well in 6-well plates containing DMEM supplemented with 10% FBS. Following exposure to the specified agents, the plates were maintained at 37 °C in a 5% CO₂ atmosphere for 10–14 days. Subsequently, the colonies were stained with 0.1% crystal violet (Sigma–Aldrich) and the quantification was performed. Three independent replicates were conducted for each experimental group.

GSH and MDA assay

The GSH concentration was assessed using the Reduced Glutathione (GSH) Content Detection Kit (BC1175, Solarbio), and the MDA content was measured using the Malondialdehyde (MDA) Content Detection Kit (BC0025, Solarbio). All procedures were conducted in accordance with the manufacturer’s guidelines.

Tissue samples

This study collected brain tissue samples from 8 trauma patients undergoing cerebral tissue resection and 47 tumor patients undergoing tumor resection at Nanfang Hospital, Southern Medical University (Guangzhou, China). The cohort included 16 grade II, 15 grade III, and 16 grade IV gliomas. Written informed consent was obtained from each patient. All specimens were pathologically confirmed and classified according to the 2021 World Health Organization Classification of Central Nervous System Tumors. Fresh samples were immediately snap-frozen and stored in liquid nitrogen. The use of human brain tumor specimens and the database was written consent and ethically approved by the Institutional Review Board at Nanfang Hospital of Southern Medical University (Guangzhou, China).

Xenograft studies

Six-week-old female BALB/c nude mice (RRID: MGI:2161072; Guangdong Provincial Medical Laboratory Animal Center) were housed under specific-pathogen-free (SPF) conditions for the intracranial implantation of GBM cells. Using a stereotactic apparatus (Stoelting), 5 × 10⁵ GBM cells suspended in 10 µL of PBS were injected into the brain of each nude mouse. From day 10 to day 18, RSL3 was administered every other day at a dose of 50 µg/g mouse body weight. Throughout the experiment, the mice had ad libitum access to food and water. On day 21, MRI (Bruker) was performed to measure intracranial tumour volume, which was calculated using the following formula: tumour volume (V) = length × width × width × 1/2. When severe neurological symptoms emerged, the mice were euthanized and perfused transcardially with heparin-saline followed by 4% PFA. The brains were then removed, fixed in 4% PFA, and processed for subsequent histological assessment. Kaplan–Meier survival curves were generated on the basis of the data. All animal experiments were approved by the Animal Experiment Center of Southern Medical University.

Statistical analysis

The data are presented as the means ± standard deviations (SDs), and all experiments were performed in triplicate. All the statistical analyses were conducted using GraphPad Prism 7.0 or R software (version 4.2.2). A P value < 0.05 was considered the threshold for significance. The level of statistical significance is indicated as follows: ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001.

Results

Expression and prognostic analysis of TEP1 in public databases

Initially, we evaluated TEP1 expression across various cancers using the TCGA database and observed that TEP1 was markedly upregulated in tumour tissues compared with normal tissues (Fig. 1A). In GBM samples from the TCGA, TEP1 exhibited copy number variations (CNVs), including heterozygous deletions (at a frequency of 29.29%), heterozygous amplifications (at a frequency of 5.03%), and homozygous deletions (at a frequency of 0.17%) (Fig. 1B). The presence and magnitude of CNVs are closely associated with the degree of tumour malignancy. For example, in endometrial cancer, an increased CNV burden is correlated with aggressive features, including advanced tumour stage and higher tumour grade, ultimately resulting in a poorer prognosis [23]. Similarly, in colorectal cancer, CNVs are linked to the adenoma-to-carcinoma transition, underscoring their role in tumour progression and metastasis [24]. In additional datasets, such as the GSE4290 and CGGA datasets, TEP1 mRNA levels were significantly elevated (Fig. 1C and D). Consistently, analysis of clinical samples from the Clinical Proteomic Tumour Analysis Consortium (CPTAC) revealed that TEP1 protein abundance was significantly greater in tumour tissues (Fig. 1E, Supplementary Fig. S1A). Furthermore, we assessed both the mRNA and protein levels of TEP1 and identified a positive correlation between them (Fig. 1F). To further investigate the relationship between TEP1 expression and clinical outcomes, we conducted Kaplan–Meier survival analyses using the TCGA and CGGA datasets, which revealed that higher TEP1 expression was associated with shorter OS (Fig. 1G and H), thereby underscoring the potential value of TEP1 as a biomarker of GBM progression. Collectively, our results demonstrate that TEP1 is upregulated in GBM and that its elevated expression is associated with a poor prognosis.

Fig. 1.

Database analysis of TEP1 expression and prognostic value. A Expression of TEP1 across multiple cancer types from the TCGA database. B Copy number variations (CNVs) of TEP1 in different tumors. The heterozygous amplification (Hate.Amp.), homozygous amplification (Homo.Amp.), Heterozygous Deletion (Hete.Del.), homozygous deletion (Homo.Del.). Expression of TEP1 in the GSE4290 (normal = 23, GBM = 157) (C) and CGGA (LGG = 632, GBM = 388) (D) datasets. E The protein abundance of TEP1 in the CPTAC dataset (normal = 10, GBM = 100). F A scatter plot demonstrating a positive correlation between TEP1 mRNA expression and TEP1 protein abundance in the CPTAC dataset. OS analysis of patients with high (red) versus low (blue) TEP1 mRNA expression based on the TCGA GBM-LGG (G) and CGGA (H) datasets. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

TEP1 expression is correlated with clinical parameters and serves as an independent predictor of survival outcomes in GBM patients

To assess the associations between TEP1 expression and clinical features in GBM patients, we analysed data from the CGGA database. The results indicated that TEP1 expression was elevated in glioma patients with IDH wild-type, 1p19q noncoding, and radiated tumors and that TEP1 levels increased with glioma malignancy. In contrast, factors such as sex, age, and treatment type (e.g., chemotherapy) did not appear to influence TEP1 expression (Fig. 2A and Supplementary Fig. S1B). Kaplan–Meier analysis of CGGA data revealed that elevated TEP1 expression was associated with a poorer prognosis for both IDH-mutant and IDH wild-type tumors. Notably, within the IDH wild-type subgroup, patients with high TEP1 expression presented the shortest OS (Fig. 2B and C). Furthermore, both univariate and multivariate Cox regression analyses identified TEP1 as an independent prognostic factor for OS in GBM patients (Fig. 2D and E). Therefore, TEP1 may serve as an independent predictor of GBM patient outcomes.

Fig. 2.

Association of TEP1 expression with clinicopathologic parameters. A Variations in TEP1 expression with respect to glioma IDH status, 1p19q status, grade and recurrence risk. OS analysis of patients in the CGGA cohort with high (red) versus low (blue) TEP1 mRNA expression and IDH mutation (B) and or IDH wild type (C). Univariate Cox analysis (D) and multivariate Cox analysis (E) of clinical information and gene expression data. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

Functional analysis of TEP1-related genes

We performed a differential expression analysis of samples with high versus low TEP1 expression and visualized the results using volcano plots (Fig. 3A, Supplementary Fig. S2A). KEGG analysis revealed that genes upregulated in the high TEP1 samples were enriched in lipid metabolism and atherosclerosis pathways, the cell cycle, apoptosis, ECM–receptor interactions, and immune response pathways (Fig. 3B). GO analysis of the DEGs was conducted to examine the related biological processes (BPs), cellular components (CCs), and molecular functions (MFs). The main enriched CC terms included focal adhesions, cell–substrate junctions, collagen-containing extracellular matrix, and vesicle lumens (Fig. 3C). The main enriched BP terms were viral responses, extracellular matrix organization, immune responses, and immunoglobulin-mediated signalling pathways (Fig. 3D). Additionally, these DEGs exhibited protease inhibitor activity and were involved in collagen binding (Fig. 3E). In contrast, KEGG and GO analyses using the TCGA database revealed the PI3K–Akt signalling pathway, as well as multiple pathways related to infection and immune responses (Supplementary Fig. S2B and S2C). GSEA further revealed that TEP1 expression was closely associated with antigen presentation and processing, the cell cycle, ferroptosis, and the PI3K signalling pathway (Supplementary Fig. S3D). Notably, inhibition of PI3K or AKT diminishes the telomere-protective effect of TRF1—an integral component of the shelterin complex—leading to increased DNA damage and fragility [25], which is consistent with the role of TEP1 in telomerase stability. In our study, gene correlation analyses revealed a significant positive association between the expression of TEP1 and the expression of numerous genes involved in ferroptosis suppression and telomere biology (Fig. 3F and Supplementary Fig. S2E). These results collectively underscore the pivotal role of TEP1 in orchestrating diverse cellular pathways linked to tumour progression, thereby highlighting its potential as a promising therapeutic target in cancer treatment.

Fig. 3.

Functional analysis of TEP1. A Volcano plot of differentially expressed genes (DEGs) between the high TEP1 and low TEP1 groups in the CGGA-GBM dataset; colour gradients from blue to red indicate increasing significance of changes in gene expression. B KEGG enrichment analysis of the DEGs; bubble size represents the number of genes enriched in each pathway, and bubble colour corresponds to the adjusted p value (adj. P) with a gradient from blue (higher adj. P) to red (lower adj. P). GO enrichment analysis of the DEGs categorized into cellular component (CC) (C), biological process (BP) (D), and molecular function (MF) (E) categories; colour intensity from light to deep indicates the magnitude of the log fold change. F Correlation heatmap depicting the relationships among genes with a colour gradient ranging from red (indicating negative correlation) to blue (indicating positive correlation); the colour intensity and size of each tile or pie segment represent the correlation strength

Immunoinfiltration analysis of GBM patients and its association with TEP1 expression

The tumour microenvironment (TME) plays a key role in the occurrence and development of glioma [26]. To further investigate the relationships between the prognostic signatures and the TME in the high-TEP1 and low-TEP1 groups, we initially employed the CIBERSORT algorithm to visualize the proportions of 22 immune cell types (Fig. 4A). We subsequently applied the ssGSEA algorithm to analyse immune cell distributions within each group. Compared with the low-TEP1 group, the high-TEP1 group presented higher percentages of 14 immune cell types—including plasmacytoid, immature, and activated DCs; multiple T-cell subsets; neutrophils; natural killer cells; and activated B cells—which is consistent with the results of our enrichment analyses (Fig. 4B). Correlation analysis further underscored these findings by revealing strong associations among these immune cells—particularly natural killer T cells and plasmacytoid DCs—suggesting an immune-mediated role of TEP1 in the TME (Fig. 4C).

Fig. 4.

Immune infiltration analysis of TEP1 in the TCGA-GBM cohort. A Stacked histogram of the proportions of 22 distinct types of infiltrating immune cells in GBM patients. B Boxplot of immune cell infiltration differences between patients with high versus low TEP1 expression. C Lollipop plot of the correlation between TEP1 expression and 28 immune cell subsets. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

TEP1 is most highly enriched in plasmacytoid dendritic cells and natural killer T cells

In this study, we analysed 40 samples from the GSE182109 dataset to identify cellular subtypes with high TEP1 expression. Each sample underwent initial quality filtering, resulting in the retention of 112,878 cells for further analysis. We subsequently employed an “anchor-based” integration method to remove batch effects and performed normalization, centring, and PCA-based dimensionality reduction; the top 15 principal components were retained. Following cluster annotation on the basis of marker genes, we employed UMAP for clustering and partitioned the 112,878 cells into 12 subpopulations, which were subsequently labelled using the SingleR package (Fig. 5A). UMAP visualization of TEP1 density revealed elevated expression in M2-type tumour-associated macrophages (TAMs), microglia, and T cells (Fig. 5B), with T cells exhibiting the highest expression (Fig. 5C).

Fig. 5.

Results of the scRNA-seq analysis of TEP1. A UMAP plot of single-cell samples from GSE1182109 clustered into 12 clusters, with cells colour-coded by annotated cell types. B UMAP plot of the expression patterns of the TEP1 marker gene. C Violin plots of TEP1 marker gene expression. D UMAP plot of T cells and dendritic cells (DCs) identified via scRNA-seq grouped into 8 distinct clusters. E Peak graph of TEP1 expression levels in annotated cells from D. F Pseudotime analysis of the developmental trajectory of TEP1 in T cells

To further assess TEP1 expression in different T-cell subsets and DCs, we performed clustering and annotation of T-cell subpopulations and plasmacytoid DCs. The results revealed that both cell types presented elevated TEP1 expression (Fig. 5D and E), which was consistent with our earlier immune infiltration findings. Additionally, we extracted T cells to delineate their developmental trajectory, revealing a gradual increase in TEP1 expression during cell maturation (Fig. 5F). Given that TEP1 is essential for maintaining telomere stability, the coelevation of TEP1 in these cell types may modulate cellular survival, proliferation, or apoptosis, ultimately contributing to GBM pathogenesis. This discovery enhances our understanding of the underlying mechanisms of GBM and reveals new avenues for potential therapeutic interventions.

Drug susceptibility analysis revealed TEP1 May be involved in ferroptosis regulation

To investigate the relationship between TEP1 expression levels and anticancer therapy response, we employed the OncoPredict algorithm to convert gene expression matrices from the CGGA and TCGA datasets into drug sensitivity matrices. We then classified GBM samples into high-TEP1 and low-TEP1 groups to compare their drug responses. Analysis of CGGA-GBM data revealed significant differences in the sensitivity to drugs targeting the cell cycle, DNA replication, mitosis, the PI3K/mTOR pathway, ERK/MAPK signalling, and protein stability/degradation (Fig. 6A and Supplementary Fig. S3A). Furthermore, we listed the drugs whose efficacy differed most between samples grouped according to enrichment of each target pathway category (Fig. 6B and Supplementary Fig. S3B). A similar pattern was observed in the TCGA-GBM dataset, with the drug pelitinib—an EGFR signalling inhibitor—also exhibiting a significant difference in efficacy between groups (Fig. 6C and D and Supplementary Fig. S3B, S3D). These findings indicate that TEP1 may be directly or indirectly involved in multiple pathways in vivo. Furthermore, by integrating CTRP and GDSC data, we identified a potential drug that may be more effective against tumors with high TEP1 levels (Fig. 6E). Notably, recent studies analysing drug–protein associations have also highlighted a correlation between dasatinib and genes [27, 28]. Additionally, we found that TEP1 correlated differently with RSL3 (a ferroptosis inducer) and N-acetylcysteine (NAC; a ferroptosis inhibitor) (Fig. 6F and Supplementary Fig. S3E), suggesting that TEP1 may be directly or indirectly involved in the regulation of ferroptosis. In summary, these results suggest that high TEP1 expression in GBM is associated with distinct drug response profiles and may directly influence ferroptosis regulation.

Fig. 6.

Drug susceptibility analysis of high and low TEP1 samples. A Volcano plot of differential drug sensitivities between the high-TEP1 and low-TEP1 groups on the basis of the GDSC-predicted drug sensitivity matrix in the CGGA-GBM dataset. B Boxplot of the IC50 distributions for these drugs in the high-TEP1 and low-TEP1 groups from the same dataset as in (A). C Volcano plot of differential drug sensitivities between the high-TEP1 and low-TEP1 groups on the basis of the GDSC-predicted drug sensitivity matrix in the TCGA-GBM dataset. D Boxplot of the IC50 distributions for these drugs in the high-TEP1 and low-TEP1 groups from the same dataset as in (C). E Integration of differential drugs based on GDSC and CTRP drug response data in TCGA-GBM samples (adjusted P-value < 0.05 and log fold-change < −0.3). F Histogram of the correlation coefficients for all gene–drug associations derived from the GDSC and CTRP drug response data for the TCGA-GBM samples. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

TEP1 expression is elevated in GBM cell lines and primary human glioma samples

To investigate the localization and expression of TEP1 in GBM, we performed immunofluorescence analysis on human GBM cell lines (A172, U87MG, LN229, T98, and U251), primary human GBM cells (P2#GBM), and three GBM tissue block specimens. The results demonstrated that TEP1 was predominantly localized in the nucleus (Fig. 7A and B). Subsequent analyses of lysates from these cell lines and Human Glioma Cell (HEB) via qRT‒PCR and Western blotting revealed significantly elevated TEP1 RNA and protein levels in GBM cells relative to HEB. Mesenchymal-type GBM cells (U87MG and LN229) presented moderately increased TEP1 expression, whereas classical-type GBM cells (T98 and U251) presented more markedly increased expression relative to that in HEB (Fig. 7C and E). Data from the CTRP drug sensitivity database indicated that, compared with T98 and U251 cells, U87MG and LN229 cells presented lower IC50 and AUC values for RSL3, suggesting that TEP1 expression may contribute significantly to RSL3 sensitivity (Fig. 7F). These results indicate that elevated TEP1 levels may contribute significantly to GBM progression and therapeutic sensitivity to agents such as RSL3. Furthermore, Western blot analysis of primary human GBM samples revealed a progressive increase in TEP1 protein levels with increasing tumour grade (Fig. 7F and G). Consistent with these findings, IHC of paraffin-embedded clinical samples—including grade II, III, and IV tumors, as well as nontumorous brain tissues—confirmed a progressive increase in TEP1 expression with tumour grade (Fig. 7I). Taken together, these findings highlight that TEP1 overexpression in GBM is associated with increased tumour grade.

Fig. 7.

Localization and expression of TEP1 in GBM. A Representative TEP1 immunostaining of GBM cell lines (A172, U87MG, LN229, T98, and U251) and P2#GBM. B Representative TEP1 immunostaining of three GBM tissue block specimens. Scale bar: 20 μm. C Relative TEP1 RNA levels in GBM cell lines and P2#GBM (n = 3). D Western blot analysis of TEP1 protein levels in lysates from GBM cell lines and P2#GBM. E Relative TEP1 protein levels in GBM cell lines and P2#GBM (n = 3). F IC50 and AUC values of RSL3 for each glioma cell line from CTRP. G Representative western blot analysis of TEP1 protein levels in human glioma tissues (n = 9) and normal brain tissues (n = 3). H Relative quantification of the western blots shown in G. I IHC analysis of TEP1 in Grade II, III, and IV tumors and in nontumorous brain tissues. Scale bar: 100 μm. All the data are presented as the means ± SDs. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

TEP1 inhibits RSL3-mediated ferroptosis in GBM cells

To explore the role of TEP1 in ferroptosis, we knocked down TEP1 in the T98G and U251 glioma cell lines and induced ferroptosis using RSL3 (Supplementary Fig. S4A). TEP1 knockdown resulted in a marked decrease in cell viability, which was reversed by the ferroptosis inhibitor deferoxamine (DFO) (Fig. 8A), suggesting that TEP1 regulates ferroptosis. To rule out other cell death pathways, TEP1-knockdown cells were treated with a panel of inhibitors, and only the ferroptosis inhibitor restored cell viability (Fig. 8B). Similarly, in TEP1-knockdown cells treated with RSL3, we observed cell shrinkage, reduced colony formation, decreased intracellular glutathione (GSH) levels, and significantly elevated malondialdehyde (MDA) levels, and these effects were partially reversed by DFO treatment (Fig. 8C and F). In contrast, overexpression of TEP1 in U87MG and LN229 cells partially restored normal cell morphology, colony formation, cell viability, and intracellular GSH levels while reducing MDA levels (Supplementary Fig. S4B-S4G). Transmission electron microscopy (TEM) revealed that TEP1 knockdown alone did not significantly alter mitochondrial morphology; however, under RSL3 treatment, mitochondria underwent substantial shrinkage, which was reversed by DFO treatment (Fig. 8G). Additionally, lipid peroxidation levels measured using the Alexa Fluor 488-A probe corroborated these observations (Fig. 8H and I). Taken together, these results demonstrate that TEP1 inhibits RSL3-mediated ferroptosis in glioma cells. Based on drug sensitivity analysis suggesting that Dasatinib is more effective against tumors with high TEP1 expression (Fig. 6E), we consequently evaluated the combined effects of RSL3 and Dasatinib in high-TEP1 GBM cell lines (T98 and U251). The results demonstrated a synergistic inhibitory effect on cell proliferation following combination treatment in both cell lines (Supplementary Fig. S4H).

Fig. 8.

TEP1 inhibits the ferroptosis of glioma cells. A Viability of T98G and U251 cells treated with or without RSL3 (0.6 µM) and deferoxamine (DFO, 100 µM) under shTEP1 or negative control conditions. B Viability of the T98G and U251 cell lines transduced with sh-NC, sh-TEP1#1, or sh-TEP1#2 in the presence of the ferroptosis inhibitor ferrostatin-1 (Fer-1, 1 µM), the apoptosis inhibitor Z-VAD-FMK (ZVAD, 1 µM), or the necroptosis inhibitor necrostatin-1 (Nec-1, 1 µM). C Representative phase-contrast microscopy images of T98G and U251 cells treated with RSL3 (2 µM) under shTEP1 or negative control conditions. Scale bar: 50 μm. Colony formation assays with corresponding colony counts (D) and intracellular levels of GSH (E) and MDA (F) under the same treatment as in (A) were quantified and are presented as histograms. G TEM images of T98G and U251 cell lines treated under the same conditions as in (A). Red arrows indicate mitochondria. Scale bars: 500 nm. H Lipid ROS production was measured in T98G and U251 cell lines treated with or without RSL3 (0.6 µM) or DFO (100 µM) under shTEP1 or negative control conditions. I Lipid ROS production was measured in U87MG and LN229 cell lines treated with or without RSL3 (0.6 µM) under oe-TEP1 or vector control conditions. All the data are presented as the means ± SDs. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

TEP1 inhibits RSL3-induced ferroptosis in vivo

To investigate the in vivo role of TEP1, U87MG-Vector and U87MG-OE-TEP1 cells were implanted into the brains of nude mice. RSL3 (50 mg/kg) was administered intraperitoneally every two days from day 10 to day 18 (Fig. 9A). Coronal MR scans performed on day 21 revealed that oe-TEP1 substantially alleviated RSL3-induced tumour shrinkage (Fig. 9B and C). Consistently, survival analysis revealed that oe-TEP1 significantly attenuated the extension of survival observed with RSL3 treatment (Fig. 9D). Furthermore, immunohistochemical analysis revealed that elevated TEP1 levels were associated with promotion of cellular proliferation and inhibition of ferroptosis (Fig. 9E and F), suggesting potential interplay between these processes. Collectively, these findings reveal TEP1 as a critical regulator of RSL3-induced ferroptosis in vivo.

Fig. 9.

TEP1 inhibits RSL3-mediated ferroptosis in vivo. A Tumour induction protocol for nude mice. B Coronal sections from MRI scans were used to evaluate tumour growth in nude mice after treatment. C Tumour volume was quantified and is presented in the histogram. D OS analysis of nude mice with in situ tumorigenesis on the basis of survival data. E IHC analysis for quantifying the expression of Ki67, TEP1, 4-HNE, and GPX4 in intracranial tumors from different groups. Scale bar: 100 μm. F Histograms of quantified IHC scores. All the data are presented as the means ± SDs. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001

Discussion

In this study, we investigated the role of TEP1 in GBM by analysing RNA expression data from the TCGA, CGGA, and GEO databases, as well as protein abundance data from the CPTAC database. Our analysis revealed that TEP1 is upregulated in GBM and that elevated TEP1 expression is significantly associated with a poor prognosis. Subsequent univariate and multivariate Cox regression analyses confirmed that TEP1 serves as an independent prognostic factor for GBM. It is important to note that these prognostic findings were derived solely from public datasets with complete survival information. We acknowledge that the lack of survival data in our internal GBM cohort precluded independent validation of this result. Previous studies have demonstrated that TEP1 is associated with an increased risk of bladder cancer [6] and breast cancer [7], and these findings are consistent with the results of the current study. Therefore, the expression pattern and prognostic value of TEP1 in GBM suggest that it may serve as a promising prognostic biomarker and play a pivotal role in clinical diagnosis and treatment.

A pharmacotranscriptomic study revealed that IDH1/2-mutant brain tumors are enriched in ferroptosis-related gene signatures, with IDH1/2-mutant low-grade gliomas (LGGs) exhibiting greater sensitivity to ferroptosis than their wild-type counterparts [29]. Additional evidence indicates that the oncometabolite 2-hydroxyglutarate (2-HG), produced by IDH1 mutants, can induce oxidative stress by impairing antioxidant defences—including reduced glutathione peroxidase (GPX) activity—thereby sensitizing cells to ferroptosis [30, 31]. Moreover, 2-HG can disrupt histone methylation patterns, drive TERT activation, and create conditions that indirectly lead to telomerase reactivation, thereby contributing to the immortalization and transformation of astrocytes [32]. In our study, TEP1 expression was associated with IDH mutation status, with significantly higher expression in IDH wild-type gliomas than in IDH-mutant gliomas. More importantly, high TEP1 expression in IDH wild-type gliomas was correlated with a poorer prognosis (Fig. 2B and C). To understand the mechanism behind this survival difference, future work should focus on delineating the molecular interplay between TEP1 and ID1 status. We also found that TEP1 expression is positively correlated with the expression of several ferroptosis-inhibiting genes, including NBPF10, RB1, and SLC40A1 (Fig. 3F). Therefore, TEP1 may act in concert with various ferroptosis-related genes to decrease ferroptosis sensitivity in GBM.

Recent advancements in cancer immunotherapy have exhibited significant success in treating advanced tumors, and a deeper understanding of tumour immune infiltration may provide ideas for improving cancer treatment [33, 34]. In lung cancer models, tumour-associated plasmacytoid dendritic cells (pDCs) within primary tumors exhibit impaired functionality and protumorigenic characteristics, which are associated with a poor prognosis [35]. In hepatocellular carcinoma and intrahepatic cholangiocarcinoma, a high density of tumour-associated pDCs is correlated with increased recurrence rates and reduced OS [36, 37]. In acute myeloid leukaemia (AML), an increase in pDCs is associated with a poor prognosis, particularly in secondary AML, where pDCs exhibit a clonal relationship with leukaemic blasts [38]. We assessed the tumour immune microenvironment using RNA-sequencing data and observed a strong positive correlation between TEP1 expression and the abundance of DCs as well as that of several T-cell subsets. Single-cell analysis revealed the highest TEP1 expression in pDCs and natural killer T cells. Consistent with our findings, elevated TEP1 levels are associated with pDC infiltration and indicate a poor prognosis.

Telomerase is a ribonucleoprotein enzyme that is crucial for maintaining telomere length. However, beyond this canonical function, telomerase—particularly its catalytic subunit, telomerase reverse transcriptase (TERT)—is involved in various signalling pathways and influences processes such as apoptosis, cell proliferation, and the stress response [39]. Additionally, studies have revealed that telomerase can interact with key intracellular pathways, including the PI3K/AKT/mTOR, NF-κB, and Wnt/β-catenin pathways [40, 41]. TEP1, a component of telomerase that binds to TERC [15], may also be involved in telomerase-related cellular physiological processes beyond telomere maintenance. In our study, functional enrichment analysis revealed that TEP1 is enriched in multiple distinct pathways. Drug sensitivity analysis also revealed associations with various pathways inhibited by drugs. Therefore, TEP1 may interact directly or indirectly with components of multiple cellular pathways. However, whether the regulation of these different cellular pathways by TEP1 is directly or indirectly dependent on TERT remains unknown. TEP1 is a key component present in both telomerase and ribosomal ribonucleoprotein complexes and is capable of stabilizing the structure and function of telomerase RNA and vRNA [4]. Instability and dysfunction of ribosomes can lead to a rapid decrease in GPX4 levels, thereby sensitizing cells to ferroptosis [16]. In our study, TEP1 knockdown led to decreased levels of GSH and GPX4 and increased levels of MDA in cells. Conversely, overexpression of TEP1 resulted in elevated levels of GSH and GPX4 and reduced levels of MDA. These findings suggest that TEP1 may reduce ferroptosis sensitivity in GBM by promoting the synthesis of GPX4.

Our study confirmed that TEP1 functions as a ferroptosis inhibitor. Our drug sensitivity analysis revealed that TEP1 expression is negatively correlated with the sensitivity to RSL3 and positively correlated with the sensitivity to NAC, indicating its complex role in modulating ferroptosis. Studies have shown that p53, a well-established tumour suppressor gene, inhibits ferroptosis through multiple mechanisms, although certain p53 mutants (e.g., p53R273H and p53R175H) can promote ferroptosis by interacting with NRF2 to suppress SLC7A11 expression [42]. Similarly, BAP1 normally inhibits ferroptosis by deubiquitinating the SLC7A11 promoter and reducing its transcription; however, in some contexts, BAP1 may alter cellular sensitivity to ferroptosis inducers via alternative mechanisms—such as changes in iron metabolism or antioxidant regulation—thereby increasing drug sensitivity when its expression is elevated [43]. Moreover, AMER1 promotes ferroptosis by interacting with SLC7A11 and the ferritin light chain (FTL) to facilitate proteolysis. Conversely, mutations in AMER1 can induce resistance to ferroptosis and promote the haematogenous metastasis of tumour cells [44]. These examples underscore that both the mutation status of genes and specific cellular contexts critically influence ferroptosis sensitivity. Notably, our study did not consider the impact of TEP1 mutation status or varying genetic background on the sensitivity to ferroptosis. Regarding the relationship between TEP1 and the response to RSL3 and NAC, our findings indicate that TEP1, in different mutation states or via alternative pathways, may regulate the expression of other ferroptosis-related genes. However, further research is warranted to determine whether TEP1 directly regulates the transcription of ferroptosis-related genes or modulates other ferroptosis pathways.

Our drug sensitivity analysis revealed that dasatinib and other agents may demonstrate enhanced efficacy against tumors with high TEP1 levels. Dasatinib is a multi-target tyrosine kinase inhibitor targeting including the Src kinase family, which has been shown to inhibit cell replication and promote cell death in various tumor cells [45]. However, dasatinib monotherapy has demonstrated limited efficacy in patients with recurrent glioblastoma [46]. Further studies have indicated that the combination of dasatinib and the WEE1 inhibitor MK-1775 exhibits potent synergistic effects in proneural glioblastoma stem cells (GSCs) [47]. Simultaneously, additional research has revealed that Src kinase targeting can overcome ferroptosis resistance to enhance radiosensitivity in glioblastoma [48]. Therefore, whether the combination of dasatinib with ferroptosis inducers could yield improved therapeutic outcomes represents a promising avenue worthy of further investigation.

In summary, our study revealed that TEP1 is upregulated in GBM, that its overexpression is correlated with a poor prognosis, and that it serves as an independent prognostic factor. Furthermore, bioinformatics analysis revealed that TEP1 may be involved in multiple distinct signalling pathways and is associated with tumour immune infiltration. Research has revealed that the telomerase inhibitor imetelstat effectively induces ferroptosis [49]. Our current study revealed that TEP1, a component of telomerase, inhibits ferroptosis.

Looking forward, several critical research directions emerge from our findings. Beyond prognosis, TEP1 should be integrated into multimodal prognostic models alongside variables like IDH status and MGMT methylation. Its role in DNA damage repair also warrants investigation into its potential as a predictor of therapeutic response. Mechanistically, the upstream drivers of TEP1 overexpression in GBM are unknown, and the downstream pathways mediating its oncogenic function require elucidation. These mechanistic insights are essential to evaluate the ultimate therapeutic question: whether targeting TEP1 could sensitize tumors to conventional therapies.

Conclusion

Taken together, our analysis confirmed that TEP1 is a robust prognostic indicator in GBM and that its overexpression is strongly correlated with poor OS and advanced pathological grade. Functional enrichment and drug sensitivity analyses revealed that TEP1 is related to diverse cellular processes, and immune infiltration analysis suggested its involvement in immunological responses, but further investigation is needed on this topic. Notably, in vitro and in vivo evidence indicates that TEP1 expression inhibits RSL3-induced ferroptosis in GBM cells. Thus, these findings offer novel insights for predicting the response to GBM therapy and provide a theoretical basis for personalized treatment strategies.

Abbreviations

GBM

Glioblastoma

TEP1

Telomere-associated protein 1

CNS

Central nervous system

TMZ

Temozolomide

CNVs

Copy number variations

vRNA

Vault RNA

ROS

Reactive oxygen species

GSH

Glutathione

MDA

Malondialdehyde

GPX4

Glutathione peroxidase 4

TCGA

The Cancer Genome Atlas

CGGA

Chinese Glioma Genome Atlas

GEO

Gene Expression Omnibus

CPTAC

Clinical Proteomic Tumor Analysis Consortium

OS

Overall survival

IDH

Isocitrate dehydrogenase

DEGs

Differentially expressed genes

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of enes and Genomes

GSEA

Gene Set Enrichment Analysis

GDSC

Genomics of Drug Sensitivity in Cancer

CTRP

Cancer Therapeutics Response Portal

CCLE

Cancer Cell Line Encyclopedia

IHC

Immunohistochemistry

DFO

Deferoxamine

TEM

Transmission electron microscopy

Author contributions

KXC, YQW & HL wrote the main manuscript text. KXC performed the data analysis, while ZCC, GPQ & YG assisted with data analysis. YQW, TL & SQZ conducted the background investigation, and KXC, QZ & YAP collected the data. KXC, ZCC, HC & RZW carried out the experiments. YTL & HL designed the project, secured funding and supervised the overall work.

Funding

This work was supported by the General Programs from the National Natural Science Foundation of China (82073305, 82373321, 81902544 and 82573580 to H. L.; 81972355 and 82373398 to Y.T.L.)

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The procedures for care and use of animals were approved by the Ethics Committee of the Nanfang Hospital, Southern Medical University and all applicable institutional and governmental regulations concerning the ethical use of animals were followed with relevant regulations and written informed consent was obtained from patients.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.