Skip to main content
NCBI home page
Discover Oncology logoLink to Discover Oncology
. 2026 Feb 7;17:411. doi: 10.1007/s12672-026-04611-2

Development and validation of a novel senescence-associated gene signature for prediction of survival and endocrine-disrupting chemicals in bladder cancer

Maohua Qin 1,#, Guofeng Xie 1,#, Haimeng Xie 1,#, Baisheng Lin 1, Zili Dai 1, Zijin Cheng 2, Li Wang 1, Jian Zhang 1,✉, Feixiang Wang 3,✉
PMCID: PMC12979719  PMID: 41653270

Abstract

Endocrine-disrupting chemicals (EDCs) are prevalent in various aspects of daily life, impacting the homeostasis of physiological processes and potentially contributing to the pathogenesis of cancer. Bladder cancer (BLCA) is the most common malignancy of the urinary system and one of the most prevalent malignancies globally. Cellular senescence is increasingly recognized as a key factor in cancer development. However, research remains limited regarding how EDCs influence BLCA through senescence-related genes (SRGs). In this study, expression data and clinical information were sourced from the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Senescence phenotype-related differentially expressed genes (DEGs) were identified. A senescence-related risk model was developed and validated by TCGA and GEO datasets. EDCs targeting DEGs, were extrapolated using the chemical toxicogenomic database (CTD). A Sankey diagram is employed to visualize the relationships between EDCs and key genes, identifying AKR1B1, EPHA3, IRF5, PGR, SREBF1 and TEAD4, as key factors in predicting cellular senescence in BLCA. Finally, molecular docking analysis was conducted to investigate the interaction between key genes and EDCs, thereby validating their feasibility. Our findings provide new insights into the role of SRGs in BLCA stratification and precision medicine. Furthermore, we elucidated the relationship between EDC-mediated cell SRGs and the pathogenesis of BLCA.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-026-04611-2.

Keywords: Bladder cancer, Endocrine-disrupting chemicals, Senescence, Molecular docking

Introduction

Bladder cancer (BLCA), the malignant tumor of the urinary system, is the 10th most common cancer and the second most prevalent urological tumor, with significant morbidity and mortality [1, 2]. Nearly 75% of patients are diagnosed with non-muscle-invasive bladder cancer (NMIBC), typically confined to the bladder lining, whereas 20% have muscle-invasive bladder cancer (MIBC), characterized by deeper tissue invasion [3]. The presence of senescent cells is positively associated with higher prevalence, and mortality of BLCA [4]. Importantly, microbial imbalances related to aging in BLCA trigger widespread metabolic changes that ultimately weaken the tumor’s immune environment, highlighting a key factor in the heightened BLCA burden in elderly populations [5]. Thus, it is urgent and crucial to explore novel molecular targets associated with invasion and distant metastasis of BLCA for enhancing the ability to predict patient prognosis and optimize stratified treatment strategies.

Endocrine-disrupting chemicals (EDCs) has been scientific investigation since 1977, and with the deepening of investigations, there has been growing concern regarding their impacts on human health [6]. EDCs are chemical substances that interfere with the normal functioning of the endocrine system, leading to detrimental health effects on the organism as a whole, its offspring, or specific populations [7]. According to statistical data, the treatment costs for diseases induced by EDCs exposure amount to approximately $340 billion dollars [8]. Previous studies have indicated that certain EDCs exhibit carcinogenic activity, particularly affecting hormone-sensitive tissues such as the breast, prostate, ovary, and thyroid [9]. Although the bladder is not recognized as a hormone-sensitive tissue, the presence of hormone receptors in the urothelium makes it a biological target for EDCs, thereby promoting the onset and progression of BLCA [10]. Exposure to bisphenol compounds has been linked to the development of cancer, with bisphenol A (BPA) and bisphenol S (BPS) from the bisphenol family being used in the production of plastics [11, 12]. These chemicals are present in various everyday products, such as water bottles and food containers [13]. Measurable concentrations of BPA have been detected in the urine of over 90% of the population, and 78% of urine samples show detectable levels of BPS [14, 15]. Urine can be stored in the bladder for several hours, significantly increasing the duration of EDCs exposure and the associated cancer risk [16, 10]. As plastics age and degrade, they leach from consumer products, leading to widespread human exposure through ingestion, inhalation, and skin contact, raising concerns about the potential link between EDCs and BLCA [17]. Additionally, studies have shown that BPA exposure is closely related to the molecular regulation of cellular senescence [18]. Exposure to EDCs has been proven to accelerate the process of cellular senescence [19]. However, the impact of EDCs on BLCA through the modulation of SRGs remains particularly scarce.

In this study, we utilize RNA transcriptome analysis and single-cell RNA sequencing analysis to comprehensively explore the expression patterns of SRGs in BLCA. Accordingly, we introduced prognostic signatures to quantify these patterns and predict clinical outcomes and responses to immunotherapy in BLCA. Additionally, multiple GEO cohorts were employed for external validation, further confirming the potential of the model as a critical biomarker for BLCA. Building upon existing clinical prediction models [20], we present a novel theoretical framework by integrating EDCs with BLCA genes involved in the regulation of cellular senescence. This approach explores the impact of EDCs on BLCA. Our findings elucidate the interactions between EDCs, genes, and BLCA, providing new insights into the prognosis and immunotherapy of BLCA. The interaction between EDCs and BLCA offers significant implications for cancer prevention strategies. Finally, through molecular docking, we further enhanced the predictive accuracy of toxicological assessments regarding the impact of EDCs on the cellular senescence process in BLCA.

Materials and methods

Data collection

BLCA tissues (n = 412) and non-cancerous tissues (n = 19) from The Cancer Genome Atlas Program-BLCA (TCGA-BLCA) were acquired from Genomic Data Commons (GDC, https://portal.gdc.cancer.gov/). GSE31684, GSE32894 and GSE48075 for BLCA and GSE78220 for PD-1immunotherapeutic efficacy were downloaded from Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/). Patients with incomplete clinicopathological information were excluded from the analysis. A curated list of senescence-associated genes (SRGs) was derived from the CellAge database (https://genomics.senescence.info/download.html), applying Tumor-Associated Genes. Genes absent in the TCGA dataset were excluded, resulting in 375 SRGs for analysis.

Consensus clustering and differential gene expression

The “limma” packages in R software (version 4.4.1) were conducted to identify shared differentially expressed genes (DEGs). There were 118 DEGs (71 up-regulated and 47 down-regulated) between normal and tumor tissues. Univariate Cox regression was employed to identify prognosis-associated DEGs among SRGs and TCGA-BLCA with FDR < 0.05, and log2|FC|> 1 [21]. Based on these SRGs, consensus clustering was performed using the “ConsensusClusterPlus” R package [22]. BLCA samples were clustered into two clusters using unsupervised clustering analysis for further analysis.

Prognostic SRGs-based signature construction

To establish an SRGs-based signature, we conducted a series of analysis involving univariate and multivariate Cox regression, as well as least absolute shrinkage and selection operator (LASSO) regression with tenfold cross-validation via the “glmnet” R package [23]. To prevent overfitting, we selected “lambda.min” for the computation. A final set of 6 genes (AKR1B1, EPHA3, IRF5, PGR, SREBF1, TEAD4) was used to construct a prognostic formula. Scores for each BLCA patient were calculated as follows: SRGscore = ∑ Coef(i) * Expr(i), where Coef(i) represents the coefficients of the genes and Expr(i) represents relative expression of genes in the cohort. Patients with BLCA were categorized into low- and high-risk subgroups based on their median risk scores. Multivariate Cox regression analysis was performed to determine whether the risk score could serve as an independent prognostic predictor in patients with BLCA.

Validation and performance assessment

Kaplan–Meier analysis was employed to compare the overall survival of high and low score subgroups. The accuracy of the signature in predicting survival at 1, 3, and 5 years was assessed using ROC curves via the “timeROC” R package [24]. The 1-, 3-, and 5-year survival predictors were generated using nomograms derived from clinicopathological factors (including age, sex, N, stage, and SRGscore). Calibration curves were generated using the “rms” R package to assess and compare the alignment between nomogram predictions and actual survival outcomes. The curves were superimposed on a reference line to demonstrate the signature’s precision. Decision curve analysis (DCA) was conducted to assess the nomogram’s clinical utility and benefits of the nomogram.

Functional enrichment analysis

Functional enrichment analyses were performed via the “clusterProfiler” R package [25], focusing on GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways. Additionally, Gene Set Enrichment Analysis (GSVA) was performed to uncover the biological processes among different SRGs subclusters by using the “GSVA” R package [26]. GSVA analysis was conducted using the gene set “c5.go.Hs.symbols.gmt” acquired from the MSigDB database. The “LIMMA” R package was used to characterize the functional annotation of SRGs.

Immune infiltration analysis

Using the Estimation of Stromal and Immune cells in Malignant Tumor tissues using Expression data (ESTIMATE algorithm), We observed differences in immune scores, stromal scores, ESTIMATE scores, and tumor purity across the samples [27]. The CIBERSORT method, implemented via the “CIBERSORT” R package was used to evaluate the relative abundance of different cell types and the single-sample gene set enrichment analysis (ssGSEA) method via the “GSVA” R package.

Prediction of immunotherapy responsiveness

The Tumor Immune Dysfunction and Exclusion (TIDE) algorithm was employed to forecast how BLCA patients would respond to ICIs therapy [28]. We also analyzed the disparities in immune checkpoint gene expression levels and TMB between high-risk and low-risk groups as potential markers for predicting immunotherapy response.BLCA mutation data were collected from the TCGA database and TMB was calculated using the “maftools” R package [29].

Single-cell RNA sequencing analysis

The transcriptional profile of bladder carcinoma cells were derived from the single-cell RNA sequencing (scRNA-seq) data of GSM5360669 and further analyzed with the “Seurat” R package [30]. To ensure high-quality scRNA-seq data, genes expressed in at least 5 cells within each sample were included in the analysis. Subsequently, low-quality cells based on specific criteria: cells with fewer than 50 expressed genes, or those with mitochondrial genome unique molecular identifiers (UMIs) exceeding 5%, were excluded. Next, data normalization was performed with the “NormalizeData” R package. The top 1500 variable features were identified for each dataset, and Principal Component Analysis (PCA) was achieved using the “FindVariableFeatures” and “RunPCA” functions.

The t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm was applied using the top 20 principal components as input for dimensionality reduction. Cell clusters were identified using the “FindClusters” function with a resolution of 0.4. Guided by known marker genes associated with various cell types, the “SingleR” tool was used to identify and refine cell type classifications. Meanwhile, the enrichment expression of TEAD4 was analyzed at the single cell level through the TISCH2 database to further validate our research findings.

Single gene analysis of TEAD4

The TEAD4 gene, selected from the prognostic model, underwent additional analysis. We used the Wilcoxon test to compare the expression levels of the TEAD4 gene between normal tissues and tumor tissues.

Tumor immune microenvironment (TIME) estimation

Single sample gene set enrichment analysis (ssGSEA) was used to quantify the immune function of the TIME. To assess the relationship between immune cell infiltration and two clusters, CIBERSORT (http://CIBERSORT.stanford.edu/) was used to estimate the relative subsets of RNA transcripts, enabling the measurement of immune cell type proportions in the samples.

EDC-gene-BLCA interaction

To elucidate the relationship between key SRGs and EDCs in BLCA, we first screened for these chemicals associated with SRGs using the Comparative Toxicogenomics Database (https://ctdbase.org/). After identifying the chemicals, we obtained the EDC data from the Endocrine Disruptor Exchange (TEDX) database (https://endocrinedisruption.org/) and explored the interactions between these chemicals and genes and constructed the EDC-Gene-BLCA network diagram.

Molecular docking and visualization

To further investigate the potential interactions between EDCs and genes associated with cellular senescence in BLCA, we retrieved the 3D chemical structure of benzo[a]pyrene from the PubChem database to study its binding sites. Subsequently, we searched for our core genes on the RCSB PDB website(https://www.rcsb.org/) and filtered the results based on “Homo sapiens” entries to obtain the required protein structures [31]. The molecular docking between benzo[a]pyrene and the core genes was performed using the CB-Dock2 website(https://cadd.labshare.cn/cb-dock2/index.php) to predict the interactions between benzo[a]pyrene and the core target proteins. At the same time, we optimized the grid box size and genetic algorithm parameters. The binding affinity between them (Vina score) was calculated, with lower values indicating stronger binding.

Statistical analysis

The Wilcox test was used for comparisons between normal and tumor tissues. The log-rank test and Kaplan-Meier method were used to assess the prognostic differences. The Kruskal-Wallis test method was used to compare the differences in data distribution of multiple genes across different clusters. The Chi-square test was used to determine the differences in immune therapy responses between groups. Univariate and multivariate Cox regression analyses were performed to evaluate the relationship between variables and BLCA patient prognosis. The Wilcoxon test was used to compare differences in TIME scores as well. The “maftools” package was utilized to characterize the mutation burden between high- and low-risk SRGs groups in the TCGA-BLCA cohort. The Pearson correlation coefficient was used to evaluate the relationship between immune cells in TEAD4 high- and low-expression groups. All data analysis was performed using R software (version 4.4.1). Perl software (v5.30.3; https://strawberryperl.com/), and SPSS (version 25.0; IBM Corp.). Data are expressed as mean ± SD. A P value < 0.05 was considered statistically significant. (*p < 0.05; **p < 0.01; ***p < 0.001; ns indicates not significant).

Results

Mutation patterns and expression of SRGs in BLCA

In this study, we selected a total of 375 tumor-related SRGs and analyzed the mutation prevalence in 412 BLCA patients. After excluding cases with incomplete clinical data, SRGs mutations were notably found in 395 patients, who were then included in further analysis. To further explore the role of SRGs in BLCA, we compared the expression of 375 SRGs between normal and bladder cancer tissues using the TCGA cohort. The analysis indicated that 71 genes had higher expression levels in normal tissues, while 47 genes were upregulated in tumor tissues (Supplemental Fig. 1 A, B). To obtain more insights into the influence of these genes on tumor initiation and progression, we performed enrichment analysis on the screened senescence-associated differentially expressed genes (DEGs) to study the significant DEGs among BLCA individuals. The results of the KEGG analysis revealed that DEGs were associated with pathways such as “cell cycle”, “cellular senescence”, “p53 signaling pathway” and “cisplatin resistance” (Supplemental Fig. 1 C). And GO analysis showed that DEGs were associated with processes such as “mitotic cell cycle transition”, “regulation of cell cycle transition”, “spindle”, “chromosome” and “DNA-binding transcription factor activity” (Supplemental Fig. 1D). These pathways are intricately linked to the process of cellular senescence.

Univariate Cox regression identified 29 SRGs associated with prognosis, classified as either “risk” or “protective” genes (Supplemental Fig. 1E). The relationships among these 29 SRGs are depicted, illustrating the corresponding expression patterns among 29 prognosis-related SRGs (Fig. 1A).

Fig. 1.

Fig. 1

Open in a new tab

Gene and Immune Infiltration Differences Across SRGs Clusters. A Prognostic Network of 29 SRGs in TCGA-BLCA. Green dots indicate favorable prognostic factors, while purple dots indicate risky prognostic factors. Lines connecting the genes represent their interactions, with thickness indicating the strength of the correlation. Red lines represent positive correlations, and blue lines represent negative correlations. B Consensus Clustering analysis of the TCGA-BLCA cohort with k = 2. C Consensus Cumulative Distribution Function (CDF) plot determined the optimal number of clusters (k = 2) for the TCGA-BLCA cohort. E Abundance of TIME-infiltrating cells in SRGs patterns. G Survival analysis of the TCGA-BLCA cohort according to SRGs patterns, including SRGs cluster A and SRGs cluster B. * p < 0.05; ** p < 0.01; *** p < 0.001

Consensus clustering on SRGs in BLCA

To identify the underlying functions of SRGs, the “ConsensusClusterPlus” R package was employed to perform consensus clustering, identifying two distinct subgroups with K = 2 as the optimal value (Fig. 1B, C). The results of Principal Component Analysis (PCA) demonstrated that patients could be divided into two clusters based on the distribution characteristics (Fig. 1D). Differential analysis of SRGs revealed significant differences in expression levels across the subclusters, with 26 SRGs showing distinct variations. Among these, 25 genes were significantly upregulated in the SRGs B cluster, including AKT3, AURKA, BAG3, CDK5R1, DUSP1, DUSP3, EPHA3, FGFR1, FSCN1, GLI1, INCENP, JUN, ME1, MMP9, MYLK, PGR, PITX1, RBM2, SRF, TAGLN, TEAD1, TEAD4, TFDP1, and TWIST2. In contrast, only IRF5 was significantly downregulated in the SRGs A cluster (Fig. 1E). To assess immune cell infiltration, we performed cellular infiltration analysis of TIME between the two clusters by using the “ssGSEA” algorithm. Cluster A showed higher abundance of immunosuppressive cells, such as regulatory T cells, whereas Cluster B demonstrated elevated expression of CD56dim NK cells, suggesting a more active and potentially effective anti-tumor immune response (Fig. 1F). Meanwhile, Kaplan-Meier survival analysis further confirmed SRGs cluster B exhibited a significant survival advantage compared to SRGs cluster A (Fig. 1G). The result indicates that higher expression of SRGs in BLCA patients is often associated with poorer prognosis.

The heatmap demonstrates that 29 SRGs exhibit differential distribution across clinical characteristics, including pathological stage, lymph node metastasis, age, and gender (Fig. 2A). By modulating SRGs activity, it is possible to alter the clinical characteristics of patients, and influence the progression of the disease [32]. To comprehensively understand the impact of these SRGs on pathway-level progression in BLCA, we performed GSVA enrichment analysis (Fig. 2B). The results revealed that Cluster A was associated with pathways such as cell motility, regulation of cell proliferation, and regulation of cell communication. This indicates that the tumor tissues in this cluster exhibit higher proliferative activity and are more prone to metastasis.

Fig. 2.

Fig. 2

Open in a new tab

Biological functions and Responsiveness to Immunotherapy of SRGs patterns in BLCA. A Heatmap of clinicopathologic features of SRGs patterns. SRGs cluster, N, age and gender were used as patient annotations. B GSVA enrichment analysis indicating biological pathways among SRGs patterns. C-F ESTIMATE scores analysis between cluster A and cluster B. ESTIMATEScore (C), ImmuneScore (D), StromalScore (E), TumorPurity (F). G-J TIDE analysis including TIDE score (G), Dysfunction score (H), Exclusion score (I), potential immunotherapy responders (J). * p < 0.05; ** p < 0.01; *** p < 0.001

Responsiveness to immunotherapy between two clusters

The TIME is crucial in the onset and development of cancer [33]. Variations in TIME among cancer patients are strongly linked to how well they respond to immunotherapy [34]. Although current research on the relationship between cellular senescence and TIME remains limited, understanding the connection between TIME and the expression patterns of different SRGs in bladder cancer is of critical importance. The “ESTIMATE” package was utilized to characterize changes in TIME components between different groups.Cluster A exhibited increased stromal, immune, and ESTIMATE scores, along with decreased tumor purity, suggesting a higher level of immune and stromal cell infiltration and a more complex tumor microenvironment (Fig. 2C-F).

The TIDE algorithm is specifically designed to predict responses to immune checkpoint blockade (ICB) therapy. A higher TIDE score indicates a lower likelihood of benefiting from ICB therapy and a greater potential for tumor immune evasion. In our analysis, Cluster A exhibited higher “TIDE scores,” “Exclusion scores,” and “Dysfunction scores”. Moreover, a greater percentage of patients in Cluster B are expected to react positively to immunotherapy. This evidence suggests that patients in Cluster B have a higher chance of benefiting from immunotherapy than those in Cluster A (Fig. 2G-J). This result could explain why Cluster A exhibits active immune cells yet has poor patient prognosis.

Establishment and validation of a senescence-based signature

Considering the impact of SRGs on prognosis and immunotherapy responsiveness in BLCA patients, researchers sought to develop a prognostic signature to uncover the underlying complexity of BLCA. Through univariate Cox regression analysis, 29 candidate prognostic genes were initially identified. These genes were further refined using the LASSO algorithm, narrowing it down to 12 key genes (Fig. 3A, B). Subsequently, a multivariate Cox regression analysis was conducted, resulting in the construction of a prognostic signature based on six genes: AKR1B1, EPHA3, IRF5, PGR, SREBF1, TEAD4. We established a prognostic signature signature with these genes, calculated as: Risk Score = 0.193*AKR1B1 + 0.245*EPHA3–0.233*IRF5 + 0.615*PGR + 0.325*SREBF1 + 0.244*TEAD4, where each term represents the product of the gene’s regression coefficient and its normalized expression value. Based on predictions from the HPA website regarding protein molecule expression levels, we have determined that EPHA3 is a prognostic marker in BLCA with high expression levels. Conversely, AKR1B1 is commonly expressed across patients (Fig. 3C). Using the median value of the risk score as a cutoff, we divided the TCGA-BLCA patients into high- and low-risk groups, and subsequently obtained their Overall Survival (OS) and Progression-Free Survival (PFS) data (Fig. 3D, E). OS and PFS were significantly lower in the high-risk group compared to the low-risk group. To assess the predictive performance of our prognostic signature, we plotted the Receiver Operating Characteristic (ROC) curves for 1, 3, and 5 years, finding values of 0.688, 0.664, and 0.713, respectively. These results suggest that the signature has good stability and reliability, demonstrating good predictive performance of our signature (Fig. 3I). Furthermore, the accuracy of our signature was validated in three independent external validation sets (GSE32894, GSE31684, and GSE48075, Fig. 3F-H), yielding satisfactory results with 3-year AUC values of 0.719, 0.617, and 0.710, respectively (Fig. 3J-L).

Fig. 3.

Fig. 3

Open in a new tab

Establishment and Internal/External Validation of the SRGs Prognostic Signature. A Identification of six prognostic biomarkers using lasso-cox regression. B Coefficients were analyzed by least absolute shrinkage and selection operator (lasso) cox regression. C Representative immunohistochemical images of EPHA3 and AKR1B1. D, E Survival prognostic analysis of Disease-Free Progression and Overall Survival in TCGA-BLCA. F-J Survival prognostic analysis of the GSE32894 (D), GSE31684 dataset (E), and GSE48075 dataset (F) in high- and low-risk groups. I-L AUC of the SRGs signature in the TCGA-BLCA cohort (I), GSE32894 dataset (J), GSE31684 dataset (K) and GSE48075 dataset (L)

Based on the median SRGscore, the risk scores of each patient in the BLCA cohort were ranked from low to high. The expression levels of the six signature genes were also displayed (Supplemental Fig. 2 A). Significantly, the expression level of the “risk” gene IRF5 was lower in this group, suggesting it may function as a protective gene (Supplemental Fig. 2B). To visually describe the relationships among clusters, risk subgroups, and survival status in BLCA, the Sankey diagram was used to visually display the attribute transitions of individual patients, and illustrated that patients with higher SRGscore are more likely to be classified into Cluster A and have poorer prognoses (Supplemental Fig. 2 C, D). Multivariate Cox regression analysis indicated that age, T stage and the SRGs signature served as independent prognostic factors (Supplemental Fig. 2E). Next, the SRGscore, pathological stage, age, and gender were integrated into a nomogram to predict the 1-, 2-, and 3-year probabilities of OS events in the TCGA-BLCA cohort. The C-indices for the 1-, 2-, and 3-year OS rates were 0.895, 0.687, and 0.63, respectively, demonstrating the robust performance of the nomogram (Supplemental Fig. 2 F, G). In the DCA, the SRGscore displayed consistent positive and large net benefits across a broad range of risk thresholds (Supplemental Fig. 2H-J). In summary, these results suggest that the prognostic signature based on the concept of cellular senescence holds significant clinical predictive value for BLCA.

SRGs signature somatic mutation profile

The TMB of each sample was calculated and analyzed. Waterfall plots were generated to display the top 20 genes with the highest mutation frequencies in high-risk and low-risk patients (Supplemental Fig. 3 A, B). The results revealed that genes were mutated more frequently in the high-risk group than in the low-risk group, with the rate of the 15th most significantly mutated gene being 17% vs. 11%. Next, the expression of immune checkpoints was evaluated in both the high-risk group and the low-risk group. The expression of 21 immune checkpoints was significantly dysregulated, with IDO1, TIGIT and CD274 significantly overexpressed in the high-risk group than in the low-risk group (Supplemental Fig. 3 C). Patients with low risk and high CD274 expression demonstrated better survival outcomes compared to those with high risk and high CD274 expression. And patients with low risk and low CD274 expression exhibited better survival compared to those with high risk and low CD274 expression (Supplemental Fig. 3D). The IDO1 and TIGIT stratification groups showed similar findings, suggesting that the expression of immune checkpoints could cause different effects on the immune microenvironment in low-risk and high-risk patients, thereby impacting their survival (Supplemental Fig. 3E, F).

SRGs patterns in the role of ICI therapy

To comprehensively assess the relationship between SRGscore and immunotherapy, we incorporated the GSE78220 cohort that evaluates the response of anti-PD-L1 immunotherapy (atezolizumab). We noticed that a higher SRGscore was associated with adverse survival outcomes post-immunotherapy. The boxplot indicated that the low SRGscore group showed improved clinical responses to anti-PD-L1 immunotherapy (Supplemental Fig. 3G). And we noticed that a higher SRGscore was associated with adverse survival outcomes post-immunotherapy (Supplemental Fig. 3H). Meanwhile, the ROC curve analysis demonstrated that the SRGscore exhibited good predictive performance for immunotherapy survival outcomes, with an AUC of 0.724 (Supplemental Fig. 3I).

Single-cell RNA-sequencing data and single gene analysis of TEAD4

The transcriptional profile of BLCA cells were derived from the scRNA-seq data of GSM5360669. Subsequently, twelve distinct cell clusters were visualized by t-SNE (Fig. 4A). The heatmap was created to illustrate the expression levels of the top five marker genes for each cluster (Fig. 4B). Additionally, the distribution of the marker genes for each cluster was visualized using scatter plots (Fig. 4C). The twelve cell clusters were categorized into eight distinct cell types: Epithelial cells, NK cell, T cells, Monocyte, Neutrophils, CMP, DC, B cell (Fig. 4D). Furthermore, the expression and distribution of the TEAD4 gene were visualized (Fig. 4E, F). Interestingly, we observed a significant enrichment of TEAD4 expression in epithelial cells. To further validate our conclusions, we performed an independent analysis using single-cell RNA sequencing data (GSE130001) from the TISCH2 database. The results consistently demonstrated a significant enrichment of TEAD4 in epithelial cells, thereby reinforcing our findings at the single-cell resolution (Supplemental Fig. 4A-C).

Fig. 4.

Fig. 4

Open in a new tab

Single-cell RNA-sequencing analysis and single gene analysis of TEAD4. A T-SNE plot of 12 cell clusters. B Heatmap depicting the top 5 marker genes for each cluster. C Scatter plot showing the distribution of the marker gene across each cluster. D T-SNE plot of all clusters with cell clustering annotations. E, F Scatter plot (E) and bubble plot (F) displaying the distribution of TEAD4 gene in clusters. G Differential expression of TEAD4 between normal and BLCA tissues. H Kaplan-Meier survival curve Illustrating TEAD4 expression between high and low risk scores. I ROC curve of TEAD4 in predicting survival time. J Correlation between various immune cells and TEAD4

The TEAD4 gene showed significantly increased expression in BLCA (Fig. 4G). Individuals with high TEAD4 expression had a poorer prognosis compared to those with low TEAD4 expression (Fig. 4H). The AUC of the ROC curve was 0.622, indicating that the TEAD4 gene offered moderate prognostic predictive accuracy (Fig. 4I). The TEAD4 gene is associated with a variety of immune cells, particularly CD4T cells, M1 Macrophages, Neutrophils (Fig. 4J). In addition, to confirm the expression correlation of TEAD4 within senescence markers, the GEPIA2 database was employed to analyze its association with key senescence markers. The results demonstrated a significant correlation between TEAD4 and these markers, supporting its potential role as a representative gene of SRGs (Supplemental Fig. 4D).

EDC-gene interaction analysis for key genes

Based on the analysis of the CTD and EDCX databases, we selected 27 EDCs associated with BLCA (Fig. 5A-F). We conducted a correlation analysis between key genes related to cellular senescence identified in BLCA and EDCs (Fig. 5G). The analysis results suggested that EDCs may play a potential role in the occurrence and development of BLCA.

Fig. 5.

Fig. 5

Open in a new tab

EDCs-Genes-BLCA interaction network analysis. A-F Interaction analysis BLCA of and EDCs. EPHA3 (A), AKR1B1 (B), IRF5 (C), PGR (D), SREBF1 (E), TEAD4 (F). G Sankey diagram displaying the EDCs-Genes-BLCA interaction network

Molecular docking of benzo[a]pyrene with core target proteins

Due to the widespread presence of benzo[a]pyrene in air, soil, and water, it is typically metabolized in the body to form active metabolites. These metabolites can react with DNA to form adducts, potentially leading to cellular mutations and other diseases, including cancer. Based on the analysis of EDCs, we identified extensive connections between benzo[a]pyrene and target genes. Therefore, molecular docking analysis was conducted with huge genes (AKR1B1, EPHA3, IRF5, PGR, SREBF1, TEAD4) to assess the interactions between benzo[a]pyrene and these targets (Fig. 6A-G). The molecular docking results demonstrate that benzo[a]pyrene exhibits strong binding to the surface cavities of AKR1B1, EPHA3, IRF5, PGR, SREBF1, and TEAD4, with binding affinities all below 6.5 kcal/mol (9.9, 7.8, 7.6, 12.6, 8.8, and 6.7 kcal/mol, respectively), indicating a high degree of binding stability. These findings suggest that under long-term outdoor exposure, benzo[a]pyrene may play a crucial promoting role in the initiation and progression of BLCA by tightly binding to core receptors. This interaction may induce alterations in intracellular signaling pathways, thereby influencing tumor development and progression. Therefore, understanding the interactions between benzo[a]pyrene and these receptors provides new avenues for the early diagnosis and treatment of BLCA.

Fig. 6.

Fig. 6

Open in a new tab

Molecular docking results of B[a]P with target proteins. A-F Molecular docking visualization of B[a]P with hug genes in BLCA. AKR1B1 (A), EPHA3 (B), IRF5 (C), PGR (D), SREBF1 (E), TEAD4 (F)

Discussion

In spite of recent advances in immunotherapy and targeted therapy, BLCA is still notorious for its high incidence and recurrence rates. Cellular senescence is a complex biological process that involves changes in gene expression, accumulation of DNA damage, loss of protein function, and cell cycle arrest [35]. Earlier research has highlighted the dual function of cellular senescence in cancer: it prevents tumor formation and restricts malignant growth in the initial phases, but as the tumor persists, it aids in tumor development, encourages invasion and migration, and inhibits immune surveillance [36]− [37]. SRGs, such as p53, RB, and PTEN, play vital importance in cancer development. Additionally, it is intimately connected with the emergence and advancement of various illnesses and functions as an efficient means for classifying cancer patients [38]. Another prognostic model involving the PRPF19 gene elaborates on the connection between cellular senescence and stemness regulation in BLCA [39]. However, the underlying mechanisms of cellular senescence in regulating BLCA phenotypes remain unclear.

In this study, we screened 29 DEGs associated with cellular senescence. Functional enrichment analysis revealed that cell cycle regulation, the p53 pathway, and platinum drug resistance are involved in the progression of BLCA mediated by SRGs. Recent studies have shown that chromosome segregation errors can trigger the p53/p21 response, leading to alterations in chromatin structure [40]. This could be crucial in the transformation of cells and the advancement of cancer, and it is strongly associated with cellular senescence. The onset of cellular senescence is a major factor in determining the response of cancer cells to chemotherapy, including platinum-based drugs [41]. Next, we developed a prognostic signature to quantify the classification of high- and low-risk groups based on the expression patterns of SRGs, which were selected through machine learning and consist of six genes (AKR1B1, EPHA3, IRF5, PGR, SREBF1, TEAD4). Significant survival differences were observed when patients were categorized based on both risk score and the expression of key immune checkpoint genes (CD274, IDO1, and TIGIT; all p < 0.001). Patients with high expression of immune checkpoints together with a high-risk score exhibited the poorest prognosis, suggesting enhanced immune evasion and aggressive tumor biology. Conversely, those with high checkpoint expression but a low-risk score demonstrated relatively favorable survival outcomes, implying that elevated checkpoint expression in this subgroup may reflect a more immunoresponsive microenvironment. Among them, TEAD4, a member of the TEADs family, exerts its function by interacting with transcriptional co-activators [42]. The transcription factor TEAD4 can influence biological processes such as cell proliferation and tissue regeneration, and it plays a significant role in cancer [43].Recent studies have found that the downregulation of the AMPKα/Yap1/TEAD4/CCNE1/2 axis can inhibit the proliferation of BLCA cells [44]. Interestingly, TEAD4 mediates epithelial-mesenchymal transition through the PI3K/AKT signaling pathway in BLCA, thereby modulating tumor progression [45].

Furthermore, we analyzed the distribution of the TEAD4 gene in BLCA at the single-cell sequencing level. The result shows that the expression of TEAD4 is significantly enriched in epithelial cells. It also demonstrated a significant effect in predicting the prognosis of BLCA. These findings suggest that investigating cellular senescence and immunotherapy in BLCA holds significant clinical potential for enhancing patient prognosis.

To deepen the genetic understanding of BLCA and broaden the application of our findings, we analyzed the EDCs associated with BLCA. Our analysis using CTD and huge SRGs in BLCA led to the creation of a chemical-gene interaction network, validating the connection between EDCs and BLCA, and identifying 27 potential EDCs that may contribute to BLCA. Chemical substances that interact with two or more genes include: sodium arsenite, quercetin, nicotine, methylcholanthrene, 1,3-butadiene, 2-naphthylamine, cadmium chloride, 4,4’-diaminodiphenylmethane, curcumin, ammonium chloride, arsenic trioxide, arsenic, arsenite, benzo(a)pyrene, beta-hexachlorocyclohexane, bisphenol A, diethylnitrosamine, ethanol, flutamide, hexachlorocyclohexane, and smoke.

Benzo[a]pyrene and its metabolites exhibit high toxicity and carcinogenicity, making them common persistent organic pollutants and EDCs [46]. It can be widely distributed in the environment and gradually accumulate in organisms, such as in the urinary system, through various pathways, thereby exerting significant environmental impacts and inducing disease development [47]− [48]. A study reported that under specific epigenetic modifications of certain proteins, bladder epithelial cells exposed to benzo[a]pyrene and its metabolite BPDE undergo an accelerated epithelial-mesenchymal transition process [44]. In our study, we found that benzo[a]pyrene interacts with the six SRGs identified in BLCA. These interactions suggest that benzo[a]pyrene, as an EDC, may regulate cellular senescence, the TIME, and cancer-related signaling pathways through SRGs, thereby promoting the progression of BLCA. The effects of benzo[a]pyrene are likely to exacerbate the TIME by interfering with SRGs. Furthermore, these interactions suggest that the identified SRGs could serve as potential biomarkers for BLCA associated with benzo[a]pyrene exposure and may represent pivotal targets for therapeutic strategies.

Limitations

While the study provides valuable insights, there are several limitations that need to be considered. Firstly, the analysis relied solely on publicly available datasets such as TCGA and GEO. While these are valuable resources, they may contain inherent biases related to sample collection, patient demographics, and sequencing platforms, which could affect the generalizability of the results. Secondly, although bioinformatic analyses and molecular docking were performed to link EDCs to key SRGs. SRGs relied solely on LASSO regression and consensus clustering, it may fail to capture critical epistatic, synergistic, or threshold-dependent relationships within the senescence-related gene (SRG) network, the relationship between EDC exposure and senescence-related transcriptional programs in BLCA is inferred from bioinformatic correlations and requires experimental validation to establish causality. Thirdly, molecular docking suggests potential binding sites for benzo[a]pyrene on core senescence-related proteins; however, the stability and biological relevance of these interactions require validation through molecular dynamics simulations and experimental assays. The use of databases like CTD and computational docking provides preliminary evidence of chemical-gene interactions but does not account for the complexity of real-life exposure scenarios, such as low-dose chronic exposure, mixture effects, or metabolic activation of pro-EDCs. Fourthly, Our scRNA-seq analysis is based on a single publicly available BLCA dataset, which limits our ability to account for interpatient heterogeneity, prospective multi-center scRNA-seq efforts or spatial profiling could confirm the presence and localization of senescence-related cell states across diverse patients. Fifthly, Other layers of regulation—such as epigenetics, post-translational modifications, or protein-level expression—were not integrated, which might provide a more comprehensive understanding of EDC effects on senescence and BLCA. Last, the risk model based on SRGs was constructed and validated bioinformatically but lacks validation in prospective clinical cohorts. Moreover, it does not incorporate actual EDC exposure levels in patients, which limits direct clinical applicability.

Conclusions

In conclusion, this study proposes a novel SRGs-based scoring signature capable of identifying distinct clusters in BLCA. The risk signature, constructed using six SRGs, provides a robust predictive tool for clinical applications. Previous studies have primarily focused on exploring gene-EDC interactions, with limited investigation into the intricate relationships among EDCs, genes, and BLCA. Our research fills this gap by addressing these interactions. Moreover, we offer a new perspective for future studies, combining single-cell analysis with EDC investigations. This approach facilitates a deeper understanding of the gene-level impacts of EDCs on disease progression and enables the identification of relevant pathways.

Supplementary Information

Supplementary Material 1 (16.8MB, tif)
Supplementary Material 2 (24.5MB, tif)
Supplementary Material 3 (26.6MB, tif)
Supplementary Material 4 (10.1MB, tif)

Acknowledgements

Not applicable.

Author contributions

Maohua Qin: Conceptualization, Data curation, Investigation, Writing – original draft. Guofeng Xie: Writing – original draft, Writing – review. Haimeng Xie: Data curation, Investigation. Baisheng Lin: Writing – original draft, Writing – review & editing. Zili Dai: Data curation, Writing – review & editing. Zijing Chen: Data curation, Formal analysis. Li Wang: Investigation, Writing – review & editing. Jian Zhang: Writing – original draft, Writing – review & editing. Feixiang Wang: Data curation, Investigation. All authors contributed to the article and approved the submitted version.

Funding

This study was supported by Grants from Guangdong and Basic and Applied Basic Research Foundation (2022A1515220011) and Educational Commission of Guangdong Province, China (2023ZDZX2050).

Data availability

All data supporting the conclusions of this article are publicly available. Specifically, transcriptome datasets can be accessed from The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/) and Gene Expression Omnibus (GSE31684, GSE32894, GSE48075 and GSE78220, http://www.ncbi.nlm.nih.gov/geo/), EDC-related targets are provided by the Comparative Toxicogenomics Database (CTD, https://ctdbase.org/), and single-cell sequencing data were obtained from the GEO repository. A curated list of senescence-associated genes (SRGs) was derived from the CellAge database (https://genomics.senescence.info/download.html/).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors have given consent for publication.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Maohua Qin, Guofeng Xie and Haimeng Xie contributed equally to this work.

Contributor Information

Jian Zhang, Email: zhangjian@gzhmu.edu.cn.

Feixiang Wang, Email: wangfx@gzhmu.edu.cn.

References

  • 1.Compérat E, Amin MB, Cathomas R, et al. Current best practice for bladder cancer: a narrative review of diagnostics and treatments. Lancet. 2022;400(10364):1712–21. 10.1016/S0140-6736(22)01188-6. [DOI] [PubMed] [Google Scholar]
  • 2.Babjuk M, Burger M, Capoun O, et al. European Association of Urology Guidelines on Non-muscle-invasive Bladder Cancer (Ta, T1, and Carcinoma in Situ). Eur Urol. 2022;81(1):75–94. 10.1016/j.eururo.2021.08.010. [DOI] [PubMed] [Google Scholar]
  • 3.Kamat AM, Hahn NM, Efstathiou JA, et al. Bladder cancer [published correction appears in Lancet. 2016;388(10061):2742. doi: 10.1016/S0140-6736(16)31776-7.]. Lancet. 2016;388(10061):2796–810. 10.1016/S0140-6736(16)30512-8. [DOI] [PubMed] [Google Scholar]
  • 4.McHugh D, Gil J. Senescence and aging: causes, consequences, and therapeutic avenues. J Cell Biol. 2018;217(1):65–77. 10.1083/jcb.201708092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Martin A, Woolbright BL, Umar S, Ingersoll MA, Taylor JA 3rd. Bladder cancer, inflammageing and microbiomes. Nat Rev Urol. 2022;19(8):495–509. 10.1038/s41585-022-00611-3. [DOI] [PubMed] [Google Scholar]
  • 6.Ahn C, Jeung EB. Endocrine-disrupting chemicals and disease endpoints. Int J Mol Sci. 2023;24(6):5342. 10.3390/ijms24065342. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.International Programme on Chemical Safety. Chapter 1: executive summary. Global assessment on the state of the science of endocrine disruptors. World Health Organization: Geneva, Switzerland,; 2002. p. 1.
  • 8.Attina TM, Hauser R, Sathyanarayana S, et al. Exposure to endocrine-disrupting chemicals in the USA: a population-based disease burden and cost analysis. Lancet Diabetes Endocrinol. 2016;4(12):996–1003. 10.1016/S2213-8587(16)30275-3. [DOI] [PubMed] [Google Scholar]
  • 9.Buha A, Matovic V, Antonijevic B, et al. Overview of cadmium thyroid disrupting effects and mechanisms. Int J Mol Sci. 2018;19(5):1501. 10.3390/ijms19051501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Pellerin E, Caneparo C, Chabaud S, Bolduc S, Pelletier M. Endocrine-disrupting effects of bisphenols on urological cancers. Environ Res. 2021;195:110485. 10.1016/j.envres.2020.110485. [DOI] [PubMed] [Google Scholar]
  • 11.Murata M, Kang JH. Bisphenol A (BPA) and cell signaling pathways. Biotechnol Adv. 2018;36(1):311–27. 10.1016/j.biotechadv.2017.12.002. [DOI] [PubMed] [Google Scholar]
  • 12.Derouiche S, Warnier M, Mariot P, et al. Bisphenol A stimulates human prostate cancer cell migration via remodelling of calcium signalling. Springerplus. 2013;2(1): 54. 10.1186/2193-1801-2-54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Shafei A, Ramzy MM, Hegazy AI, et al. The molecular mechanisms of action of the endocrine disrupting chemical bisphenol A in the development of cancer. Gene. 2018;647:235–43. 10.1016/j.gene.2018.01.016. [DOI] [PubMed] [Google Scholar]
  • 14.Li S, Wang B, Tang Q, Liu J, Yang X. Bisphenol A triggers proliferation and migration of laryngeal squamous cell carcinoma via GPER mediated upregulation of IL-6. Cell Biochem Funct. 2017;35(4):209–16. 10.1002/cbf.3265. [DOI] [PubMed] [Google Scholar]
  • 15.Song P, Fan K, Tian X, Wen J. Bisphenol S (BPS) triggers the migration of human non-small cell lung cancer cells via upregulation of TGF-β. Toxicol In Vitro. 2019;54:224–31. 10.1016/j.tiv.2018.10.005. [DOI] [PubMed] [Google Scholar]
  • 16.Lewis SA. Everything you wanted to know about the bladder epithelium but were afraid to ask. Am J Physiol Renal Physiol. 2000;278(6):F867-74. 10.1152/ajprenal.2000.278.6.F867. [DOI] [PubMed] [Google Scholar]
  • 17.Baneshi M, Tonney-Gagne J, Halilu F, et al. Unpacking phthalates from obscurity in the environment. Molecules. 2023;29(1):106. 10.3390/molecules29010106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Prabhu NB, Adiga D, Kabekkodu SP, Bhat SK, Satyamoorthy K, Rai PS. Bisphenol A exposure modulates reproductive and endocrine system, mitochondrial function and cellular senescence in female adult rats: a hallmarks of polycystic ovarian syndrome phenotype. Environ Toxicol Pharmacol. 2022;96:104010. 10.1016/j.etap.2022.104010. [DOI] [PubMed] [Google Scholar]
  • 19.Shang Y, Chen K, Ni H, et al. Environmentally relevant concentrations of perfluorobutane sulfonate impair locomotion behaviors and healthspan by downregulating mitophagy in C. elegans. J Hazard Mater. 2024;480:135938. 10.1016/j.jhazmat.2024.135938. [DOI] [PubMed] [Google Scholar]
  • 20.Dai Z, Lin B, Cao Y, et al. Melatonin reverses EGFR-TKI therapeutic resistance by modulating crosstalk between circadian-related gene signature and immune infiltration patterns in patients with COVID-19 and lung adenocarcinoma. Comput Biol Med. 2024;180:108937. 10.1016/j.compbiomed.2024.108937. [DOI] [PubMed] [Google Scholar]
  • 21.Ritchie ME, Phipson B, Wu D, et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015. 10.1093/nar/gkv007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572–3. 10.1093/bioinformatics/btq170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33(1):1–22. [PMC free article] [PubMed] [Google Scholar]
  • 24.Blanche P, Dartigues JF, Jacqmin-Gadda H. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Stat Med. 2013;32(30):5381–97. 10.1002/sim.5958. [DOI] [PubMed] [Google Scholar]
  • 25.Wu T, Hu E, Xu S, et al. ClusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation (Camb). 2021;(3): 100141. 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14(1):7. 10.1186/1471-2105-14-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Yoshihara K, Shahmoradgoli M, Martínez E, et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat Commun. 2013;4:2612. 10.1038/ncomms3612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Jiang P, Gu S, Pan D, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med. 2018;24(10):1550–8. 10.1038/s41591-018-0136-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Mayakonda A, Lin DC, Assenov Y, Plass C, Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res. 2018;28(11):1747–56. 10.1101/gr.239244.118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Hao Y, Hao S, Andersen-Nissen E, et al. Integrated analysis of multimodal single-cell data. Cell. 2021;184(13):3573-3587.e29. 10.1016/j.cell.2021.04.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Berman HM, Westbrook J, Feng Z, et al. The protein data bank. Nucleic Acids Res. 2000;28(1):235–42. 10.1093/nar/28.1.235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Zheng X, Liu Z, Zhong J, et al. Downregulation of HINFP induces senescence-associated secretory phenotype to promote metastasis in a non-cell-autonomous manner in bladder cancer. Oncogene. 2022;41(28):3587–98. 10.1038/s41388-022-02371-1. [DOI] [PubMed] [Google Scholar]
  • 33.Lee YC, Lam HM, Rosser C, Theodorescu D, Parks WC, Chan KS. The dynamic roles of the bladder tumour microenvironment. Nat Rev Urol. 2022;19(9):515–33. 10.1038/s41585-022-00608-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Binnewies M, Roberts EW, Kersten K, et al. Understanding the tumor immune microenvironment (TIME) for effective therapy. Nat Med. 2018;24(5):541–50. 10.1038/s41591-018-0014-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Kumari R, Jat P. Mechanisms of cellular senescence: cell cycle arrest and senescence associated secretory phenotype. Front Cell Dev Biol. 2021;9:645593. 10.3389/fcell.2021.645593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Lau L, David G. Pro- and anti-tumorigenic functions of the senescence-associated secretory phenotype. Expert Opin Ther Targets. 2019;23(12):1041–51. 10.1080/14728222.2019.1565658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Eggert T, Wolter K, Ji J, et al. Distinct functions of senescence-associated immune responses in liver tumor surveillance and tumor progression. Cancer Cell. 2016;30(4):533–47. 10.1016/j.ccell.2016.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Zeng Q, Gong Y, Zhu N, Shi Y, Zhang C, Qin L. Lipids and lipid metabolism in cellular senescence: emerging targets for age-related diseases. Ageing Res Rev. 2024;97:102294. 10.1016/j.arr.2024.102294. [DOI] [PubMed] [Google Scholar]
  • 39.Wang Y, Wang J, He J, et al. Comprehensive analysis of PRPF19 immune infiltrates, DNA methylation, senescence-associated secretory phenotype and ceRNA network in bladder cancer. Front Immunol. 2023;14:1289198. 10.3389/fimmu.2023.1289198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Hervé S, Scelfo A, Bersano Marchisio G, et al. Chromosome mis-segregation triggers cell cycle arrest through a mechanosensitive nuclear envelope checkpoint. Nat Cell Biol. 2025;27(1):73–86. 10.1038/s41556-024-01565-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Paffenholz SV, Salvagno C, Ho YJ, et al. Senescence induction dictates response to chemo- and immunotherapy in preclinical models of ovarian cancer. Proc Natl Acad Sci U S A. 2022;119(5):e2117754119. 10.1073/pnas.2117754119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Pobbati AV, Hong W. Emerging roles of TEAD transcription factors and its coactivators in cancers. Cancer Biol Ther. 2013;14(5):390–8. 10.4161/cbt.23788. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Huh HD, Kim DH, Jeong HS, Park HW. Regulation of TEAD transcription factors in cancer biology. Cells. 2019;8(6):600. 10.3390/cells8060600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Wu Y, Zheng Q, Li Y, et al. Metformin targets a YAP1-TEAD4 complex via AMPKα to regulate CCNE1/2 in bladder cancer cells. J Exp Clin Cancer Res. 2019;38(1):376. 10.1186/s13046-019-1346-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Chi M, Liu J, Mei C, et al. TEAD4 functions as a prognostic biomarker and triggers EMT via PI3K/AKT pathway in bladder cancer. J Exp Clin Cancer Res. 2022;41(1):175. 10.1186/s13046-022-02377-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Dai Y, Xu X, Huo X, Faas MM. Effects of polycyclic aromatic hydrocarbons (PAHs) on pregnancy, placenta, and placental trophoblasts. Ecotoxicol Environ Saf. 2023. 10.1016/j.ecoenv.2023.115314. [DOI] [PubMed] [Google Scholar]
  • 47.Sun K, Song Y, He F, Jing M, Tang J, Liu R. A review of human and animals exposure to polycyclic aromatic hydrocarbons: health risk and adverse effects, photo-induced toxicity and regulating effect of microplastics. Sci Total Environ. 2021;773:145403. 10.1016/j.scitotenv.2021.145403. [DOI] [PubMed] [Google Scholar]
  • 48.Martin-Tornero E, Luque-Uría A, Durán-Merás I, Espinosa-Mansilla A. A novel analytical methodology for the determination of hydroxy polycyclic aromatic hydrocarbons in breast and cow milk samples. J Chromatogr B Analyt Technol Biomed Life Sci. 2020;1136:121912. 10.1016/j.jchromb.2019.121912. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (16.8MB, tif)
Supplementary Material 2 (24.5MB, tif)
Supplementary Material 3 (26.6MB, tif)
Supplementary Material 4 (10.1MB, tif)

Data Availability Statement

All data supporting the conclusions of this article are publicly available. Specifically, transcriptome datasets can be accessed from The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/) and Gene Expression Omnibus (GSE31684, GSE32894, GSE48075 and GSE78220, http://www.ncbi.nlm.nih.gov/geo/), EDC-related targets are provided by the Comparative Toxicogenomics Database (CTD, https://ctdbase.org/), and single-cell sequencing data were obtained from the GEO repository. A curated list of senescence-associated genes (SRGs) was derived from the CellAge database (https://genomics.senescence.info/download.html/).

Articles from Discover Oncology are provided here courtesy of Springer

RESOURCES