Multi-omics integration identifies prognostic genes linked to adaptive immunity response and radiosensitivity in cervical cancer

Rui Quan; Jinbao Wang; Guanghui Cheng

doi:10.1016/j.tranon.2026.103029

. 2026 Sep 21;73:103029. doi: 10.1016/j.tranon.2026.103029

Multi-omics integration identifies prognostic genes linked to adaptive immunity response and radiosensitivity in cervical cancer

Rui Quan ¹, Jinbao Wang ^1,¹, Guanghui Cheng ^1,^⁎

PMCID: PMC13629342 PMID: 42767096

Highlights

•
Constructed a two-gene (ITM2A/TCL1A) prognostic model linking adaptive immunity and radiosensitivity in cervical cancer.
•
Identified ITM2A and TCL1A as radiosensitizers; their downregulation promotes radioresistance.
•
Risk groups differ in immune infiltration, chemokine/TCR signaling, and immunotherapy response.
•
Single-cell profiling revealed ITM2A/TCL1A involvement in mast cell differentiation.
•
Model predicts immunotherapy benefit (high-risk) and CDK4/6 inhibitor sensitivity (low-risk).

Keywords: Cervical cancer, Adaptive immune response, Radiosensitivity, Single-cell RNA sequencing

Abstract

Background

Radiotherapy is central to cervical cancer (CC) treatment, yet radiosensitivity remains a key determinant of prognosis. Although adaptive immune response (AIR) influences tumor progression, its role in modulating CC radiosensitivity and the underlying molecular mechanisms remain largely unexplored. This study aims to elucidate the link between AIR and radiosensitivity, and to identify novel therapeutic targets.

Methods

Multi-omics data were integrated to identify AIR-related prognostic genes via differential expression, weighted gene co-expression network analysis (WGCNA), univariate Cox, and least absolute shrinkage and selection operator (LASSO) analyses. A risk model was constructed and validated. Functional enrichment, immune infiltration, and drug sensitivity were assessed. Clinical validation used immunohistochemistry, and a radioresistant cell line was established. The expression levels of the key genes were measured by quantitative reverse transcription-polymerase chain reaction (qRT-PCR) and Western blotting; colony formation assays evaluated radiosensitivity following their knockdown.

Results

Two genes (ITM2A and TCL1A) were identified to construct a prognostic risk model. High-risk groups exhibited distinct immune pathways, reduced immune infiltration, and altered drug sensitivity. Both genes correlated with mast cell differentiation. Notably, ITM2A and TCL1A were downregulated in radioresistant phenotypes, and their knockdown significantly enhanced post-irradiation colony formation versus controls, indicating a radiosensitizing role.

Conclusions

ITM2A and TCL1A are identified as novel AIR-related radiosensitivity genes. Our risk model may serve as a potential prognostic stratification tool for CC, offering insights into combinatorial immunotherapy and radiosensitization strategies.

Introduction

Cervical cancer (CC) ranks as the fourth most common malignancy in women globally, with approximately 600,000 new cases and 340,000 deaths annually [1]. Persistent high-risk human papillomavirus (HPV) infection, particularly types 16 and 18, is the primary cause, responsible for roughly 70% of cases [2,3]. Despite screening advances, CC incidence and mortality remain high, with a disproportionately heavy burden in less-developed regions [[4], [5], [6], [7]]. Radiotherapy is one of the main treatment modalities for cervical cancer, but tumor-intrinsic radioresistance limits its efficacy [8]. Although concurrent chemoradiotherapy (CRT) improves clinical outcomes, a substantial proportion of patients ultimately develop disease progression or recurrence [9]. Therefore, elucidating the mechanisms of radiosensitivity is crucial for prognosis prediction and personalized therapy.

The adaptive immune system evolved to complement the limited recognition capacity of innate immunity [10]. Lymphocytes, including T cells mediating cellular immunity and B cells producing antibodies, are its principal effectors [11]. Adaptive immunity serves as a central determinant in both the development and treatment of CC. Its dysregulation acts as a key pathogenic driver, which is prominently exemplified by HPV's ability to subvert the nuclear factor kappa-B (NF-κB) pathway to evade immune clearance, thereby facilitating persistent infection and subsequent carcinogenesis. This mechanism constitutes the central pathological link between adaptive immunity and cervical cancer initiation [12,13].CC cells further shape an immunosuppressive microenvironment, for instance by generating neutrophils that suppress T-cell function [14]. Additionally, adaptive immunity interacts with CRT in a manner that serves as a critical determinant of clinical response. Neoadjuvant immune checkpoint blockade can enhance anti-tumor T-cell responses and improve efficacy compared to concurrent strategies [15]. Therefore, elucidating the molecular mechanisms underlying the association between adaptive immunity and sensitivity to radiotherapy and chemotherapy in cervical cancer is of great significance for improving patient prognosis.

In recent years, single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool for dissecting the cellular composition of the tumor microenvironment (TME) and revealing intercellular interactions [16,17]. In CC, scRNA-seq has identified key therapy-induced processes, such as innate immune activation and major histocompatibility complex class II upregulation [18], and revealed cellular heterogeneity among epithelial, fibroblast, and CD8⁺ - cell subsets [19].

Therefore, this study integrated bulk RNA-seq and scRNA-seq data from multiple databases, including The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO), and European Molecular Biology Laboratory, to explore various aspects such as metabolic pathway dysregulation, signaling pathway dysregulation, and changes in cell subsets. Additionally, the study identified prognostic genes for CC, investigated the immune characteristics and immunotherapy responsiveness associated with these genes, and validated these findings in the TCGA cohort. Subsequently, the above findings was further validated at the clinical level by constructing prognostic models based on immunohistochemical staining of cervical cancer tissue microarrays. Finally, the reliability of the study’s analytical results was validated by measuring the expression levels of key genes in parental cells and radiation-resistant cells using quantitative reverse transcription-polymerase chain reaction (qRT-PCR) and western blot analysis. This study also provides new insights into the adaptive immune responses underlying CC’s response to radiotherapy and offers novel approaches for the treatment of CC.

Materials and methods

Collecting data

In this study, RNA sequencing, clinical, and survival data of the cervical cancer dataset (TCGA-CESC) were extracted from the TCGA database (https://portal.gdc.cancer.gov/). The TCGA-CESC dataset consisted of 291 tumor specimens and 3 normal cervical tissue samples as controls. First, the 291 CC samples in this dataset were divided at a 7:3 ratio into a TCGA-CESC training set (204 CC samples) and a TCGA-CC validation set (87 CC samples). Second, 47 normal cervical tissue samples were downloaded from the Genotype-Tissue Expression (GTEx) database (https://gtexportal.org/home/) and then merged with the original 3 normal cervical tissue samples in the TCGA-CESC dataset; this combined set served as the control group corresponding to the TCGA-CESC training set (51 samples total). Finally, the batch effect between CC samples and control samples in the TCGA-CESC training set was eliminated using the edgeR package (v3.42.4) [20], yielding the TCGA-CC training set for subsequent analyses, which comprised 204 CC samples and 51 control samples. The GSE56363 dataset (platform: GPL4133) was downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) as a radiochemotherapy sensitivity dataset. This dataset consisted of 12 complete response and 9 non-complete response CC tumor samples. Additionally, the GSE208653 scRNA-seq dataset (platform: GPL24676) comprised 3 CC tumor and 2 normal cervical tissue specimens. In addition, 765 adaptive immune response-related genes (AIRRGs) were downloaded from the Molecular Signatures Database (MSigDB) (https://www.gsea-msigdb.org/gsea/msigdb) (Supplementary Table S1).

Weighted gene co-expression network analysis (WGCNA)

The AIRRG score for CC samples in the TCGA-CC_training set was calculated via single-sample gene set enrichment analysis (ssGSEA) using the GSVA package (v1.53.28) [21]. Patients were divided into high- and low-AIRRG score groups based on an optimal cutoff value determined by the surv_cutpoint function from the survminer package (v0.4.9) [22], followed by Kaplan-Meier (K-M) survival analysis with the log-rank test (P < 0.05) to explore the CC survival association between high- and low-AIRRG score groups.

WGCNA analysis was performed on CC samples from the TCGA-CC_training set using the WGCNA package (v1.73) [23]. First, CC samples were clustered via the hclust function, and outlier samples were inspected for potential removal based on clustering results. Next, the pickSoftThreshold function was applied to determine the soft threshold for the clustered samples, ensuring that gene interactions maximally conformed to a scale-free network distribution. The soft threshold (power) was finally determined based on the screening criteria of a scale-free topology with R² > 0.85 and a connectivity <200. Then, the adjacency matrix between genes was calculated, and it was converted into a topological overlap matrix using the blockwiseModules function to measure the similarity among genes. Furthermore, genes were clustered via the hierarchical clustering method, and co-expression modules were identified using the dynamic pruning method. Following this, the AIRRG score was treated as a trait using the labeledHeatmap function, and Pearson correlation analysis was performed between these scores and the module eigengenes of co-expression modules via the cor function (|correlation (cor)| > 0.3, P < 0.05). Genes in the module with the highest absolute correlation with AIRRG score were selected as adaptive immune response module-related genes (AIRMRGs).

Identification, functional enrichment, and protein-protein interaction (PPI) analysis of cervical cancer differentially expressed radiochemotherapy sensitivity-related adaptive immune response-related genes (CC-DE-RS-AIRRGs)

Radiochemotherapy sensitivity-related differentially expressed genes (RS-DEGs) between complete response and non-complete response (complete response vs. non-complete response) CC samples in the GSE56363 dataset were obtained by means of the limma package (v3.56.2) [24] (P < 0.05 and |log2 fold change (log2FC)| > 0.5). Meanwhile, in the TCGA-CC training set, CC differentially expressed genes (CC-DEGs) between CC and control samples (CC vs. control) were identified using the DESeq2 package (v1.40.2) [25], with the criteria of P 〈 0.05 and |log₂FC| 〉 0.5. The ggplot2 package (v3.5.1) [26] was employed to generate volcano plots, and heatmaps were plotted using the ComplexHeatmap package (v2.16.0) [27]. The top 10 up- and down-regulated RS-DEGs/CC-DEGs were presented separately, ranked in descending order based on log₂FC values. Afterwards, RS-DEGs, CC-DEGs, and AIRMRGs were intersected utilizing the ggvenn package (v0.1.9) (https://CRAN.R-project.org/package=ggvenn) to obtain CC-DE-RS-AIRRGs.

To assess the functions linked to CC-DE-RS-AIRRGs, the clusterProfiler package (v4.10.1) [28] was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses on these genes (P < 0.05).

The PPI retrieval tool of the Search Tool for the Retrieval of Interacting Genes (STRING) database (http://string-db.org) was used to analyze the protein interactions in CC-DE-RS-AIRRGs (confidence score = 0.15), and a PPI network was constructed using the STRINGdb package (v2.12.1) [29] (score_threshold=150).

Identification of prognostic genes

To identify prognostic genes, this study employed patient samples containing survival data from the TCGA-CC training set. First, based on CC-DE-RS-AIRRGs, univariate Cox regression was applied using the coxph function of the survival package (v3.7–0) [30] to screen genes significantly associated with CC survival (hazard ratio (HR) ≠ 1, P < 0.2). Next, the proportional hazards (pH) hypothesis test was performed for genes significantly associated with CC survival via the cox.zph function to identify candidate prognostic genes (P > 0.05). Subsequently, the glmnet package (v4.1–8) [31] was used to perform least absolute shrinkage and selection operator (LASSO) regression analysis on candidate prognostic genes to reduce the risk of overfitting. The parameters were set as nfold = 5. Through cross-validation, the optimal lambda value (lambda.min) was determined, and in the model constructed with the optimal lambda value, genes with non-zero coefficients were regarded as prognostic genes.

Assessment and verification of risk models

In the TCGA-CC training set and TCGA-CC validation set, risk scores for each CC patient were calculated based on prognostic genes using the randomForestSRC package (v3.3.1) [32,33].

To categorize CC patients into high-risk samples (HRS) and low-risk samples (LRS), optimal cut-off values were identified using the surv_cutpoint function. CC patients with survival data in the TCGA-CC training set and TCGA-CC validation set were stratified into HRS and LRS based on these cut-offs, and risk score curves and survival status distribution plots for CC patients were generated using the ggplot2 (v3.5.1) package. The expression profiles of prognostic genes between the HRS and LRS groups were further visualized using heatmaps.

To evaluate the overall survival (OS) difference between HRS and LRS and the predictive ability of the risk model, based on the data of CC patients with complete survival information in TCGA-CC training set and TCGA-CC validation set, first, the K-M survival curves of CC patients between HRS and LRS were plotted using the survival package (v3.7–0) (log-rank test, P < 0.05). Subsequently, the receiver operating characteristic (ROC) curves for 1-year, 3-year, and 5-year OS of CC patients were plotted using the timeROC package (v0.4) [34]. The effectiveness of the risk model was assessed by the area under the curve (AUC) within different time intervals (AUC > 0.6).

To further evaluate the discrimination ability and internal stability of the risk model, 1000 Bootstrap resampling iterations (with replacement, sample size equal to that of the training set) were performed in the TCGA-CC training set. After each resampling, a random survival forest model was refitted based on the prognostic genes, and its concordance index (C-index) was calculated. The histogram and density curve of the 1000 C-index values were plotted using the ggplot2 package, with the mean and 95% confidence interval marked, to assess the predictive robustness of the model.

Clinical features and functional analyses

CC patients in the TCGA-CC training set were subgrouped by clinical features: age (≤ 65 years, > 65 years), stage (stageI, stageII, stageIII, stageIV), M stage (M0, M1), N stage (N0, N1, N2), and T stage (T1, T2, T3, T4). The Wilcoxon test and Kruskal-Wallis test (P < 0.05) were used to compare the significance of differences in risk scores across subgroups.

To understand the disparities in the biological functions of HRS and LRS, the "c2.cp.kegg.v7.4.symbols.gmt" gene set from the MSigDB (http://www.gsea-msigdb.org/gsea/msigdb/index.jsp) was used as the reference. Differential analysis between HRS and LRS in the TCGA-CC training set was performed using the DESeq2 package (v1.40.2), and log₂FC values were calculated from the results and sorted in descending order. Gene Set Enrichment Analysis (GSEA) was conducted via the clusterProfiler package (v4.10.1) with thresholds of |normalized enrichment score (NES)| > 1, false discovery rate (FDR) < 0.25, and P < 0.05. Subsequently, for the TCGA-CC training set, the activity levels of 14 tumor-related pathways were calculated using the PROGENy package (v1.22.0) [35], including androgen, epidermal growth factor receptor, estrogen, hypoxia, Janus kinase-signal transducer and activator of transcription (JAK-STAT), mitogen-activated protein kinase, NF-κB, p53, phosphoinositide 3-kinase, transforming growth factor beta, tumor necrosis factor alpha, TNF-related apoptosis-inducing ligand, vascular endothelial growth factor (VEGF), and Wnt pathways. To ensure the accuracy of the calculation results, the gene expression matrix was first normalized using variance-stabilizing normalization, and then the Wilcoxon test was performed to analyze the differences in the activity levels of the aforementioned pathways between the HRS and LRS groups (P < 0.05).

In order to elucidate the potential functional pathways of prognostic genes in CC, for CC patients with complete survival information in the TCGA-CC training set, the "c2.cp.kegg.v7.4.symbols.gmt" gene set from MSigDB (http://www.gsea-msigdb.org/gsea/msigdb/index.jsp) was used as the reference. Spearman correlation analysis was performed between each prognostic gene and all other genes using the cor function to obtain correlation coefficients, which were then sorted in descending order. GSEA pathway enrichment analysis was conducted via the clusterProfiler package (v4.10.1) based on the ranking results (|NES| > 1, FDR < 0.25, and P < 0.05).

Immune microenvironment and immunotherapy analyses

To investigate the infiltration distribution of immune cells in the HRS and LRS groups, based on the CC patient samples with survival information in the TCGA-CC training set, deconvolution P-values for each sample were first obtained using the CIBERSORT package (v0.1.0) [36] combined with the Monte Carlo sampling method. Samples with P ≥ 0.05 were excluded. For the valid samples after screening, the abundances of 22 types of immune cells [37] in the HRS and LRS groups were further evaluated. Subsequently, the Wilcoxon test was used to explore the differences in infiltration of the 22 cell types between HRS and LRS samples in the TCGA-CC training set (P < 0.05). Cells with marked differences between the HRS and LRS groups were defined as CC-differential cells (CC-DICs). Moreover, in the CC patient samples with survival information from the TCGA-CC training set, Spearman correlation analysis among the CC-DICs was performed using the psych package (v2.4.6.26) [38] (|cor| > 0.3 and P < 0.05).

To evaluate the immunotherapy response of patients in the HRS and LRS groups, the Tumor Immune Dysfunction and Exclusion (TIDE) scores of CC patients in the training set were obtained from the TIDE database (http://tide.dfci.harvard.edu/), and the differences in TIDE scores between the HRS and LRS groups were compared using the Wilcoxon test (P < 0.05). Subsequently, a machine learning scoring model was constructed based on the representative gene set of Immune Phenotype Score (IPS) (https://tcia.at/home), with a score range of 0 to 10, which was used to predict patients’ response to immune checkpoint inhibitors [39]. Furthermore, the Wilcoxon test was performed again to compare the differences in IPS between the HRS and LRS groups (P < 0.05).

Drug sensitivity and drug prediction analyses

To further investigate effective chemotherapy drugs for CC patients, in the samples of patients with survival information in the TCGA-CC training set, 204 chemotherapy drugs for CC were obtained from the Genomics of Drug Sensitivity in Cancer (GDSC) database (https://www.cancerrxgene.org/). The half-maximal inhibitory concentration (IC₅₀) values of the chemotherapy drugs in CC samples were calculated using the pRRophetic package (v 0.5) [40], and differences in IC₅₀ values of the chemotherapy drugs between HRS and LRS groups were compared using the Wilcoxon test (P < 0.05).

To explore the drugs associated with prognostic genes, the drugs associated with prognostic genes were predicted using the Drug-Gene Interaction Database (https://dgidb.org/).

Chromosome localization and regulatory network analyses

To analyze the locations of prognostic genes on human chromosomes, the distribution of prognostic genes across chromosomes was analyzed using the RCircos package (v1.2.2) [41].

To understand the upstream regulatory factors associated with prognostic genes, the microRNAs (miRNAs) targeting prognostic genes were first predicted using the starBase (http://starbase.sysu.edu.cn/index.php) and the miRDB (https://mirdb.org/). The intersection of the miRNAs predicted by the two databases was obtained to identify key miRNAs. Subsequently, the long non-coding RNAs (lncRNAs) targeting key miRNAs were predicted via the StarBase database (http://starbase.sysu.edu.cn/index.php). Finally, the messenger RNA (mRNA)-miRNA-lncRNA regulatory network was constructed using Cytoscape software (v3.10.2) [42].

scRNA-seq analysis

To obtain high-quality single-cell data and identify cell clusters, the Seurat package (v5.1.0) [43] was used for the GSE208653 dataset (CC and normal samples). Seurat objects were created with parameters min.cells = 3 and min.features = 200, followed by filtering cells with percent.mt < 20%, nFeature_RNA (genes per cell) between 200 and 9000, and nCount_RNA (total gene expression per cell) lower than 100,000 to retain high-quality cells.

To screen for highly variable genes (HVGs), first, the filtered single-cell data were batch-corrected using the HarmonyIntegration function. Subsequently, the integrated data were normalized with the LogNormalize function; specifically, the gene expression level of each cell was first normalized by the total expression level of the cell, then multiplied by a scaling factor (default value: 10,000), and finally log-transformed. After normalization was completed, the variance-stabilizing transformation method in the FindVariableFeatures function was applied to screen the genes with the highest expression variability based on the relationship between the mean and variance of gene expression, which were defined as HVGs for subsequent analyses. Finally, the screening results were visualized using the LabelPoints function, and the top 10 genes with the highest variability were labeled in the visualization plot.

For single-cell data dimensionality reduction and clustering, the expression matrix of the aforementioned HVGs was first scaled with the ScaleData function to eliminate biases from differing gene expression measurement units. Principal component analysis (PCA) was then performed on the scaled data using RunPCA, and statistically significant principal components (PCs) (P < 0.05) were screened and retained via the JackStrawPlot function. Based on these PCs, cell subsets were identified sequentially with FindNeighbors and FindClusters (resolution = 0.1), followed by nonlinear dimensionality reduction and clustering visualization of distinct cell clusters using the Uniform Manifold Approximation and Projection (UMAP) algorithm.

To annotate the aforementioned cell clusters into distinct cell types (DCTs), the specifically expressed genes of each cell cluster in the GSE208653 dataset were identified using the FindAllMarkers function (logfc.threshold=0.25, min.pct=0.1, only.pos=TRUE). Combined with the marker genes from the literature [44], the cell clusters were manually annotated as DCTs. Furthermore, the DimPlot function was used to visualize the expression patterns of the marker genes in each DCT. After that, a bar diagram was formed using the ggplot2 package (v3.5.1) to present the proportion and quantity of DCTs in each sample.

In order to evaluate the expression levels of prognostic genes in DCTs, differences in the expression levels of prognostic genes between CC and normal samples within DCTs were compared using the Wilcoxon test (P < 0.05). DCTs with significant differences in prognostic gene expression between the two groups were designated as key cells (KCs).

Cell communication, cell cycle, and pseudotime trajectory analyses

To investigate the interactions between KCs and other DCTs, in the GSE208653 dataset, cell-cell communication analysis was performed on the annotated DCTs using the CellChat package (v1.6.1) [45]. After operations such as CellChat object creation and data preprocessing, the cell communication network was inferred to visualize the interaction count and interaction strength between different cell types. Subsequently, bar charts and heatmaps were also used to visualize the differences in biological signaling pathways between the CC group and the normal group.

To determine the cell cycle stages of the annotated DCTs, the CellCycleScoring function of the Seurat package (v5.1.0) was used to assign cell cycle scores to DCTs, thereby labeling cells in the S, G1, and G2/M phases. Subsequently, the Wilcoxon test was performed to analyze the differences in cell cycle scores between CC samples and normal samples (P < 0.05).

To explore key-cell heterogeneity, PCA was conducted on KCs from the GSE208653 dataset via the RunPCA function, with statistically significant principal components retained for downstream analyses. Small key-cell clusters were identified using the FindNeighbors and FindClusters functions (resolution = 0.1), followed by UMAP-based clustering to generate key-cell clusters. Marker genes of each cluster were detected with the FindAllMarkers function (min.pct = 0.5, logfc.threshold = 0.25, test.use = "auc") and compared against cell-type markers in the CellMarker database (http://117.50.127.228/CellMarker/) to define key-cell subtypes.

To investigate the dynamic changes in gene expression of individual cells during key-cell state transitions, pseudotime trajectory analysis was performed on KCs using the Monocle2 package (v2.28.0) [46]. Subsequently, the DifferentialGeneTest function was used to analyze the kinetic characteristics of gene expression during cell differentiation, thereby identifying the key differentially expressed genes involved in the differentiation process. After that, based on the node division results of the pseudotime trajectory, the pseudotime change trends of all KCs were visualized, and the distribution characteristics of key-cell subtypes at different differentiation stages were displayed. Finally, the orderCells function was used to analyze the expression distribution of prognostic genes in different differentiation states of KCs.

Clinical samples

For this clinical validation component, cervical cancer tissue microarrays were utilized. The tumor tissues originated from the pathological specimens of 55 cervical cancer patients who underwent curative radiotherapy at our center, and this study was approved by the Ethics Committee of China-Japan Union Hospital of Jilin University (Approval No.:2025,052,103, approved on [2025– 05– 21]). The immunohistochemical (IHC) staining of the tissue microarrays was performed by Beijing Zhichao Weiye Biotechnology Co., Ltd. All patients received curative radiotherapy doses as indicated, and the response to radiotherapy was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST). Patients achieving complete response (CR) were classified into the radiosensitive group, while those with non-complete response (NCR) were assigned to the radioresistant group. IHC staining was performed using the TCL1A antibody (Immunoway, USA) and the ITM2A antibody (Immunoway, USA). The interpretation of immunohistochemical staining results was facilitated by Wuhan Servicebio Technology Co., Ltd. The Histochemistry score (H-score) was employed to quantify the expression levels of the target proteins in the cervical cancer tissue microarray. This method converts both the proportion of positively stained areas and the staining intensity within each section into corresponding numerical values, enabling a comprehensive semi-quantitative analysis that accounts for both the extent and intensity of positive immunostaining. The collection and reporting of data from this cervical cancer tissue microarray were approved by the Ethics Committee of China-Japan Union Hospital of Jilin University.

Cell lines and cell culture

The human cervical cancer cell line SiHa was obtained from the Shanghai Institute of Life Sciences, Chinese Academy of Sciences (Shanghai, China) and authenticated by short tandem repeat profiling. Cells were cultured in Dulbecco's modified Eagle medium (ZQXZBIO, China) supplemented with 10% fetal bovine serum (ZQXZBIO, China) and 1% penicillin-streptomycin solution (Hyclone, USA), and maintained at 37 °C in a humidified incubator with 5% CO₂.

Based on parental SiHa cells, the radioresistant cell line R-SiHa was established by a fractionated increasing-dose irradiation protocol, relying solely on irradiation pressure for selection. SiHa cells in the logarithmic growth phase were seeded into 6-cm dishes, and when the confluence reached approximately 70%, they were exposed to X-ray irradiation. The cells were sequentially irradiated with 2, 4, 6, and 8 Gy. Immediately after each irradiation, the culture medium was replaced with fresh complete medium, and the next irradiation was performed once the cells recovered. After completion of the first increasing-dose cycle (2–8 Gy), a second identical cycle of 2, 4, 6, and 8 Gy was repeated, giving a total of two cycles and a cumulative irradiation dose of approximately 40 Gy. The entire screening and culture procedure lasted about 8 weeks. After irradiation, cells that retained proliferative capacity were continuously cultured under standard conditions, and a stably growing radioresistant cell line was obtained approximately 2 weeks later and designated R-SiHa.

Ionizing radiation

Cells were seeded at an appropriate density in cell culture dishes. A 1.5 cm thick tissue compensation pad was placed under the dishes, and the cells were irradiated using a medical linear accelerator (Elekta, Sweden) that delivered 2, 4, 6 and 8MV X-rays at a vertical incidence. The dose rate was set at 200 cGy/min, and the source-to-skin distance was 100 cm, following the same procedure as described above.

Colony formation assay

Cells in the logarithmic growth phase were digested and resuspended, followed by cell counting and dilution to an appropriate density. Subsequently, 400, 800, 1200, 1600 and 2000 cells were seeded into 6-well plates, corresponding to irradiation doses of 0, 2, 4, 6 and 8 Gy, respectively. After complete adherence in a 37 °C incubator, the cells were subjected to different treatment conditions and then continuously cultured in the incubator. Following 14 days of culture, the cells were fixed with methanol on ice and subsequently stained with crystal violet dye (Solarbio, China) at room temperature. Finally, colonies were counted under a microscope, with a cluster of no fewer than 50 cells scored as one colony. Based on the colony counts obtained from three repeated experiments, survival curves were fitted using GraphPad Prism 10 according to the single-hit multi-target model formula:y = 1−(1−exp(−k × x))^N.

Cell viability assay

Cells were seeded into 96-well plates at a density of 3 × 10³ cells per well and cultured overnight. Subsequently, the cells were irradiated with a dose of 6 Gy. At different time points, cell viability was assessed using the Cell Counting Kit-8 (CCK-8) assay kit (Beyotime, China), and the absorbance was measured at 450 nm using a microplate reader (Bio-radimark, USA).

Western blot experiment

Cells were lysed with radioimmunoprecipitation assay buffer(ZQXZBIO, China) containing protease and phosphatase inhibitors (Roche,USA) to extract total cellular proteins. The protein samples from different groups were denatured by boiling and then separated by molecular weight using sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels (Immunoway, USA). Following electrophoresis, the proteins were transferred onto nitrocellulose membranes (Roche, USA) assembled in a sandwich structure with the gels. Membranes were then blocked with 5% skim milk (Beyotime, China). Subsequently, the membranes were incubated overnight at 4 °C with primary antibodies against TCL1A, ITM2A, and β-actin (ImmunoWay, USA). The next day, the membranes were incubated with secondary antibodies (CST, USA) for 1 hour at room temperature. Finally, the protein bands were visualized using a two-color infrared laser imaging system (LI-COR Odyssey, USA) with appropriate parameter settings. All experiments were independently repeated three times.

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

Total RNA was extracted using the Super FastPure Cell RNA Isolation Kit (Vazyme, China), and the RNA concentration was measured. Subsequently, RNA was reverse transcribed into complementary DNA (cDNA) using a reverse transcription kit (TAKARA, Japan). Quantitative PCR was performed using SYBR Green PCR Master Mix (Roche, USA) with specific primers (Supplementary Table S2) and cDNA templates. After centrifugation, quantitative real-time PCR was performed using a real-time PCR system (Eppendorf, Germany) to obtain cycle threshold values for each group. Relative RNA levels were calculated using the 2^−ΔΔCt method, with GAPDH serving as the internal reference gene for normalization.

Small interfering RNA (siRNA) transient transfection

All siRNA sequences were designed and synthesized by Suzhou GenePharma Co., Ltd. (China). Cells were seeded in six-well plates and cultured for 24 h to reach approximately 50% confluence. siRNA transfections were carried out using the siRNA-mate plus transfection reagent (GenePharma, China). Protein expression levels were assessed at 48–96 h post-transfection.

Statistical analysis

All experiments were repeated at least three times, and the experimental values are presented as the mean ± standard deviation.Data were analyzed and processed using R software (v4.3.3) and GraphPad Prism 10. scRNA-seq analysis was performed with the Seurat package (v5.1.0). For intergroup comparisons, the Wilcoxon test was used, and a P < 0.05 was regarded as statistically significant.

Results

Identification of 1668 AIRMRGs in CC

The K-M curve showed that CC patients in the TCGA-CC_training set were divided into high- and low-AIRRG score groups based on the optimal cutoff value (0.972) of the AIRRG score. A significant survival difference was observed between the two groups, with the overall survival (OS) of patients in the low-AIRRG score group being significantly poorer than that in the high-AIRRG score group (P = 0.0076) (Fig. 1a). Then, WGCNA clustering analysis showed that no outlier samples were present in the TCGA-CC_training set (Fig. 1b). The value of β was 5 based on R² > 0.85 (Fig. 1c). The analysis identified 9 co-expression modules (Fig. 1d). Among them, the MEyellow module exhibited the highest positive correlation with the AIRRG score (cor = 0.90, P = 7e-74) (Fig. 1e), and the yellow module, which contained 1668 genes, was designated as AIRMRGs (Supplementary Table S3).

Fig. 1 — **Identification of adaptive immune response module-related genes (AIRMRGs). a:** Survival curves of high- and low-score groups for adaptive immune response-related genes (AIRRGs). b: Sample clustering analysis. c: Soft threshold selection for weighted gene co-expression network analysis (WGCNA) modules. d: Cluster dendrogram of co-expression modules. e: Correlation heatmap between co-expression modules and AIRRGs scores (color intensity indicates correlation strength; darker red represents stronger positive correlation, while darker blue represents stronger negative correlation).

Identification, functional, and PPI of 9 CC-DE-RS-AIRRGs in CC

In total, 3290 RS-DEGs were identified between complete response and non-complete response (complete response vs. non-complete response) CC samples in the GSE56363 dataset. Among them, 1732 RS-DEGs were up-regulated and 1558 RS-DEGs were down-regulated in complete response CC samples (Supplementary Table S4). Additionally, 2903 CC-DEGs were identified between CC and control samples in the TCGA-CC training set. Among them, 1726 CC-DEGs were up-regulated and 1177 CC-DEGs were down-regulated in CC samples (Supplementary Table S5). Volcano plots were used to label the top 10 up/down-regulated RS-DEGs/CC-DEGs ranked by log₂FC (Fig. 2a). In the meantime, a heatmap depicted the expression levels of the top 10 up/down-regulated RS-DEGs/CC-DEGs (Fig. 2b). Subsequently, 9 CC-DE-RS-AIRRGs were identified by intersecting the 3290 RS-DEGs, 2903 CC-DEGs, and 1668 AIRMRGs (Fig. 2c).

Fig. 2 — Identification, function, and protein-protein interaction (PPI) analysis of cervical cancer differentially expressed radiochemotherapy sensitivity-related adaptive immune response-related genes (CC-DE-RS-AIRRGs). a: Volcano plots of radiochemotherapy sensitivity-related differentially expressed genes (RS-DEGs) from the GSE56363 dataset (left) and cervical cancer differentially expressed genes (CC-DEGs) from the TCGA-CC training set (right) (x-axis: log2 fold change [log2FC]; y-axis: -log10[P]). b: Heatmaps of RS-DEGs from the GSE56363 dataset (left) and CC-DEGs from the TCGA-CC training set (right) (x-axis: samples; y-axis: differentially expressed genes; each dot represents the expression level of a gene; darker red indicates higher expression). c: Venn diagram of the intersection among RS-DEGs, CC-DEGs, and adaptive immune response module-related genes (AIRMRGs). **d-e:** Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) [[92], [93], [94]] enrichment analyses of CC-DE-RS-AIRRGs (x-axis: -log10 of the enrichment P value; y-axis: enriched terms; in panel d, colors indicate the Gene Ontology categories Biological Process, Cellular Component, and Molecular Function; in panel e, color indicates Kyoto Encyclopedia of Genes and Genomes pathways). f: PPI interaction network of CC-DE-RS-AIRRGs (circles represent genes; orange nodes indicate interacting genes, while gray nodes indicate isolated genes; line thickness represents interaction strength).

The 9 CC-DE-RS-AIRRGs were involved in 155 Gene Ontology terms (including 114 Biological Process terms, 5 Cellular Component terms, and 36 Molecular Function terms) and 8 KEGG pathways (P < 0.05) (Fig. 2d-e, Supplementary Table S6). Specifically, these GO terms included sphingolipid metabolic process and locomotory behavior in the Biological Process category, external side of plasma membrane and nuclear inner membrane in the Cellular Component category, as well as phosphoric diester hydrolase activity and glycerophospholipid flippase activity in the Molecular Function category. These KEGG pathways included morphine addiction, neuroactive ligand signaling, calcium signaling pathway, and others. Following this, the PPI network consisted of 9 genes and 5 interactions (Fig. 2f). For example, there were interactions between GNG7 and PDE1B, and between TCL1A and ABCB1.

Development and verification of risk models in CC

The 4 genes (TCL1A, ABI3BP, GNG7, and ITM2A) were identified by univariate Cox analysis as being linked to CC survival (HR ≠ 1, P < 0.2) (Fig. 3a). Among these, TCL1A, ABI3BP, GNG7, and ITM2A were protective factors for CC survival (HR < 1). Then, all 4 genes passed the pH hypothesis test as candidate prognostic genes (P > 0.05) (Fig. 3b). After that, 2 prognostic genes (TCL1A and ITM2A) were obtained when lambda.min = 0.02 (Fig. 3c).

Subsequently, CC samples in the TCGA-CC training set (optimal cut-off value = 16.32) and TCGA-CC validation set (optimal cut-off value = 15.49) were stratified into HRS and LRS samples using the optimal cut-off value of risk scores. Among them, the TCGA-CC training set contained 76 high-risk and 128 low-risk samples, while the TCGA-CC validation set included 42 high-risk and 45 low-risk samples.

In the TCGA-CC training set and TCGA-CC validation set, risk curve analyses and survival status distributions demonstrated that the mortality rate in the LRS was markedly lower than that in the HRS, and survival outcomes for the HRS were significantly inferior compared with those of the LRS (Fig. 3d). Concurrently, in the TCGA-CC training set and TCGA-CC validation set, heatmaps depicted that TCL1A and ITM2A were found to be lowly expressed in HRS (Fig. 3e). Thereafter, in both the TCGA-CC training set (P < 0.0001) and TCGA-CC validation set (P = 0.00076), K-M survival analysis revealed that the OS rate of patients in the HRS was markedly lower than that of those in the LRS (Fig. 3f). ROC curve analysis showed that in the TCGA-CC training set and TCGA-CC validation set, the AUC values of this prognostic model for predicting 1-year, 3-year, and 5-year OS of CC patients were all greater than 0.7, suggesting that the model had a certain discriminative ability (Fig. 3g). This observation underscored the model's capacity to prognosticate outcomes for CC patients.

We further performed 1000 Bootstrap resampling iterations in the TCGA-CC training set. After each resampling, a random survival forest model was refitted and the C-index was calculated. The results showed that the mean C-index of the 1000 iterations was 0.809, with a 95% confidence interval of [0.761, 0.859]. The histogram and density curve of the C-index distribution demonstrated that the vast majority of C-index values were concentrated around 0.80, indicating a relatively concentrated overall distribution (Supplementary Fig. S1). These results suggested that the risk model constructed based on ITM2A and TCL1A exhibited good and stable discrimination ability in the internal data, alleviating, to some extent, the potential risk of overfitting resulting from the limited sample size and the lenient threshold used in gene screening.

Clinical feature and functional enrichment in CC

Clinical feature analysis showed that there were significant differences in risk scores among the stage subgroups (P = 0.0088) and T stage subgroups (P = 0.00758) (Fig. 4a). These results indicated that stage and T stage might play potential roles in the prognosis of CC patients.

The GSEA results showed that there was a significant enrichment of 50 pathways (such as T cell receptor signaling pathway and chemokine signaling pathway) in both the HRS and LRS (|NES| > 1, FDR < 0.25, and P < 0.05) (Fig. 4b, Supplementary Table S7). In addition, a significant difference in the tumor-related VEGF pathway was identified between the HRS and LRS groups (P < 0.05) (Fig. 4c). The differential pathway enrichment results between the HRS and LRS groups provided potential reference information for the prognosis assessment of CC patients.

Based on the screening thresholds of P < 0.05, FDR < 0.25, and |NES| > 1, ITM2A and TCL1A were significantly enriched in 35 and 29 pathways, respectively (Fig. 4d, Supplementary Table S8). Among these pathways, 20 were significantly co-enriched in ITM2A and TCL1A, including the T cell receptor signaling pathway and chemokine signaling pathway, among others. The pathway enrichment characteristics of ITM2A and TCL1A were identified as predictive indicators for the efficacy of immunotherapy and targeted therapy.

Immune cells and immunotherapy of risk model

Immunoinfiltration analysis revealed that 8 immune cells exhibited significantly different infiltration levels between HRS and LRS (P < 0.05). Specifically, the infiltration levels of naive B cells, resting dendritic cells, M1 macrophages, and resting mast cells were significantly decreased in HRS compared to LRS, while the infiltration levels of activated dendritic cells, activated mast cells, resting natural killer (NK) cells, and T follicular helper cells were significantly increased (P < 0.05) (Fig. 5a-b). Additionally, a significant negative correlation was detected between activated dendritic cells and M1 macrophages (cor = −0.467, P < 0.001), while a strong positive correlation was found between activated mast cells and resting NK cells (cor = 0.481, P < 0.001) (Fig. 5c, Supplementary Table S9). The differential characteristics of the immune microenvironment and the association rules among immune cells between the HRS and LRS groups of cervical cancer were clearly revealed by these results.

Fig. 5 — **Immune cells and immunotherapy of risk model. a:** Bar plot of immune cell infiltration proportions between high-risk samples (HRS) and low-risk samples (LRS) (x-axis: samples; y-axis: immune cell types; each dot represents the abundance of a cell type; darker red indicates higher abundance). b: Differential analysis of immune cell infiltration between HRS and LRS groups. c: Correlation heatmap of cervical cancer-differential immune cells (CC-DICs) (color indicates correlation strength; darker color represents stronger correlation; red indicates negative correlation, blue indicates positive correlation). d: Differential analysis of Tumor Immune Dysfunction and Exclusion (TIDE) scores between HRS and LRS groups. e: Differential analysis of Immune Phenotype Score (IPS) scores between HRS and LRS groups.

The immunotherapeutic responses uncovered that there were significant differences in the TIDE, ips_ctla4_neg_pd1_pos, and ips_ctla4_pos_pd1_pos scores between the HRS and LRS groups (P < 0.05) (Fig. 5d-e). Notably, TIDE, ips_ctla4_neg_pd1_pos, and ips_ctla4_pos_pd1_pos scores in the HRS were significantly lower than those in the LRS. These results indicated that the efficacy of monotherapy with immune checkpoint inhibitors alone was limited, and the adoption of an immune combination strategy might break through immune tolerance.

Drug of risk model and prognostic genes in CC

The drug sensitivity analysis identified that the IC₅₀ values of 166 drugs exhibited significant differences between HRS and LRS. Among the top 5 chemotherapeutic agents ranked by ascending P-values, BMS, elephantin, PRIMA-1MET, ribociclib, and SB216763 had their IC₅₀ values in the LRS notably lower than those in the HRS (P < 0.05) (Fig. 6, Supplementary Table S10). These results provided direct evidence for the selection of clinical treatment regimens; specifically, the five drugs could be preferentially used as the first-line chemotherapy regimen for patients in the LRS group.

Fig. 6 — **Drugs related to prognostic genes and risk model. a:** Drug sensitivity analysis between high-risk samples (HRS) and low-risk samples (LRS).

In total, 2 drugs associated with the TCL1A were predicted, including estradiol valerate and exemestane. These results suggested that the two drugs could be selectively administered in clinical practice based on patients’ molecular characteristics, including hormone receptor expression levels and TCL1A expression status.

Chromosomal localization and miRNA-lncRNA interaction pairs of 2 prognostic genes in CC

Chromosomal localization analysis showed that ITM2A was mainly located on chromosome X and TCL1A was mainly located on chromosome 14 (Fig. 7a). These results helped to reveal the origin of abnormal gene expression in CC and improve the theoretical system of the molecular regulatory network in CC.

In total, 4 key miRNAs were predicted via the ITM2A (Fig. 7b). For example, ITM2A was regulated by hsa-miR-296– 3p, hsa-miR-3613– 5p, hsa-miR-485– 3p, and hsa-miR-539– 3p Subsequently, a total of 13 lncRNAs were predicted via these 4 key miRNAs. For instance, NEAT1 was found to bind to hsa-miR-296– 3p, hsa-miR-3613– 5p, hsa-miR-485– 3p, and hsa-miR-539– 3p (Fig. 7c). These results indicated that the expression level of the target gene ITM2A could be indirectly modulated by regulating molecules such as hsa-miR-296– 3p, thereby inhibiting the proliferation, invasion, and metastasis abilities of cervical cancer cells. For TCL1A, under the same prediction and intersection screening criteria, no key miRNAs were identified. This was possibly attributed to the lack of conserved miRNA binding sites in the TCL1A transcript or the incomplete annotation of miRNA regulatory relationships for this gene in the current databases.

Identification of mast cells in CC

The GSE208653 dataset included 46,689 cells and 20,601 genes (Supplementary Fig. S2a). After filtering, the dataset retained 42,380 cells and 20,601 genes (Supplementary Fig. S2b). The data were standardized to extract 2000 HVGs (Supplementary Fig. S2c), and 15 cell clusters were obtained by clustering analysis of the top 30 PCs (Supplementary Fig. S2d-f). The 15 cell clusters were categorized into 9 DCTs: endothelial cells, natural killer T cells, B lymphocytes (B cells), plasma cells, mast cells, fibroblasts, myeloid cells, epithelial cells, and neutrophils (Fig. 8a). The violin diagram showed that the marker genes of each DCT had a relatively high level of specific expression (Fig. 8b). The bar diagram displayed a larger proportion and quantity of NK T cells and epithelial cells in each sample (Fig. 8c).

Fig. 8 — **Identification of mast cells. a:** Uniform Manifold Approximation and Projection (UMAP) plot of annotated cell types. b: Violin plot of average expression of marker genes. c: Bar plot of the proportion and count of each cell type in each sample. d: Differential expression analysis of prognostic genes in annotated cell types.

The Wilcoxon test demonstrated that significant differences in the expression level of ITM2A existed between CC and normal samples in mast cells, NK T cells, epithelial cells, and B cells, and significant differences in the expression level of TCL1A were observed between CC and normal samples in mast cells and plasma cells (P < 0.05) (Fig. 8d). The above results showed that both ITM2A and TCL1A were differentially expressed between CC and normal samples in mast cells; therefore, mast cells were regarded as KCs.

Cell communication and cell cycle of mast cells in CC

Cell communication analysis revealed distinct differences in interaction frequency and intensity between the CC group and the normal group: the number of cellular interactions in the normal group (5390) was significantly lower than that in the CC group (6513), and the overall communication intensity of the CC group (0.128) was higher than that of the normal group (0.067) (Fig. 9a). Meanwhile, the communication network diagram also showed that the communication intensity of the CC group was higher than that of the normal group; for instance, an interaction was found between mast cells and mast cells (Fig. 9b). Bar charts showed differences in the information flow of signaling pathways between the CC group and the normal group: information flow in most pathways was more active in the CC group, whereas that in a small number of pathways was more active in the normal group, which reflected the differences in cell communication patterns between the two groups. For example, IL16 and GDF accounted for a larger proportion of information flow and also exhibited stronger information flow in the CC group (Supplementary Fig. S3a). Heatmaps demonstrated that intercellular signaling pathway communication in the CC group was more active and intense, while that in the normal group was relatively moderate. For instance, OX40 signaling was more active and intense in mast cells of the CC group (Supplementary Fig. S3b). The aforementioned results markers that the intensity of cell communication and the activation levels of key signaling pathways could serve as important supplementary indicators for evaluating the prognosis of CC patients.

Fig. 9 — **Cell communication and cell cycle analysis of mast cells. a:** Bar plot of the number and intensity of cell communication in CC and normal groups. b: Network diagrams of the differential interaction number (left) and differential interaction strength (right) between the cervical cancer and normal groups in the GSE208653 single-cell RNA-sequencing dataset. c: Cell cycle scores of different cell types.

Cell cycle analysis showed that all annotated DCTs were distributed across the S, G1, and G2/M phases. Among mast cells, significant differences were observed in the S and G2/M phase scores between the CC group and the normal group, with the scores of the S and G2/M phases being lower in the CC samples (P < 0.05) (Fig. 9c). These results suggested that the proliferative activity of mast cells in the tumor microenvironment was significantly inhibited.

Pseudotime trajectory of mast cells in CC

Based on resolution (0.1) and top 10 PCs, mast cell data were clustered into 3 cell clusters (Supplementary Fig. S4), which were further annotated into 3 mast subtypes, including MAST 1, MAST 2, and MAST 3 (Fig. 10a).

In the pseudotime trajectory, mast cells differentiated from right to left over time, differentiating into 7 different states (Fig. 10b-c). Among them, state 1 was primarily observed in the early stage of mast cell differentiation, while states 6 and 7 were mainly identified in the terminal stage of mast cell differentiation. Meanwhile, all three mast cell subtypes were involved throughout the entire process of cell differentiation (Fig. 10d). Subsequently, during mast cell development, the expression level of ITM2A remained unchanged initially, then increased gradually, decreased slowly afterward, and finally rose slowly again, whereas that of TCL1A remained basically unchanged throughout the process (Fig. 10e). These results suggested that the expression level of ITM2A affected mast cell differentiation, thereby contributing to CC progression.

Correlation of tcl1a and itm2a expression with radiotherapy prognosis in cervical cancer patients

In this study, a tissue microarray was constructed comprising specimens from 55 patients with cervical squamous cell carcinoma who underwent curative radiotherapy at our department between January 2018 and December 2022. All patients were pathologically confirmed as having squamous cell carcinoma and received no neoadjuvant chemotherapy or surgical intervention prior to radiotherapy. Treatment consisted of external beam radiotherapy (with or without concurrent platinum-based chemotherapy) followed by brachytherapy, with a prescribed dose of ≥85 Gy delivered to 90% of the high-risk clinical target volume, achieving a curative dose to the target region. The total treatment duration was controlled within 56 days. One patient was excluded from immunohistochemical analysis due to tissue autolysis; the baseline demographic and clinical characteristics of the remaining 54 patients are summarized in Table 1.

Table 1.

Patient and tumor characteristics.

Characteristics	n(%)
Median age	57.0 (interquartile range: 50.8–64.0)
International Federation of Gynecology and Obstetrics (FIGO) stage (2018)
IB₁	1(1.9%)
IB₂	1(1.9%)
IIA₁	3(5.6%)
IIA₂	2(3.7%)
IIB	10(18.5%)
IIIA	1(1.9%)
IIIB	11(20.3%)
IIIC₁	17(31.4%)
IIIC₂	2(3.7%)
IVA	4(7.4%)
IVB	2(3.7%)
Short-term efficacy
CR	17(31.5%)
NCR	37(68.5%)

Open in a new tab

Abbreviations: CR, complete response; NCR, non-complete response; IQR, interquartile range.

Short-term efficacy was evaluated according to the RECIST 1.1. CR was defined as the disappearance of all target lesions. All other responses, including partial response, progressive disease, and stable disease, were classified as NCR.

A total of 66 tissue cores were obtained from the 55 patients with cervical squamous cell carcinoma. Among these, 55 cores were derived from pre-treatment biopsy specimens, while the remaining 11 cores (C2, C7, D7, D9, E2, E4, E10, F3, F9, G1, G5) were obtained from specimens collected during radiotherapy. One core (B4) exhibited tissue autolysis and was excluded. Consequently, IHC analysis was ultimately performed on 54 pre-treatment cores from 54 patients (Fig. 11a, Fig. 11b).

Fig. 11 — **Analysis of the correlation between TCL1A and ITM2A expression levels and radiotherapy prognosis in patients with cervical squamous cell carcinoma using tissue microarrays. a:** Representative immunohistochemical images showing TCL1A expression in the cervical squamous cell carcinoma tissue microarray. Scale bar = 200 μm. b: Representative immunohistochemical images showing ITM2A expression in the cervical squamous cell carcinoma tissue microarray. Scale bar = 200 μm. c: Histochemistry scores (H-scores) of TCL1A(P = 0.0029) and ITM2A(P = 0.0399) in the complete response (CR) and non-complete response (NCR) groups of patients with cervical squamous cell carcinoma. (*P < 0.05, **P < 0.01).

Based on the RECIST 1.1 criteria, 17 of the 54 patients (31.5%) achieved CR and were assigned to the CR group, while the remaining 37 patients (68.5%) who did not achieve CR were assigned to the NCR group. Further analysis revealed that the pre-treatment H-scores for TCL1A(P < 0.01) and ITM2A(P < 0.05) were significantly higher in the CR group compared to the NCR group (Fig. 11c).These findings suggest that TCL1A expression in cervical cancer is associated with enhanced radiosensitivity and an improved prognosis.

Establishment and validation of a radioresistant cervical cancer cell line

To investigate the association of TCL1A and ITM2A with radiotherapy resistance in CC, the R-SiHa cell line was established from parental SiHa cells using the fractionated dose-escalation irradiation protocol described in the Materials and Methods (Fig. 12a). After cumulative irradiation of approximately 40 Gy and continuous selection, stably growing R-SiHa cells were obtained for subsequent validation of the radioresistant phenotype.

Fig. 12 — **Establishment and validation of a radioresistant cervical cancer cell line. a:** Procedure for establishing the radioresistant subline R-SiHa. b: Clonogenic assay comparing the survival fractions of parental SiHa cells and radioresistant R-SiHa cells after irradiation (6 Gy). c: Cell viability of SiHa and R-SiHa cells measured on days 2 and 3 after 6 Gy irradiation. (*P < 0.05, ***P < 0.001).

Clonogenic assay results demonstrated that under identical irradiation conditions (2, 4, and 6 Gy), the survival fraction of R-SiHa cells was significantly higher than that of the parental SiHa cells (P < 0.001) (Fig. 12b). According to the fitting results of the single-hit multi-target model, a higher D₀ value was associated with a significant decrease in radiosensitivity. As shown in Table 2, the D₀ value of the R-SiHa group was significantly increased, indicating that this cell line had acquired radiation resistance, resulting in higher survival under the same radiation dose.Furthermore, cell viability assays (CCK-8) performed at multiple time points post-irradiation (day 2 and day 3) revealed significantly higher viability in R-SiHa cells compared to SiHa cells (P < 0.05) (Fig. 12c). These results collectively indicate the successful establishment of the R-SiHa cell line, which exhibits a pronounced radioresistant phenotype.

Table 2.

Fitting Results of Cell Survival Curves for the Single-Hit Multi-Target Model.

Group	D₀	N	Dq
SiHa	1.343	2.259	1.100
R-SiHa	1.999	1.502	0.814

Open in a new tab

Abbreviations: D0, mean lethal dose; Dq, quasi-threshold dose; N, extrapolation number; R-SiHa, radioresistant SiHa cells.

Low expression of tcl1a and itm2a proteins in the radioresistant cervical cancer cell line

The aforementioned analyses identified TCL1A and ITM2A as protective factors for CC survival, with low expression in the high-risk group associated with poor prognosis. In this context, R-SiHa can be considered analogous to the high-risk group. Subsequently, we validated these findings at both the protein and mRNA levels. Western blot analysis revealed that the expression levels of TCL1A and ITM2A were lower in R-SiHa cells than in SiHa cells (Fig. 13a). Consistently, qRT-PCR analysis showed that the mRNA levels of TCL1A and ITM2A were significantly lower in R-SiHa cells compared to SiHa cells (P < 0.05) (Fig. 13b). These results are concordant with the trends observed in previous analyses. Therefore, our findings validate that low expression of TCL1A and ITM2A may be associated with radiation resistance in cervical cancer cells.

Fig. 13 — **TCL1A and ITM2A proteins are expressed at low levels in radioresistant cervical cancer cell lines. a:** Western blot analysis was performed to examine the differential expression of TCL1A and ITM2A proteins in SiHa and R-SiHa cells. b: Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was conducted to determine the differences in TCL1A and ITM2A messenger RNA (mRNA) levels between SiHa and R-SiHa cells. (*P < 0.05).

Knockdown of tcl1a and itm2a reduces radiosensitivity of cervical cancer cells

In this study, we investigated the effects of TCL1A and ITM2A knockdown on the radiosensitivity of cervical cancer cells. To ensure adequate knockdown efficiency for subsequent experiments, we performed transient knockdown of TCL1A and ITM2A proteins in SiHa cells using two independent siRNA sequences per gene. The protein-level knockdown efficiencies were confirmed by Western blotting for both siRNAs (Fig. 14a,b). Colony formation assays revealed that at an irradiation dose of 8 Gy, the clonogenic survival rates of SiHa cells in the TCL1A- and ITM2A-knockdown groups were significantly higher than those in the control groups (Fig. 14c,d). These results preliminarily suggested that knockdown of TCL1A or ITM2A may reduce the radiosensitivity of cervical cancer cells.

Fig. 14 — **Knockdown of TCL1A and ITM2A reduces radiosensitivity of cervical cancer cells. a:** Western blot analysis of the knockdown efficiency following transient TCL1A knockdown in SiHa cells. b: Western blot analysis of the knockdown efficiency following transient ITM2A knockdown in SiHa cells. c: Colony formation assay showing the clonogenic survival of SiHa cells after TCL1A knockdown at irradiation doses of 0 Gy and 8 Gy. d: Colony formation assay showing the clonogenic survival of SiHa cells after ITM2A knockdown at irradiation doses of 0 Gy and 8 Gy. (*P < 0.05, **P < 0.01).

Discussion

CC, the fourth most common malignancy in women, is primarily treated with radiotherapy; however, its associated toxicities often worsen prognosis, and current supportive care remains limited [[47], [48], [49], [50], [51]]. Persistent high-risk HPV infection is the predominant etiological factor for cervical cancer and contributes to its pathogenesis by modulating adaptive immunity, particularly by affecting effector T-cell infiltration and regulatory T-cell accumulation. This immune involvement may also influence responses to immune checkpoint inhibitors combined with radiotherapy [15,52,53]. Elucidating the link between radiosensitivity and adaptive immunity is crucial for optimizing treatment. Meanwhile, in recent years, some researchers have attempted to construct prognostic models for cervical cancer from an immunological perspective. For example, one study developed a 15-gene signature using WGCNA and LASSO regression, which was found to be associated with programmed death-ligand 1 expression and immune infiltration [54]. Another study established a 4-gene model based on immune cell infiltration-based grouping and indicated that patients in the low-risk group might be more sensitive to immune checkpoint inhibitors [55]. Additionally, a research team constructed a prognostic model using 29 pairs of immune-related genes [56]. Moreover, the systemic immune-inflammation index, derived from peripheral blood inflammatory markers, has been validated as an independent prognostic factor for cervical cancer [57]. Collectively, these studies have confirmed the value of immune status in the prognosis of cervical cancer from various perspectives. However, most of them focus on global immune infiltration or systemic inflammation levels, with limited integration of two dimensions that directly influence clinical treatment response—namely, adaptive immune responses and radiosensitivity—and lack an in-depth dissection of the functional roles of key genes in specific cell subsets. In contrast, this study innovatively integrates adaptive immunity-related genes with radiosensitivity data. Through multi-omics integration, we constructed a two-gene risk model comprising ITM2A and TCL1A. Using single-cell RNA sequencing, we revealed the dynamic expression characteristics of these two genes during mast cell differentiation. Furthermore, through radioresistant cell lines and clinical samples, we validated their association with the radioresistance phenotype. This work provides a novel addition to the immunoprophylactic research landscape of cervical cancer.

T-cell leukemia/lymphoma 1A (TCL1A) activates the protein kinase B (Akt)/mechanistic target of rapamycin (mTOR) pathway and regulates autophagy by promoting Akt phosphorylation [58]. Concurrently, it enhances T cell receptor (TCR) signaling kinase activity and amplifies transcriptional responses of JAK-STAT or nuclear factor of activated T cells pathways, thereby sensitizing immune signals [59]. Furthermore, its genotype regulates the expression of cytokines and chemokines that bridge innate and adaptive immunity, influencing T and B cell function [60,61]. Intratumoral TCL1A+ B cells are also critical for controlling CC progression [62], suggesting that TCL1A may participate in CC pathogenesis through the modulation of adaptive immunity.

Similarly, integral membrane protein 2A (ITM2A), a target gene of the T cell-regulating transcription factor GATA-3 [63], is preferentially expressed in T-lineage hematopoietic cells. Its expression is induced during positive selection, a major histocompatibility complex-mediated core step in thymic T cell maturation [64,65], indicating deep involvement in adaptive immune regulation. Dysregulation of ITM2A expression is associated with tumors [66]. In CC, its upregulation enhances cisplatin sensitivity via the Notch pathway [67]. Moreover, as a survival-protective factor, its low expression in CC tissues is closely linked to CC development, progression, and prognosis [67].

Notably, the pathological microenvironment of CC is closely associated with adaptive immune dysregulation and aberrant signaling pathways. High expression of programmed cell death protein 1 (PD-1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) in CC tissues can block TCR signaling, leading to T cell exhaustion [68]. Simultaneously, the hypoxic microenvironment and abnormal chemokine expression (e.g., upregulation of CCR4, activation of the CXCL12/CXCR7 axis) drive CC proliferation and metastasis. However, CXCL12 also exhibits tumor-suppressive functions [[69], [70], [71]], suggesting that an imbalance in TCR and chemokine signaling is a crucial potential mechanism for CC progression and immune evasion. Additionally, the status of immune cell infiltration influences CC prognosis. For instance, activated mast cells promote CC invasion and dissemination and correlate with poor prognosis [72,73]. Conversely, replication factor C subunit 4 exerts anti-tumor effects by influencing the immune microenvironment; its high expression negatively correlates with activated mast cell and resting NK cell infiltration and indicates a favorable prognosis [74]. Aligning with these findings, our study identified TCL1A and ITM2A as survival-protective factors in CC, with their low expression in the high-risk group may be associated with poor prognosis.

Our GSEA results revealed that the high- and low-risk groups were significantly enriched in pathways such as the chemokine signaling pathway and the T-cell receptor signaling pathway. The chemokine signaling pathway, a core pathway regulating chemotaxis, immune cell infiltration, and tumor microenvironment remodeling [75], plays a critical dual role in the malignant progression and immune evasion of cervical cancer. Studies have identified, through sequencing analysis of cervical cancer tissues and normal cervical tissues, that various key chemokines and their receptors are aberrantly overexpressed in cancerous tissues, and that their expression levels are positively correlated with tumor International Federation of Gynecology and Obstetrics (FIGO) stage, lymph node metastasis, and lymphovascular invasion [76]. Additionally, the chemokine signaling pathway can accelerate the cell cycle progression of cervical cancer cells, inhibit apoptosis, and mediate tumor angiogenesis by activating classic oncogenic pathways such as JAK-STAT and extracellular signal-regulated kinase, thereby providing microenvironmental support for the malignant proliferation of cervical cancer [77]. Further studies have confirmed that the development of radiotherapy resistance in cervical cancer is highly correlated with the immunosuppressive microenvironment mediated by aberrant chemokine signaling pathway activity [78].

Furthermore, the results of this study demonstrate that the two prognostic genes, ITM2A and TCL1A, are also enriched in the chemokine signaling pathway, and that their expression levels are negatively correlated with the degree of radiotherapy resistance in cervical cancer. This further indicates that the chemokine signaling pathway is an important indicator for prognosis assessment.

Immunotherapy has emerged as a pivotal direction in cancer treatment, and its efficacy in CC patients, along with the identification of benefiting populations, has garnered significant attention. The TIDE score and IPS are key indicators for assessing immunotherapy sensitivity in tumor patients [79]. Furthermore, prognostic indicators based on metabolic genes have stratified CC patients into distinct subtypes, wherein the metabolism-related high-risk group demonstrated poorer overall prognosis alongside lower TIDE scores, indicating potential benefits from immunotherapy for this subgroup [80]. Additionally, studies have found that patients in immune cell-related low-risk groups exhibited higher ips_ctla4_pos_pd1_pos, ips_ctla4_pos_pd1_neg, ips_ctla4_neg_pd1_pos, and ips_ctla4_neg_pd1_neg scores, suggesting that low-risk patients may respond more favorably to immunotherapy [81]. This study further confirms that TIDE scores and certain IPS scores in the high-risk CC group were significantly lower than those in the low-risk group. Taken together with existing evidence, lower TIDE scores and specific IPS subtypes may indicate increased sensitivity to immunotherapy in CC patients, suggesting that patients in the high-risk group in this study could potentially benefit from immunotherapy.

The present study further revealed that the IC50 value of Ribociclib (a cyclin-dependent kinase 4/6 [CDK4/6] inhibitor) was significantly lower in the low-risk group of CC patients compared to the high-risk group. A lower IC50 value indicates higher cellular sensitivity to the drug, suggesting that CC patients in the low-risk group may be more sensitive to Ribociclib treatment. CDK4/6 inhibitors combined with endocrine therapy have become a first-line standard treatment regimen for certain tumors. Among these, Ribociclib, a selective CDK4/6 inhibitor, has demonstrated clear potential in the treatment of CC. Existing studies have confirmed that Ribociclib combined with endocrine therapy significantly improves progression-free survival in patients compared to endocrine therapy alone [82]. Additionally, Ribociclib effectively inhibits the proliferation and induces apoptosis of human CC cells in both in vitro and in vivo experiments, exerting direct anti-tumor effects [83]. Research on Abemaciclib, another CDK4/6 inhibitor, provides further insight into the underlying mechanisms, showing that this drug enhances the sensitivity of HPV-negative CC to chemotherapy by specifically inhibiting the CDK4/6-retinoblastoma protein-E2F (Rb-E2F) and mTOR pathways [84]. This suggests that CDK4/6 inhibitors may exert therapeutic effects by regulating key signaling pathways. Integrating existing research findings, it can be inferred that Ribociclib may exert its therapeutic effects by inhibiting CC cell proliferation and inducing apoptosis, while potentially modulating relevant signaling pathways such as the CDK4/6-Rb-E2F pathway.

Mast cells (MCs), as innate immune cells, have traditionally been recognized for their central roles in allergic and inflammatory responses. They are widely distributed throughout various tissues and can secrete a plethora of vasoactive mediators and pro-inflammatory factors [85]. Recent studies have revealed that MCs are also deeply involved in adaptive immune regulation, coordinating various complex cellular processes, such as promoting the trafficking of immune cells to sites of infection, an effect that can be inhibited by IL-37 [86,87]. Consequently, their role within the TME has garnered increasing attention. In CC, the association between MCs and tumor progression has been substantiated by multiple studies: research has identified a transformative tumor-associated MC subset, C2 ALOX5+ MCs, within CC, which is positioned at a critical juncture of tumor differentiation and directly influences CC progression [88]. Radiotherapy significantly alters the cellular composition and gene expression within the TME of CC patients, notably upregulating MCs, which subsequently participate in multiple cellular axes within the CC ecosystem post-radiotherapy, playing key roles in tumor immunosuppression and stromal remodeling [89]. The present study further elucidates the potential molecular mechanisms and differentiation characteristics underlying the association between MCs and CC. ITM2A and TCL1A exhibited significant differential expression in MCs between CC and normal samples. Compared to normal samples, MCs in CC samples showed significantly reduced S phase and G2/M phase scores, suggesting that the cell cycle progression of MCs may be inhibited in CC.

These findings suggest that MCs may influence the progression of CC through multiple mechanisms, including the secretion of bioactive mediators, the involvement of specific subsets (such as C2 ALOX5⁺ MCs) in differentiation, and the dynamic regulation of ITM2A expression to modulate their functional state. Among these, ITM2A and TCL1A show promise as potential targets for regulating MC function and intervening in the progression of CC.

MCs, as innate immune cells, traditionally participate mainly in allergic and inflammatory responses and can secrete various vasoactive mediators and pro-inflammatory factors [85]. Recent studies further indicate that MCs can also participate in reshaping the adaptive immune microenvironment by regulating immune-cell recruitment and intercellular signaling [86,87]. In cervical cancer, a specific C2 ALOX5+ tumor-associated MC subpopulation may participate in CC progression [88], and radiotherapy can also alter the composition and function of MCs in the tumor microenvironment, allowing them to participate in post-radiotherapy immunosuppression and stromal remodeling [89]. In this study, MCs were identified as key cells showing differential expression of both ITM2A and TCL1A, and altered cell-communication patterns, cell-cycle status, and pseudotime expression of ITM2A were observed in MCs from the cervical cancer group. Taken together, these findings suggest that ITM2A and TCL1A may participate in remodeling the post-radiotherapy immune microenvironment by affecting MCs and their signaling interactions with other immune cells, thereby indirectly influencing the radiosensitivity of cervical cancer. Existing studies also show that TCL1A is closely associated with immune-cell function, and that intratumoral TCL1A+ B cells are associated with better survival outcomes in cervical cancer [62]. ITM2A is also involved in immune-cell development and regulation, as well as treatment-response regulation [90,91], providing a certain biological basis for this potential microenvironment-related mechanism. On the other hand, in this study, knockdown of ITM2A or TCL1A in SiHa epithelial tumor cells markedly increased colony-forming ability after irradiation, and both genes were downregulated in R-SiHa cells, suggesting that they may also directly participate in maintaining the intrinsic radiosensitive phenotype of tumor cells. Therefore, the current results cannot clearly distinguish whether the radiosensitizing effects of ITM2A and TCL1A mainly arise from tumor-cell-intrinsic survival regulation or are driven by MC-mediated changes in the immune microenvironment; the MC-related mechanism is still mainly inferred from single-cell association results and previous studies. Future studies should further clarify the relative contributions and potential connections between these two mechanisms through MC-specific gene intervention, co-culture of MCs with cervical cancer cells, conditioned-medium treatment, and in vivo experiments under radiotherapy conditions.

Certainly, it must be acknowledged that this study has several limitations. First, the downstream signalling pathways of ITM2A and TCL1A and their functional coupling with adaptive immunity remain to be elucidated. Second, the association with mast cells is only descriptive and lacks direct functional validation, and its generalisability to other subtypes or broader populations is uncertain. Third, although internal bootstrap validation yielded favourable discrimination, the relatively small sample sizes of the training and validation datasets and the necessarily lenient screening thresholds may still introduce false positives and overfitting. To address these issues, we plan to perform mechanistic studies (e.g., co-immunoprecipitation and mast cell co-culture) to clarify the underlying molecular events, conduct in vivo experiments to verify the biological relevance, and carry out external validation using multi-centre cohorts and prospectively collected clinical data to assess the model's generalisability and clinical utility.

Conclusions

This study identified two prognostic genes (ITM2A and TCL1A) associated with radiotherapy sensitivity and adaptive immune response in CC using public databases. Based on these prognostic genes, risk models related to radiosensitivity and adaptive immune response were constructed and validated for the first time to predict the prognosis of CC patients treated with radiotherapy, aiming to improve prognostic outcomes and inform clinical treatment strategies. However, it must be acknowledged that the underlying mechanisms involving these key genes remain to be elucidated. We will continue to investigate these key genes through further cellular and in vivo experiments to validate their functions and mechanisms in cervical cancer.

Data availability

The datasets analysed during the current study are available in the Gene Expression Omnibus (GEO) repository at https://www.ncbi.nlm.nih.gov/geo/, reference numbers GSE56363, GSE208653; and The Cancer Genome Atlas (TCGA) repository at https://portal.gdc.cancer.gov/, reference numbers TCGA-CC and TCGA-CESC; and the Genotype-Tissue Expression (GTEx) database at https://gtexportal.org/home/.

Funding statement

This work was partially supported by grants from the Project of Science and Technology Department of Jilin Province (grant number 20230203082SF); Horizontal Project of Jilin University [grant numbers 2019YX435, 2019155, 2023YX0056].

Ethics declarations

The study was conducted in accordance with the Declaration of Helsinki, and approved by the ethics committee in China-Japan Union Hospital of Jilin University(Approval No.:2025052103, approved on [2025-05-21]). Informed consents have been obtained from the patients.

Consent to participate/Consent to publish

All patients provided written informed consent at the time of clinical sample collection for experiments, ensuring that the research process complied with ethical norms and that patients' rights and wishes were fully respected.

CRediT authorship contribution statement

Rui Quan: Writing – original draft, Visualization, Software, Methodology, Investigation, Formal analysis, Data curation. Jinbao Wang: Writing – review & editing, Validation, Supervision, Formal analysis. Guanghui Cheng: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Funding acquisition, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We sincerely thank the Scientific Research Center of China-Japan Union Hospital of Jilin University for its support and assistance throughout this study. We also thank the other staff from the Departments of Radiation Oncology and Nursing for their valuable help during the completion of this work.

Footnotes

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.tranon.2026.103029.

Supplementary Fig. S1. Bootstrap internal validation results

Supplementary Fig. S2. Quality control, dimensionality reduction, and clustering analysis of single-cell data. a-b: nFeature_RNA, nCount_RNA, and percent.mt of single-cell data before and after quality control. c: Screening of highly variable genes (HVGs). d: JackStraw analysis. e: Scree plot of principal component analysis (PCA) linear dimensionality reduction across dimensions. f: Uniform Manifold Approximation and Projection (UMAP) plot of cell clustering.

Supplementary Fig. S3. a: Bar plot of the relative information flow proportion and intensity between CC samples and normal samples. b: Differences in signal patterns of each cell type between CC samples and normal samples (rows: cell types; columns: signaling factors; each value represents the communication signal intensity of the signaling factor in the corresponding cell type.)

Supplementary Fig. S4. Uniform Manifold Approximation and Projection (UMAP) dimensionality reduction and clustering analysis of mast cells.

Supplementary Fig. S5. Unedited Western blot image.

Supplementary Table S1. 765 adaptive immune response-related genes (AIRRGs).

Supplementary Table S2. Primers for quantitative reverse transcription-polymerase chain reaction (qRT-PCR), small interfering RNA (siRNA) sequences, and antibody information.

Supplementary Table S3. 1,668 adaptive immune response module-related genes (AIRMRGs).

Supplementary Table S4. 3,290 radiochemotherapy sensitivity-related differentially expressed genes (RS-DEGs).

Supplementary Table S5. 2,903 cervical cancer differentially expressed genes (CC-DEGs).

Supplementary Table S6. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional analyses of cervical cancer differentially expressed radiochemotherapy sensitivity-related adaptive immune response-related genes (CC-DE-RS-AIRRGs).

Supplementary Table S7. Gene set enrichment analysis (GSEA) pathways between high-risk samples (HRS) and low-risk samples (LRS).

Supplementary Table S8. Gene set enrichment analysis (GSEA) pathways of prognostic genes.

Supplementary Table S9. Correlation coefficients among cervical cancer-differential immune cells (CC-DICs).

Supplementary Table S10. Drug of risk model.

Contributor Information

Rui Quan, Email: quanrui24@mails.jlu.edu.cn.

Jinbao Wang, Email: jinbao@jlu.edu.cn.

Guanghui Cheng, Email: chenggh@jlu.edu.cn.

Appendix. Supplementary materials

mmc1.jpg^{(548.4KB, jpg)}

mmc2.jpg^{(1.6MB, jpg)}

mmc3.jpg^{(1.4MB, jpg)}

mmc4.jpg^{(546.4KB, jpg)}

mmc5.pdf^{(586.9KB, pdf)}

mmc6.xlsx^{(18KB, xlsx)}

mmc7.docx^{(17.8KB, docx)}

mmc8.xlsx^{(29.9KB, xlsx)}

mmc9.xlsx^{(349.1KB, xlsx)}

mmc10.xlsx^{(338KB, xlsx)}

mmc11.xlsx^{(24.7KB, xlsx)}

mmc12.xlsx^{(28.1KB, xlsx)}

mmc13.xlsx^{(20.7KB, xlsx)}

mmc14.xlsx^{(10.1KB, xlsx)}

mmc15.xlsx^{(26.5KB, xlsx)}

References

1.Anusha T., Brahman P.K., Sesharamsingh B., Lakshmi A., Bhavani K.S. Electrochemical detection of cervical cancer biomarkers. Clin. Chim. Acta. 2025;567 doi: 10.1016/j.cca.2024.120103. [DOI] [PubMed] [Google Scholar]
2.Molina Mariano A, Steenbergen Renske D M, Pumpe Anna, Kenyon Angelique N, Melchers Willem JG. HPV integration and cervical cancer: a failed evolutionary viral trait. Trends Mol. Med. 2024;30:890–902. doi: 10.1016/j.molmed.2024.05.009. [DOI] [PubMed] [Google Scholar]
3.Arbyn Marc, Xu Lan, Simoens Cindy, Martin-Hirsch Pierre Pl. Prophylactic vaccination against Human papillomaviruses to prevent cervical cancer and its precursors. Cochrane Database Syst. Rev. 2018;2020 doi: 10.1002/14651858.CD009069.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Wentzensen N., Arbyn M. HPV-based cervical cancer screening- facts, fiction, and misperceptions. Prev. Med. 2017;98:33–35. doi: 10.1016/j.ypmed.2016.12.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Bardakos N.V. CORR Insights®: is cemented or cementless femoral stem fixation more durable in patients older than 75 years of age? A comparison of the best-performing stems. Clin. Orthop. Relat. Res. 2018;476:1438–1440. doi: 10.1097/CORR.0000000000000353. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Small W., Bacon M.A., Bajaj A., Chuang L.T., Fisher B.J., Harkenrider M.M., Jhingran A., Kitchener H.C., Mileshkin L.R., Viswanathan A.N., et al. Cervical cancer: a global health crisis. Cancer. 2017;123:2404–2412. doi: 10.1002/cncr.30667. [DOI] [PubMed] [Google Scholar]
7.Bedell S.L., Goldstein L.S., Goldstein A.R., Goldstein A.T. Cervical cancer screening: past, present, and future. Sex. Med. Rev. 2020;8:28–37. doi: 10.1016/j.sxmr.2019.09.005. [DOI] [PubMed] [Google Scholar]
8.Li X., Liu H., Zhou X., Zhou Y., Zhang Y., Liou Y.-L., Zeng M., Zhu H. PAX1 Hypomethylation as a prognostic biomarker for radioresistance of cervical cancer. Clin. Epigenetics. 2023;15:123. doi: 10.1186/s13148-023-01538-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Wang Yanhong, Ouyang Yi, Cao Xinping, Cai Qunrong. Identifying hub genes for chemo-radiotherapy sensitivity in cervical cancer: a Bi-dataset in Silico analysis. Discov. Oncol. 2024;15 doi: 10.1007/s12672-024-01328-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Cooper M.D., Alder M.N. The evolution of adaptive immune systems. Cell. 2006;124:815–822. doi: 10.1016/j.cell.2006.02.001. [DOI] [PubMed] [Google Scholar]
11.Bonilla F.A., Oettgen H.C. Adaptive Immunity. J. Allergy Clin. Immunol. 2010;125:S33–S40. doi: 10.1016/j.jaci.2009.09.017. [DOI] [PubMed] [Google Scholar]
12.Tilborghs S., Corthouts J., Verhoeven Y., Arias D., Rolfo C., Trinh X.B., van Dam P.A. The role of nuclear factor-kappa B signaling in Human cervical cancer. Crit. Rev. Oncol. Hematol. 2017;120:141–150. doi: 10.1016/j.critrevonc.2017.11.001. [DOI] [PubMed] [Google Scholar]
13.Walboomers J.M., Jacobs M.V., Manos M.M., Bosch F.X., Kummer J.A., Shah K.V., Snijders P.J., Peto J., Meijer C.J., Muñoz N. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. 1999;189:12–19. doi: 10.1002/(SICI)1096-9896(199909)189:1<12::AID-ATH431>3.0.CO;2-F. [DOI] [PubMed] [Google Scholar]
14.Lecointre M., Guillot J., Marcone R., Ozdoganlar D., Cayatte M., Jaensson Gyllenbäck E., Liberg D., Fournier N., Homicsko K., Hanahan D. HPV16-Expressing tumors release multiple IL1 ligands to orchestrate systemic immunosuppression whose disruption enables efficacy of a therapeutic vaccine. Cancer Discov. 2025;15:1458–1483. doi: 10.1158/2159-8290.CD-25-0382. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Mayadev J., Zamarin D., Deng W., Lankes H.A., Pesci G., Kim H., Chino J.P., Banbury B., Sherry N., Sharon E., et al. Neoadjuvant or concurrent atezolizumab with chemoradiation for locally advanced cervical cancer: a randomized phase I trial. Nat. Commun. 2025;16:553. doi: 10.1038/s41467-024-55200-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Wagner J., Rapsomaniki M.A., Chevrier S., Anzeneder T., Langwieder C., Dykgers A., Rees M., Ramaswamy A., Muenst S., Soysal S.D., et al. A single-cell atlas of the tumor and immune ecosystem of Human breast cancer. Cell. 2019;177:1330–1345.e18. doi: 10.1016/j.cell.2019.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Olalekan S., Xie B., Back R., Eckart H., Basu A. Characterizing the tumor microenvironment of metastatic ovarian cancer by single-cell transcriptomics. Cell. Rep. 2021;35 doi: 10.1016/j.celrep.2021.109165. [DOI] [PubMed] [Google Scholar]
18.Liu C., Li X., Huang Q., Zhang M., Lei T., Wang F., Zou W., Huang R., Hu X., Wang C., et al. Single-cell RNA-sequencing reveals radiochemotherapy-induced innate immune activation and MHC-II upregulation in cervical cancer. Signal. Transduct. Target. Ther. 2023;8:44. doi: 10.1038/s41392-022-01264-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Li C., Wu H., Guo L., Liu D., Yang S., Li S., Hua K. Single-cell transcriptomics reveals cellular heterogeneity and molecular stratification of cervical cancer. Commun. Biol. 2022;5:1208. doi: 10.1038/s42003-022-04142-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Chen Y., Lun A.T.L., Smyth G.K. From reads to genes to pathways: differential expression analysis of RNA-seq experiments using rsubread and the edgeR quasi-likelihood pipeline. F1000Res. 2016;5:1438. doi: 10.12688/f1000research.8987.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Hänzelmann Sonja, Castelo Robert, Guinney Justin. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinform. 2013;14 doi: 10.1186/1471-2105-14-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Zhou Jing, Guo Li, Wang Yuzhen, Li Lina, Guo Yahuan, Duan Lian, Jiao Mi, Xi Pan, Wang Pei. Development and validation of a risk prognostic model based on the H. Pylori infection phenotype for stomach adenocarcinoma. Heliyon. 2024;10 doi: 10.1016/j.heliyon.2024.e36882. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Langfelder Peter, Horvath Steve. WGCNA: an R package for weighted correlation network analysis. BMC Bioinform. 2008;9 doi: 10.1186/1471-2105-9-559. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Ritchie Matthew E, Phipson Belinda, Wu Di, Hu Yifang, Law Charity W, Shi Wei, Smyth Gordon K. Limma Powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43 doi: 10.1093/nar/gkv007. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Gustavsson Emil K, Zhang David, Reynolds Regina H, Garcia-Ruiz Sonia, Ryten Mina. Ggtranscript: an R package for the visualization and interpretation of transcript isoforms using Ggplot2. Bioinformatics. 2022;38:3844–3846. doi: 10.1093/bioinformatics/btac409. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Gu Zuguang. Complex heatmap visualization. iMeta. 2022;1 doi: 10.1002/imt2.43. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Wu T., Hu E., Xu S., Chen M., Guo P., Dai Z., Feng T., Zhou L., Tang W., Zhan L., et al. Clusterprofiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021;2 doi: 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Szklarczyk Damian, Gable Annika L, Nastou Katerina C, Lyon David, Kirsch Rebecca, Pyysalo Sampo, Doncheva Nadezhda T, Legeay Marc, Fang Tao, Bork Peer, Jensen Lars J, von Mering Christian. The STRING Database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res. 2021;49:D605–D612. doi: 10.1093/nar/gkaa1074. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Zhang Zhongheng. Semi-parametric regression model for Survival data: graphical visualization with R. Ann. Transl. Med. 2016;4:461. doi: 10.21037/atm.2016.08.61. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Friedman Jerome, Hastie Trevor, Tibshirani Rob. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 2010;33 [PMC free article] [PubMed] [Google Scholar]
32.Zhao Pengfei, Zhen Hongchao, Zhao Hong, Huang Yongjie, Cao Bangwei. Identification of hub genes and potential molecular mechanisms related to radiotherapy sensitivity in rectal cancer based on multiple datasets. J. Transl. Med. 2023;21 doi: 10.1186/s12967-023-04029-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Lu Mengjiao, Yu Xialei, Hu Jingyan, Wang Jiajing, Wang Taozuo. Cytotoxic T-lymphocytes in acute myeloid leukemia: monitoring prognosis and guiding treatment choice. J. Gene Med. 2024;26 doi: 10.1002/jgm.3587. [DOI] [PubMed] [Google Scholar]
34.Blanche Paul, Dartigues Jean-François, Jacqmin-Gadda Hélène. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Stat. Med. 2013;32:5381–5397. doi: 10.1002/sim.5958. [DOI] [PubMed] [Google Scholar]
35.Schubert Michael, Klinger Bertram, Klünemann Martina, Sieber Anja, Uhlitz Florian, Sauer Sascha, Garnett Mathew J, Blüthgen Nils, Saez-Rodriguez Julio. Perturbation-response genes reveal signaling footprints in cancer gene expression. Nat. Commun. 2018;9 doi: 10.1038/s41467-017-02391-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Profiling tumor infiltrating immune cells with CIBERSORT. Methods Mol. Biol. 2018;1711 doi: 10.1007/978-1-4939-7493-1_12. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Li Weihang, Ding Ziyi, Zhang Huan, Shi Quan, Wang Dong, Zhang Shilei, Xu Songjie, Gao Bo, Yan Ming. The roles of blood lipid-metabolism genes in immune infiltration could promote the development of IDD. Front. Cell. Dev. Biol. 2022;10 doi: 10.3389/fcell.2022.844395. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Robles-Jimenez L.E., Aranda-Aguirre E., Castelan-Ortega O.A., Shettino-Bermudez B.S., Ortiz-Salinas R., Miranda M., Li X., Angeles-Hernandez J.C., Vargas-Bello-Pérez E., Gonzalez-Ronquillo M. Worldwide traceability of antibiotic residues from livestock in wastewater and soil: a systematic review. Animals. 2021;12:60. doi: 10.3390/ani12010060. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Lai J., Lin X., Zheng H., Xie B., Fu D. Characterization of stemness features and construction of a stemness subtype classifier to predict survival and treatment responses in lung squamous cell carcinoma. BMC Cancer. 2023;23:525. doi: 10.1186/s12885-023-10918-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Geeleher P., Cox N., Huang R.S. pRRophetic: An R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS One. 2014;9 doi: 10.1371/journal.pone.0107468. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.An J., Lai J., Sajjanhar A., Batra J., Wang C., Nelson C.C. J-circos: an interactive circos plotter. Bioinformatics. 2015;31:1463–1465. doi: 10.1093/bioinformatics/btu842. [DOI] [PubMed] [Google Scholar]
42.Franz M., Lopes C.T., Fong D., Kucera M., Cheung M., Siper M.C., Huck G., Dong Y., Sumer O., Bader G.D. Cytoscape.Js 2023 update: a graph theory library for visualization and analysis. Bioinformatics. 2023;39 doi: 10.1093/bioinformatics/btad031. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Satija R., Farrell J.A., Gennert D., Schier A.F., Regev A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 2015;33:495–502. doi: 10.1038/nbt.3192. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Guo C., Qu X., Tang X., Song Y., Wang J., Hua K., Qiu J. Spatiotemporally deciphering the mysterious mechanism of persistent HPV-induced malignant transition and immune remodelling from HPV-infected normal cervix, precancer to cervical cancer: integrating single-cell RNA-sequencing and spatial transcriptome. Clin. Transl. Med. 2023;13 doi: 10.1002/ctm2.1219. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Jin S., Guerrero-Juarez C.F., Zhang L., Chang I., Ramos R., Kuan C.-H., Myung P., Plikus M.V., Nie Q. Inference and analysis of cell-cell communication using CellChat. Nat. Commun. 2021;12:1088. doi: 10.1038/s41467-021-21246-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Fang Z., Li J., Cao F., Li F. Integration of scRNA-seq and bulk RNA-seq reveals molecular characterization of the immune microenvironment in acute pancreatitis. Biomolecules. 2022;13:78. doi: 10.3390/biom13010078. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Abu-Rustum N.R., Yashar C.M., Arend R., Barber E., Bradley K., Brooks R., Campos S.M., Chino J., Chon H.S., Crispens M.A., et al. NCCN Guidelines® insights: cervical Cancer, version 1.2024. J. Natl. Compr. Canc Netw. 2023;21:1224–1233. doi: 10.6004/jnccn.2023.0062. [DOI] [PubMed] [Google Scholar]
48.Schubert M., Bauerschlag D.O., Muallem M.Z., Maass N. Alkatout, I. Challenges in the diagnosis and individualized treatment of cervical cancer. Med. (Kaunas) 2023;59:925. doi: 10.3390/medicina59050925. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Burmeister C.A., Khan S.F., Schäfer G., Mbatani N., Adams T., Moodley J., Prince S. Cervical cancer therapies: current challenges and future perspectives. Tumour Virus Res. 2022;13 doi: 10.1016/j.tvr.2022.200238. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Puspitasari I.M., Legianawati D., Sinuraya R.K., Suwantika A.A. Cost-effectiveness analysis of chemoradiation and radiotherapy treatment for stage IIB and IIIB cervical cancer patients. Int. J. Womens Health. 2021;Volume 13:221–229. doi: 10.2147/IJWH.S289781. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Palagudi M., Para S., Golla N., Meduri K.C., Duvvuri S.P., Vityala Y., Sajja D.C., Damineni U. Adverse effects of cancer treatment in patients with cervical cancer. Cureus. 2024;16 doi: 10.7759/cureus.54106. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Litwin Tamara R, Irvin Sarah R, Chornock Rebecca L, Sahasrabuddhe Vikrant V, Stanley Margaret, Wentzensen Nicolas. Infiltrating T-cell markers in cervical carcinogenesis: a systematic review and meta-analysis. Br. J. Cancer. 2021;124:831–841. doi: 10.1038/s41416-020-01184-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Marcucci Fabrizio, Rumio Cristiano. The tumor-promoting effects of the adaptive immune system: a cause of hyperprogressive disease in cancer? Cell. Mol. Life Sci. 2F: CMLS. 2021;78:853–865. doi: 10.1007/s00018-020-03606-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Su R., Jin C., Bu H., Xiang J., Zhou L., Jin C. Development and validation of an immune-related prognostic signature in cervical cancer. Front. Oncol. 2022;12 doi: 10.3389/fonc.2022.861392. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Mei J., Xing Y., Lv J., Gu D., Pan J., Zhang Y., Liu J. Construction of an immune-related gene signature for prediction of prognosis in patients with cervical cancer. Int. Immunopharmacol. 2020;88 doi: 10.1016/j.intimp.2020.106882. [DOI] [PubMed] [Google Scholar]
56.Nie H., Bu F., Xu J., Li T., Huang J. 29 Immune-related genes pairs signature predict the prognosis of cervical cancer patients. Sci. Rep. 2020;10 doi: 10.1038/s41598-020-70500-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Huang H., Liu Q., Zhu L., Zhang Y., Lu X., Wu Y., Liu L. Prognostic value of preoperative systemic immune-inflammation index in patients with cervical cancer. Sci. Rep. 2019;9:3284. doi: 10.1038/s41598-019-39150-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Correction to: TCL1A acts as a tumour suppressor by modulating gastric cancer autophagy via miR-181a-5p-TCL1A-akt/mTOR-c-MYC loop. Carcinogenesis. 2023;44:535. doi: 10.1093/carcin/bgad035. [DOI] [PubMed] [Google Scholar]
59.Oberbeck S., Schrader A., Warner K., Jungherz D., Crispatzu G., von Jan J., Chmielewski M., Ianevski A., Diebner H.H., Mayer P., Ados A Kondo, Wahnschaffe L., Braun T., Müller T.A., Wagle P., Bouska A., Neumann T., Pützer S., Varghese L., Pflug N., Thelen M., Makalowski J., Riet N., Göx H J M, Rappl G., Altmüller J., Kotrová M., Persigehl T., Hopfinger G., Hansmann M L, Schlößer H., Stilgenbauer S., Dürig J., Mougiakakos D., von Bergwelt-Baildon M., Roeder I., Hartmann S., Hallek M., Moriggl R., Brüggemann M., Aittokallio T., Iqbal J., Newrzela S., Abken H., Herling M. Noncanonical effector functions of the T-memory-like T-PLL cell are shaped by cooperative TCL1A and TCR signaling. Blood. 2020;136:2786–2802. doi: 10.1182/blood.2019003348. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Ho M.-F., Ingle J.N., Bongartz T., Kalari K.R., Goss P.E., Shepherd L.E., Mushiroda T., Kubo M., Wang L., Weinshilboum R.M. TCL1A Single-nucleotide polymorphisms and estrogen-mediated toll-like receptor-MYD88-dependent nuclear factor-κb activation: single-nucleotide polymorphism- and selective estrogen receptor modulator-dependent modification of inflammation and immune response. Mol. Pharmacol. 2017;92:175–184. doi: 10.1124/mol.117.108340. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Ad L. The role of chemokines in linking innate and adaptive immunity. Curr. Opin. Immunol. 2002;14:129–135. doi: 10.1016/s0952-7915(01)00308-9. [DOI] [PubMed] [Google Scholar]
62.Punt S., Corver W.E., van der Zeeuw S.A.J., Kielbasa S.M., Osse E.M., Buermans H.P.J., de Kroon C.D., Jordanova E.S., Gorter A. Whole-transcriptome analysis of flow-sorted cervical cancer samples reveals that B cell expressed TCL1A is correlated with improved survival. Oncotarget. 2015;6:38681–38694. doi: 10.18632/oncotarget.4526. [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Tai T.-S., Pai S.-Y., Ho I.-C. Itm2a, a target gene of GATA-3, plays a minimal role in regulating the development and function of T cells. PLoS One. 2014;9 doi: 10.1371/journal.pone.0096535. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Spashett R., Fernie G., Reid I.C., Cameron I.M. MADRS symptom subtypes in ECT-treated depressed patients: relationship to response and subsequent ECT. J. ECT. 2014;30:227–231. doi: 10.1097/YCT.0000000000000091. [DOI] [PubMed] [Google Scholar]
65.Kisielow Pawel, Teh Hung Sia, Blüthmann Horst, von Boehmer Harald. Positive selection of antigen-specific T cells in Thymus by restricting MHC molecules. Nature. 1988;335:730–733. doi: 10.1038/335730a0. [DOI] [PubMed] [Google Scholar]
66.Park J.-S., Jeon E.K., Chun S.H., Won H.S., Lee A., Hur S.Y., Lee K.H., Bae S.-N., Yoon S.-C., Hong S.H. ERCC1 (Excision Repair Cross-Complementation Group 1) Expression as a predictor for response of neoadjuvant chemotherapy for FIGO stage 2B uterine cervix cancer. Gynecol. Oncol. 2011;120:275–279. doi: 10.1016/j.ygyno.2010.10.034. [DOI] [PubMed] [Google Scholar]
67.Bonifazi Gianluca, Lista Luca, Menasce Dario, Mezzetto Mauro, Pedrini Daniele, Spighi Roberto, Zoccoli Antonio. Study on the effects of the restrictive measures for containment of the COVID-19 pandemic on the reproduction number [Formula: see Text] in Italian regions. Eur. Phys. J. Plus. 2021;136 doi: 10.1140/epjp/s13360-021-02076-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Saglam O., Conejo-Garcia J. PD-1/PD-L1 immune checkpoint inhibitors in advanced cervical cancer. Integr. Cancer Sci. Ther. 2018;5 doi: 10.15761/ICST.1000272. [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Liu L.-B., Xie F., Chang K.-K., Shang W.-Q., Meng Y.-H., Yu J.-J., Li H., Sun Q., Yuan M.-M., Jin L.-P., et al. Chemokine CCL17 induced by hypoxia promotes the proliferation of cervical cancer cell. Am. J. Cancer Res. 2015;5:3072–3084. [PMC free article] [PubMed] [Google Scholar]
70.The CXCL12/CXCR7 signalling axis promotes proliferation and metastasis in cervical cancer. Med. Oncol. 2021;38 doi: 10.1007/s12032-021-01481-2. [DOI] [PubMed] [Google Scholar]
71.Yadav Suresh Singh, Prasad Shyam Babu, Prasad Chandra Bhushan, Pandey Lakshmi Kant, Pradhan Satyajit, Singh Sunita, Narayan Gopeshwar. CXCL12 Is a key regulator in tumor microenvironment of cervical cancer: an in vitro study. Clin. Exp. Metastasis. 2016;33:431–439. doi: 10.1007/s10585-016-9787-9. [DOI] [PubMed] [Google Scholar]
72.Rudolph M Isolde, Boza Yadira, Yefi Roger, Luza Sandra, Andrews Edilia, Penissi Alicia, Garrido Pablo, Rojas I Gina. The influence of mast cell mediators on migration of SW756 cervical carcinoma cells. J. Pharmacol. Sci. 2008;106:208–218. doi: 10.1254/jphs.fp0070736. [DOI] [PubMed] [Google Scholar]
73.Wang J., Li Z., Gao A., Wen Q., Sun Y. The prognostic landscape of tumor-infiltrating immune cells in cervical cancer. Biomed. Pharmacother. 2019;120 doi: 10.1016/j.biopha.2019.109444. [DOI] [PubMed] [Google Scholar]
74.Huang Bifen, Zheng Jianqing, Chen Bizhen, Wu Min, Xiao Lihua. Analysis of the correlation between RFC4 expression and tumor immune microenvironment and prognosis in patients with cervical cancer. Front. Genet. 2025;16 doi: 10.3389/fgene.2025.1514383. [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Sayegh Hiba Al, Assi Zahraa, Kanaan Amjad, El Sibai Mirvat. CXCL1 in triple‑negative breast cancer: mechanisms, challenges, and therapeutic opportunities (Review) Oncol. Rep. 2026;55:1–14. doi: 10.3892/or.2026.9091. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Song G., Luo J., Zou S., Lou F., Zhang T., Zhu X., Yang J., Wang X. Molecular classification of Human papillomavirus-positive cervical cancers based on immune signature enrichment. Front. Public. Health. 2022;10 doi: 10.3389/fpubh.2022.979933. [DOI] [PMC free article] [PubMed] [Google Scholar]
77.Wang L., Wang L., Cao Y., Wang Q., Sun X. A chemokine-based prognostic model featuring CXCL8, ITGA5, BACE2, CCR7, CERS4, and MEI1 for cervical cancer. Eur. J. Med. Res. 2025;31:123. doi: 10.1186/s40001-025-03688-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
78.Bao Qiufang, Tang Wenlu, Tang Wenzhong, Zhu Xiaohong, Hong Linliang, Zhang Bin. Identification of immune-related cervical cancer prognostic biomarkers and construction of prognostic model based on tumor microenvironment. Eur. J. Med. Res. 2025;30 doi: 10.1186/s40001-025-02515-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Chen Qian, Deng Dongfeng, Zhu Hong, Li Shan. Integrative analysis of bulk multiomics and single-cell transcriptomics elucidates age-related molecular landscape alterations in cervical cancer. Ann. Surg. Oncol. 2025;32:7000–7013. doi: 10.1245/s10434-025-17464-w. [DOI] [PubMed] [Google Scholar]
80.Ma Boyi, Ren Chenlu, Yin Yadong, Zhao Shuhua, Li Jia, Yang Hong. Immune cell infiltration and prognostic index in cervical cancer: insights from metabolism-related differential genes. Front. Immunol. 2024;15 doi: 10.3389/fimmu.2024.1411132. [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Wang N., Nanding A., Jia X., Wang Y., Yang C., Fan J., Dong A., Zheng G., Ma J., Shi X., et al. Mining of immunological and prognostic-related biomarker for cervical cancer based on immune cell signatures. Front. Immunol. 2022;13 doi: 10.3389/fimmu.2022.993118. [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Ramírez Stieben L.A., Vargas M.C., Polillo D.C., Lufft K., Saldías P.R., Bedini I. Metastasis of breast cancer to the thyroid gland. Med. (B Aires) 2024;84:741–745. [PubMed] [Google Scholar]
83.Xiong Y., Li T., Assani G., Ling H., Zhou Q., Zeng Y., Zhou F., Zhou Y. Ribociclib, a selective cyclin D kinase 4/6 inhibitor, inhibits proliferation and induces apoptosis of Human cervical cancer in Vitro and in Vivo. Biomed. Pharmacother. 2019;112 doi: 10.1016/j.biopha.2019.108602. [DOI] [PubMed] [Google Scholar]
84.Liu Yuan, Zhao Runsheng, Fang Shanshan, Li Quan, Jin Yiqiang, Liu Bo. Abemaciclib sensitizes HPV-negative cervical cancer to chemotherapy via specifically suppressing CDK4/6-Rb-E2F and mTOR pathways. Fundam. Clin. Pharmacol. 2021;35:156–164. doi: 10.1111/fcp.12574. [DOI] [PubMed] [Google Scholar]
85.Theoharides T.C., Valent P., Akin C. Mast cells, mastocytosis, and related disorders. N Engl. J. Med. 2015;373:163–172. doi: 10.1056/NEJMra1409760. [DOI] [PubMed] [Google Scholar]
86.E C.Y., Cai Y., Sun B.Z., Guan L.Y., Jiang T. Hepatic insulin-like growth factor receptor is upregulated by activation of the GSK3B-FOXO3 pathway after partial hepatectomy. J. Biol. Regul. Homeost. Agents. 2017;31:549–555. [PubMed] [Google Scholar]
87.Jain R., Tikoo S., Weninger W. Mast cell granules: modulating adaptive immune response remotely. J. Allergy Clin. Immunol. 2019;143:1731–1733. doi: 10.1016/j.jaci.2018.11.029. [DOI] [PubMed] [Google Scholar]
88.Zhao F., Hong J., Zhou G., Huang T., Lin Z., Zhang Y., Liang L., Tang H. Elucidating the role of tumor-associated ALOX5+ mast cells with transformative function in cervical cancer progression via single-cell RNA sequencing. Front. Immunol. 2024;15 doi: 10.3389/fimmu.2024.1434450. [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Gu Meijia, He Ti, Yuan Yuncong, Duan Suling, Li Xin, Shen Chao. Single-cell RNA sequencing revealed changes in the tumor microenvironment induced by radiotherapy for cervical cancer and the molecular mechanism of mast cells in immunosuppression. Funct. Integr. Genom. 2025;25 doi: 10.1007/s10142-025-01564-7. [DOI] [PubMed] [Google Scholar]
90.Li Y., Wang J., Gao C., Hu Q., Mao X. Integral membrane protein 2A enhances sensitivity to chemotherapy via notch signaling pathway in cervical cancer. Bioengineered. 2021;12:10183–10193. doi: 10.1080/21655979.2021.2001218. [DOI] [PMC free article] [PubMed] [Google Scholar]
91.Kirchner J., Bevan M.J. ITM2A Is induced during thymocyte selection and T cell activation and causes downregulation of CD8 when overexpressed in CD4(+)CD8(+) double positive thymocytes. J. Exp. Med. 1999;190:217–228. doi: 10.1084/jem.190.2.217. [DOI] [PMC free article] [PubMed] [Google Scholar]
92.Kanehisa M., Furumichi M., Sato Y., Matsuura Y., Ishiguro-Watanabe M. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53:D672–D677. doi: 10.1093/nar/gkae909. [DOI] [PMC free article] [PubMed] [Google Scholar]
93.Kanehisa M., Goto S. KEGG: kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 2000;28:27–30. doi: 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
94.Kanehisa M. Toward understanding the origin and evolution of cellular organisms. Protein Sci. 2019;28:1947–1951. doi: 10.1002/pro.3715. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

mmc1.jpg^{(548.4KB, jpg)}

mmc2.jpg^{(1.6MB, jpg)}

mmc3.jpg^{(1.4MB, jpg)}

mmc4.jpg^{(546.4KB, jpg)}

mmc5.pdf^{(586.9KB, pdf)}

mmc6.xlsx^{(18KB, xlsx)}

mmc7.docx^{(17.8KB, docx)}

mmc8.xlsx^{(29.9KB, xlsx)}

mmc9.xlsx^{(349.1KB, xlsx)}

mmc10.xlsx^{(338KB, xlsx)}

mmc11.xlsx^{(24.7KB, xlsx)}

mmc12.xlsx^{(28.1KB, xlsx)}

mmc13.xlsx^{(20.7KB, xlsx)}

mmc14.xlsx^{(10.1KB, xlsx)}

mmc15.xlsx^{(26.5KB, xlsx)}

Data Availability Statement

[bib0001] 1.Anusha T., Brahman P.K., Sesharamsingh B., Lakshmi A., Bhavani K.S. Electrochemical detection of cervical cancer biomarkers. Clin. Chim. Acta. 2025;567 doi: 10.1016/j.cca.2024.120103. [DOI] [PubMed] [Google Scholar]

[bib0002] 2.Molina Mariano A, Steenbergen Renske D M, Pumpe Anna, Kenyon Angelique N, Melchers Willem JG. HPV integration and cervical cancer: a failed evolutionary viral trait. Trends Mol. Med. 2024;30:890–902. doi: 10.1016/j.molmed.2024.05.009. [DOI] [PubMed] [Google Scholar]

[bib0003] 3.Arbyn Marc, Xu Lan, Simoens Cindy, Martin-Hirsch Pierre Pl. Prophylactic vaccination against Human papillomaviruses to prevent cervical cancer and its precursors. Cochrane Database Syst. Rev. 2018;2020 doi: 10.1002/14651858.CD009069.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0004] 4.Wentzensen N., Arbyn M. HPV-based cervical cancer screening- facts, fiction, and misperceptions. Prev. Med. 2017;98:33–35. doi: 10.1016/j.ypmed.2016.12.040. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0005] 5.Bardakos N.V. CORR Insights®: is cemented or cementless femoral stem fixation more durable in patients older than 75 years of age? A comparison of the best-performing stems. Clin. Orthop. Relat. Res. 2018;476:1438–1440. doi: 10.1097/CORR.0000000000000353. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0006] 6.Small W., Bacon M.A., Bajaj A., Chuang L.T., Fisher B.J., Harkenrider M.M., Jhingran A., Kitchener H.C., Mileshkin L.R., Viswanathan A.N., et al. Cervical cancer: a global health crisis. Cancer. 2017;123:2404–2412. doi: 10.1002/cncr.30667. [DOI] [PubMed] [Google Scholar]

[bib0007] 7.Bedell S.L., Goldstein L.S., Goldstein A.R., Goldstein A.T. Cervical cancer screening: past, present, and future. Sex. Med. Rev. 2020;8:28–37. doi: 10.1016/j.sxmr.2019.09.005. [DOI] [PubMed] [Google Scholar]

[bib0008] 8.Li X., Liu H., Zhou X., Zhou Y., Zhang Y., Liou Y.-L., Zeng M., Zhu H. PAX1 Hypomethylation as a prognostic biomarker for radioresistance of cervical cancer. Clin. Epigenetics. 2023;15:123. doi: 10.1186/s13148-023-01538-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0009] 9.Wang Yanhong, Ouyang Yi, Cao Xinping, Cai Qunrong. Identifying hub genes for chemo-radiotherapy sensitivity in cervical cancer: a Bi-dataset in Silico analysis. Discov. Oncol. 2024;15 doi: 10.1007/s12672-024-01328-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0010] 10.Cooper M.D., Alder M.N. The evolution of adaptive immune systems. Cell. 2006;124:815–822. doi: 10.1016/j.cell.2006.02.001. [DOI] [PubMed] [Google Scholar]

[bib0011] 11.Bonilla F.A., Oettgen H.C. Adaptive Immunity. J. Allergy Clin. Immunol. 2010;125:S33–S40. doi: 10.1016/j.jaci.2009.09.017. [DOI] [PubMed] [Google Scholar]

[bib0012] 12.Tilborghs S., Corthouts J., Verhoeven Y., Arias D., Rolfo C., Trinh X.B., van Dam P.A. The role of nuclear factor-kappa B signaling in Human cervical cancer. Crit. Rev. Oncol. Hematol. 2017;120:141–150. doi: 10.1016/j.critrevonc.2017.11.001. [DOI] [PubMed] [Google Scholar]

[bib0013] 13.Walboomers J.M., Jacobs M.V., Manos M.M., Bosch F.X., Kummer J.A., Shah K.V., Snijders P.J., Peto J., Meijer C.J., Muñoz N. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. 1999;189:12–19. doi: 10.1002/(SICI)1096-9896(199909)189:1<12::AID-ATH431>3.0.CO;2-F. [DOI] [PubMed] [Google Scholar]

[bib0014] 14.Lecointre M., Guillot J., Marcone R., Ozdoganlar D., Cayatte M., Jaensson Gyllenbäck E., Liberg D., Fournier N., Homicsko K., Hanahan D. HPV16-Expressing tumors release multiple IL1 ligands to orchestrate systemic immunosuppression whose disruption enables efficacy of a therapeutic vaccine. Cancer Discov. 2025;15:1458–1483. doi: 10.1158/2159-8290.CD-25-0382. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0015] 15.Mayadev J., Zamarin D., Deng W., Lankes H.A., Pesci G., Kim H., Chino J.P., Banbury B., Sherry N., Sharon E., et al. Neoadjuvant or concurrent atezolizumab with chemoradiation for locally advanced cervical cancer: a randomized phase I trial. Nat. Commun. 2025;16:553. doi: 10.1038/s41467-024-55200-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0016] 16.Wagner J., Rapsomaniki M.A., Chevrier S., Anzeneder T., Langwieder C., Dykgers A., Rees M., Ramaswamy A., Muenst S., Soysal S.D., et al. A single-cell atlas of the tumor and immune ecosystem of Human breast cancer. Cell. 2019;177:1330–1345.e18. doi: 10.1016/j.cell.2019.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0017] 17.Olalekan S., Xie B., Back R., Eckart H., Basu A. Characterizing the tumor microenvironment of metastatic ovarian cancer by single-cell transcriptomics. Cell. Rep. 2021;35 doi: 10.1016/j.celrep.2021.109165. [DOI] [PubMed] [Google Scholar]

[bib0018] 18.Liu C., Li X., Huang Q., Zhang M., Lei T., Wang F., Zou W., Huang R., Hu X., Wang C., et al. Single-cell RNA-sequencing reveals radiochemotherapy-induced innate immune activation and MHC-II upregulation in cervical cancer. Signal. Transduct. Target. Ther. 2023;8:44. doi: 10.1038/s41392-022-01264-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0019] 19.Li C., Wu H., Guo L., Liu D., Yang S., Li S., Hua K. Single-cell transcriptomics reveals cellular heterogeneity and molecular stratification of cervical cancer. Commun. Biol. 2022;5:1208. doi: 10.1038/s42003-022-04142-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0020] 20.Chen Y., Lun A.T.L., Smyth G.K. From reads to genes to pathways: differential expression analysis of RNA-seq experiments using rsubread and the edgeR quasi-likelihood pipeline. F1000Res. 2016;5:1438. doi: 10.12688/f1000research.8987.2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0021] 21.Hänzelmann Sonja, Castelo Robert, Guinney Justin. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinform. 2013;14 doi: 10.1186/1471-2105-14-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0022] 22.Zhou Jing, Guo Li, Wang Yuzhen, Li Lina, Guo Yahuan, Duan Lian, Jiao Mi, Xi Pan, Wang Pei. Development and validation of a risk prognostic model based on the H. Pylori infection phenotype for stomach adenocarcinoma. Heliyon. 2024;10 doi: 10.1016/j.heliyon.2024.e36882. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0023] 23.Langfelder Peter, Horvath Steve. WGCNA: an R package for weighted correlation network analysis. BMC Bioinform. 2008;9 doi: 10.1186/1471-2105-9-559. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0024] 24.Ritchie Matthew E, Phipson Belinda, Wu Di, Hu Yifang, Law Charity W, Shi Wei, Smyth Gordon K. Limma Powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43 doi: 10.1093/nar/gkv007. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0025] 25.Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0026] 26.Gustavsson Emil K, Zhang David, Reynolds Regina H, Garcia-Ruiz Sonia, Ryten Mina. Ggtranscript: an R package for the visualization and interpretation of transcript isoforms using Ggplot2. Bioinformatics. 2022;38:3844–3846. doi: 10.1093/bioinformatics/btac409. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0027] 27.Gu Zuguang. Complex heatmap visualization. iMeta. 2022;1 doi: 10.1002/imt2.43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0028] 28.Wu T., Hu E., Xu S., Chen M., Guo P., Dai Z., Feng T., Zhou L., Tang W., Zhan L., et al. Clusterprofiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021;2 doi: 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0029] 29.Szklarczyk Damian, Gable Annika L, Nastou Katerina C, Lyon David, Kirsch Rebecca, Pyysalo Sampo, Doncheva Nadezhda T, Legeay Marc, Fang Tao, Bork Peer, Jensen Lars J, von Mering Christian. The STRING Database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res. 2021;49:D605–D612. doi: 10.1093/nar/gkaa1074. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0030] 30.Zhang Zhongheng. Semi-parametric regression model for Survival data: graphical visualization with R. Ann. Transl. Med. 2016;4:461. doi: 10.21037/atm.2016.08.61. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0031] 31.Friedman Jerome, Hastie Trevor, Tibshirani Rob. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 2010;33 [PMC free article] [PubMed] [Google Scholar]

[bib0032] 32.Zhao Pengfei, Zhen Hongchao, Zhao Hong, Huang Yongjie, Cao Bangwei. Identification of hub genes and potential molecular mechanisms related to radiotherapy sensitivity in rectal cancer based on multiple datasets. J. Transl. Med. 2023;21 doi: 10.1186/s12967-023-04029-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0033] 33.Lu Mengjiao, Yu Xialei, Hu Jingyan, Wang Jiajing, Wang Taozuo. Cytotoxic T-lymphocytes in acute myeloid leukemia: monitoring prognosis and guiding treatment choice. J. Gene Med. 2024;26 doi: 10.1002/jgm.3587. [DOI] [PubMed] [Google Scholar]

[bib0034] 34.Blanche Paul, Dartigues Jean-François, Jacqmin-Gadda Hélène. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Stat. Med. 2013;32:5381–5397. doi: 10.1002/sim.5958. [DOI] [PubMed] [Google Scholar]

[bib0035] 35.Schubert Michael, Klinger Bertram, Klünemann Martina, Sieber Anja, Uhlitz Florian, Sauer Sascha, Garnett Mathew J, Blüthgen Nils, Saez-Rodriguez Julio. Perturbation-response genes reveal signaling footprints in cancer gene expression. Nat. Commun. 2018;9 doi: 10.1038/s41467-017-02391-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0036] 36.Profiling tumor infiltrating immune cells with CIBERSORT. Methods Mol. Biol. 2018;1711 doi: 10.1007/978-1-4939-7493-1_12. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0037] 37.Li Weihang, Ding Ziyi, Zhang Huan, Shi Quan, Wang Dong, Zhang Shilei, Xu Songjie, Gao Bo, Yan Ming. The roles of blood lipid-metabolism genes in immune infiltration could promote the development of IDD. Front. Cell. Dev. Biol. 2022;10 doi: 10.3389/fcell.2022.844395. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0038] 38.Robles-Jimenez L.E., Aranda-Aguirre E., Castelan-Ortega O.A., Shettino-Bermudez B.S., Ortiz-Salinas R., Miranda M., Li X., Angeles-Hernandez J.C., Vargas-Bello-Pérez E., Gonzalez-Ronquillo M. Worldwide traceability of antibiotic residues from livestock in wastewater and soil: a systematic review. Animals. 2021;12:60. doi: 10.3390/ani12010060. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0039] 39.Lai J., Lin X., Zheng H., Xie B., Fu D. Characterization of stemness features and construction of a stemness subtype classifier to predict survival and treatment responses in lung squamous cell carcinoma. BMC Cancer. 2023;23:525. doi: 10.1186/s12885-023-10918-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0040] 40.Geeleher P., Cox N., Huang R.S. pRRophetic: An R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS One. 2014;9 doi: 10.1371/journal.pone.0107468. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0041] 41.An J., Lai J., Sajjanhar A., Batra J., Wang C., Nelson C.C. J-circos: an interactive circos plotter. Bioinformatics. 2015;31:1463–1465. doi: 10.1093/bioinformatics/btu842. [DOI] [PubMed] [Google Scholar]

[bib0042] 42.Franz M., Lopes C.T., Fong D., Kucera M., Cheung M., Siper M.C., Huck G., Dong Y., Sumer O., Bader G.D. Cytoscape.Js 2023 update: a graph theory library for visualization and analysis. Bioinformatics. 2023;39 doi: 10.1093/bioinformatics/btad031. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0043] 43.Satija R., Farrell J.A., Gennert D., Schier A.F., Regev A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 2015;33:495–502. doi: 10.1038/nbt.3192. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0044] 44.Guo C., Qu X., Tang X., Song Y., Wang J., Hua K., Qiu J. Spatiotemporally deciphering the mysterious mechanism of persistent HPV-induced malignant transition and immune remodelling from HPV-infected normal cervix, precancer to cervical cancer: integrating single-cell RNA-sequencing and spatial transcriptome. Clin. Transl. Med. 2023;13 doi: 10.1002/ctm2.1219. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0045] 45.Jin S., Guerrero-Juarez C.F., Zhang L., Chang I., Ramos R., Kuan C.-H., Myung P., Plikus M.V., Nie Q. Inference and analysis of cell-cell communication using CellChat. Nat. Commun. 2021;12:1088. doi: 10.1038/s41467-021-21246-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0046] 46.Fang Z., Li J., Cao F., Li F. Integration of scRNA-seq and bulk RNA-seq reveals molecular characterization of the immune microenvironment in acute pancreatitis. Biomolecules. 2022;13:78. doi: 10.3390/biom13010078. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0047] 47.Abu-Rustum N.R., Yashar C.M., Arend R., Barber E., Bradley K., Brooks R., Campos S.M., Chino J., Chon H.S., Crispens M.A., et al. NCCN Guidelines® insights: cervical Cancer, version 1.2024. J. Natl. Compr. Canc Netw. 2023;21:1224–1233. doi: 10.6004/jnccn.2023.0062. [DOI] [PubMed] [Google Scholar]

[bib0048] 48.Schubert M., Bauerschlag D.O., Muallem M.Z., Maass N. Alkatout, I. Challenges in the diagnosis and individualized treatment of cervical cancer. Med. (Kaunas) 2023;59:925. doi: 10.3390/medicina59050925. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0049] 49.Burmeister C.A., Khan S.F., Schäfer G., Mbatani N., Adams T., Moodley J., Prince S. Cervical cancer therapies: current challenges and future perspectives. Tumour Virus Res. 2022;13 doi: 10.1016/j.tvr.2022.200238. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0050] 50.Puspitasari I.M., Legianawati D., Sinuraya R.K., Suwantika A.A. Cost-effectiveness analysis of chemoradiation and radiotherapy treatment for stage IIB and IIIB cervical cancer patients. Int. J. Womens Health. 2021;Volume 13:221–229. doi: 10.2147/IJWH.S289781. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0051] 51.Palagudi M., Para S., Golla N., Meduri K.C., Duvvuri S.P., Vityala Y., Sajja D.C., Damineni U. Adverse effects of cancer treatment in patients with cervical cancer. Cureus. 2024;16 doi: 10.7759/cureus.54106. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0052] 52.Litwin Tamara R, Irvin Sarah R, Chornock Rebecca L, Sahasrabuddhe Vikrant V, Stanley Margaret, Wentzensen Nicolas. Infiltrating T-cell markers in cervical carcinogenesis: a systematic review and meta-analysis. Br. J. Cancer. 2021;124:831–841. doi: 10.1038/s41416-020-01184-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0053] 53.Marcucci Fabrizio, Rumio Cristiano. The tumor-promoting effects of the adaptive immune system: a cause of hyperprogressive disease in cancer? Cell. Mol. Life Sci. 2F: CMLS. 2021;78:853–865. doi: 10.1007/s00018-020-03606-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0054] 54.Su R., Jin C., Bu H., Xiang J., Zhou L., Jin C. Development and validation of an immune-related prognostic signature in cervical cancer. Front. Oncol. 2022;12 doi: 10.3389/fonc.2022.861392. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0055] 55.Mei J., Xing Y., Lv J., Gu D., Pan J., Zhang Y., Liu J. Construction of an immune-related gene signature for prediction of prognosis in patients with cervical cancer. Int. Immunopharmacol. 2020;88 doi: 10.1016/j.intimp.2020.106882. [DOI] [PubMed] [Google Scholar]

[bib0056] 56.Nie H., Bu F., Xu J., Li T., Huang J. 29 Immune-related genes pairs signature predict the prognosis of cervical cancer patients. Sci. Rep. 2020;10 doi: 10.1038/s41598-020-70500-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0057] 57.Huang H., Liu Q., Zhu L., Zhang Y., Lu X., Wu Y., Liu L. Prognostic value of preoperative systemic immune-inflammation index in patients with cervical cancer. Sci. Rep. 2019;9:3284. doi: 10.1038/s41598-019-39150-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0058] 58.Correction to: TCL1A acts as a tumour suppressor by modulating gastric cancer autophagy via miR-181a-5p-TCL1A-akt/mTOR-c-MYC loop. Carcinogenesis. 2023;44:535. doi: 10.1093/carcin/bgad035. [DOI] [PubMed] [Google Scholar]

[bib0059] 59.Oberbeck S., Schrader A., Warner K., Jungherz D., Crispatzu G., von Jan J., Chmielewski M., Ianevski A., Diebner H.H., Mayer P., Ados A Kondo, Wahnschaffe L., Braun T., Müller T.A., Wagle P., Bouska A., Neumann T., Pützer S., Varghese L., Pflug N., Thelen M., Makalowski J., Riet N., Göx H J M, Rappl G., Altmüller J., Kotrová M., Persigehl T., Hopfinger G., Hansmann M L, Schlößer H., Stilgenbauer S., Dürig J., Mougiakakos D., von Bergwelt-Baildon M., Roeder I., Hartmann S., Hallek M., Moriggl R., Brüggemann M., Aittokallio T., Iqbal J., Newrzela S., Abken H., Herling M. Noncanonical effector functions of the T-memory-like T-PLL cell are shaped by cooperative TCL1A and TCR signaling. Blood. 2020;136:2786–2802. doi: 10.1182/blood.2019003348. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0060] 60.Ho M.-F., Ingle J.N., Bongartz T., Kalari K.R., Goss P.E., Shepherd L.E., Mushiroda T., Kubo M., Wang L., Weinshilboum R.M. TCL1A Single-nucleotide polymorphisms and estrogen-mediated toll-like receptor-MYD88-dependent nuclear factor-κb activation: single-nucleotide polymorphism- and selective estrogen receptor modulator-dependent modification of inflammation and immune response. Mol. Pharmacol. 2017;92:175–184. doi: 10.1124/mol.117.108340. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0061] 61.Ad L. The role of chemokines in linking innate and adaptive immunity. Curr. Opin. Immunol. 2002;14:129–135. doi: 10.1016/s0952-7915(01)00308-9. [DOI] [PubMed] [Google Scholar]

[bib0062] 62.Punt S., Corver W.E., van der Zeeuw S.A.J., Kielbasa S.M., Osse E.M., Buermans H.P.J., de Kroon C.D., Jordanova E.S., Gorter A. Whole-transcriptome analysis of flow-sorted cervical cancer samples reveals that B cell expressed TCL1A is correlated with improved survival. Oncotarget. 2015;6:38681–38694. doi: 10.18632/oncotarget.4526. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0063] 63.Tai T.-S., Pai S.-Y., Ho I.-C. Itm2a, a target gene of GATA-3, plays a minimal role in regulating the development and function of T cells. PLoS One. 2014;9 doi: 10.1371/journal.pone.0096535. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0064] 64.Spashett R., Fernie G., Reid I.C., Cameron I.M. MADRS symptom subtypes in ECT-treated depressed patients: relationship to response and subsequent ECT. J. ECT. 2014;30:227–231. doi: 10.1097/YCT.0000000000000091. [DOI] [PubMed] [Google Scholar]

[bib0065] 65.Kisielow Pawel, Teh Hung Sia, Blüthmann Horst, von Boehmer Harald. Positive selection of antigen-specific T cells in Thymus by restricting MHC molecules. Nature. 1988;335:730–733. doi: 10.1038/335730a0. [DOI] [PubMed] [Google Scholar]

[bib0066] 66.Park J.-S., Jeon E.K., Chun S.H., Won H.S., Lee A., Hur S.Y., Lee K.H., Bae S.-N., Yoon S.-C., Hong S.H. ERCC1 (Excision Repair Cross-Complementation Group 1) Expression as a predictor for response of neoadjuvant chemotherapy for FIGO stage 2B uterine cervix cancer. Gynecol. Oncol. 2011;120:275–279. doi: 10.1016/j.ygyno.2010.10.034. [DOI] [PubMed] [Google Scholar]

[bib0067] 67.Bonifazi Gianluca, Lista Luca, Menasce Dario, Mezzetto Mauro, Pedrini Daniele, Spighi Roberto, Zoccoli Antonio. Study on the effects of the restrictive measures for containment of the COVID-19 pandemic on the reproduction number [Formula: see Text] in Italian regions. Eur. Phys. J. Plus. 2021;136 doi: 10.1140/epjp/s13360-021-02076-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0068] 68.Saglam O., Conejo-Garcia J. PD-1/PD-L1 immune checkpoint inhibitors in advanced cervical cancer. Integr. Cancer Sci. Ther. 2018;5 doi: 10.15761/ICST.1000272. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0069] 69.Liu L.-B., Xie F., Chang K.-K., Shang W.-Q., Meng Y.-H., Yu J.-J., Li H., Sun Q., Yuan M.-M., Jin L.-P., et al. Chemokine CCL17 induced by hypoxia promotes the proliferation of cervical cancer cell. Am. J. Cancer Res. 2015;5:3072–3084. [PMC free article] [PubMed] [Google Scholar]

[bib0070] 70.The CXCL12/CXCR7 signalling axis promotes proliferation and metastasis in cervical cancer. Med. Oncol. 2021;38 doi: 10.1007/s12032-021-01481-2. [DOI] [PubMed] [Google Scholar]

[bib0071] 71.Yadav Suresh Singh, Prasad Shyam Babu, Prasad Chandra Bhushan, Pandey Lakshmi Kant, Pradhan Satyajit, Singh Sunita, Narayan Gopeshwar. CXCL12 Is a key regulator in tumor microenvironment of cervical cancer: an in vitro study. Clin. Exp. Metastasis. 2016;33:431–439. doi: 10.1007/s10585-016-9787-9. [DOI] [PubMed] [Google Scholar]

[bib0072] 72.Rudolph M Isolde, Boza Yadira, Yefi Roger, Luza Sandra, Andrews Edilia, Penissi Alicia, Garrido Pablo, Rojas I Gina. The influence of mast cell mediators on migration of SW756 cervical carcinoma cells. J. Pharmacol. Sci. 2008;106:208–218. doi: 10.1254/jphs.fp0070736. [DOI] [PubMed] [Google Scholar]

[bib0073] 73.Wang J., Li Z., Gao A., Wen Q., Sun Y. The prognostic landscape of tumor-infiltrating immune cells in cervical cancer. Biomed. Pharmacother. 2019;120 doi: 10.1016/j.biopha.2019.109444. [DOI] [PubMed] [Google Scholar]

[bib0074] 74.Huang Bifen, Zheng Jianqing, Chen Bizhen, Wu Min, Xiao Lihua. Analysis of the correlation between RFC4 expression and tumor immune microenvironment and prognosis in patients with cervical cancer. Front. Genet. 2025;16 doi: 10.3389/fgene.2025.1514383. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0075] 75.Sayegh Hiba Al, Assi Zahraa, Kanaan Amjad, El Sibai Mirvat. CXCL1 in triple‑negative breast cancer: mechanisms, challenges, and therapeutic opportunities (Review) Oncol. Rep. 2026;55:1–14. doi: 10.3892/or.2026.9091. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0076] 76.Song G., Luo J., Zou S., Lou F., Zhang T., Zhu X., Yang J., Wang X. Molecular classification of Human papillomavirus-positive cervical cancers based on immune signature enrichment. Front. Public. Health. 2022;10 doi: 10.3389/fpubh.2022.979933. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0077] 77.Wang L., Wang L., Cao Y., Wang Q., Sun X. A chemokine-based prognostic model featuring CXCL8, ITGA5, BACE2, CCR7, CERS4, and MEI1 for cervical cancer. Eur. J. Med. Res. 2025;31:123. doi: 10.1186/s40001-025-03688-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0078] 78.Bao Qiufang, Tang Wenlu, Tang Wenzhong, Zhu Xiaohong, Hong Linliang, Zhang Bin. Identification of immune-related cervical cancer prognostic biomarkers and construction of prognostic model based on tumor microenvironment. Eur. J. Med. Res. 2025;30 doi: 10.1186/s40001-025-02515-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0079] 79.Chen Qian, Deng Dongfeng, Zhu Hong, Li Shan. Integrative analysis of bulk multiomics and single-cell transcriptomics elucidates age-related molecular landscape alterations in cervical cancer. Ann. Surg. Oncol. 2025;32:7000–7013. doi: 10.1245/s10434-025-17464-w. [DOI] [PubMed] [Google Scholar]

[bib0080] 80.Ma Boyi, Ren Chenlu, Yin Yadong, Zhao Shuhua, Li Jia, Yang Hong. Immune cell infiltration and prognostic index in cervical cancer: insights from metabolism-related differential genes. Front. Immunol. 2024;15 doi: 10.3389/fimmu.2024.1411132. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0081] 81.Wang N., Nanding A., Jia X., Wang Y., Yang C., Fan J., Dong A., Zheng G., Ma J., Shi X., et al. Mining of immunological and prognostic-related biomarker for cervical cancer based on immune cell signatures. Front. Immunol. 2022;13 doi: 10.3389/fimmu.2022.993118. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0082] 82.Ramírez Stieben L.A., Vargas M.C., Polillo D.C., Lufft K., Saldías P.R., Bedini I. Metastasis of breast cancer to the thyroid gland. Med. (B Aires) 2024;84:741–745. [PubMed] [Google Scholar]

[bib0083] 83.Xiong Y., Li T., Assani G., Ling H., Zhou Q., Zeng Y., Zhou F., Zhou Y. Ribociclib, a selective cyclin D kinase 4/6 inhibitor, inhibits proliferation and induces apoptosis of Human cervical cancer in Vitro and in Vivo. Biomed. Pharmacother. 2019;112 doi: 10.1016/j.biopha.2019.108602. [DOI] [PubMed] [Google Scholar]

[bib0084] 84.Liu Yuan, Zhao Runsheng, Fang Shanshan, Li Quan, Jin Yiqiang, Liu Bo. Abemaciclib sensitizes HPV-negative cervical cancer to chemotherapy via specifically suppressing CDK4/6-Rb-E2F and mTOR pathways. Fundam. Clin. Pharmacol. 2021;35:156–164. doi: 10.1111/fcp.12574. [DOI] [PubMed] [Google Scholar]

[bib0085] 85.Theoharides T.C., Valent P., Akin C. Mast cells, mastocytosis, and related disorders. N Engl. J. Med. 2015;373:163–172. doi: 10.1056/NEJMra1409760. [DOI] [PubMed] [Google Scholar]

[bib0086] 86.E C.Y., Cai Y., Sun B.Z., Guan L.Y., Jiang T. Hepatic insulin-like growth factor receptor is upregulated by activation of the GSK3B-FOXO3 pathway after partial hepatectomy. J. Biol. Regul. Homeost. Agents. 2017;31:549–555. [PubMed] [Google Scholar]

[bib0087] 87.Jain R., Tikoo S., Weninger W. Mast cell granules: modulating adaptive immune response remotely. J. Allergy Clin. Immunol. 2019;143:1731–1733. doi: 10.1016/j.jaci.2018.11.029. [DOI] [PubMed] [Google Scholar]

[bib0088] 88.Zhao F., Hong J., Zhou G., Huang T., Lin Z., Zhang Y., Liang L., Tang H. Elucidating the role of tumor-associated ALOX5+ mast cells with transformative function in cervical cancer progression via single-cell RNA sequencing. Front. Immunol. 2024;15 doi: 10.3389/fimmu.2024.1434450. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0089] 89.Gu Meijia, He Ti, Yuan Yuncong, Duan Suling, Li Xin, Shen Chao. Single-cell RNA sequencing revealed changes in the tumor microenvironment induced by radiotherapy for cervical cancer and the molecular mechanism of mast cells in immunosuppression. Funct. Integr. Genom. 2025;25 doi: 10.1007/s10142-025-01564-7. [DOI] [PubMed] [Google Scholar]

[bib0090] 90.Li Y., Wang J., Gao C., Hu Q., Mao X. Integral membrane protein 2A enhances sensitivity to chemotherapy via notch signaling pathway in cervical cancer. Bioengineered. 2021;12:10183–10193. doi: 10.1080/21655979.2021.2001218. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0091] 91.Kirchner J., Bevan M.J. ITM2A Is induced during thymocyte selection and T cell activation and causes downregulation of CD8 when overexpressed in CD4(+)CD8(+) double positive thymocytes. J. Exp. Med. 1999;190:217–228. doi: 10.1084/jem.190.2.217. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0092] 92.Kanehisa M., Furumichi M., Sato Y., Matsuura Y., Ishiguro-Watanabe M. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53:D672–D677. doi: 10.1093/nar/gkae909. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0093] 93.Kanehisa M., Goto S. KEGG: kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 2000;28:27–30. doi: 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0094] 94.Kanehisa M. Toward understanding the origin and evolution of cellular organisms. Protein Sci. 2019;28:1947–1951. doi: 10.1002/pro.3715. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Multi-omics integration identifies prognostic genes linked to adaptive immunity response and radiosensitivity in cervical cancer

Rui Quan

Jinbao Wang

Guanghui Cheng

Highlights

Abstract

Background

Methods

Results

Conclusions

Introduction

Materials and methods

Collecting data

Weighted gene co-expression network analysis (WGCNA)

Identification, functional enrichment, and protein-protein interaction (PPI) analysis of cervical cancer differentially expressed radiochemotherapy sensitivity-related adaptive immune response-related genes (CC-DE-RS-AIRRGs)

Identification of prognostic genes

Assessment and verification of risk models

Clinical features and functional analyses

Immune microenvironment and immunotherapy analyses

Drug sensitivity and drug prediction analyses

Chromosome localization and regulatory network analyses

scRNA-seq analysis

Cell communication, cell cycle, and pseudotime trajectory analyses

Clinical samples

Cell lines and cell culture

Ionizing radiation

Colony formation assay

Cell viability assay

Western blot experiment

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

Small interfering RNA (siRNA) transient transfection

Statistical analysis

Results

Identification of 1668 AIRMRGs in CC

Fig. 1.

Identification, functional, and PPI of 9 CC-DE-RS-AIRRGs in CC

Fig. 2.

Development and verification of risk models in CC

Fig. 3.

Clinical feature and functional enrichment in CC

Fig. 4.

Immune cells and immunotherapy of risk model

Fig. 5.

Drug of risk model and prognostic genes in CC

Fig. 6.

Chromosomal localization and miRNA-lncRNA interaction pairs of 2 prognostic genes in CC

Fig. 7.

Identification of mast cells in CC

Fig. 8.

Cell communication and cell cycle of mast cells in CC

Fig. 9.

Pseudotime trajectory of mast cells in CC

Fig. 10.

Correlation of tcl1a and itm2a expression with radiotherapy prognosis in cervical cancer patients

Table 1.

Fig. 11.

Establishment and validation of a radioresistant cervical cancer cell line

Fig. 12.

Table 2.

Low expression of tcl1a and itm2a proteins in the radioresistant cervical cancer cell line

Fig. 13.

Knockdown of tcl1a and itm2a reduces radiosensitivity of cervical cancer cells

Fig. 14.

Discussion

Conclusions

Data availability

Funding statement

Ethics declarations

Consent to participate/Consent to publish

CRediT authorship contribution statement

Declaration of competing interest

Acknowledgments

Footnotes

Contributor Information

Appendix. Supplementary materials

References

Associated Data

Supplementary Materials

Data Availability Statement