Characterization of T cell markers in endometrial carcinoma through single-cell RNA sequencing Authors: Menghuang Zhao, Lulu Zhou Published: 2025 ABSTRACT AbstractBackgroundEndometrial carcinoma is a leading gynecological malignancy worldwide and is frequently associated with poor prognosis in advanced-stage patients. Immunotherapy has emerged as a promising treatment modality in endometrial cancer. T cells are recognized as vital mediators of immunosurveillance and cancer eradication.MethodsIn this study, we leveraged single-cell RNA sequencing data from the GSE173682 dataset to identify T cell marker genes. Additionally, we collected bulk RNA sequencing data and clinical data from 544 endometrial cancer patients to construct a predictive model. We further explored the correlations between this model and immunotherapy response, drug sensitivity, and mutation status.ResultsWe developed an eight-gene prediction signature based on T cell marker genes, which demonstrated high accuracy in predicting the prognosis of endometrial cancer patients. The area under the receiver operating characteristic curve was 0.838 at 5 years. This T cell-related signature was strongly associated with immune-related function scores and showed potential for predicting responses to both immunotherapy and chemotherapy in endometrial cancer patients.ConclusionWe have established a robust prognostic model based on T cell genes. This model may contribute to the development of more precise therapeutic strategies in the management of endometrial cancer, though further validation is needed.Supplementary InformationThe online version contains supplementary material available at 10.1007/s12672-025-04289-y. Introduction Endometrial cancer (EC) ranks among the most frequently diagnosed gynecological malignancies globally, with an estimated 420,242 new cases and 97,704 fatalities reported in 2022 [1]. The incidence and mortality rates of endometrial cancer continue to rise steadily. Over the past 30 years, the overall incidence has surged by 132%, potentially attributable to the increased prevalence of risk factors such as obesity and an aging population [2]. Given that the heterogeneity of endometrial cancer is a pivotal factor influencing cancer progression and overall survival, which also leads to varying responses among individual patients to the same treatment [3], it is imperative to employ single-cell RNA sequencing (scRNA-seq) technology to identify effective prognostic biomarkers. In recent years, immunotherapy has demonstrated significant benefits in the treatment of endometrial cancer [4]. The dynamic crosstalk between neoplastic cells and their immune microenvironment significantly influences endometrial cancer progression, metastasis, and therapeutic outcomes [4]. As key effectors of antitumor immunity, T cells play a pivotal role in immunosurveillance and tumor elimination. Imbalanced regulation or a deficiency of T cells in tumors can result in immunotherapy resistance [5]. Understanding these interactions is crucial for developing personalized treatment strategies that can enhance the immune response and overcome resistance mechanisms in endometrial cancer. Our investigation identified T cell-related markers in endometrial cancer through single-cell analysis. We developed a T cell marker gene signature (TCMGS) and evaluated its prognostic potential for endometrial cancer. We also examined its impact on the response to immunotherapy. Materials and methods Data source The GEO database was utilized to gather scRNA-seq data for five endometrial carcinoma samples from GSE173682 [6]. Additionally, the TCGA database includes 544 endometrial cancer patients, for whom bulk RNA-sequencing data and clinical statistics were examined to provide insights into endometrial carcinoma. Analysis of scRNA-seq data The scRNA-seq data were analyzed utilizing the R packages “Seurat” and “SingleR” [7, 8]. The quality control for scRNA-seq data began by excluding clusters with less than three cell counts, cells with fewer than 200 detected genes, and cells with more than 10% mitochondrial genes. Data normalization was performed using the “NormalizeData” function, followed by batch effect correction implemented through the “Harmony” package. To visualize cellular heterogeneity, we performed dimensionality reduction using T-distributed stochastic neighbor embedding (t-SNE), which provided a two-dimensional representation of distinct cell populations. Differential gene expression analysis between clusters was conducted using the “FindAllMarkers” function, with stringent thresholds set at |log(fold change)| > 1 and an adjusted P-value < 0.05 for identifying cluster-specific marker genes. Cell type annotation was subsequently performed by referencing the Human Primary Cell Atlas through the “SingleR” package, enabling precise classification of distinct cellular subpopulations within the identified clusters. Construction of a prognostic model To evaluate the predictive significance of T cell marker genes in TCGA endometrial carcinoma cohorts, we initially conducted univariate Cox regression analysis, selecting genes with significant associations (P < 0.05) with overall survival. To minimize overfitting and select the most robust prognostic genes, we employed the LASSO method with Cox proportional hazards regression. This approach is extensively utilized for regression analysis involving high-dimensional variables. The optimal parameter was determined through tenfold cross-validation using the “cv.glmnet” function. The prognostic gene signature was identified based on a list of non-zero coefficients. These selected genes were further incorporated into a stepwise multivariate Cox regression model using SPSS software, where risk scores were computed as follows: risk-score = Σi Coefficient(mRNAi) × Expression(mRNAi). The prognostic significance of these gene signatures was further confirmed. Patients were categorized into low- and high-risk groups according to the median risk score threshold. The predictive performance of the T-cell marker gene signature was validated through the area under the curve (AUC) using the “survivalROC” package. In the Kaplan–Meier survival analysis, the log-rank test from the “survminer” R package was utilized to evaluate the statistical significance of differences between the survival curves. Functional enrichment analysis The online software Database for Annotation, Visualization, and Integrated Discovery (DAVID) [9] was utilized for Gene Ontology (GO) enrichment analysis. GO terms are categorized into three main domains: Biological Process (BP), Cellular Component (CC), and Molecular Function (MF). The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database [10] comprises a collection of manually drawn pathway maps that detail molecular interactions and reaction networks. In this study, KEGG pathway enrichment analysis was conducted using DAVID. P < 0.05 was considered to indicate a significant difference for GO and KEGG pathways. Identification of immune characteristics for the risk model and assessment of immunotherapy To characterize the immune landscape within the tumor microenvironment (TME) of endometrial cancer, we performed single sample gene set enrichment analysis (ssGSEA) on the TCGA dataset, quantifying the enrichment scores of 28 distinct immune cell populations. This approach enabled the systematic evaluation of immune cell infiltration patterns based on specific gene expression signatures. Comparative analysis of immune infiltration profiles between risk groups was conducted using t-test to determine statistical significance. Given the critical role of immune checkpoint molecules in modulating therapeutic responses, we specifically examined the expression patterns of key regulators, including CTLA4 and PD-1, across defined risk strata. The Cancer Immunome Atlas (TCIA) was employed to quantify immune scores (IPS) in endometrial cancer specimens, thereby predicting clinical responses to anti-PD1, anti-CTLA4 monotherapies, and their combination therapy [11]. The differential sensitivity of endometrial cancer to immunotherapy between high- and low-risk groups was further investigated. These analyses provided insights into the potential immunological determinants of clinical outcomes in different risk categories. Drug sensitivity analysis The R package ‘oncoPredict’ was utilized to calculate the sensitivity scores of the drugs in the Genomics of Drug Sensitivity in Cancer (GDSC) database [12]. The t-test was employed to compare the differences in chemotherapy sensitivity between the high- and low-risk groups. Mutation analysis To delineate the genomic landscape across different risk stratifications, we systematically analyzed somatic mutation profiles obtained from TCGA datasets. Mutation frequencies were calculated for each gene within the cohort, followed by comprehensive visualization of mutational patterns through waterfall plots generated using the “maftools” package. Tumor Mutational Burden (TMB) was quantified and stratified based on mean values, with subsequent survival analysis performed through Kaplan-Meier analysis to evaluate prognostic implications. Additionally, patients in the high- and low-risk groups were compared with respect to Tumor Mutational Burden. Result Identifying T-cell marker gene expression profiles Figure 1A presents the spectrum of detected gene counts, sequencing depth, and mitochondrial content percentages for each sample. The scRNA-seq data derived from five patients were screened, and the spatial distribution of cells was annotated (Fig. 1B). Principal Component Analysis (PCA) was applied to reduce data dimensions and discern 14 cell clusters based on the 2,000 most variable genes (Fig. 1C). Utilizing the Human Primary Cell Atlas as a reference, cells in clusters 1 and 10 were classified as T cells (Fig. 1D). Ultimately, applying a threshold of |logFC| > 1 and an adjusted P-value of 0.05, we identified 1130 T-cell marker genes associated with endometrial carcinoma. Identification of T cell-related markers in endometrial cancer. A The spectrum of gene counts, sequencing depth, and mitochondrial content percentages in five endometrial cancer samples. B The T-SNE showed the distribution of cells in five endometrial cancer samples. C A total of 14 clusters of all samples were identified. D A total of 9 subtypes of cells were identified based on SingleR annotation methods Constructing and evaluating a prognostic model In our study, the TCGA endometrial carcinoma cohort was used to conduct univariate Cox regression analysis. Leveraging the 1130 T-cell marker genes, we developed a predictive signature. Among these, 247 genes demonstrated a significant association with overall survival (Table S1). Utilizing the optimal lambda value and corresponding coefficients, LASSO analysis narrowed the list down to 29 genes (Fig. 2A). Through multivariate Cox regression analysis, we ultimately established a predictive model based on eight genes: TFPI, MGAT4A, RAB32, DRAM1, DUSP6, DAPK3, NRIP1 and LTB (Table 1). The risk score is calculated as follows: Risk score =(0.391 * TFPI expression + 0.529 * MGAT4A expression + 0.351 * RAB32 expression − 0.474 * DRAM1 expression − 0.254 * DUSP6 expression − 0.566 * DAPK3 expression − 0.602*NRIP1 expression − 0.238 * LTB expression). Patients were categorized into low- and high-risk groups based on the median risk score, ranking from low to high. The Kaplan–Meier survival analysis revealed that high-risk patients experienced significantly worse overall survival compared to low-risk patients (Fig. 2B). To assess the predictive accuracy of the signature for 5-year overall survival (OS), we employed receiver operating characteristic (ROC) curves. The AUC value for this interval was 0.838 (Fig. 2C). Building and evaluating the prognostic model. A LASSO regression of overall survival related genes. B The survival curve of endometrial cancer with high- and low-risk score in TCGA. C The ROC curve analysis Table 1Multivariate cox regression analysis of overall survivalGene symbolCoefficientHazard ratio (HR)P valueTFPI0.3911.4790.001MGAT4A0.5291.698< 0.001RAB320.3511.420.006DRAM1− 0.4740.6220.001DUSP6− 0.2540.7760.015DAPK3− 0.5660.5680.001NRIP1− 0.6020.548< 0.001LTB− 0.2380.7880.003 Multivariate cox regression analysis of overall survival Functional enrichment analysis The outcomes of the Gene Ontology analysis indicated that the genes were predominantly linked to Biological Process functions such as T cell activation (p = 4.58E−07), apoptotic process (P = 6.67E−06) and positive thymic T cell selection (p = 1.18E−05) (Fig. 3A). Regarding Cellular Component, the GO results revealed that genes were mainly concentrated in extracellular exosome (P = 4.44E−14), focal adhesion (P = 4.82E−09) as well as alpha-beta T cell receptor complex (P = 2.07E−07) (Fig. 3B). In Molecular Function terms, genes were implicated in protein binding (P = 9.16E−16), identical protein binding (P = 3.86E−06) and RNA binding (P = 6.33E−05) (Fig. 3C). The KEGG pathway enrichment analysis showed that genes were significantly enriched in pathways related to endometrial cancer, PD-L1 expression and PD-1 checkpoint pathway in cancer, T cell receptor signaling pathway, as well as the cell cycle (Fig. 3D). The results of the GO enrichment analysis and KEGG enrichment analysis. A Biological process. B Cellular component. C Molecular function. D KEGG enrichment analysis Immune cell infiltration analysis and immunotherapy sensitivity analysis Within the TCGA endometrial cancer cohort, we assessed the infiltration levels of 28 immune cell types utilizing ssGSEA. Our findings revealed that the high-risk subgroup exhibited markedly elevated proportions of Effector memory CD4 T cell. In contrast, the low-risk subgroup was characterized by a higher abundance of Activated B cell, Activated CD8 T cell, Activated dendritic cell, and Central memory CD4 T cell (Fig. 4A). Considering the pivotal role of immune checkpoint inhibitors in immunotherapy, we evaluated the expression levels of commonly expressed immune checkpoint-related genes across distinct risk stratifications. Our investigation revealed significantly elevated expression levels of CTLA4 and PD-1in the low-risk patient cohort compared to their high-risk counterparts (Fig. 4B). Immune cell infiltration analysis. A Comparative analysis of immune infiltration profiles between high-risk and low-risk groups. B Immune checkpoint variations across groups at high- and low-risk. ns: P > 0.05, *: P ≤ 0.05, **: P ≤ 0.01, ***: P ≤ 0.001,****: P ≤ 0.0001 We further investigated the differential sensitivity of endometrial cancer to immunotherapy between high- and low-risk groups. The low-risk group demonstrated a significantly higher relative probability of response to anti-CTLA4 and anti-PD1 immunotherapy compared to the high-risk group, indicating superior therapeutic efficacy in this subgroup (Fig. 5A–C). Immunotherapy sensitivity analysis. A CTLA4-negative and PD1-positive response difference. B CTLA4-positive and PD1-negative response difference. C CTLA4-positive and PD1-positive response difference Drug sensitivity analysis Chemotherapy represents a crucial treatment modality for endometrial carcinoma. Cisplatin, paclitaxel, and docetaxel are frequently employed chemotherapeutic agents in the management of endometrial carcinoma. We assessed the IC50 values of these drugs in both high- and low-risk score groups. As depicted in Fig. 6A and C, the IC50 values for cisplatin, paclitaxel, and docetaxel were significantly higher in the high-risk score group compared to the low-risk score group (all P < 0.05). This finding indicates that patients with a low-risk score exhibit a more favorable response to chemotherapy for endometrial carcinoma. The IC50 values of chemotherapeutic agents in different risk score group. A Cisplatin. B Paclitaxel. C Docetaxel Gene mutation analysis Figure 7A presents the comprehensive mutation profile of endometrial carcinoma. The genetic mutations of the top 15 genes are illustrated in Fig. 7B. The genetic mutations of the T-cell marker genes are depicted in Fig. 7C. NRIP1 was found to be mutated in 11% of the samples. Within the TCGA cohort, we explored the relationship between tumor mutational burden and risk score. The results indicated that high-risk patients had significantly lower TMB compared to low-risk patients (Fig. 7D). According to the Kaplan–Meier analysis based on the mean TMB values, the high-TMB group demonstrated superior overall survival compared to the low-TMB group (Fig. 7E). Gene mutation analysis. A General description of the mutation landscape. B Top 15 mutated genes. C The genetic mutations of the eight T-cell marker genes. D TMB expression varies across samples at low- and high-risk. (E) Survival curves for TMB high- and low- group Discussion Studies have demonstrated that cells and molecules within the tumor immune microenvironment (TIME) influence cancer progression and metastasis by modulating the balance between cytotoxic reactions and immune suppression. The heterogeneity of TIME is a critical determinant of tumor prognosis [13]. Similar to many cancers, TIME is instrumental in the progression of endometrial cancer [14]. Upon invasion by endometrial tumor cells, both innate and adaptive immune responses are triggered and engage in a “combat mode”. Dendritic cells (DCs) that have engulfed tumor antigens migrate to the lymph nodes, where T cells, upon antigen stimulation, become activated and target foreign elements within the body [15]. Among these, CD8+ cytotoxic T cells (CTLs) are particularly significant, as they can directly launch a potent attack on endometrial carcinoma cells. Concurrently, CD4+ helper T (Th) cells can induce specific B cell responses, initiating both humoral and cytotoxic immune reactions [16–18]. This study revealed that most immune cells are highly expressed in the low-risk group, including activated CD8 T cells, activated B cells, and activated dendritic cells. Tumor-infiltrating lymphocytes (TILs) have been recognized as prognostic indicators for various cancers, such as melanoma, colorectal cancer, ovarian cancer, and endometrial carcinoma [19]. Research indicates a positive correlation between the presence of TILs, particularly CD8+ cytotoxic T cells, and the prognosis of endometrial carcinoma patients, while the number of TILs in late-stage endometrial cancer patients is significantly reduced [20, 21]. In summary, the tumor immune microenvironment of endometrial cancer is characterized by the infiltration of both innate and adaptive immune cells with anti-tumor capabilities. The ultimate outcome of the interaction between these immune cells and tumor cells will determine the tumor’s fate. If endometrial cancer cells manage to create an immunosuppressive microenvironment through various pathways, they can evade immune surveillance and enter an escape phase, potentially leading to metastasis [19]. The significance of immunotherapy lies in its ability to activate and enhance the body’s immune response, reverse the immunosuppressive state, and thereby eliminate cancer cells, improving patient prognosis [22]. Our research found that the low-risk group exhibits higher expression of immune cells and demonstrates a more favorable prognosis. Chemotherapy stands as a vital treatment approach for endometrial cancer. Drugs such as cisplatin, paclitaxel, and docetaxel are frequently employed in chemotherapy regimens. Our research indicated that a low-risk score was associated with a more favorable response to chemotherapy in endometrial cancer patients. This finding could potentially elucidate, in part, the improved prognosis observed in the low-risk group. Tumor Mutational Burden indicates alterations within the tumor cell genome. The DNA damage repair (DDR) system is essential for preserving genomic integrity in cancer cells. Dysregulation of DDR pathways drives the accumulation of DNA errors and elevates tumor mutational burden. There is a direct correlation between TMB and the quantity of neoantigens, mutations can lead to the presentation of new antigens on the tumor cell surface via MHCI molecules, which in turn can provoke an immune response against the tumor [23]. Exogenous DDR inhibition or intrinsic DDR defects have been shown to upregulate immune checkpoint protein expression in cancer cells [24]. TMB has emerged as a significant biomarker for forecasting the effectiveness of immune checkpoint inhibitors (ICIs) treatment, a fact corroborated by prior research in melanoma and lung cancer immunotherapy [25]. Moreover, it has been established that tumors with high TMB, microsatellite instability-high (MSI-H), and DNA polymerase epsilon (POLE) mutations are more likely to respond favorably to ICIs [26, 27]. Clinical outcomes of PD-1/PD-L1 blockade in endometrial cancer demonstrate a strong association with microsatellite instability (MSI) status. The KEYNOTE-158 study reported a 48% objective response rate in MSI-high advanced endometrial cancer patients treated with pembrolizumab (a humanized monoclonal anti-PD-1 antibody) [28]. According to our study, TMB is more frequently observed in the low-risk score group. The low-risk group exhibited significantly higher expression of CTLA4 and PD-1 compared to the high-risk group, and demonstrated greater sensitivity to anti-CTLA4 and anti-PD-1 immunotherapy. Individuals with a higher mutational burden generally exhibit a more positive prognosis. Single-cell RNA sequencing technology, which examines various cell subpopulations, has recently gained recognition as a promising tool for investigating tumor heterogeneity and identifying potential therapeutic targets. In this study, we utilized scRNA-seq analysis to construct a prognostic model for endometrial carcinoma based on eight T cell marker genes: TFPI, MGAT4A, RAB32, DRAM1, DUSP6, DAPK3, NRIP1, and LTB. TFPI is a plasma serine protease inhibitor that plays a vital role in regulating blood coagulation by inhibiting the initiation complex TF/FVIIa/Factor Xa (FXa) [29]. TFPI is also expressed by cancer cells and has been shown to have non-hemostatic functions [30, 31]. In breast cancer cells, TFPI has demonstrated multifaceted anti-tumor activities, including suppression of angiogenesis [32], inhibition of cellular migration [33], and modulation of metastatic potential, proliferation, and apoptosis [34, 35]. Conversely, Fisher et al. revealed a contrasting pro-migratory function of TFPI in bladder tumor cells expressing TF in the presence of exogenous factor VIIa, supporting an oncogenic role of TFPI [36]. Elevated TFPI levels was observed in hepatocellular carcinoma tissues relative to normal tissue, with its overexpression demonstrating a significant correlation with poor clinical outcomes in hepatocellular carcinoma patients [37]. Moreover, TFPI expression exhibited significant correlations with immune system activity and tumor-infiltrating immune cells across breast cancer, melanoma, and papillary thyroid carcinoma [35, 38, 39]. In our study, high TFPI expression was indicative of a poorer outcome in endometrial carcinoma patients. Prior research in endometrial cancer has indicated that MGAT4A functions as a tumor promoter. More precisely, in the presence of galectin-9, MGAT4A augments the proliferation and invasiveness of endometrial cancer cells by facilitating glucose metabolism. Additionally, there is a significant correlation between MGAT4A overexpression and poorly differentiated histological subtypes, as well as unfavorable prognoses [40]. In endometrial cancer with TP53 mutations, the p53 mutants fail to suppress MGAT4A expression, resulting in a poorer prognosis for patients [40]. Our study also revealed that elevated MGAT4A expression is associated with a poor prognosis. RAB proteins are part of the Ras superfamily and are classified as small GTPases with molecular weights between 20 and 30 kD. RAB32, a member of this family, is pivotal in processes related to the biogenesis of autophagic intracellular vesicle trafficking, apoptosis, and lysosome-associated organelles [41–43]. Rab32 also plays a crucial role in host immune defense against pathogenic infections [44]. Emerging evidence indicates that downregulation of RAB32 significantly impairs the growth potential of ovarian cancer cells [45]. In the context of hepatocellular carcinoma, recent investigations by Zheng He et al. have revealed that RAB32 serves as a functional mediator of miR-30c-5p-associated cell proliferation and invasion [46]. RAB32 expression is notably upregulated in patients with glioma and in glioma cell lines. The expression level exhibited a positive correlation with the glioma grade and functioned as an independent prognostic indicator. In vitro studies have shown that RAB32 knockdown inhibits glioblastoma cell invasion, migration, and proliferation via the JAK/STAT3 signaling pathway [47]. In addition, RAB32 expression significantly correlated with immune cell infiltration and immune checkpoint molecule expression in the glioblastoma tumor microenvironment [48].Our study found that high-risk group with elevated RAB32 expression demonstrated significantly diminished responsiveness to immunotherapy. And high RAB32 expression was associated with a poorer outcome in endometrial carcinoma patients. As a critical regulator of autophagic processes, DRAM1 exhibits frequent downregulation across multiple malignancies [49]. In the presence of growth factors, DRAM1 modulates the activation of the IGF-1 receptor and suppresses the downstream PI3K-AKT-mTOR pathway, leading to the activation of autophagy and inhibition of cell proliferation across different types of human cancers [50]. It has been reported that DRAM1 levels are reduced in non-small cell lung cancer and are inversely correlated with EGFR levels. Furthermore, overexpression of DRAM1 has been shown to inhibit the proliferation, invasion, migration, and epithelial-mesenchymal transition (EMT) of non-small cell lung cancer cell lines with mutant EGFR, both in vitro and in vivo [51]. DRAM1 has also been identified as a robust prognostic biomarker for predicting survival outcomes in endometrial carcinoma [52]. Our research corroborates these findings, revealing a similar association. DRAM1 also contributed to host immune defense against pathogens [53]. In oral squamous cell carcinoma, DRAM1 expression was significantly correlated with T-cell infiltration [54]. Our study revealed that higher DRAM1 expression in the low-risk group was associated with increased abundance of activated B cells, activated CD8⁺ T cells, activated dendritic cells, and central memory CD4⁺ T cells. DUSP6 functions as a specific phosphatase that catalyzes the dephosphorylation of both tyrosine and threonine residues on ERK1/2, leading to the inactivation of this critical MAPK–ERK1/2 pathway [55]. Accumulated evidence indicates that DUSP6 is downregulated in multiple malignancies, including lung cancer, ovarian cancer, nasopharyngeal carcinoma, and esophageal squamous cell carcinoma [56]. However, there is a discrepancy regarding the expression of DUSP6 in endometrial cancer. Prior studies have suggested that DUSP6 is commonly overexpressed in endometrial adenocarcinoma [57]. DUSP6 knockdown has been shown to reduce cell metastasis and invasion, while DUSP6 overexpression inhibits apoptosis under serum-free conditions [58]. Patients with high DUSP6 expression had significantly shorter progression-free survival and overall survival [58]. In contrast, research by Ming-Jun Fan reported opposite results, observing downregulation of DUSP6 in endometrial adenocarcinoma compared to normal controls [55]. Elevated levels of DUSP6 markedly curtailed the growth, invasive capacity, and migratory potential of tumor cells, and suppressed EMT-associated markers. Conversely, reducing DUSP6 expression led to opposing effects. The inhibitory impact of DUSP6 on endometrial adenocarcinoma is mediated through the suppression of the ERK signaling pathway. Within our research, we observed that high expression of DUSP6 correlated with a favorable prognosis. In addition, utilizing mice lacking DUSP6 (DUSP6−/−), Wei et al. revealed its critical role in modulating T cell receptor signaling, thereby regulating follicular helper T cell differentiation and T cell metabolism [59]. Our findings showed that elevated DUSP6 expression in the low-risk subgroup is associated with enhanced immune cell infiltration and improved responsiveness to immunotherapy. The level of DAPK3 expression demonstrated a strong correlation with both the histological grade and the depth of myometrial invasion in cases of endometrial cancer. DAPK3 was identified as a critical regulator of anti-tumor immunity via the stimulator of interferon genes (STING) pathway of cytosolic DNA sensing. DAPK3 coordinated STING post-translational modifications to activate tumor-intrinsic innate immunity and enforce immune surveillance. DAPK3 deficiency in IFN-β-producing tumors resulted in accelerated tumor growth and impaired immune infiltration of dendritic cells and cytotoxic lymphocytes [60]. Furthermore, Kaplan-Meier survival analysis revealed that patients with low DAPK3 expression exhibited a shorter overall survival compared to those with high DAPK3 expression [61]. Consistent with these findings, our investigation yielded analogous results. Through comprehensive genomic analysis, Ferreira et al. identified NRIP1 as the predominant mutation hotspot in endometrial cancer cell lines, with its alterations strongly associated with microsatellite instability-mediated dysregulation of estrogen receptor signaling pathways [62]. Beyond its role in endometrial cancer, NRIP1 has been shown to preferentially interact with ERβ isoforms and exhibit tumor suppressive functions in both ovarian and colorectal malignancies [63, 64]. Moreover, NRIP1 expression exhibited a strong positive correlation with the infiltration of macrophages, dendritic cells, CD8+ T cells, CD4+ T cells, and neutrophils [65]. Our research revealed that high NRIP1 expression correlates with more abundant immune cell infiltration and favorable prognosis. LTB, a type II membrane protein from the TNF family, is primarily expressed on B and T lymphocytes [66]. LTB serves as the ligand for the lymphotoxin-β receptor (LTBR), and the LTBR signaling pathway is known to augment the effector functions of T cells [67]. In our study, we found that LTB, as a T cell marker, was indicative of a favorable prognosis in endometrial carcinoma. Conclusions In our study, we initially conducted an extensive analysis of single-cell RNA sequencing data from patients with endometrial carcinoma to pinpoint prognostic signatures and assess the tumor immune microenvironment. We established a T cell marker gene signature to create a survival model for evaluating the progression of endometrial carcinoma. Our findings suggest that the T cell marker gene signature could potentially serve as a novel target for prognostic assessment and therapeutic intervention in the future management of endometrial carcinoma progression. Limitations While this study offers significant insights, several limitations warrant consideration. Firstly, the retrospective nature of the TCGA-derived data used for model construction may introduce inherent biases, necessitating validation through larger, independent cohorts and prospective clinical studies to confirm its robustness and translational potential. Secondly, while our computational analyses identified promising biomarkers, their biological relevance and mechanistic roles require further experimental validation through systematic in vitro and in vivo investigations to fully elucidate their functional significance in disease pathogenesis. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. Acknowledgements Not applicable. Author contributions [MZ]: Conceptualization, methodology, software, data curation, formal analysis, writing—original draft. [LZ]: methodology, resources, supervision, validation, investigation, writing—review & editing. All authors reviewed the manuscript. Funding No funding was received for this study. Data availability The data analyzed in this study were obtained from the TCGA (http://cancergenome.nih.gov/) and GEO (http://www.ncbi.nlm.nih.gov/geo/) databases. Declarations Ethics approval and consent to participate The data for this study were obtained from publicly available databases. Thus, no specific ethical approval was required for this work. As all data were obtained from publicly available datasets and no new patients were enrolled, individual consent to participate was not applicable to this study. Consent for publication All authors have read and approved the final manuscript. No individual person’s data is included in this study, and therefore no consent for publication is required. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The data analyzed in this study were obtained from the TCGA (http://cancergenome.nih.gov/) and GEO (http://www.ncbi.nlm.nih.gov/geo/) databases.