Abstract
Zinc finger protein (ZNF) family is the largest transcription factor family in the human genome. Studies have shown that the aberrant expression of ZNF had a potential role in tumorigenesis. However, the role of ZNF family genes in uterine corpus endometrial carcinoma (UCEC) is still not systematically understood. Transcriptomic profiles and clinical data of endometrial carcinoma were obtained from TCGA and GEO databases. Based on differentially expressed ZNF genes, unsupervised clustering was employed to stratify samples into distinct subtypes, followed by enrichment analysis to compare gene expression patterns and pathway alterations across subgroups. Patients were subsequently divided into training and test cohorts, through univariate COX regression, LASSO (least absolute shrinkage and selection operator) regression, and multivariate COX regression analyses, we identified prognostic ZNF genes in UCEC. A risk scoring model was established based on these prognostic ZNF genes, and its predictive performance was validated using extensive clinical data. The model was further evaluated for its associations with tumor microenvironment, immune infiltration, immunotherapy response, somatic mutations, and drug sensitivity. Additionally, single-cell RNA sequencing data and the human protein atlas (HPA) database were utilized to investigate the cellular-level functions and impacts of ZNF genes. We stratified UCEC patients into three subtypes based on differentially expressed ZNF genes, which exhibited distinct prognostic outcomes and pathway enrichment profiles. Eight prognostic ZNF genes were identified and incorporated into a ZNF scoring system named ZNF score to quantify patient risk. The ZNF score integrated with clinical characteristics demonstrated robust predictive performance in UCEC patients. Immune infiltration analysis demonstrated a significant increase in M1 Macrophages and M2 Macrophages abundance within the high-risk group. Drug sensitivity analysis identified potential therapeutic agents. Single-cell analysis, immunohistochemistry demonstrated and data-independent acquisition based quantitative proteomic analysis lower expression of most ZNF genes in tumor tissues compared to normal tissues, with predominant distribution in endothelial and epithelial cells. This study represents the first investigation utilizing ZNF genes to determine prognostic outcomes in UCEC patients. Our findings shed light on the potential of the ZNF score as a tool to evaluate ZNF expression patterns, immune cell infiltration, response to pharmacotherapy, clinicopathological features, and survival outcomes in UCEC. This may provide the more effective guide to select immunotherapeutic strategies of UCEC in the future.