Senescent epithelial cells remodel the tumor microenvironment and drive early LUAD progression: a multiomics risk model and single-cell analysis

Lung adenocarcinoma (LUAD) remains a major cause of cancer-related mortality, and there are currently few reliable biomarkers available for accurate prognosis and effective targeted therapy. Accumulating evidence demonstrates that senescent cells play an important role in tumor progression and immune evasion; nevertheless, their specific contribution to LUAD pathogenesis has not yet been fully elucidated. Accordingly, a five-gene senescence-related risk model comprising FGF2, GAPDH, CCNA2, ENO1, and DKK1 was established using the Least Absolute Shrinkage and Selection Operator regression applied to bulk RNA sequencing data obtained from The Cancer Genome Atlas (TCGA). Patients stratified as high risk by this model exhibited significantly poorer overall survival and progression-free survival, accompanied by marked activation of pathways associated with immune infiltration, epithelial–mesenchymal transition, and extracellular matrix remodeling. Integrative single-cell RNA sequencing analysis further revealed a distinct epithelial subpopulation (E9) defined by preferential activation of the senescence-associated gene DKK1. This subpopulation, which emerged from integrative single-cell transcriptomic analysis, exhibited pronounced senescence-associated characteristics, significantly increased cellular stemness, and extensive intercellular communication capacity, and uniquely expressed COL17A1 together with transcriptional programs that were strongly associated with epidermal development. Pseudotime analyses consistently positioned E9 cells at an early stage of tumor evolution. At the functional level, DKK1 contributed to senescence-associated phenotypic features, thereby promoting tumor cell migration and metastatic potential. From a clinical perspective, patients with concurrent high levels of COL17A1 and DKK1 expression experienced significantly worse clinical outcomes. Collectively, these findings identify a senescence-driven epithelial subpopulation that contributes to LUAD progression via DKK1-mediated activation of aging-related pathways and highlight DKK1 as a potential therapeutic target.