Introduction: The connection between aging and cancer is complex. Previous research has highlighted the association between the aging process of lung adenocarcinoma (LUAD) cells and the immune response, yet there remains a gap in confirming this through single-cell data validation. Here, we aim to develop a novel aging-related prognostic model for LUAD, and verify the alterations in the genome and immune microenvironment linked to cellular senescence.
Methods: We integrated a comprehensive collection of senescence genes from the GenAge and CellAge databases and employed the least absolute shrinkage and selection operator (LASSO) Cox analysis to construct and validate a novel prognostic model for LUAD. This model was then utilized to examine the relationship between aging, tumor somatic mutations, and immune cell infiltration. Additionally, we explored the heterogeneity of senescence and intercellular communication within the LUAD tumor microenvironment (TME) through single-cell transcriptomic data analysis.
Results: By exploring the expression profiles of 586 cellular senescence-related genes in 428 LUAD patients, we constructed an aging-related genes (ARGs) risk model included 10 ARGs and validated it as an independent prognostic predictor for LUAD patients. Notably, patients with low aging scores (LAS group) exhibited better survival, lower tumor mutation burden (TMB), lower somatic mutation frequency, lower tumor proliferation rate, and an immune activated phenotype compared to patients with high aging scores (HAS group). While the HAS group was enriched in tumor cells and showed a lower infiltration of CD8-CCR7, CD8- CXCL13, CD8-GNLY, FCGR3A NK cells, XCL1 NK cells, plasma cell (PC) and other immune subsets. Furthermore, the SPP1 and TENASCIN pathways, associated with tumor immune escape and tumor progression, were also enriched in the HAS group. Additionally, our study also indicated that senescence levels were heterogeneous in the LUAD tumor microenvironment (TME), especially with tumor cells in the LAS group showing higher age scores compared to those in the HAS group.
Conclusions: Collectively, our findings underscore that ARRS through ARGs serves as a robust biomarker for the prognosis in LUAD.
Keywords: cellular senescence; heterogeneity; lung adenocarcinoma; machine learning; tumor microenvironment.
Copyright © 2024 Ru, Cui, Wu, Tan, An, Wu, Ma, Hao, Xiao, Bai, Liu, Xia and Zhao.