A new prognostic model for lung adenocarcinoma according propionate metabolism related genes: a comprehensive bioinformatic study

BACKGROUND: The prognostic mechanisms of lung adenocarcinoma (LUAD) remain unclear, while the propionate metabolic pathway has been implicated in promoting tumor growth across multiple cancer types. This study aims to elucidate the mechanistic basis by which the propionate pathway influences LUAD progression at the genetic level. METHODS: The TCGA-LUAD cohort was retrieved from The Cancer Genome Atlas (TCGA), and LUAD-related datasets (GSE13213, GSE72079) were obtained from the Gene Expression Omnibus (GEO). Differentially expressed genes (DEGs) between LUAD and normal tissues were first identified, followed by intersection with propionate metabolism-related genes (PMRGs) to derive DE-PMRGs. After partitioning these genes into training and validation sets, a prognostic risk model was constructed via univariate Cox regression and LASSO regression analysis, which was validated in independent cohorts. Patients were stratified into high- and low-risk groups based on the risk model, followed by gene set variation analysis (GSVA), immune microenvironment profiling, and chemosensitivity prediction. RESULTS: A total of 166 DE-PMRGs were identified by intersecting 4,403 DEGs with 531 PMRGs. A risk model was constructed using five characteristic genes (LDHA, KYNU, SLC2A1, CFTR, MAOB) via univariate Cox and LASSO analyses. GSVA revealed 18 activated pathways in the high-risk group (e.g., heme metabolism, P53 signaling), versus 14 pathways in the low-risk group (e.g., E2F targets, mTORC1 signaling). Significant differences were observed in 14 immune cell types (e.g., eosinophils, neutrophils) and 4 immune checkpoints (PDCD1LG2, CD274, CD27, IDO1) between risk groups. LDHA, KYNU, and SLC2A1 were significantly positively correlated with activated CD4 + T cells, γδ T cells, and memory B cells, while CFTR and MAOB were associated with 9 immune cell types (e.g., activated B cells, eosinophils). Eight chemotherapeutic agents were identified to correlate with risk scores via drug sensitivity analysis. CONCLUSION: This study identifies five propionate metabolism-related genes (LDHA, KYNU, SLC2A1, CFTR, MAOB) that may influence LUAD prognosis, providing a scientific foundation for further mechanistic investigations and potential clinical applications. (Liu C, He L, Peng Z, Luo J, A New Prognostic Model for Lung Adenocarcinoma According Propionate Metabolism Related Genes: A Comprehensive Bioinformatic Study, Abstract Book of MEDLIFE2024 & ICBLS2024 (ISBN:979-8-88599-099-8), 2024.).