Abstract / Summary
Lung adenocarcinoma (LUAD) is characterized by high incidence and low survival rate. The molecular network of ferroptosis in lung adenocarcinoma requires further investigation, and this study aims to develop a ferroptosis-related prognostic model and identify the potential biomarkers. Ferroptosis-related genes (FRGs) in the GSE115002, GSE31210, and TCGA-LUAD datasets were analyzed using differential expression analysis, weighted gene co-expression network analysis (WGCNA), least absolute shrinkage and selection operator (LASSO) regression, and multivariate analysis. Prognosis-associated candidate biomarkers were screened, based on which a prognostic signature model was constructed. Correlations between the expression of key genes and tumor microenvironment as well as genetic mutation heterogeneity were analyzed. Single-cell RNA sequencing (scRNA-seq) analysis was performed using the GSE189357 dataset to explore the expression characteristics and inferred biological correlations of the candidate genes. We identified 16 candidate intersecting FRGs in the GSE115002 cohort via differential expression analysis and WGCNA for LASSO and multivariate analysis. An eight-gene prognostic signature consisting of RGS4, SLC1A4, SC5D, DHCR7, SLC25A1, DDIT4, SLC7A11, and SLC7A5 was developed in the TCGA-LUAD cohort and validated in the GSE31210 cohort. RGS4 expression was associated with favorable prognosis and with differences in immune and inflammatory profiles in LUAD, while its differential expression was associated with distinct landscapes of mutated genes. Single-cell sequencing analysis of GSE189357 suggested that RGS4 expression was predominantly observed in fibroblasts. RGS4-enriched fibroblast subclusters showed transcriptional features suggestive of inflammatory signaling, cytoskeletal organization, and extracellular matrix remodeling. This study developed an FRG-based prognostic signature for LUAD survival prediction with favorable predictive performance. Public single-cell transcriptomic analysis suggested fibroblast enrichment of RGS4, supporting its potential relevance to the LUAD tumor microenvironment. RGS4 was identified as a prognosis-associated gene, and its potential as a prognostic biomarker requires future experimental and clinical cohort validation.