Identification and validation of selenium metabolism-related genes in lung adenocarcinoma prognosis using bioinformatics analysis

Yun Chen¹Ping Li²Yang Wang¹Shuai Shen¹Ni Chen¹Hao Peng¹Zheyuan Xu^3*

• ¹The First People's Hospital of Yunnan Province, Kunming, China
• ²Affiliated Calmette Hospital of Kunming Medical University, Kunming, China
• ³The First People’s Hospital of Yunnan Province, Kunming, China

Background: The disruption of selenium metabolism has been associated with tumor progression. However, the prognostic significance and underlying molecular mechanisms of selenium metabolism in lung adenocarcinoma (LUAD) remain inadequately understood. This study primarily aimed to identify and validate prognostic genes related to selenium metabolism in LUAD patients. Methods: Transcriptomic datasets from patients diagnosed with LUAD were meticulously analyzed to identify differentially expressed genes associated with selenium metabolism. The genes selected for the prognostic risk model were determined through various analyses, including differential gene expression assessment, univariate and multivariate Cox proportional hazards regression analyses, as well as other relevant analytical methods. A systematic approach was employed for functional enrichment analysis, characterization of the immune microenvironment, somatic mutation analysis, and evaluation of drug sensitivity to elucidate the mechanisms linked to prognostic genes and risk categories. Finally, a reverse transcription quantitative PCR(RT-qPCR) assay was conducted to validate the expression levels of the identified prognostic genes. Results: F2, GPX3, KMO, and KYNU were identified as prognostic genes for establishing a risk model. The functions of these LUAD prognostic genes were influenced by DNA replication pathways, cell cycle regulation, and quiescent CD4 memory T cells. In the high-risk group(HRG), KEAP1, TTN, and USH2A exhibited the highest mutation rate at 48%, while TTN had an even higher mutation rate of 52% in the low-risk group (LRG). Within the HRG cohort, both cisplatin and gemcitabine demonstrated significant sensitivity. Ultimately, RT-qPCR findings corroborated results obtained from bioinformatics analyses; specifically compared to normal samples: GPX3, KMO, KYNU showed significant downregulation in LUAD tissues while F2 was found to be upregulated in LUAD. Conclusion: This study identified four prognostic genes in LUAD and examined their associated mechanisms of action, which may contribute to the development of novel treatment strategies. The integration of immune characterization with drug sensitivity analysis offers valuable insights for stratified therapy.