A machine learning-derived hypoxia- and lactylation-associated gene signature for prognostic stratification and immune landscape characterization in lung adenocarcinoma.

Lung cancer is one of the most frequently diagnosed cancers and the leading cause of cancer-related death worldwide. Hypoxia and histone lactylation are emerging regulators of tumor progression and immune escape. However, their concurrent enrichment and immune-associated features in lung adenocarcinoma (LUAD) remain unclear. In this study, we employed a combination of hypoxia- and lactylation-related genes to construct a new prognostic model using LASSO, XGBoost, and Random Forest algorithms. Single-cell RNA-seq (scRNA-seq) data were integrated to assess cell-type-specific expression. Functional enrichment, immune infiltration, TIDE-based immunotherapy response prediction and drug sensitivity analyses were performed to elucidate the biological characteristics of the subtypes and to explore their potential clinical and therapeutic implications. Promotion of lung cancer cell proliferation by the core gene polyadenylate-binding protein 1 (PABPC1) was verified by qRT-PCR, CCK-8, colony formation, wound healing and Transwell assays in two LUAD cell lines. In conclusion, we successfully established a new hypoxia- and lactylation-related gene-based model with value for predicting the prognosis of LUAD patients. By integrating multiple bioinformatic analyses and combining them with cell-based experimental validation, we found that cytoplasmic PABPC1 is a potential prognostic marker for lung cancer.