Single-cell dissection and multi-cohort validation identify a hypoxia-related prognostic signature with experimental verification in lung adenocarcinoma.

Background: Lung adenocarcinoma (LUAD) shows marked molecular heterogeneity and diverse clinical outcomes. Tumor hypoxia is a key microenvironmental factor that promotes metabolic reprogramming, invasive phenotypes, therapeutic resistance, and immune remodeling. However, the cellular distribution of hypoxia signaling in LUAD and its value for robust prognostic stratification remain incompletely understood. Methods: We integrated single-cell RNA sequencing data, multi-cohort bulk transcriptomic data, and clinical follow-up information to investigate hypoxia-related features in LUAD. At the single-cell level, cells from a Gene Expression Omnibus LUAD cohort were annotated and hypoxia scores were calculated using ssGSEA. At the bulk level, weighted gene co-expression network analysis and differential expression analysis were combined to identify hypoxia-related key genes. Multiple algorithms were compared to construct a prognostic model, which was further validated in external cohorts. Clinical associations, pathway enrichment, immune infiltration, mutational features, and drug sensitivity were also analyzed. In addition, the top contributing gene was functionally validated Results: Single-cell analysis identified nine major cell populations and revealed marked heterogeneity in hypoxia signaling across cell types. Hypoxia scores were highest in macrophages and fibroblasts, intermediate in epithelial and endothelial cells, and generally low in lymphocyte-related populations. In bulk analyses, 103 hypoxia-related key genes were identified. A prognostic model derived from these genes effectively stratified patients into high-risk and low-risk groups in the TCGA training cohort and two independent external validation cohorts. High-risk patients had significantly poorer survival and more advanced clinical features. Functional analyses showed that the high-risk group was enriched for proliferation- and invasion-related programs, whereas the low-risk group showed enrichment in metabolism- and homeostasis-related pathways. Immune, mutational, and drug sensitivity analyses further supported distinct biological characteristics between risk groups. Moreover, overexpression of VWA1, the key gene with the highest contribution to the model, significantly promoted A549 cell proliferation and migration. Conclusion: This study systematically characterized the cellular heterogeneity of hypoxia signaling in LUAD and established a cross-cohort validated hypoxia-related prognostic model. The hypoxia-associated high-risk state was closely linked to poor prognosis and distinct functional and microenvironmental features. These findings provide new evidence for prognostic assessment and individualized therapeutic exploration in LUAD.