Integration of Single-Cell and Bulk Transcriptomes to Identify a Poor Prognostic Tumor Subgroup to Predict the Prognosis of Patients with Early-stage Lung Adenocarcinoma.

Single-cell RNA sequencing (scRNA-seq) has emerged as a pivotal technology for investigating novel therapeutic targets in cancer. Despite its significance, there remains a scarcity of studies utilizing this technology to address treatment strategies specifically tailored for early-stage lung adenocarcinoma (LUAD). Consequently, this study aimed to investigate the tumor microenvironment (TME) characteristics and develop a prognostic model for early-stage LUAD. The markers identifying cell types were obtained from the CellMarker database and published research. The SCEVAN package was employed for identifying malignant lung epithelial cells. Single-cell downstream analyses were conducted using the SCP package, encompassing gene set enrichment analysis, enrichment analysis, pseudotime trajectory analysis, and differential expression analysis. Calibration curves, receiver operating characteristic curves, and decision curve analysis were employed to assess the performance of the prognostic model for LUAD. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR), western blot, cell transfection, cell proliferation, and cell invasion assays were performed to validate the expression and biological function. Seven cell types were distinguished in the scRNA-seq dataset through the utilization of cell markers documented in published literature. Four subpopulations of early-stage LUAD tumor cells exhibited a high degree of heterogeneity. The prognostic model constructed by and showed a great prediction for distinguishing the early-stage LUAD and normal tissues. The validation of and expression levels was carried out through both RT-qPCR and western blot analyses. Eventually, experiments, including CCK8, colony formation, EdU, and transwell assays, confirmed that and could promote LUAD cell proliferation and migration. Our study provided a comprehensive characterization of the TME in LUAD through integrative single-cell and bulk transcriptomic analyses. We identified dynamic transitions from normal epithelial cells to tumor cells, revealing the heterogeneity and evolution of malignant LUAD cells. The novel prognostic model based on KRT8 and PERP demonstrated robust predictive performance, offering a promising tool for early-stage LUAD risk stratification. Functional experiments further confirmed that KRT8 and PERP promote tumor proliferation and migration, providing new insights into their roles as therapeutic targets.