Background Lung cancer, particularly the non‐small cell lung cancer (NSCLC) subtypes lung squamous cell carcinoma (LUSC) and lung adenocarcinoma (LUAD), exhibits high heterogeneity and high mortality. This study aimed to explore their tumor microenvironment (TME) features, cellular interactions, and potential therapeutic targets. Methods Using scRNA‐seq datasets (GSE200972, GSE117570, and GSE127465) and TCGA bulk RNA‐seq data, we performed cell clustering, pseudotime trajectory, cell–cell communication, and survival analyses. Batch correction and quality control were applied first, followed by cell type annotation with SingleR, copy number variation inference with InferCNV, and intercellular signaling investigation with CellChat. Results Four epithelial signatures (S1–S4) with distinct gene expression profiles were identified, with S3 specific to LUSC and correlated with high malignancy. Pseudotime analysis revealed distinct differentiation trajectories: S2→S1→S3/S4 in LUSC and S4→S1→S2 in LUAD. In LUSC, S3 interacted with macrophages via the SPP1 and MIF ligand–receptor pairs, involving the PI3K‐Akt pathways; in LUAD, S4 communicated with neutrophils through MIF , linked to interferon‐related pathways. Macrophages played a central role in the TME, with SPP1-CD44 as a key ligand–receptor pair in LUSC and RESISTIN-CAP1 in LUAD. Additionally, CD44 and CD74 expression correlated with prognosis in LUSC and LUAD, respectively. Conclusion This study highlights subtype‐specific epithelial signatures, identifies key signaling pathways (e.g., MIF ), and pinpoints candidate therapeutic targets ( CD44 , CD74 ). These discoveries shed new light on the distinct pathogenic mechanisms of LUSC and LUAD and provide actionable insights to facilitate the clinical translation of subtype‐specific personalized immunotherapies.
Zhang et al. (2026) studied this question.