Integrative multi-omics identifies endoplasmic reticulum stress-related prognostic biomarkers and CD99-mediated cell–cell communication in lung adenocarcinoma

BACKGROUND: Endoplasmic reticulum stress (ERS) plays a crucial role in tumor progression, yet its impact on lung adenocarcinoma (LUAD) heterogeneity and cell-cell communication (CCC) is not fully understood. Understanding how ERS shapes the tumor microenvironment (TME) is critical for the development of targeted therapies. This study investigated the impact of ERS on LUAD progression, immune evasion, and drug resistance by identifying ERS-related prognostic biomarkers and analyzing their effects on TME remodeling and CCC networks. METHODS: Single-cell transcriptomics, bulk RNA sequencing, and spatial transcriptomics were integrated to identify ERS signatures and key genes in LUAD, investigating their effects on the TME, chemotherapy sensitivity, immune response, and CCC. Consensus clustering based on the expression of ERS-related genes classified LUAD cases into two subtypes. A machine-learning-based prognostic model was constructed. CCC networks were analyzed to identify critical mediators of ERS-driven interactions. Immunohistochemistry and qPCR were used to verify the correlation between the key genes and the occurrence and prognosis of LUAD. RESULTS: Patients with LUAD were classified into two subtypes according to the expression of ERS-related genes. Subtype 1 was associated with significantly worse prognosis, differences in chemotherapy sensitivity, and altered immune checkpoint expression. Univariate Cox regression revealed 53 prognostic genes. A machine-learning-based prognostic model demonstrated robust predictive ability and identified 29 ERS-related key genes. These genes significantly affected chemotherapy sensitivity and immune checkpoint expression. CD99-CD99 interactions were identified as key mediators of ERS-related CCC. The pan-cancer analysis revealed that key genes influence cell proliferation and anti-tumor immunity across various cancer types. CONCLUSION: This study comprehensively analyzes ERS heterogeneity in LUAD, identifying novel key genes involved in the regulation of tumor proliferation, chemotherapy sensitivity, anti-tumor immunity, TME remodeling, and CCC networks. Furthermore, it highlights the critical role of CD99 -mediated CCC in ERS, offering potential biomarkers and therapeutic targets for LUAD.