Abstract
In addition to embryonic development, wound healing, and tissue fibrosis, epithelial-mesenchymal transition (EMT) is another process that enhances tumor invasiveness and metastatic activity. Exosomes transport a variety of bioactive components between cells and are crucial for cell‒cell communication in multiple complex biological processes, including cancer. Although a few studies have shown that exosomes encapsulate microRNAs that induce a pro-EMT tumor microenvironment, a systematic survey of potential EMT-related regulators in lung cancer exosomes is still lacking. To identify exosome-related EMT signals that could be employed for precise cancer diagnosis, we used a computational approach to generate a list of candidates EMT regulators and performed experimental validation in lung cancer cell lines. Particularly, we focused on exosome-derived differentially expressed genes that were not previously reported to be associated with lung cancer. We identified 25 exosome-derived protein coding regulators associated with EMT with aberrant transcript expression in both lung squamous cell carcinoma and lung adenocarcinoma. By focusing on clinical features such as survival time, smoking status, tumor purity, and primary tumor subtypes, we found that these 25 genes are important for lung cancer development based on a combined cohort of 9781 lung cancer samples from 24 independent genomics studies. By validating two examples of upregulated and downregulated exosome-derived regulators, we confirmed that TLN1 is a potential oncogene in lung cancer progression, which suggests that it may serve as a diagnostic marker. In summary, our results provide a potential exosome-based biomarker for cancer diagnosis that could be used as a therapeutic tool to control the occurrence of EMT and affect cancer progression.