Long-term exposure to carbon nanotubes (CNTs) has been reported to induce malignant transformation. This study aimed to screen candidate biomarkers for monitoring occupational workers to prevent the development of lung cancer. mRNA (GSE56104) and methylation (GSE153246) profiles of lung epithelial BEAS-2B cells exposed to malignant transformation dose of single-walled CNTs or control medium were downloaded from Gene Expression Omnibus database. A total of 1513 differentially expressed genes (DEGs) and 912 differentially methylated genes (DMGs) were identified using LIMMA method. The weighted correlation network analysis identified blue and turquoise modules were associated with malignant transformation of BEAS-2B cells, 124 DMGs of which were overlapped with DEGs. The mRNA and methylation levels of four methylation-driven DEGs were validated in both lung adenocarcinoma (LUAD) and squamous cell carcinomas (LUSC) of The Cancer Genome Atlas dataset and they were associated with overall survival of LUAD patients. Downregulation of PXMP4 and MCOLN2, while upregulation of MET was confirmed in both LUSC and LUAD via Human Protein Atlas analysis. PXMP4 and MET protein levels were also supported in the proteomic analysis of LUAD. Receiver operating characteristic (ROC) curve analysis showed the combination of four genes may be the optimal biomarker for predicting lung cancer, with the area under ROC curve >0.9. Function analysis revealed BARX2 may interact with CCND1 to regulate cell cycle; MET and PXMP4/MCOLN2 may positively correlate with CCR5/IL-6 or GATA3/HLA-DPB1/HLA-DPA1 to involve immune regulation. In conclusion, these four methylation-driven genes may be candidate prognostic and diagnostic biomarkers for single-walled CNT-related lung cancer.
Keywords: Bioinformatics; Carbon nanotubes; Lung cancer; Malignant transformation.
Copyright © 2020 Elsevier Inc. All rights reserved.