A 17 gene panel for non‐small‐cell lung cancer prognosis identified through integrative epigenomic‐transcriptomic analyses of hypoxia‐induced epithelial–mesenchymal transition. Issue 7 (29th May 2019)

• Record Type: Journal Article
• Authors: Chen, Yue‐Lei; Zhang, Yihe; Wang, Junwen; Chen, Na; Fang, Weiying; Zhong, Jianing; Liu, Yi; Qin, Rui; Yu, Xinxin; Sun, Zhongsheng; Gao, Fei
• Is Part Of: Molecular oncology. Volume 13:Issue 7(2019)
• Page Start: 1490
• Page End: 1502
• Publication Date: 2019-05-29
• Subjects: biomarker -- EMT -- epigenetics -- metastasis -- non‐small‐cell lung cancer; Cancer -- Molecular aspects -- Periodicals; 616.994005
• DOI: 10.1002/1878-0261.12491
• Languages: English
• ISSNs: 1574-7891

Abstract

Abstract : As a critical feature of the tumor microenvironment, hypoxia is known to be a potent inducer of tumor metastasis, and it has been proposed that the initial steps in metastasis involve epithelial–mesenchymal transition (EMT). The strong correlation among hypoxia, EMT, and metastasis suggests that integrative assessment of gene expression and the DNA modification program of hypoxia‐induced EMT via high‐throughput sequencing technologies may increase our understanding of the molecular basis of tumor invasion and metastasis. Here, we present the genomewide transcriptional and epigenetic profiles of non‐small‐cell lung cancer (NSCLC) cells under normoxic and hypoxic conditions. We demonstrate that hypoxia induces EMT along with dynamic alterations of transcriptional expression and epigenetic modifications in both A549 and HCC827 cells. After training using a dataset from patients with invasive and noninvasive lung adenocarcinomas with an artificial neural network algorithm, a characteristic 17‐gene panel was identified, consisting of genes involved in EMT, hypoxia response, glycometabolism, and epigenetic modifications. This 17‐gene signature clearly stratified NSCLC patients with significant differences in overall survival across three independent datasets. Our study may be suitable as a basis for further selection of gene signatures to potentially guide prognostic stratification in patients with NSCLC. Abstract : A 17‐gene panel was identified with an artificial neural network algorithm, involving genes identified from a hypoxia‐induced cell model of epithelial–mesenchymal transition. This 17‐gene signature clearly stratified NSCLC patients with significant differences in overall survival and thereafter may be suitable as a basis for further selection of gene signatures to potentially guide prognostic stratification in patients with NSCLC.