Research Paper Volume 11, Issue 22 pp 10316—10337

Multi-omics analysis reveals epithelial-mesenchymal transition-related gene FOXM1 as a novel prognostic biomarker in clear cell renal carcinoma

Jing Song 1, , Fangzhou Song 2, , Kun Liu 2, , Wanfeng Zhang 1, , Ruihan Luo 1, , Yongyao Tang 2, , Longke Ran 1, ,

  • 1 Department of Bioinformatics, The Basic Medical School of Chongqing Medical University, Chongqing 400016, China
  • 2 Molecular and Tumor Research Center, Chongqing Medical University, Chongqing 400016, China

received: June 14, 2019 ; accepted: November 8, 2019 ; published: November 19, 2019 ;

https://doi.org/10.18632/aging.102459
How to Cite

Copyright © 2019 Song et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Abstract

Identification of novel clinical biomarker in clear cell renal carcinoma (ccRCC) is warranted. Integrating transcriptome (n=1669), DNA methylation (n=577) and copy number data (n=832), we developed a method to identify driver biomarkers by analyzing the omics-level dynamics of Epithelial-Mesenchymal Transition (EMT)-related genes in ccRCC. We first identified 504 expression dynamic changed genes involved in ccRCC-associated key pathways such as EMT, cell cycle, EGFR and PI3K/AKT signaling. Further analysis identified 229 (90 gene promoters) aberrant expression quantitative trait methylation (eQTM) and 256 genes with expression quantitative trait copy number (eQTCN) alterations. Among them, FOXM1 was affected by both eQTM and eQTCN. FOXM1 copy number amplification (115/500, 23% of patients), occurred in an amplified peak in chromosome 12q13.3, was enriched in late-stage ccRCC samples and was associated with worse survival. FOXM1-overexpressed pT3 patients with distant metastasis showed ~25% shorter overall survival in both training (log-rank P=0.006) and validation (log-rank P=0.018) cohorts. The eQTM-gene hybrid signature (cg00044170 and FOXM1), superior to either gene expression or DNA methylation alone, showed great potential in diagnosing localized ccRCC in training (area under curve = 0.958) and validation datasets. FOXM1 could be a novel prognostic biomarker and shed light for early diagnosis at molecular level in ccRCC.

Abbreviations

ccRCC: clear cell renal carcinoma; EMT: Epithelial–mesenchymal transition; CDH1: E-cadherin; VIM: vimentin; GDC: Genomic Data Commons; GTEx: Genotype-Tissue Expression; GEO: Gene Expression Omnibus; CN: copy number; BH: Benjamini-Hochberg; FDR: false discovery rate; eQTM: expression quantitative trait methylation; eQTCN: expression quantitative trait copy number; OR: odds ratio; EDCG: expression dynamic changed gene; OS: overall survival; PFS: progression-free survival; AUC: area under curve.