Novel Candidate Key Drivers in the Integrative Network of Genes, MicroRNAs, Methylations, and Copy Number Variations in Squamous Cell Lung Carcinoma

Abstract

The mechanisms of lung cancer are highly complex. Not only mRNA gene expression but also microRNAs, DNA methylation, and copy number variation (CNV) play roles in tumorigenesis. It is difficult to incorporate so much information into a single model that can comprehensively reflect all these lung cancer mechanisms. In this study, we analyzed the 129 TCGA (The Cancer Genome Atlas) squamous cell lung carcinoma samples with gene expression, microRNA expression, DNA methylation, and CNV data. First, we used variance inflation factor (VIF) regression to build the whole genome integrative network. Then, we isolated the lung cancer subnetwork by identifying the known lung cancer genes and their direct regulators. This subnetwork was refined by the Bayesian method, and the directed regulations among mRNA genes, microRNAs, methylations, and CNVs were obtained. The novel candidate key drivers in this refined subnetwork, such as the methylation of ARHGDIB and HOXD3, microRNA let-7a and miR-31, and the CNV of AGAP2, were identified and analyzed. On three large public available lung cancer datasets, the key drivers ARHGDIB and HOXD3 demonstrated significant associations with the overall survival of lung cancer patients. Our results provide new insights into lung cancer mechanisms.

Network pharmacology of cancer: From understanding of complex interactomes to the design of multi-target specific therapeutics from nature

“…The methylation status of HOXD3, ARHGDIB, AGAP2 genes was reported to be involved in cell signaling and cytoskeletal organization in squamous cell lung carcinoma. Furthermore, the methylation status of these genes predicted the prognosis of patients (91).…”

The impact of the Cancer Genome Atlas on lung cancer

“…[49] Another study evaluated a wide range of novel cancer drivers for LUSC and found that patients with high expression of the ARHGDIB gene had good prognoses while patients with high expression of the HOXD3 gene had poor prognoses. [50] Another paper studied whether the addition of multiple genomic components to existing prognostic models for LUSC improved predictive powers and found no significant improvement, suggesting that measurements at the transcription level may be more predictive of clinical outcomes than measurements at the DNA\ epigenetic level for LUSC. [51]…”

Genome-wide analysis of microRNA and mRNA expression signatures in cancer

“…We found that approximately 1.3% of the predicted miRNA targets formed reciprocal "up-down" or "down-up" expression relationships with miRNAs, suggesting that the modulation of miRNA-mRNA interactions could be promising biomarkers for cancers. Integrated expression analysis of miRNAs and mRNAs using microarray technology have revealed signatures implicated in ovarian cancer [165,166] , breast cancer [167][168][169][170][171] , metastatic osteosarcoma [172] , gastric cancer [173,174] , pancreatic cancer [175] , oral squamous cell carcinoma [176] , lung cancer [177][178][179] , prostate cancer [180,181] , follicular thyroid tumors [182] , retinoblastoma [183] , etc. The miRNA/mRNA interactome not only includes the interactions between an miRNA and its targets but also comprises the TF-miRNA network.…”