Integrated Network Analysis Reveals FOXM1 and MYBL2 as Key Regulators of Cell Proliferation in Non-small Cell Lung Cancer

Abstract

Background: Loss of control on cell division is an important factor for the development of non-small cell lung cancer (NSCLC), however, its molecular mechanism and gene regulatory network are not clearly understood. This study utilized the systems bioinformatics approach to reveal the "driver-network" involve in tumorigenic processes in NSCLC. Methods: A meta-analysis of gene expression data of NSCLC was integrated with protein-protein interaction (PPI) data to construct an NSCLC network. MCODE and iRegulone were used to identify the local clusters and its upstream transcription regulators involve in NSCLC. Pair-wise gene expression correlation was performed using GEPIA. The survival analysis was performed by the Kaplan-Meier plot. Results: This study identified a local "driver-network" with highest MCODE score having 26 up-regulated genes involved in the process of cell proliferation in NSCLC. Interestingly, the "driver-network" is under the regulation of TFs FOXM1 and MYBL2 as well as miRNAs. Furthermore, the overexpression of member genes in "driver-network" and the TFs are associated with poor overall survival (OS) in NSCLC patients. Conclusion: This study identified a local "driver-network" and its upstream regulators responsible for the cell proliferation in NSCLC, which could be promising biomarkers and therapeutic targets for NSCLC treatment.

Keywords: gene expression; gene network; meta-analysis; non-small cell lung cancer; systems bioinformatics.

Figure 1

Functional annotation of up-regulated genes in NSCLC (A); and down-regulated genes in NSCLC compared to control (B). GO, Gene Ontology; BP, Biological Processes; MF, Molecular Function; CC, Cell Component; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Figure 2

NSCLC network showing protein-protein integration network in NSCLC. Red node indicates up-regulated; while blue node indicates down-regulated mRNAs in NSCLC compared to normal. Size of the node is based upon degree of connectivity of the node. Edges in the network represent direct interactions between nodes.

Figure 3

Regulators of gene cluster 1–5. Each column indicates gene in a cluster, while each row indicates TF identified by iRegulone (A–E). Up-regulated DEGs in the cluster is red with positive log2FC; while down-regulated DEGs is blue with negative log2FC. TF binding with the mRNA is in purple, while non-binding in cyan. “NaN” If the log2FC is not available in our list of DEGs. (F) Venn diagram showing common TFs regulating different clusters.

Figure 4

(A) Correlation analysis of expression of genes in Cluster 1 and its TFs. Expression of gene is on Y-axis while TF is on X-axis. (B) Overall survival analysis in NSCLC patients using Kaplan-Meier plots for genes of Cluster 1 and associated TFs.

Figure 5

miRNA network of Cluster 1 showing miRNAs targeting mRNAs and TFs of Cluster 1. Red node indicates up-regulated; while the blue node indicates down-regulated expression in NSCLC compared to normal. Size of the node is based upon degree of connectivity of the node. Nodes shape with triangle, round, and diamond represent TFs, mRNAs, and miRNAs, respectively.

Figure 6

(A) OncoPrint of genes in Cluster 1 and associated TFs alteration in NSCLC. Lollipop plot with distribution of mutations in NSCLC across protein domains of (B) ASPM; (C) NUF2; and (D) CENPF.