A network-based approach to identify DNA methylation and its involved molecular pathways in testicular germ cell tumors

Abstract

Background: Testicular germ cell tumors (TGCT) is the most common testicular malignancy threaten young male reproductive health. This study aimed to identify aberrantly methylated-differentially expressed genes and pathways in TGCT by comprehensive bioinformatics analysis. Methods: Data of gene expression microarrays (GSE3218, GSE18155) and gene methylation microarrays (GSE72444) were collected from GEO database. Integrated analysis acquired aberrantly methylated-genes. Functional and pathway enrichment analysis were performed using DAVID database. Protein-protein interaction (PPI) network was constructed by STRING and App Mcode was used for module analysis. GEPIA platform and DiseaseMeth database were used for confirming the expression and methylation levels of hub genes. Finally, Human Protein Atlas database was performed to evaluate the prognostic significance. Results: Totally 604 hypomethylation-high expression and 147 hypermethylation-low genes were identified. The high expressed genes were enriched in biological processes of cell proliferation and migration. The top 8 hub genes of PPI network were GAPDH, VEGFA, PTPRC, RIPK4, MMP9, CSF1R, KRAS and FN1. After validation in GEPIA platform, all hub genes were elevated in TGCT tissues. Only MMP9, CSF1R and PTPRC showed hypomethylation-high expression status, which predicted the poor outcome of TGCT patients. Conclusion: Our study indicated possible aberrantly methylated-differentially expressed genes and pathways in TGCT by bioinformatics analysis, which may provide novel insights for unraveling pathogenesis of TGCT.

Keywords: bioinformatics analysis; genetic etiology; methylation; testicular germ cell tumors.

Fig 1

The integrated analysis of abnormal DNA methylation genes in TGCT. The gene counting numbers were summarized in gene expression datasets (GSE3218 blue and GSE18155 red) and gene methylation dataset (GSE72444 green), respectively. Left panel (A) was represented the hypomethylation and up-regulated genes, while right panel (B) represented the hypermethylation and down-regulated genes. These two graphs were drawn by Bioinformatics & Evolutionary Genomics.

Fig 2

GO and KEGG pathway enrichment analysis. GO analysis covered three groups, including biological process (A), cellular component (B) and molecular function (C). (D) The most significantly enriched molecular pathways of the hypomethylation-upregulated and hypermethylation-downregulated genes were figured out by KEGG analysis. The size of each circle indicates the counting number on each part, while the color represents the P-value of the enrichment analysis.

Fig 3

Protein-protein interaction (PPI) network complex and top three modules of the hypomethylation-upregulated and hypermethylation-downregulated genes. (A) The PPI networks. The size of node indicated the degree value. Yellow nodes represented hub genes. The thickness of edges represented correlation degree. (B-D) Module 1 consists of 30 nodes, module 2 consists of 35 nodes and module 3 consists of 58 nodes.

Fig 4

The mRNA expression levels of 8 hub genes in TGCT. The published online data of gene mRNA expression level were analyzed by GEPIA platform. These 8 hub genes were all up-regulated in the TGCT tissues comparing with the normal tissues, including GAPDH, VEGFA, PTPRC, RIPK4, MMP9, CSF1R, KRAS, FN1. *P<0.05.

Fig 5

Methylation analysis of 8 hub genes in TGCT. The promoters of MMP9, CSF1R and PTPRC were hypomethylated. No obvious change was observed in methylation levels of KRAS and RIPK4, whereas GAPDH, VEGFA and FN1 showed gene hypermethylation in TGCT.

Fig 6

The prognostic significance of hub genes in TGCT. Higher levels of MMP9, CSF1R and PTPRC were associated with poorer overall survival time, respectively.