Co-expression network analysis identifies key modules and hub genes implicated in esophageal squamous cell cancer progression

Esophageal squamous cell cancer (ESCC) is one of the most common malignant tumors with high mortality in China, whereas the underlying molecular mechanisms of ESCC are largely unknown. In this study, we aimed to identify the abnormally expressed genes which may play crucial roles in ESCC progression. We performed a weighted gene co-expression network analysis with 1494 differentially expressed genes identified in GSE53624 dataset, and found that the turquoise and black modules were highly correlated with tumor grade of ESCC. Gene ontology enrichment analysis showed that the turquoise module was associated with development and differentiation, and the black module was involved in cell cycle related processes. A total of 3 hub genes: PXMP2, TMEM45B and NEK2 were identified in two modules. Gene expression of 3 genes was associated with overall survival of ESCC patients. Receiver operating characteristic curve analysis showed that all of 3 genes could effectively distinguish ESCC from normal samples in GSE53624 dataset as well as in GSE53622 and TCGA datasets. Gene set enrichment analysis showed that PXMP2 and TMEM45B might be linked to lipid metabolism, and NEK2 was probably related to cell cycle and DNA repair. In addition, both PXMP2 and TMEM45B showed significant hypermethylation of the promoter regions in ESCC, whereas the promoter region of NEK2 was hypomethylated. These findings indicate that PXMP2, TMEM45B and NKE2 may contribute to ESCC progression.