Introduction: This work carried out comprehensive bioinformatics analysis and explored potential actions of COL22A1 in hepatocellular carcinoma (HCC) pathophysiology.
Methods: Microarray datasets including GSE159220, GSE104766, GSE104310 and GSE33294 were retrieved for bioinformatics analysis. Overlapped differentially expressed genes (DEGs) among GSE159220, GSE104766, GSE104310 and GSE33294 were first subjected to functional annotation analysis. Protein–protein interaction (PPI) network was constructed to reveal the hub genes in the network. The miRNAs-mRNA from hub genes were also constructed. Prognostic analysis of patients with HCC was undertaken to evaluate prognostic values of these hub genes. Functional assays were employed to determine actions of COL22A1 in the HCC cell progression.
Results: Ninety-one common up-regulated DEGs and 55 common down-regulated DEGs among GSE159220, GSE104766, GSE104310 and GSE33294 were identified. Functional annotation analysis revealed that these common DEGs may be correlated with the extracellular matrix. The protein–protein interaction analysis further revealed 20 hub genes. Among these hub genes, high expression levels of COL2A1, COL22A1, HAPLN1, PTHLH and SOX2 showed positive associatin with shorter overall survival (OS) of HCC patients, while low expression levels of COL17A1, REN and TRH exhibited correlation with shorter OS of HCC patients. Furthermore, the COL22A1 was overexpressed in HCC tissues when compared to normal ones, and the promoter methylation level of COL22A1 was significantly lower in the HCC tissues when compared to normal ones. Validation assays deciphered that COL22A1 silencing suppressed HCC cell invasion, migration and epithelial mesenchymal transition.
Conclusion: This study identified several hub genes closely correlated with HCC pathogenesis, and COL22A1 may be involved in the HCC metastasis, but this still requires further mechanistic studies.
Keywords: hepatocellular carcinoma, differentially expressed genes, hub genes, COL22A1, prognosis