Abstract
METHODS: We compared differentially expressed genes (DGEs) in HBV-positive and -negative tumor samples and conducted a Spearman correlation study between the DGEs and HBV titers within The Cancer Genome Atlas (TCGA). Moreover, we validated the results of our in-house samples. RESULTS: In this study, we discovered a series of genes that correlated statistically with HBV infection based on the TCGA database. These genes were related to increased inflammation and some oncogenic signaling pathways via Gene Set Enrichment Analysis (GSEA). PDZK1 is an ideal gene, which mostly relates positively to HBV infection; moreover, it is overexpressed in human HCC, especially in those HBV-infected HCCs. After analyzing the TCGA data and performing a verification study using our own samples, PDZK1 expression was investigated to be significantly associated with PI3K-Akt signaling and fatty acid metabolism. Further, single-sample GSEA analysis of tumor immune cell infiltration gene sets revealed that high PDZK1expression in HCC tissues was significantly associated with increased tumor-associated macrophages (TAMs) and regulatory T cells(Tregs). CONCLUSIONS: PDZK1 is an HBV infection-related gene, which plays oncogenic roles, possibly due to enhancing PI3K-Akt, fatty acid usage in tumor cells and TAMs, and Treg-induced immunosuppression.