Over-expression of RRM2 predicts adverse prognosis correlated with immune infiltrates: A potential biomarker for hepatocellular carcinoma

Abstract

Background: Ribonucleotide reductase regulatory subunit M2 (RRM2) has been reported to be an oncogene in some malignant tumors, such as lung adenocarcinoma, oral squamous cell carcinoma, glioblastoma, and breast cancer. However, the clinical significance of RRM2 in hepatocellular carcinoma has been less studied. The aim of this study was to assess the importance of RRM2 in hepatocellular carcinoma (HCC) based on the Cancer Genome Atlas (TCGA) database.

Methods: The RRM2 expression levels and clinical features were downloaded from the TCGA database. Immunohistochemistry results between tumor tissues and normal tissues were downloaded from the Proteinatlas database. Meanwhile, the expression levels of RRM2 in tumor and paraneoplastic tissues were further verified by qRT-PCR and Western Blotting. Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) and protein-protein-interactions (PPI) network were constructed to analyze RRM2-related downstream molecules. In addition, RRM2 expression-related pathways performed by gene set enrichment analysis (GSEA). Association analysis of RRM2 gene expression and immune infiltration was performed by single-sample GSEA (ssGSEA).

Results: The RRM2 expression level in tumor tissues was higher than normal tissues (P <0.001). The elevated expression of RRM2 in HCC was significantly correlated with T stage (P <0.05), pathologic stage (P <0.05), tumor status (P <0.05), histologic grade (P<0.001), and AFP (P <0.001). HCC with higher RRM2 expression was positively associated with worse OS (overall survival), PFS (progression-free survival), and DSS (disease-specific survival). In the univariate analysis, the expression of RRM2, T stage, M stage, pathologic stage, and tumor status were negatively correlated with OS (P <0.05). Further analysis using multivariate Cox regression showed that tumor status (P<0.01) and RRM2 expression (P<0.05) were independent prognostic factors of OS in HCC. GO/KEGG analysis showed that the critical biological process (chromosome condensation and p53 signaling pathway) might be the possible function mechanism in promoting HCC. Moreover, GSEA showed that several pathways were enriched in RRM2 high-expression samples, including PD-1 signaling, cell cycle, P27 pathway, and T cell receptor signaling pathway. RRM2 was significantly correlated with the infiltration level of CD8 T cells, Cytotoxic cells, DCs, Neutrophils, NK cells, and T helper cells (P <0.05).

Conclusion: Over-expression of RRM2 predict adverse prognosis and is correlated with immune infiltrates in HCC. RRM2 may be a significant molecular biomarker for HCC diagnosis and prognosis.

Keywords: The Cancer Genome Atlas; bioinformatics analysis; biomarker; hepatocellular carcinoma; ribonucleotide reductase regulatory subunit M2.

Figure 1

Relationship between RRM2 expression and hepatocellular carcinoma (HCC). (A) Differential expression of RRM2 between tumor tissues and normal tissues. (B) Differential expression of RRM2 between tumor tissues and matched para-cancerous tissues. (C) The mRNA level of RRM2. (D) The protein level of RRM2 in 5 paired of HCC tissues and adjacent tissues using Western blotting. (E) The results of Immunohistochemistry between the normal tissues and the tumor tissues. (F) The correlation analysis of mRNA expression levels between RRM2 and NCAPG was performed. (G) Diagnostic value of RRM2 expression in HCC. *** represents P <0.001.

Figure 2

Correlation between RRM2 expression and clinical characteristics. Relationship of RRM2 expression with (A) T stage, (B) Pathologic stage, (C) Tumor status, (D) Histologic grade, and (E) AFP level. *represents P <0.05, ** represents P <0.01, *** represents P <0.001.

Figure 3

Increased expression of RRM2 indicated poor prognosis in HCC. Of the 374 cases, patients with high expression of RRM2 had significantly (A) shorter OS (P=0.003), (B) DSS (P=0.003), and (C) PFI (P=0.001); (D) A nomogram that integrate RRM2 and other prognostic factors in HCC from TCGA data; (E) The calibration curve of the nomogram.

Figure 4

Network construction for RRM2 correlated genes in HCC. (A) Top 5 genes of positively or negatively correlated with RRM2 were shown in Heatmap. (B) GO analysis and KEGG pathway reveal the underlying mechanism of RRM2 in the promotion of HCC. (C) The PPI network of RRM2 interaction partners generated by STRING and Cytoscape. The color represents the degree score.

Figure 5

GSEA enrichment analysis results. GSEA results showed that (A) SIGNALING BY RHO FTPASES, (B) GPCR ligand binding, (C) PD1 signaling, (D) Cell cycle, (E) DNA replication, (F) T cell receptor signaling pathway, (G) P27 pathway, (H) RB pathway, (I) SRCRPTP pathway. FDR, false discovery rate; NES, normalized Enrichment Score.

Figure 6

Relationship between RRM2 expression and immune infiltration. (A) Relationship between RRM2 expression and immune cells. (B–M) Further analysis in different immune cell subpopulations in the high- and low-expression groups of RRM2. *represents P <0.05, ** represents P <0.01, *** represents P <0.001. NS, Non Significance.