Identification of hub genes-based predictive model in hepatocellular carcinoma by robust rank aggregation and regression analysis

VHL-dependence of EHHADH Expression in a Human Renal Cell Carcinoma Cell Line

The von Hippel-Lindau tumor suppressor gene (VHL) is mutated in up to 90% of clear cell renal cell carcinoma (ccRCC) cases, thus playing a key role in ccRCC pathogenesis. ccRCC can be classified as a metabolic disease in which alterations in fatty acid metabolism facilitate cancer cell proliferation. Enoyl-CoA hydratase and 3-hydroxyacyl CoA dehydrogenase (EHHADH) is an enzyme involved in peroxisomal fatty acid degradation. It is primarily expressed in renal proximal tubule cells, presumably the origin of ccRCC. Although EHHADH is still a relatively unexplored gene, it is known to be differentially expressed in several tumors. In this study, analysis of several databases revealed that EHHADH expression is downregulated in ccRCC samples compared to healthy kidney samples. Moreover, cell culture experiments were performed to investigate the relationship between EHHADH and VHL at the gene and protein level. qPCR and Western blot analyses using the human ccRCC cell line RCC4 revealed that EHHADH is expressed in a VHL-dependent manner. RCC4 cells reconstituted with VHL show significantly higher EHHADH mRNA and protein levels than VHL-deficient RCC4 control cells. These results indicate that the downregulation of EHHADH in ccRCC reported may be due to the loss of VHL function. This study is the first to molecularly characterize EHHADH, a key enzyme in peroxisomal ß-oxidation, in relation to VHL, suggesting a potential pathogenic interaction that is worthy of further investigation.

Prognosis-Related Molecular Subtypes and Immune Features Associated with Hepatocellular Carcinoma

Bioinformatics tools were used to identify prognosis-related molecular subtypes and biomarkers of hepatocellular carcinoma (HCC). Differential expression analysis of four datasets identified 3330 overlapping differentially expressed genes (DEGs) in the same direction in all four datasets. Those genes were involved in the cell cycle, FOXO signaling pathway, as well as complement and coagulation cascades. Based on non-negative matrix decomposition, two molecular subtypes of HCC with different prognoses were identified, with subtype C2 showing better overall survival than subtype C1. Cox regression and Kaplan-Meier analysis showed that 217 of the overlapping DEGs were closely associated with HCC prognosis. The subset of those genes showing an area under the curve >0.80 was used to construct random survival forest and least absolute shrinkage and selection operator models, which identified seven feature genes (SORBS2, DHRS1, SLC16A2, RCL1, IGFALS, GNA14, and FANCI) that may be involved in HCC occurrence and prognosis. Based on the feature genes, risk score and recurrence models were constructed, while a univariate Cox model identified FANCI as a key gene involved mainly in the cell cycle, DNA replication, and mismatch repair. Further analysis showed that FANCI had two mutation sites and that its gene may undergo methylation. Single-sample gene set enrichment analysis showed that Th2 and T helper cells are significantly upregulated in HCC patients compared to controls. Our results identify FANCI as a potential prognostic biomarker for HCC.

Hub genes associated with immune cell infiltration in breast cancer, identified through bioinformatic analyses of multiple datasets

OBJECTIVE

The aim of this study was to identify hub genes associated with immune cell infiltration in breast cancer through bioinformatic analyses of multiple datasets.

METHODS

Nonparametric (NOISeq) and robust rank aggregation-ranked parametric (EdgeR) methods were used to assess robust differentially expressed genes across multiple datasets. Protein-protein interaction network, GO, KEGG enrichment, and sub-network analyses were performed to identify immune-associated hub genes in breast cancer. Immune cell infiltration was evaluated with the CIBERSORT, XCELL, and TIMER methods. The association between the hub gene-based risk signature and survival was determined through Kaplan-Meier survival analysis, multivariate Cox analysis, and a nomogram with external verification.

RESULTS

We identified 163 robust differentially expressed genes in breast cancer through applying both nonparametric and parametric methods to multiple GEO (n = 2,212) and TCGA (n = 1,045) datasets. Integrated bioinformatic analyses further identified 10 hub genes: CXCL10, CXCL9, CXCL11, SPP1, POSTN, MMP9, DPT, COL1A1, ADAMDEC1, and RGS1. The 10 hub-gene-based risk signature significantly correlated with the prognosis of patients with breast cancer. Moreover, these hub genes were strongly associated with the extent of infiltration of CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and myeloid dendritic cells into breast tumors.

CONCLUSIONS

Integrated analyses of multiple databases led to the discovery of 10 robust hub genes that together may serve as a risk factor characteristic of the immune microenvironment in breast cancer.