Abstract
BACKGROUND: Studies have demonstrated that propionate metabolism-related genes (PMRGs) are associated with cancer progression. PMRGs are not known to be involved in Hepatocellular carcinoma (HCC). METHODS: In this study, The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases were accessed for HCC-related transcriptome data and clinical information. First, DE-PMRGs were derived by intersecting PMRGs and DEGs between HCC tissues and normal controls. The clusterProfiler R package was then used to enrich DE-PMRGs. In addition, biomarkers of HCC were identified, and a prognostic model was developed. Using functional analysis and tumor microenvironment analysis, new insights were obtained into HCC. The expression of biomarkers was validated using quantitative real-time polymerase chain reaction (qRT-PCR). RESULTS: 132 DE-PMRGs were obtained by intersecting 3690 DEGs and 291 PMRGs. Steroid and organic acid metabolism were associated with these genes. For the construction of the risk model for HCC samples, five biomarkers were identified, including Acyl-CoA dehydrogenase short chain (ACADS), CYP19A1, formiminotransferase cyclodeaminase (FTCD), glucose-6-phosphate dehydrogenase (G6PD), and glutamic-oxaloacetic transaminase (GOT2). ACADS, FTCD, and GOT2 were positive factors, whereas CYP19A1 and G6PD were negative. HCC patients with AUC greater than 0.6 were predicted to survive 1/2/3/4/5 years, indicating decent efficiency of the model. The probability of 1/3/5-survival for HCC was also predicted by the nomogram using the risk score, pathologic T stage, and cancer status. Moreover, functional enrichment analysis revealed the high-risk genes were associated with invasion and epithelial-mesenchymal transition. Significantly, immune cell infiltration and immune checkpoint expression were linked to HCC development. CONCLUSION: This study identified five biomarkers of propionate metabolism that can predict HCC prognosis. This finding may provide a deeper understanding of PMRG function in HCC.