To construct and validate a risk score model of angiogenesis-related genes to predict the prognosis of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors with high morbidity and mortality worldwide. Angiogenesis is essential for HCC progression and metastasis. Some angiogenesis-related genes promote this process, whereas other antiangiogenic genes inhibit HCC growth and metastasis. Therefore, finding new potential biomarkers for HCC prognosis prediction and treatment is essential. Public RNAseq and clinical data from TCGA and GEO database, download angiogenesis-related genes from the GeneCards, MSigDB database, through the single factor analysis of Cox, LASSO build risk score-Cox regression analysis model and external validation verified from the GEO. Cox regression analysis, Kaplan Meier (KM) curve, ROC curve, and decision-curve analysis will be used to evaluate and examine the risk score prediction effect of the model. GSVA analysis was used to assess the variation of gene sets between groups, and ClBERSOFT, ESTIMATE, and TIMER databases were used to analyze the immune infiltration in the single-cell level analysis of gene expression differences between cells. Finally, in the three pairs of HCC tissues and tissue adjacent to carcinoma by real-time fluorescent quantitative PCR (qRT_PCR) and western blotting (WB) to evaluate angiogenesis-related genes (ATP2A3 AEBP1 PNMA1, PLAT) expression level in HCC, and AEBP1 was knocked out in HCCLM3 cells, which is to study AEBP1 biological function in HCC. We established a prognostic risk assessment model based on 13 significant genes associated with HCC prognosis by Cox analysis and LASSO-Cox regression analysis. The median was used to divide these patients into high-risk and low-risk groups, and the prognosis of the high-risk group was worse than that of the low-risk group. Through the multivariate Cox regression analysis, it was found that the risk score was an independent predictor of overall survival (OS). The GSVA analysis suggested that the predicted high-risk population showed higher activity in the purine, pyrimidine, and riboflavin metabolic pathways. Compared with the low-risk group, the tumor microenvironment in the high-risk group showed a reduction in the number of cells promoting anti-tumor immunity and an increase in the number of cells inhibiting anti-tumor immunity, as well as a reduction in overall immune infiltration and matrix components. On the single-cell level, it was confirmed that the key genes (AEBP1, ATP2A3, PLAT, and PNMA1) expressed differently between liver cancer and adjacent tissue cell groups. Finally, qRT_PCR and WB results showed that ATP2A3, AEBP1, PNMA1, and PLAT were highly expressed in liver cancer tissue compared to adjacent tissue, and the proliferation, migration, and invasion of HCCLM3 cells were inhibited after knocking out AEBP1. We constructed novel risk score models as prognostic biomarkers for HCC, which has the potential to guide the development of more personalized treatment strategies for HCC patients. In addition, AEBP1 is a potential therapeutic target for HCC.