AUTHOR=Han Cuifang , Chen Jiaru , Huang Jing , Zhu Riting , Zeng Jincheng , Yu Hongbing , He Zhiwei 

TITLE=Single-cell transcriptome analysis reveals the metabolic changes and the prognostic value of malignant hepatocyte subpopulations and predict new therapeutic agents for hepatocellular carcinoma

JOURNAL=Frontiers in Oncology

VOLUME=Volume 13 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1104262

DOI=10.3389/fonc.2023.1104262

ISSN=2234-943X

ABSTRACT=Background: A hallmark of cancer is the reprogramming of energy metabolism which causes rapid cell growth and proliferation and the development of hepatocellular carcinoma (HCC) is often associated with extensive metabolic disturbances. Single cell RNA sequencing provides a better understanding of cellular behavior in the context of complex tumor microenvironments by analyzing individual cell populations.
Methods: Using The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) database single cell RNA sequencing and clinical data to investigate the metabolic pathways in hepatocellular carcinoma malignant epithelial subsets. Principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) analysis were applied to identify six cell subpopulations, namely, T/NK cells, hepatic epithelial cells, macrophages, endothelial cells, fibroblasts, and B cells. The gene set enrichment analysis (GSEA) was performed to explore the existence of pathway heterogeneity across different cell subpopulations. Univariate Cox analysis was used to screen genes differentially related to The Overall Survival in TCGA-LIHC patients based on scRNA-seq and bulk RNA-seq datasets, and LASSO analysis was used to select significant predictors for incorporation into multivariate Cox regression. Connectivity Map (CMap) was applied to analysis drug sensitivity of risk models and targeting of potential compounds in high risk groups. 
Results: Analysis of TCGA-LIHC survival data revealed the molecular markers associated with HCC prognosis, including MARCKSL1, SPP1, BSG, CCT3, LAGE3, KPNA2, SF3B4, GTPBP4, PON1, CFHR3, and CYP2C9. The RNA expression of 11 prognosis-related differentially expressed genes in normal human hepatocyte cell line MIHA and HCC cell lines HCC-LM3 and HepG2 were compared by qPCR. CPTAC database analysis and immunohistochemical analysis using HPA confirmed higher KPNA2, LAGE3, SF3B4, CCT3 and GTPBP4 protein expression and lower CYP2C9 and PON1 protein expression in HCC tissues. The results of drug sensitivity analysis and target compound screening for risk models shows that mercaptopurine is a potential anti-HCC drug.
Conclusion: The prognostic genes associated with metabolic changes in a hepatocyte subpopulation and comparison of liver malignancy cells to normal liver cells may provide insight into the metabolic characteristics of HCC and the potential prognostic biomarkers of tumor-related genes and contribute to developing new treatment strategies for individuals.