Single-cell transcriptome analysis reveals the key genes associated with macrophage polarization in liver cancer

The aim of this study was to reveal the key genes associated with macrophage polarization in liver cancer.

Data were downloaded from the Gene Expression Omnibus (GEO) and the Cancer Genome Atlas databases (TCGA). R package Seurat 4.0 was used to preprocess the downloaded single-cell sequencing data, principal component analysis, and clustering. R package SingleR was used to annotate cell types and calculate macrophage polarization scores. Spearman correlation analysis was performed to obtain key genes highly correlated with macrophage polarization in liver cancer. The Tumor IMmune Estimation Resource algorithm was used to analyze the correlation between genes and the infiltration level of macrophages. Finally, the prognostic model was constructed based on 6 macrophage polarization-related genes by multivariate Cox regression analysis. Kaplan-Meier curves and receiver operating characteristic curves validated the prognostic value of the prognostic model.

Two thousand highly variable genes were obtained after the normalization of single-cell profiles. In all, 16 principal components and 15 cell clusters were obtained. Monocytes and macrophages were the main immune cells in the microenvironment of liver cancer tissues. Macrophage polarization scores showed that cluster 5 had the highest degree of polarization. Spearman analysis yielded that a total of 6 key genes associated with macrophage polarization (CD53, TGFBI, S100A4, pyruvate kinase M, LSP1, SPP1), and Tumor IMmune Estimation Resource analysis showed that 6 key genes were significantly positively correlated with macrophage infiltration levels. The model constructed by 6 key genes could effectively evaluate the prognosis of patients with liver cancer.

The key genes associated with macrophage polarization, namely CD53, TGFBI, S100A4, pyruvate kinase M, LSP1, and SPP1, may be potential therapeutic targets for liver cancer.