Construction of TME and Identification of crosstalk between malignant cells and macrophages by SPP1 in hepatocellular carcinoma

Abstract

Liver cancer accounts for 6% of all malignancies causing death worldwide, and hepatocellular carcinoma (HCC) is the most common histological type. HCC is a heterogeneous cancer, but how the tumour microenvironment (TME) of HCC contributes to the progression of HCC remains unclear. In this study, we investigated the immune microenvironment by multiomics analysis. The tumour immune infiltration characteristics of HCC were determined at the genomic, epigenetic, bulk transcriptome and single-cell levels by data from The Cancer Genome Atlas portal and the Gene Expression Omnibus (GEO). An epigenetic immune-related scoring system (EIRS) was developed to stratify patients with poor prognosis. SPP1, one gene in the EIRS system, was identified as an immune-related predictor of poor survival in HCC patients. Through receptor-ligand pair analysis in single-cell RNA-seq, SPP1 was indicated to mediate the crosstalk between HCC cells and macrophages via SPP1-CD44 and SPP1-PTGER4 association. In vitro experiments further validate SPP1 can trigger the polarization of macrophages to M2-phenotype tumour-associated macrophages (TAMs).

Keywords: Crosstalk; Hepatocellular carcinoma (HCC); Prognosis; SPP1; Tumour microenvironment (TME); Tumour-associated macrophages (TAMs).

Fig. 1

Construction of the TME in HCC. a Comparison of the distributions of immune scores between tumours in different TNM stages. b Kaplan–Meier curve showing the association between overall survival and immune scores in HCC patients. c Immune cell population distribution in the high- and low-immunity cohorts. d Statistical analysis of the immune cell population in the high- and low-immunity cohorts. e Volcano plot indicating the DEGs between the high- and low-immunity cohorts. f ssGSEA revealing the association between significant hallmarks and immune score. g GO analysis. The x axis indicates the overlapping numbers between each GO term and DEGs. The colour of the bars represents the adjusted p values (with FDR correction)

Fig. 2

DEGs and related pathway changes in the immune infiltration-dependent status. a Heatmap displaying the expression changes of chemotactic factors. b Heatmap displaying the expression changes of genes involved in the B7-CD28 family and TNF family. c Metagenes related to IgG, interferon, LCK, MHC-II and STAT1. d Genes involved in MHC-II family and MHC-II-related pathway alterations

Fig. 3

Genomic landscape of HCC in the high-immunity and low-immunity cohorts. a, b OncoPlot showing the mutation distribution of the top 15 most frequently mutated genes in the high-immunity and low-immunity cohorts. The upper panel displays the mutation frequency of each HCC sample. The middle panel displays the types of mutations. The bar plot on the left indicates the frequency and mutation type of genes mutated in the low-immunity and high-immunity cohorts, respectively. The lower panel displays the clinical features (tumour stage, grade and immunity cohort) of each sample. The bottom panel shows the frequency and distribution of mutation types in each patient. c Mutation numbers of DELs, INSs and SNPs in the high- and low-immunity cohorts. d Forest plot displaying the top 3 most significantly differentially mutated genes between the two cohorts. e Comparison of the rate of the 10 most frequent gene mutations in the low- and high-immunity cohorts. f Heatmap displaying the mutually co-occurring and exclusive mutations of the top 20 frequently mutated genes in the high- and low-immunity groups. The colour and star symbol in each cell represent the statistical significance of the exclusivity or co-occurrence for each pair of genes

Fig. 4

Construction of an epigenetic immune-related scoring system (EIRS). a Volcano plot showing the differentially methylated genes (DMGs). b GO analysis of DMGs. c Venn plot indicating the key immune-related genes. d LASSO Cox regression. e Coefficient for each gene in the EIRS signature in the LASSO Cox model. f Genes with a significant association with OS in the EIRS signature and the HR value. g Kaplan–Meier curve showing the association between overall survival and EIRS values in HCC patients. h PD1, PD-L1 and CTLA4 expression in the high-EIRS cohort and low-EIRS cohort. i Association between EIRS value and immune cell abundance. The colour and star symbol represent the statistical significance for Pearson’s coefficient. The cross symbol indicates insignificance. j Association of EIRS and CD8 + T cell. k Association of EIRS and M0 macrophages

Fig. 5

SPP1 acts as an immune-related prognostic factor in HCC. a Immunofluorescence staining of SPP1 in SK cells. b Relative mRNA level between SK cells and SPP^KD SK cells. c Wound-healing assay of SK cells and SPP^KD SK cells. d Quantification of the wound-healing assay between SK cells and SPP^KD SK cells. e Migration assay of SK cells and SPP^KD SK cells. f Quantification of the migration assay between SK cells and SPP^KD SK cells. (E) Migration assay of SK cells and SPP^KD SK cells. g Colony formation assay of SK cells and SPP^KD SK cells. h Quantification of the colony formation assay between SK cells and SPP^KD SK cells. i, j SPP1 expression in tumour and normal tissues according to IHC staining and HE staining. k Association of SPP1 protein and RFS in HCC patients. l Association of SPP1 protein and OS in HCC patients

Fig. 6

Single-cell transcriptome analysis for tumour heterogeneity in HCC patients. a Volcano plot indicating the DEGs between the SPP1 high-expression and SPP1 low-expression groups. B GSEA between the SPP1 high-expression and SPP1 low-expression groups. c Association of M0 macrophage abundance with SPP1, CSF1 and CSF1R expression. d UMAP showing the distribution of various cell types. e Hierarchical clustering of cell types. f UMAP indicating SPP1 expression in different cell types. g UMAP showing the distribution of various cell types, including SPP1 + malignant cells and SPP1-malignant cells. h Histogram showing the expression level of SPP1. i Heatmap indicating the number of LR pairs between different cell types. j Heatmap indicating the ligand-receptor pairs between different cells