N6-Methyladenosine Modification Participates in the Progression of Hepatitis B Virus-Related Liver Fibrosis by Regulating Immune Cell Infiltration

Abstract

Aim: N6-methyladenosine (m6A) modification has been demonstrated to play an important part in hepatitis B virus (HBV) infection and immune response. This study aims to further investigate whether m6A modification plays an important role in the progression of HBV-related liver fibrosis through the regulation of immune cell infiltration.

Methods: In this study, 124 chronically HBV infected cases were obtained from the Gene Expression Omnibus database. In total, 489 m6A-and-stage related genes were selected to be associated with the m6A modification and the stage of liver fibrosis. Based on these genes, we identified two distinct gene clusters, gene clusterA and gene clusterB. The immune characteristics of the two clusters were comprehensively compared. The m6A-S score was constructed as quantification of individual m6A status. The correlations between m6A regulators and infiltrating immune cells were examined and compared in different pairs of groups with various m6A traits.

Results: Biological functions, immune cell infiltration, and cytokines expression were compared between the two gene clusters proving that the gene clusterB was more immune active and had a more advanced liver fibrosis stage. The m6A-S score was associated with immune infiltration and the progression of liver fibrosis. Five different grouping conditions with different m6A traits were set up. According to the intersection of significant genes and cells, ALKBH5 interacting with macrophage and WTAP interacting with nature killer T cells may be key points in the progress of liver fibrosis.

Conclusions: N6-methyladenosine modification is closely related to the immune cell infiltration and the fibrosis stage of chronic HBV-infected liver tissue. It provides us a better understanding of the progression of liver cirrhosis via evaluating the m6A modification pattern and immune infiltration characteristics.

Keywords: N6-methyladenosine; hepatitis B virus; immune cell infiltration; immunity; liver fibrosis.

Figure 1

Identification of N6-methyladenosine (m6A) modification patterns. (A–D) Consensus clustering of the 28 m6A regulators for k = 2–5. (E) Principal component analysis (PCA) for the expression profiles of m6A regulators showed a remarkable separation between the two m6A patterns. (F) Heatmap of the 28 m6A regulators and clinical traits in m6A-Pattern I and m6A-Pattern II. (G) Correlations among the m6A regulators. (H) Gene ontology analysis of m6A-and-stage related genes.

Figure 2

Identification of m6A-and-stage related gene subtypes. (A–D) Consensus clustering of the 489 m6A-and-stage related genes for k = 2–5. (E) Heatmap of the 489 m6A-and-stage related genes and clinical traits in gene clusterA and gene clusterB. (F) Expression heatmap of the 28 m6A regulators in gene clusterA and gene clusterB. (G) The heatmap of gene set variation analysis (GSVA) enrichment analysis was used to visualize biological pathways between gene clusterA and gene clusterB. (H) Different immune cell infiltration of two gene clusters (I) Expression of cytokines between the two gene clusters (***p < 0.001; **p < 0.01; *p < 0.05).

Figure 3

Construction of m6A-S score. (A) Difference of m6A-S score between two m6A modification patterns. (B) Difference of m6A-S score between two gene clusters. (C) Comparison of m6A-S scores among patients in different liver fibrosis stages. (D) Correlations between m6A-S score and infiltrating immune cells. (E) The receiving-operating characteristic (ROC) curve analysis of the m6A-S score were used to differentiate the mild and severe stage of liver fibrosis. The area under the curve was 0.820 with p < 0.001 and the 95% CI was 0.754–0.902. (F) The proportion of disease stages in high and low m6A-S score group. (G) Sankey diagram showing the relationship among m6A patterns, gene clusters, m6A-S score groups, and fibrosis severity.

Figure 4

Screening for critical genes and immune cells. The correlation between each infiltrating immune cell type and each m6A regulator was analyzed using Pearson's correlation analyses and compared in different pairs of subgroups. (A) Patients of m6A-Pattern I. (B) Patients of m6A-Pattern II. (C) Patients of gene clusterA. (D) Patients of gene clusterB (***p < 0.001; **p < 0.01; *p < 0.05).