New roles of N6-methyladenosine methylation system regulating the occurrence of non-alcoholic fatty liver disease with N6-methyladenosine-modified MYC

Abstract

Non-alcoholic fatty liver disease (NAFLD) has become a major chronic disease in contemporary society, affected by N6-methyladenosine (m6A) RNA methylation, one of the most common RNA modifications. Compared with healthy control, m6A RNA methyltransferase 3 (METTL3) and METTL14 increased, while Wilms tumor 1-associated protein (WTAP) and RNA-binding motif protein 15 (RBM15) decreased significantly in NAFLD, and the m6A demethylases fat mass and obesity-associated protein (FTO) elevated. Meanwhile, the m6A binding proteins, YT521-B homology (YTH) domain-containing 1 (YTHDC1), YTHDC2, insulin-like growth factor 2 mRNA binding protein 1 (IGF2BP1), heterogeneous nuclear ribonucleoprotein C (HNRNPC), and HNRNPA2B1 were decreased, while eukaryotic translation initiation factor 3 subunit H (EIF3H) was increased significantly. All these changes of m6A regulators had significant differences between healthy control and NAFLD, but no differences between the NAFL and NASH group. The expression level of RBM15, HNRNPC, and HNRNPA2B1 were related to body fat index. RBM15, YTHDC2, HNRNPC, HNRNPA2B1, and EIF3H were related to steatosis. Also, KIAA1429 and YTH domain family 1 (YTHDF1) were related to lobular inflammation. Taken together, m6A regulators were involved in the occurrence of NAFLD. More importantly, abnormal MYC was determined as a key link to m6A regulation of NAFLD. The higher MYC mRNA level was accompanied by higher HDL cholesterol and unsaturated fatty acid proportions, as well as lower fat mass, glucose, and transaminase. Taken together, dysregulation of m6A methylation caused steatosis and fibrosis, affecting the occurrence of NAFLD, and MYC might be its potential target.

Keywords: MYC; N6-methyladenosine (m6A); RNA methylation; non-alcoholic fatty liver disease (NAFLD); non-alcoholic steatohepatitis (NASH).

FIGURE 1

Overview of expression of m6A regulators in NAFLD. (A) Expression levels of m6A regulators in liver tissues. The t-test analyzed the difference in the expression levels of m6A regulators between the healthy and NAFLD tissues. (B) Boxplot showed the association of expression with clinical stages for m6A regulators. Turkey’s post hoc test was used for comparison between groups to analyze the difference in the expression of m6A regulators among different clinical stages of NAFLD. The results of significant difference analysis: *p < 0.05; **p < 0.01; and ***p < 0.001.

FIGURE 2

Correlations between the expression of m6A regulators and BMI as well as waist. Pearson’s correlation was performed to calculate the correlation between the expression of m6A regulators and BMI as well as waist. The results of significant difference analysis: ns ≥ 0.05; *p < 0.05; **p < 0.01; and ***p < 0.001.

FIGURE 3

Correlations between m6A regulators and steatosis as well as inflammation. Pearson’s correlation was performed to calculate the correlation between the expression of m6A regulators and hepatic steatosis. Spearman’s correlation was performed to calculate the correlation between the expression of m6A regulators as well as lobular inflammation. The results of significant difference analysis: ns ≥ 0.05; *p < 0.05; **p < 0.01; and ***p < 0.001.

FIGURE 4

Correlations of levels of m6A regulators with immune infiltration by combination with a single-cell dataset. (A) Distinct cell clusters were revealed in healthy human liver. (B) Heatmap showed the expression of the m6A regulator in each cell. (C) Histogram showed the relative percentage of different types of cells in each sample of GSE89632. (D) t-tests were performed to analyze the difference in cell percent between healthy control and NAFLD. The results of significant difference were showed as *p < 0.05; **p < 0.01; or ***p < 0.001. (E) Pearson’s test was conducted to analyze the correlations of the m6A regulators’ expression level with the cell percent. The ratio of correlation was shown in color. Blue was a positive correlation and red was negative correlation. Color darker, circle bigger implied a stronger correlation. Demerit marks showed p > 0.05.

FIGURE 5

Explore the coexpression relationship of m6A regulators as well as the target genes and enrichment pathways of m6A related. (A) Pearson’s test was conducted to analyze the coexpression relationship of m6A regulators. (B) Intersection of KEGG enrichment pathways, which were analyzed by GSEA according to stratify by the median expression level of RBM15, YTHDC2, HNRNPA2B1, and HNRNPC, was visualized as a Venn diagram. (C) Intersection of DEGs which were screened according to stratify by the median expression level of RBM15, YTHDC2, HNRNPA2B1, and HNRNPC, was visualized as a Venn diagram. (D) PPI network was construed by common DEGs as well as RBM15, YTHDC2, HNRNPA2B1, and HNRNPC, with the disconnected proteins being hidden. Wider lines indicated stronger evidence of protein interaction.

FIGURE 6

MYC expression profile in NAFLD. (A) Difference was analyzed in the expression of MYC between different clinical stages of NAFLD. The ANOVA was performed to the difference among the multiple clinical stages. Turkey’s post hoc test was used for comparison between groups. (B) Correlations between the expression of MYC and patients’ clinical characteristics. The correlations were all analyzed by Pearson’s test except for fibrosis, lobular inflammation, and ballooning. The correlations between the MYC expression and the characteristic of fibrosis, lobular inflammation, and ballooning were conducted by Spearman’s test. The results of significant difference analysis: ns ≥ 0.05; *p < 0.05; **p < 0.01; and ***p < 0.001.

FIGURE 7

MYC expression profile in NAFLD by using four datasets. (A) PCA showed the datasets of GSE164770, GSE37031, GSE48452, and GSE63067 were merged and adjusted in batches. (B) -test was conducted to analyze the difference in the expression of MYC between healthy controls and NAFLD. (C) Pearson’s test was performed to analyze the correlation between MYC and m6A regulators. The results of significant difference analysis: ns ≥ 0.05; *p < 0.05; **p < 0.01; and ***p < 0.001.