Dysregulated m6A modification promotes lipogenesis and development of non-alcoholic fatty liver disease and hepatocellular carcinoma

Abstract

Type 2 diabetes mellitus (DM2) is associated closely with non-alcoholic fatty liver disease (NAFLD) by affecting lipid metabolism, which may lead to non-alcoholic steatohepatitis (NASH), fibrosis, and hepatocellular carcinoma (HCC). N⁶-methyladenosine (m6A) RNA methylation is an important epigenetic regulation for gene expression and is related to HCC development. We developed a new NAFLD model oriented from DM2 mouse, which spontaneously progressed to histological features of NASH, fibrosis, and HCC with high incidence. By RNA sequencing, protein expression and methylated RNA immunoprecipitation (MeRIP)-qPCR analysis, we found that enhanced expression of ACLY and SCD1 in this NAFLD model and human HCC samples was due to excessive m6A modification, but not elevation of mature SREBP1. Moreover, targeting METTL3/14 in vitro increases protein level of ACLY and SCD1 as well as triglyceride and cholesterol production and accumulation of lipid droplets. m6A sequencing analysis revealed that overexpressed METTL14 binds to mRNA of ACLY and SCD1 and alters their expression pattern. Our findings demonstrate a new NAFLD mouse model that provides a study platform for DM2-related NAFLD and reveals a unique epitranscriptional regulating mechanism for lipid metabolism via m6A-modified protein expression of ACLY and SCD1.

Keywords: ACLY; METTL14; SCD1; lipid metabolism; type 2 diabetes mellitus.

Graphical abstract

Figure 1

NAFLD mice spontaneously develop steatosis, cirrhosis, and HCC (A) Representative photos for H&E, Masson, and Sirius red staining of liver tissues from 18-month-old control and TAN (Tmem30a-associated NAFLD) mice. (B) Quantitative analysis of collagen volume fraction and positive areas of Sirius red staining in the corresponding liver slides (n = 8). The data are represented as means ± SEM. (C) Percentage of mice with steatosis, cirrhosis, and HCC for control and TAN mice. (D) HE staining of liver sections from TAN mice showing steatosis, lobular inflammation, and hepatocyte ballooning with Mallory-Denk bodies. (E and F) Representative immunofluorescent images and quantification of CD3⁺ or CD68⁺ cells (green), TNF-α (red) staining, and DAPI (blue) in hepatic sections from indicated mice (8 months, n = 6). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 2

Increased cellular stress aggravates DNA damage and promotes cell apoptosis and proliferation in NAFLD mice (A) Representative Oil Red O staining images and quantification of lipid droplet of liver sections for 8-month-old mice (n = 5). (B) The production of ROS in liver was measured by DHE staining with green color and representative images for Nrf2 (red) by immunofluorescent staining from indicated mice (8 months, n = 4). (C) Immunofluorescent images of BiP, PDI (green), and DAPI (blue) in hepatic sections for indicated groups (8 months). (D) Immunoblot pictures and quantitative results for γ-H2AX and H2AX in liver tissues from indicated mice (8 months, n = 3). (E and F) Immunofluorescence images and quantitative analysis of Tunel staining and proliferating cell nuclear antigen (PCNA) for indicated mice. All data are represented as means ± SEM. CTL, control. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. CTL, control.

Figure 3

Enhanced expression of FA synthases accounts for de novo FA synthesis in NAFLD mice (A) Expression levels of lipid biosynthetic enzymes detected by transcriptomic analyses of hepatic tissues from indicated mice (8 months, n = 5). (B and C) The relative mRNA levels of lipid biosynthetic enzymes Srebp1 and Srebp2 detected by qRT-PCR in hepatic tissues from indicated mice (8 months, n = 5). (D) Representative immunoblots and quantification of lipid biosynthetic enzymes for mouse liver lysates (S, steatosis; C, cirrhosis; H, hepatocellular carcinoma). Protein and mRNA expression data were normalized according to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (mean ± S.D., n = 3). CTL, control. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 4

METTL3/14-induced upregulation of ACLY and SCD1 accounts for lipid production in HCC cells (A and B) Representative immunoblots of indicated proteins separately in METTL14- or METTL3-overexpressed LM3 cells and 97H cells with transient METTL14 or METTL3 knockdown. Protein expression data were normalized according to GAPDH levels (mean ± S.D., n = 3). (C and D) Representative images and quantification of lipid droplets in METTL14- or METTL3-overexpressed LM3 cells and 97H cells with METTL14 or METTL3 knockdown by Oil Red O assay. (E) Cellular triglycerides and cholesterol content were measured in LM3 cells expressing empty vector (control), METTL3, and METTL14 and were detected in 97H cells expressing scramble siRNA, METTL3, and METTL14 siRNA. Data are presented as means ± SD and are representative of 3 independent experiments. GADPH, glyceraldehyde-3-phosphate dehydrogenase; OE, overexpressed. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 5

Characterization of m6A modification and identification of ACLY and SCD1 is the target of METTL3/14 (A) Dot blot assays for relative m6A levels on poly-A selected RNA (200 and 100 ng) from LM3 cells with overexpressed-METTL14 and 97H cells with transient METTL14 knockdown (mean ± SD, n = 3). (B) Top sequence motif identified from MeRIP sequencing peaks in control and METTL14-overexpressed LM3 cells. (C) Distribution of elevated m6A peaks generated by METTL14 upregulation across all mRNAs. (D) A Venn diagram was generated from the gene sets enriched for transcripts that were substantially altered after METTL14 overexpression (RNA sequencing), along with those enriched for m6A-modified transcripts (m6A sequencing). Twelve genes were selected according to the overlaps. (E) MeRIP sequencing of the distribution of m6A peaks along ACLY and SCD1 mRNA in METTL14-overexpressed LM3 cells. (F) MeRIP-qPCR analysis of fragmented ACLY and SCD1 mRNA in METTL14-overexpressed LM3 cells. OE, overexpressed. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 6

Upregulation of METTL3/14 accounts for enhanced m6A modification of ACLY and SCD1 (A) The m6A level of poly(A) + RNAs isolated from total RNA of control and TAN liver tissues was indicated by m6A dot blot. (B–E) Representative immunoblots and quantification of METTL3 and METTL14 protein and mRNA level. Protein and mRNA expression data were normalized according to GAPDH (mean ± S.D., n = 3). (F) MeRIP analysis followed by qRT-PCR was applied to assess the m6A modification of ACLY and SCD1 in liver tissues from control and TAN mice. The enrichment of m6A in each group was calculated by m6A-IP/input and IgG-IP/input. CTL, control. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 7

The augmentation of ACLY and SCD1 is correlated with their m6A modification in HCC patients (A) Representative IHC images of METTL14, METTL3, ACLY, and SCD1 staining in human HCC tumor (T) or their adjacent tissues (P). (B) Quantification of METTL14, METTL3, ACLY, and SCD1 expression in 80 pairs of human HCC samples. (C) Correlation analysis of METTL14, METTL3, ACLY, and SCD1 expression in 80 pairs of human HCC samples. (D) The m6A level of poly(A) + RNAs isolated from total RNA of 80 pairs of human HCC samples was indicated by m6A dot blot. (E) Western blot analysis of METTL14, METTL3, ACLY, and SCD1 expression in human HCC tumor (T) and their adjacent tissues (P). (F) MeRIP analysis followed by qRT-PCR was applied to assess the m6A modification of ACLY and SCD1 in human HCC samples. The enrichment of m6A in each group was calculated by m6A-IP/input and IgG-IP/input. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; ns = not significant.