GR-mediated FTO transactivation induces lipid accumulation in hepatocytes via demethylation of m6A on lipogenic mRNAs

Abstract

Chronic stress or excessive exposure to glucocorticoids (GC) contributes to the pathogenesis of non-alcoholic fatty liver disease (NAFLD). Glucocorticoid receptor (GR) mediates the action of GC, but its downstream signalling is not fully understood. Fat mass and obesity associated (FTO) protein and its demethylation substrate N6-methyladenosine (m⁶A) are both reported to participate in the regulation of lipid metabolism, yet it remains unknown whether they are involved in GC-induced hepatic lipid accumulation as new components of GR signalling. In this study, we use both in vivo and in vitro models of GC-induced hepatic lipid accumulation and demonstrate that the activation of lipogenic genes and accumulation of lipid in liver cells are mediated by GR-dependent FTO transactivation and m⁶A demethylation on mRNA of lipogenic genes. Targeted mutation of m⁶A methylation sites and FTO knockdown further validated the role of m⁶A on 3'UTR of sterol regulatory element-binding transcription factor 1 and stearoyl-CoA desaturase mRNAs. Finally, FTO knockdown significantly alleviated dexamethasone-induced fatty liver in mice. These results demonstrate a role of GR-mediated FTO transactivation and m⁶A demethylation in the pathogenesis of NAFLD and provide new insight into GR signalling in the regulation of fat metablism in the liver.

Keywords: FTO; Fatty liver; chicken primary hepatocyte; glucocorticoid receptor; lipogenesis; m6A.

Graphical abstract

Figure 1.

Characterization of CORT-induced fatty liver and hepatic expression of lipogenic genes and GR. (A) Plasma content of TG. (B) Hepatic content of TG. (C) Histological sections stained with oil-red, scale bar, 100 um. (D) Hepatic mRNA abundance of genes involved in lipogenesis. (E) Protein expression of GR, SREBF1 and SCD. Values are means ± SEM, *P < 0.05, **P < 0.01 compared with CON (n = 8).

Figure 2.

Hepatic protein content of FTO and profile of RNA m⁶A methylation in chicken liver. (A-B) Total mRNA methylation in chicken liver were measured by slot blotting and liquid chromatography/tandem triple-quadrupole mass spectrometry, respectively. (C-D) FTO mRNA and protein levels in the liver in response to CORT challenge. H1 were used as a loading control, Values are means ± SEM, *P < 0.05, **P < 0.01, compared with CON (n = 8). (E) Venn diagram showing the overlap of m⁶A peaks in CON (7089) and CORT (6932) group. There are 5336 common peaks between these two groups. (F-G) Graphical representation of frequency of m⁶A peaks in five non-overlapping segments in CON and CORT group (TSS: 200 nucleotides downstream of the TSS, stop codon: a 400 nucleotide window centred on the stop codon). m⁶A peaks were most abundant in CDS and stop codon segments. (H-I) Sequence logo representing the most common consensus motif (RRm⁶ACU) in the m⁶A peaks in CON and CORT group. The consensus sequence was detected by DREME (version: 4.10.2), using the 101 nucleotides centred on the summits of called original narrow peaks (n = 3).

Figure 3.

Effect of CORT injection on lipogenic genes mRNA m⁶A levels. Peak calling and differential analysis, peaks were considered if their MACS-assigned FDR value was less than 0.05 with absolute value of log2-fold-change > 1. Schematic diagram showing the mRNA sequences of chicken SREBF1, ACACA, FASN and SCD genes. The y-coordinate represents the absolute number of m⁶A sites, which was not normalized with the mRNA abundance. The x-coordinate represents the mRNA region of each peaks. The bars under each gene plot represent the location of each peak according to genomic DNA sequence. The value of log2-fold-change, and the position of the peaks on mRNAs are shown on the right side of each panel. The red boxes indicate significantly different m⁶A peaks on Srebf1, Acaca, Fasn and Scd mRNA. (A-D) Absolute number of m⁶A sites on Srebf1, Acaca, Fasn and Scd mRNA in CON and CORT groups (n = 3).

Figure 4.

Mapping and functional validation of m⁶A sites in lipogenic genes. (A) The mRNA expression of Srebf1, Fans, Acaca and Scd in Hepa1-6 cells treated with scramble control or FTO shRNAs. (B) The protein content of SREBF1 and SCD in Hepa1-6 cells treated with scramble control or FTO shRNAs. (C) The percentage of Fasn, Acaca and Scd mRNA with methylation in Hepa1-6 cells treated with scramble control or FTO shRNAs. (D-E) Western blotting of PEGFP-SREBF1 and PEGFP-SCD in 293 T seeded in 12-well plates and transfected with 50 ng of SREBF1-CDS or SCD-CDS plasmid and treated with siCtrl or FTO siRNAs (n = 3). (F-G) Relative activity of the wild-type or mutant SREBF1 or SCD 3′ UTR firefly luciferase reporter in 293 T cells treated with siCtrl or FTO siRNAs. Values are means ± SEM, *P < 0.05, **P < 0.01, NS, no significant change. (n = 6).

Figure 5.

Characterization of TG accumulation in primary chicken hepatocytes. (A) Primary hepatocytes stained with nile red, scale bar, 100 um. (B) TG content in in CON, DEX or OA/DEX treated hepatocytes. (C) Gr mRNA expression. (D) GR protein content in the nuclear lysates in CON, DEX or OA/DEX treated hepatocytes. (E) FTO mRNA and (F) protein expression in CON, DEX or OA/DEX treated hepatocytes. (G) The mRNA expression of Srebf1, Fasn, Acaca and Scd in CON, DEX or OA/DEX treated hepatocytes. (H-I) Protein content of SREBF1 and SCD in CON, DEX or OA/DEX treated hepatocytes. Tubulinα was used as a loading control. Values are means ± SEM, *P < 0.05, **P < 0.01 (n = 3).

Figure 6.

Role of GR in the regulation of FTO expression. (A) Representative images of hepatocytes labelled with GR (green). Nuclei were counterstained with DAPI (blue), scale bar, 20 um. (B-C) Protein content of GR and FTO in the nuclear lysates in VEH, OA/DEX or OA/DEX+RU486 treated hepatocytes. (D) GR binding to FTO gene promoter regions in CON or OA/DEX treated hepatocytes. (E-F) GR mRNA and protein content in DF-1 cells transfected with the empty plasmid PEGFP-N1 (EV) or the recombinant PEGFP-N1 plasmid expressing GR (GR⁺). (G-H) FTO mRNA and protein expression in DF-1 cells transfected with the empty plasmid PEGFP-N1 (EV) or the recombinant PEGFP-N1 plasmid expressing GR (GR⁺). (I) 293 T cells seeded in 12-well plates were transfected with the PGL3 luciferase vector fused with FTO promoter. The FTO promoter reporter were determined 48 hr in the present of empty PEGFP-N1 plasmid (EV) or recombinant PEGFP-N1 plasmid expressing GR (GR⁺) treatment. The FTO promoter activity was also normalized by co-transfection with the b-actin-Renilla luciferase reporter. Values are means ± SEM, *P < 0.05, **P < 0.01 (n = 3).

Figure 7.

Role of FTO in the regulation of lipogenic gene expression. (A) Primary hepatocytes stained with nile red, scale bar, 100 um. (B) TG content in CON, OA/DEX or OA/DEX+MA treated hepatocytes. (C-E) Protein content of SREBF1, SCD and GR in CON, OA/DEX or OA/DEX+MA treated hepatocytes. (F-G) Protein content of SREBF1 and SCD in CON, OA/DEX or OA/DEX+RU486 treated hepatocytes. Values are means ± SEM, *P < 0.05, **P < 0.01 (n = 3).

Figure 8.

Alterations of m⁶A on mRNA of lipogenic genes. (A-D) The percentage of Srebf1, Fasn, Acaca and Scd mRNA with methylation in CON, OA/DEX or OA/DEX+MA treated hepatocytes. Values are means ± SEM, *P < 0.05, **P < 0.01 (n = 3).