E2F1 mediates sustained lipogenesis and contributes to hepatic steatosis

Abstract

E2F transcription factors are known regulators of the cell cycle, proliferation, apoptosis, and differentiation. Here, we reveal that E2F1 plays an essential role in liver physiopathology through the regulation of glycolysis and lipogenesis. We demonstrate that E2F1 deficiency leads to a decrease in glycolysis and de novo synthesis of fatty acids in hepatocytes. We further demonstrate that E2F1 directly binds to the promoters of key lipogenic genes, including Fasn, but does not bind directly to genes encoding glycolysis pathway components, suggesting an indirect effect. In murine models, E2F1 expression and activity increased in response to feeding and upon insulin stimulation through canonical activation of the CDK4/pRB pathway. Moreover, E2F1 expression was increased in liver biopsies from obese, glucose-intolerant humans compared with biopsies from lean subjects. Finally, E2f1 deletion completely abrogated hepatic steatosis in different murine models of nonalcoholic fatty liver disease (NAFLD). In conclusion, our data demonstrate that E2F1 regulates lipid synthesis and glycolysis and thus contributes to the development of liver pathology.

Figure 9. E2f1 deletion in the db/db mouse model protects against hepatic steatosis.

(A) Images of db/db E2f1^–/– compared with db/db E2F1^+/+ mouse livers. (B) Liver weight expressed as a percentage of the total mass for the indicated genotypes (n = 5–9). (C) Quantification of liver TG in the indicated genotypes (n = 5–9). (D) Liver FAME analysis of db/+ E2f1^+/+, db/db E2f1^+/+, and db/db E2f1^–/– mice (n = 4–7). (E) Desaturation index corresponded to the ratio of monounsaturated fatty acid:oleate C18:1/saturated fatty acid:palmitate C16 plus stearate C18 (MUFA/SFA) ratio. (F) Representative H&E and Oil red O staining of liver sections from db/+ E2f1^+/+, db/db E2f1^+/+, and db/db E2f1^–/– mice (original magnification, ×200). *P < 0.05, by 2-tailed, unpaired t test.

Figure 8. E2f1 deletion in db/db mouse model decreases glycolytic and lipogenic programs.

(A) Plasma insulin levels in db/+ E2f1^+/+, db/db E2f1^+/+, and db/db E2f1^–/– mice (n = 5–9). (B) Ser473 phosphorylation of AKT in the livers of db/db E2f1^+/+ and db/db E2f1^–/– mice. (C) Relative liver mRNA expression levels of the indicated genes in the annotated mouse genotypes (n = 4–6). (D) Western blot analyses of the expression of the indicated proteins in the livers of db/+ E2f1^+/+, db/db E2f1^+/+, and db/db E2f1^–/– mice. *P < 0.05, by 2-tailed, unpaired t test.

Figure 7. E2F1 gene expression is increased in obese mice and humans.

(A) Relative expression of E2f1 mRNA in the livers of mice after 16 or 33 weeks on a high-fat diet. (B) E2f1 mRNA expression in the livers of control db/+ and db/db mice. (C) Ser780 phosphorylation of RB1 in the livers of db/db mice. (D) Relative expression of E2F1 mRNA in livers of lean subjects and obese, glucose-intolerant patients. * P < 0.05 and ** P < 0.002 compared with control, by 2-tailed, unpaired t test.

Figure 6. Insulin regulates E2F1 activity through RB1 and CDK4.

(A) Relative E2f1 mRNA expression levels in the livers of mice after a 24-hour fast, followed by an 18-hour refeeding. (B) Protein expression analyses of the indicated cellular fractions in the livers of mice under the same conditions as in A. (C) E2F reporter activity (E2FRE-TK-luc) in response to insulin in HEK293T cells. (D) Mouse Fasn promoter activity in HepG2 cells transfected with empty vector, E2F1, or RB1. (E) Ser780 phosphorylation of RB1 in the livers of fasted and refed mice. (F) Ser780 phosphorylation of RB1 in hepatocytes after 1 hour of insulin stimulation. (G) RB1 co-IP with E2F1 after 1 hour of insulin stimulation in HepG2 hepatocytes. (H) Relative mRNA expression of the indicated genes in hepatocytes treated with adenovirus expressing sh-control or sh-CDK4 and stimulated for 24 hours with G5 or G25i. (I) Protein levels of ACACA, FASN, and CDK4 in hepatocytes treated with adenovirus expressing sh-control or sh-CDK4 and stimulated for 24 hours with G5 or G25i. The experiments were performed at least 3 times. *P < 0.05, by 2-tailed, unpaired t test.

Figure 5. Lipogenic genes are bona fide E2F1 targets.

(A) E2F1 ChIP on hepatocytes expressing Ad-E2F1 demonstrated E2F1 binding to the Acaca, Fasn, Scd1, Srebp1c, and Chrebp promoters, but not the Slc2a2 or Gck promoter (n = 5). (B) Luciferase-based reporter activity of the indicated promoters in HepG2 cells transfected with empty vector or the E2F1-HA expression vector. (C) Representative role of E2F1 in the transcriptional control of glycolysis and lipogenesis. E2F1 directly controlled Srebp1c, Chrebp, Acaca, Fasn, and Scd1 gene expression and indirectly controlled Slc2a2, Gck, and Pklr via SREBP1c and CHREBP. SREBP1c and CHREBP also participated in the control of Acaca, Fasn, and Scd1 gene expression. (D) E2F-TK-luc reporter activity in HepG2 hepatocytes transfected with empty vector, E2F1, or USF1. (E) Mouse Fasn promoter activity in hepatocytes transfected with empty vector, E2F1 or USF1. (F) Co-IP experiment on E2F1 and USF1 in HepG2 hepatocytes. Cells were transfected with E2F1-HA and Flag-USF1 as indicated, and protein was immunoprecipitated with a Flag Ab.

Figure 4. Glucose and lipid metabolism is impaired in E2f1^–/– hepatocytes.

(A) ECAR of E2f1^+/+ and E2f1^–/– hepatocytes after glucose treatment using a Seahorse analyzer (2 independent experiments, each with 8 technical replicates). (B) Glycogen content in E2f1 ^+/+ and E2f1 ^–/– hepatocytes treated for 24 hours with G5 or G25i. (C) Quantification of the incorporation of C14-labeled acetate in the TG fraction in hepatocytes treated for 24 hours with G5 or G25i from the indicated genotypes, as a measure of lipogenesis. (D) Representative Oil Red O staining of E2f1^+/+ versus E2f1^–/– hepatocytes (original magnification, ×200). Unless otherwise specified, all experiments represent the average of 3 independent experiments. *P < 0.05 compared with control, by 2-tailed, unpaired t test.

Figure 3. Expression of glycolytic and lipogenic genes is impaired in E2f1^–/– hepatocytes.

(A) Relative mRNA expression of relevant glycolytic and lipogenic genes in primary hepatocytes from E2f1^+/+ and E2f1^–/– mice treated for 24 hours with 5 mM low glucose alone (G5); 5 mM low glucose and 100 nM insulin (G5i); or 25 mM high glucose and 100 nM insulin (G25i). (B) Western blot analysis shows expression levels of the indicated proteins in E2f1^+/+ and E2f1^–/– hepatocytes treated for 24 hours with G5 or G25i. All experiments represent the average of 3 independent experiments. *P < 0.05 compared with control, by 2-tailed, unpaired t test.

Figure 2. Liver-specific deletion of E2f1 leads to decreased glycolytic and lipogenic programs.

(A) Relative mRNA levels of E2f1 and relevant glycolytic and lipogenic genes in the livers E2f1^fl/fl versus E2F1 LKO mice on a chow diet or after a 2-week high-fructose diet (n = 5 per group). (B) Liver weight, liver TG content, and plasma TG levels of E2f1^fl/fl versus E2F1 LKO mice after 2 weeks on a high-fructose diet (n = 5 per group). *P < 0.05, by 2-tailed, unpaired t test.

Figure 1. E2f1^–/– mice show a decrease in liver glucose and lipid metabolism.

(A) Relative mRNA levels of E2f1 and relevant glycolytic and lipogenic genes in the livers of E2f1^+/+ versus E2f1^–/– mice (n = 5 per group) under fed conditions. HFruct, high-fructose. (B) Western blot analyses of the expression of the indicated proteins in E2f1^+/+ and E2f1^–/– mice. (C) Liver glycogen content, liver TG content, and plasma TG levels in fed E2f1^+/+ and E2f1^–/– mice. (C) Liver TG content in E2f1^+/+ and E2f1^–/– mice subjected to a 9-week high-sucrose diet (6 mice per group). (D) Plasma TG levels after overnight fasting of E2f1^+/+ and E2f1^–/– mice subjected to a 9-week high-sucrose diet (6 mice per group). *P < 0.05, by 2-tailed, unpaired t test.