A critical role for ChREBP-mediated FGF21 secretion in hepatic fructose metabolism

Abstract

Objective: Increased fructose consumption is a contributor to the burgeoning epidemic of non-alcoholic fatty liver disease (NAFLD). Recent evidence indicates that the metabolic hormone FGF21 is regulated by fructose consumption in humans and rodents and may play a functional role in this nutritional context. Here, we sought to define the mechanism by which fructose ingestion regulates FGF21 and determine whether FGF21 contributes to an adaptive metabolic response to fructose consumption.

Methods: We tested the role of the transcription factor carbohydrate responsive-element binding protein (ChREBP) in fructose-mediated regulation of FGF21 using ChREBP knockout mice. Using FGF21 knockout mice, we investigated whether FGF21 has a metabolic function in the context of fructose consumption. Additionally, we tested whether a ChREBP-FGF21 interaction is likely conserved in human subjects.

Results: Hepatic expression of ChREBP-β and Fgf21 acutely increased 2-fold and 3-fold, respectively, following fructose gavage, and this was accompanied by increased circulating FGF21. The acute increase in circulating FGF21 following fructose gavage was absent in ChREBP knockout mice. Induction of ChREBP-β and its glycolytic, fructolytic, and lipogenic gene targets were attenuated in FGF21 knockout mice fed high-fructose diets, and this was accompanied by a 50% reduction in de novo lipogenesis a, 30% reduction VLDL secretion, and a 25% reduction in liver fat compared to fructose-fed controls. In human subjects, serum FGF21 correlates with de novo lipogenic rates measured by stable isotopic tracers (R = 0.55, P = 0.04) consistent with conservation of a ChREBP-FGF21 interaction. After 8 weeks of high-fructose diet, livers from FGF21 knockout mice demonstrate atrophy and fibrosis accompanied by molecular markers of inflammation and stellate cell activation; whereas, this did not occur in controls.

Conclusions: In summary, ChREBP and FGF21 constitute a signaling axis likely conserved in humans that mediates an essential adaptive response to fructose ingestion that may participate in the pathogenesis of NAFLD and liver fibrosis.

Keywords: ChREBP; FGF21; Fructose; Lipogenesis; NAFLD.

Figure 1

Fructose regulates Fgf21 in a ChREBP-dependent manner. A) Increased hepatic Fgf21 mRNA expression in mice consuming HFrD for 4 weeks. (n = 5–7/group) B) Fgf21 serum levels in mice consuming HFrD for 4 weeks are also elevated. (n = 5–7/group) C) Adipose Fgf21 gene expression in mice as in A. *P < 0.05 vs Chow, &P < 0.05 vs HDD. D) Hepatic gene expression in mice 30 min after fructose gavage (n = 4–6/group) and E) serum FGF21 levels 1 h after fructose gavage (n = 4–6/group) in WT and ChREBP KO mice. *P < 0.05 versus WT water (Wa), #P < 0.05 versus WT fructose (Fr). Data are shown as the mean ± SEM.

Figure 2

Fgf21 participates in fructose-mediated induction of ChREBP and DNL. A) HFrD diet mediated induction of hepatic ChREBPβ and its transcriptional targets are diminished in FGF21 KO mice. B) Circulating glucose levels are lower in FGF21 KO mice consuming HFrD. C) Attenuated induction of multiple enzymes regulating DNL in the livers of HFrD fed FGF21 KO mice. (A–B, WT Chow & FGF21 KO Chow n = 5/group, WT HFrD & FGF21 KO HFrD n = 7/group) D) In vivo rates of DNL are reduced in FGF21 KO mice (WT Chow n = 4, FGF21 KO Chow n = 5, WT HFrD & FGF21 KO HFrD n = 6/group). VLDL secretion is attenuated in FGF21 KO mice consuming E) HFrD (WT HFrD n = 8, FGF21 KO HFrD n = 9) and F) Chow diet (WT Chow n = 5, FGF21 KO Chow n = 5). This is, in part, underscored by G) reduced expression of enzymes regulating VLDL assembly (WT Chow & FGF21 KO Chow n = 5/group, WT HFrD & FGF21 KO HFrD n = 7/group). Taken together these impairments lead to a net reduction in H) hepatic triglyceride content. Data are shown as the mean ± SEM. *P < 0.05, Chow vs Fructose for each genotype. #P < 0.05, Fructose WT vs Fructose KO.

Figure 3

In humans, circulating Fgf21 correlates with DNL. Graph showing regression analysis between basal circulating FGF21 levels and basal rates of de novo lipogenesis in human subjects (n = 14).

Figure 4

Fructose consumption causes liver disease in the absence of Fgf21. A) Elevated serum FGF21 levels after 8 weeks of HFrD consumption. B) Fructose-mediated induction of hepatic ChREBP-β mRNA and ChREBP gene targets are diminished in FGF21 KO mice. C) Body weights and D) markers of hepatic inflammation and fibrosis in the livers of HFrD FGF21 KO mice and controls (A–D, WT Chow n = 5, KO Chow n = 6, WT HFrD & KO HFrD n = 9/group). Representative histology showing E) localization of lipid to zone 3 in FGF21-KO mice and F) increased fibrosis by Sirius stain in FGF-21 KO mice compared to WT when consuming a high fructose diet. Data are shown as the mean ± SEM. *P < 0.05, Compared to chow within genotype. #P < 0.05, compared to WT within diet.