Deletion of hepatic carbohydrate response element binding protein (ChREBP) impairs glucose homeostasis and hepatic insulin sensitivity in mice

Tara Jois, Weiyi Chen, Victor Howard, Rebecca Harvey, Kristina Youngs, Claudia Thalmann, Pradip Saha, Lawrence Chan, Michael A. Cowley, Mark W. Sleeman

Fibroblast growth factor 21 (FGF21) has emerged as a key regulator of the metabolic response to fasting. Endoplasmic reticulum (ER) stress is a well-established inducer of hepatic FGF21 expression. X-box binding protein 1 (XBP1) has been strongly implicated in regulating hepatic lipid and glucose metabolism, making it an intriguing candidate for mediating the effect of ER stress on FGF21 expression. To directly determine whether hepatic Xbp1 is required for induction of hepatic Fgf21 in vivo, Olivares and Henkel subjected mice bearing a hepatocyte-specific deletion of Xbp1 to fasting, a ketogenic diet, or pharmacologic ER stress, all potent stimuli of Fgf21 expression. By this, they provide definitive evidence that hepatic Xbp1 is not required for induction of hepatic Fgf21.

Objective: Carbohydrate response element binding protein (ChREBP) is a transcription factor that responds to glucose and activates genes involved in the glycolytic and lipogenic pathways. Recent studies have linked adipose ChREBP to insulin sensitivity in mice. However, while ChREBP is most highly expressed in the liver, the effect of hepatic ChREBP on insulin sensitivity remains unknown. To clarify the importance of hepatic ChREBP on glucose homeostasis, we have generated a knockout mouse model that lacks this protein specifically in the liver (Liver-ChREBP KO).

Methods: Using Liver-ChREBP KO mice, we investigated whether hepatic ChREBP deletion influences insulin sensitivity, glucose homeostasis and the development of hepatic steatosis utilizing various dietary stressors. Furthermore, we determined gene expression changes in response to fasted and fed states in liver, white, and brown adipose tissues.

Results: Liver-ChREBP KO mice had impaired insulin sensitivity as indicated by reduced glucose infusion to maintain euglycemia during hyperinsulinemic-euglycemic clamps on both chow (25% lower) and high-fat diet (33% lower) (p < 0.05). This corresponded with attenuated suppression of hepatic glucose production. Although Liver-ChREBP KO mice were protected against carbohydrate-induced hepatic steatosis, they displayed worsened glucose tolerance. Liver-ChREBP KO mice did not show the expected gene expression changes in liver in response to fasted and fed states. Interestingly, hepatic ChREBP deletion also resulted in gene expression changes in white and brown adipose tissues, suggesting inter-tissue communication. This included an almost complete abolition of BAT ChREBPβ induction in the fed state (0.15-fold) (p = 0.015) along with reduced lipogenic genes. In contrast, WAT showed inappropriate increases in lipogenic genes in the fasted state along with increased PEPCK1 in both fasted (3.4-fold) and fed (5.1-fold) states (p < 0.0001).

Conclusions: Overall, hepatic ChREBP is protective in regards to hepatic insulin sensitivity and whole body glucose homeostasis. Hepatic ChREBP action can influence other peripheral tissues and is likely essential in coordinating the body's response to different feeding states.