Exploring the impacts of ketogenic diet on reversible hepatic steatosis: initial analysis in male mice

Abstract

Metabolic dysfunction-associated fatty liver disease (MAFLD) is the most common chronic liver disease. Ketogenic diet (KD), a diet with very low intake in carbohydrates, gained popularity as a weight-loss approach. However, in mice models, it has been reported that an excess exposition of dietary fat induces hepatic insulin resistance and steatosis. However, data published is inconsistent. Herein, we investigated in a mouse model, the metabolic effects of KD and its contribution to the pathogenesis of NALFD. Mice were exposed to KD or CHOW diet for 12 weeks while a third group was exposed to KD for also 12 weeks and then switched to CHOW diet for 4 weeks to determine if we can rescue the phenotype. We evaluated the effects of diet treatments on fat distribution, glucose, and insulin homeostasis as well as hepatic steatosis. Mice fed with KD developed glucose intolerance but not insulin resistance accompanied by an increase of inflammation. KD-fed mice showed an increase of fat accumulation in white adipose tissue and liver. This effect could be explained by an increase in fat uptake by the liver with no changes of catabolism leading to MAFLD. Interestingly, we were able to rescue the phenotype by switching KD-fed mice for 4 weeks on a CHOW diet. Our studies demonstrate that even if mice develop hepatic steatosis and glucose intolerance after 12 weeks of KD, they do not develop insulin resistance and more importantly, the phenotype can be reversed by switching the mice from a KD to a CHOW.

Keywords: glucose intolerance; hepatic steatosis; inflammation; insulin resistance; ketogenic diet; rescue.

Figure 1

Metabolic profile and insulin sensitivity of mice fed ad libitum for 12 weeks with either CHOW or ketogenic diets (KD). (A) Evaluation of body weight, daily food intake plasma β-hydroxybutyrate and fasting glucose and insulin. (B) GTT (glucose tolerance test) was measured at week 9 and ITT (insulin tolerance test) was measured at week 10 after the beginning of the diet. AUC represents the calculated area under the curve. (C) Total and phosphorylated Akt and P70S6K levels measured in mice livers. Representative Western blots are depicted. α-tubulin was used as control (n = 3). Results are presented as mean +/− SEM (n = 3). ns, non-significant, *p-value <0.05, **p-value <0.01, and ****p-value <0.0001.

Figure 2

Adipose tissue and liver sections analysis in mice fed ad libitum for 12 weeks with either CHOW diet or Ketogenic Diet. (A) Representative images of white adipose tissue (WAT) distribution. Red arrows indicate subcutaneous WAT while blue arrows show visceral WAT. Scale Bar = 20 mm (Pictures in Supplementary Data). Quantifications were performed using the Particle Analysis function of ImageJ. Liver triglyceride levels. (B) Representative images of RedOil and eosin/hematoxyl staining of hepatic tissues sections. Scale bar = 70 μm. Quantification of total, average size area and percentage area of lipid droplets in liver sections were performed using ImageJ. Data are the mean +/− SEM. (n = 3). ns, non-significant, *p-value <0.05 and **p-value <0.01.

Figure 3

Hepatic expression of lipid metabolic genes and inflammatory cytokines in liver of mice fed ad libitum for 12 weeks with either CHOW diet or Ketogenic Diet. (A) Protein expressions of SREBP-1, phospho-ACC, CPT1 and mRNA expression of Scd1, Pparα and Cd36 are presented. α-TUBULINE and Hprt1 were used, respectively, as control. (B) Protein and mRNA expression of TNF-α (evaluated by ELISA) and Ifn-β and Il-4 are presented. Hprt1 was used as control for qPCR. Data are shown as mean +/− SEM. n = 3. Student’s t-test: *p-value <0.05, **p-value <0.01, ***p-value <0.001, and ****p-value <0.0001.

Figure 4

Effect of 4 weeks of CHOW diet after 12 weeks of Ketogenic Diet. (A) Evaluation of body weight, plasma β-hydroxybutyrate and fasting glucose, GTT was measured at week 3 and ITT was measured at week 4 after the switch of the diet. AUC represents the calculated area under the curve. (B) Representative images of WAT distribution. Red arrows indicate subcutaneous WAT while blue arrows show visceral WAT. Scale Bar = 20 mm (Pictures in Supplementary Data). Quantification was performed using the Particle Analysis function of ImageJ. Liver triglyceride levels. (C) Representative images of RedOil and eosin∕hematoxyl staining of hepatic tissues sections. Scale bar = 70 μm. Quantification of total, average size area and percentage area of lipid droplets in liver sections were performed using ImageJ. mRNA expression of Scd1, Cd36, Ifn-β, and Il-4 are presented. Hprt1 were used as control. Data are shown as mean +/− SEM. n = 3. Student’s t-test: *p-value <0.05, **p-value <0.01, ***p-value <0.001, and ****p-value <0.0001.