Livers with constitutive mTORC1 activity resist steatosis independent of feedback suppression of Akt

Abstract

Insulin resistance is an important contributing factor in non-alcoholic fatty liver disease. AKT and mTORC1 are key components of the insulin pathway, and play a role in promoting de novo lipogenesis. However, mTORC1 hyperactivity per se does not induce steatosis in mouse livers, but instead, protects against high-fat diet induced steatosis. Here, we investigate the in vivo mechanism of steatosis-resistance secondary to mTORC1 activation, with emphasis on the role of S6K1-mediated feedback inhibition of AKT. Mice with single or double deletion of Tsc1 and/or S6k1 in a liver-specific or whole-body manner were generated to study glucose and hepatic lipid metabolism between the ages of 6-14 weeks. Following 8 weeks of high-fat diet, the Tsc1-/-;S6k1-/- mice had lower body weights but higher liver TG levels compared to that of the Tsc1-/- mice. However, the loss of S6k1 did not relieve feedback inhibition of Akt activity in the Tsc1-/- livers. To overcome Akt suppression, Pten was deleted in Tsc1-/- livers, and the resultant mice showed improved glucose tolerance compared with the Tsc1-/- mice. However, liver TG levels were significantly reduced in the Tsc1-/-;Pten-/- mice compared to the Pten-/- mice, which was restored with rapamycin. We found no correlation between liver TG and serum NEFA levels. Expression of lipogenic genes (Srebp1c, Fasn) were elevated in the Tsc1-/-;Pten-/- livers, but this was counter-balanced by an up-regulation of Cpt1a involved in fatty acid oxidation and the anti-oxidant protein, Nrf2. In summary, our in vivo models showed that mTORC1-induced resistance to steatosis was dependent on S6K1 activity, but not secondary to AKT suppression. These findings confirm that AKT and mTORC1 have opposing effects on hepatic lipid metabolism in vivo.

Figure 1. The Tsc1^fl/fl;Alb^Cre;S6k1-/- mice develop normally and have increased insulin sensitivity.

A) Total body weights of male littermates fed either NCD or HFD over the study period. *, p<0.05 comparing S6k1-/- with Tsc1-/-. B) Box-and-whisker plots of liver weights (absolute and relative) and epididymal white adipose tissue weights following 8 weeks of NCD or HFD. *, p<0.05 compared to NCD within genotype, **, p<0.05 compared to control within diet. C) Glucose tolerance and insulin sensitivity tests. *, p<0.05 comparing Tsc1-/-;S6k1-/- to control. D) Fasting blood glucose and plasma insulin levels 30 minutes following glucose administration. N = 4–7 mice per group.

Figure 2. S6k1 deletion promotes steatosis in Tsc1-null hepatocytes.

A) Liver triglyceride (TG) levels in male mice fed NCD (black) and HFD (gray). *, p<0.05 compared to respective NCD; **, p<0.01 compared to respective NCD; ***, p = 0.03 comparing Tsc1-/- vs. Tsc1-/-;S6k1-/- on HFD. N = 4–7 mice per group. B) Examples of H&E histology of livers showing varying degrees of macrovesicular steatosis under NCD and HFD conditions. Original magnifications: 40x and 400x. Note the lack of large lipid droplets in HFD-fed Tsc1-/- livers.

Figure 3. Deletion of S6k1 in Tsc1-/- livers does not relieve Akt suppression.

A) Akt and mTORC1 signaling in livers under NCD and HFD conditions. Akt activity is indicated by the phosphorylation of Ser473 and Thr308 and that of its substrate, Pras40(Thr246), whereas mTORC1 activity is reflected in 4E-BP1(Ser65) phosphorylation. Note that the S6k1-/- mice do not express S6k as expected. B) Graph shows levels of Akt phosphorylation relative to total Akt expression based on densitometry (Image J) of the blots shown in A. C) Expression of genes relevant to lipid metabolism based on qRT-PCR of indicated liver samples. *, p<0.05 compared to respective controls within diet (NCD or HFD); •, p<0.05 NCD of the same genotype. N = 4–7 mice per group.

Figure 4. Co-deletion of Pten and Tsc1 in hepatocytes leads to hepatomegaly and hypoglycemia.

A) Comparisons of total body, liver and epididymal WAT weights. *, p<0.05 compared to control or where indicated by lines. **, p<0.0001 compared to all other groups. B) Glucose tolerance test with ‘area under the curve’ graph. *, p<0.05 comparing Tsc1-/- with Pten-/- and Tsc1-/-;Pten-/-. C) Insulin sensitivity test, fasting plasma insulin and fasting glucose levels. *, p<0.01 between Pten-/- and Tsc1-/-. **, p<0.01 compared to control. ***, p<0.01 compared to all groups. N = 3–6 mice per group.

Figure 5. Pten deletion does not promote steatosis in the Tsc1-/- livers.

A) Liver TG content and B) Plasma TG. *, p<0.01 compared to all other groups. **, p<0.05 compared to control. N = 4–6 mice per group. C) Liver histology by H&E and Oil red ‘O’ staining. Mice were fed NCD and fasted overnight before sacrifice. Original magnifications: 400x and 40x. fl/fl, floxed control. D) Non-esterified fatty acids levels in freshly drawn blood samples. *, p<0.05 compared with control group. N = 3–7 per group.

Figure 6. Rapamycin increases steatosis in Tsc1-/- and Tsc1-/-;Pten-/- livers.

A) H&E photomicrographs of representative livers treated with vehicle (DMSO) or rapamycin. Magnification: 400x except for Tsc1-/-;Pten-/- samples showing both 40x and 400x. B) Western blot analyses of liver lysates from two sets of treated mice. Note lack of change in Akt phosphorylation following two weeks of rapamycin. mTORC2 activity is indicated by p-Akt(Ser473) and p-NDRG1(Thr346) expression. fl/fl, floxed controls. TP, Tsc1-/-;Pten-/-. C) Expression of genes relevant to lipid metabolism based on qRT-PCR of indicated liver samples. *, p<0.05 compared to control. •, p<0.05 compared to Tsc1-/-. N = 3–6 mice per group. D) Western blot analyses of markers of metabolic, energy and autophagic stress. E) Model highlighting the in vivo effects of Akt-mTORC1-S6K1 in lipid metabolism. See text for details.