Estrogen treatment after ovariectomy protects against fatty liver and may improve pathway-selective insulin resistance

Abstract

Pathway-selective insulin resistance where insulin fails to suppress hepatic glucose production but promotes liver fat storage may underlie glucose and lipid abnormalities after menopause. We tested the mechanisms by which estrogen treatment may alter the impact of a high-fat diet (HFD) when given at the time of ovariectomy (OVX) in mice. Female C57BL/6J mice underwent sham operation, OVX, or OVX with estradiol (E2) treatment and were fed an HFD. Hyperinsulinemic-euglycemic clamps were used to assess insulin sensitivity, tracer incorporation into hepatic lipids, and liver triglyceride export. OVX mice had increased adiposity that was prevented with E2 at the time of OVX. E2 treatment increased insulin sensitivity with OVX and HFD. In sham and OVX mice, HFD feeding induced fatty liver, and insulin reduced hepatic apoB100 and liver triglyceride export. E2 treatment reduced liver lipid deposition and prevented the decrease in liver triglyceride export during hyperinsulinemia. In mice lacking the liver estrogen receptor α, E2 after OVX limited adiposity but failed to improve insulin sensitivity, to limit liver lipid deposition, and to prevent insulin suppression of liver triglyceride export. In conclusion, estrogen treatment may reverse aspects of pathway-selective insulin resistance by promoting insulin action on glucose metabolism but limiting hepatic lipid deposition.

FIG. 1.

Estrogen treatment at the time of OVX reduces diet-induced obesity. A: OVX led to increased body weight with HFD. Mice treated with E2 at the time of OVX did not gain weight with HFD. B: Adiposity was increased with HFD in sham mice and after OVX. Adiposity was reduced with E2 treatment. Letter B indicates baseline before HFD-feeding. Letter E indicates end point after HFD. C: Plasma cholesterol levels. D: Plasma triglyceride levels. E: HFD increased fasting insulin levels in sham and OVX mice, and E2 treatment reduced plasma insulin. F: Fasting glucose levels were not different between groups. *P < 0.05. For panel B, differences from baseline were defined by Student t test. For C–F, differences between groups were determined by ANOVA followed by Tukey post hoc tests.

FIG. 2.

Estrogen treatment at the time of OVX reduces pathway-selective insulin resistance. A: Schematic of the hyperinsulinemic-euglycemic clamp study with triglyceride (TG) and glucose tracers. B: Euglycemia was maintained at ∼150 mg/dL during the clamp. C: GIR to maintain euglycemia. D: Insulin sensitivity index. E: Compared with baseline (no insulin), insulin failed to suppress EndoRa after OVX (plus insulin), but insulin did suppress EndoRa after OVX with E2 treatment. F: Total plasma triglyceride during the clamp study. G: Plasma ¹⁴C-triglyceride–specific activity (SA). *P < 0.05. Differences between groups were determined by ANOVA followed by Tukey post hoc tests.

FIG. 3.

Estrogen treatment at the time of OVX reduces fatty liver. A: Oil red O staining of liver neural lipid droplets. B–E: Liver lipids were extracted with the Folch method, and neutral lipids were separated by TLC as described in research design and methods. Triglyceride (TG) (B) and diacylglycerol (DAG) (D) were quantified using an enzymatic assay; ¹⁴C-triglyceride (C) and ¹⁴C-diacylglycerol (E) were quantified by scintillation counting. *P < 0.05. Differences between groups were determined by ANOVA followed by Tukey post hoc tests. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 4.

Estrogen treatment blocks insulin (Ins)-mediated dephosphorylation of ACC and insulin reduction of apoB100 in the liver. Livers from mice in a cohort that was fasted but not clamped (−) and after hyperinsulinemic clamp study (+) were used for protein extraction and Western blotting. A: pACC and total ACC. Expression of actin and Panseau S staining were used as loading controls. ACC expression and the ratio of pACC to ACC were quantified in B and C. D: Western blot for liver apoB100, MTP, and PLTP. Expression of apoB100 (E), MTP (F), and PLTP (G) was quantified as well. *P < 0.05. Differences (+ or − insulin) were defined by Student t test. Differences between groups were determined by ANOVA followed by Tukey post hoc tests.

FIG. 5.

Estrogen treatment reduces body weight but does not prevent pathway-selective insulin (Ins) resistance in LKO mice. A: ERα protein amounts were decreased specifically in liver of LKO mice but not in other tissues. B: Weight gain with HFD was increased after OVX and prevented with E2 treatment in LKO mice and their wild-type (WT) littermates. C: OVX led to increased adiposity with HFD, which was prevented with E2 treatment in LKO mice and their littermates. Letter B indicates baseline before HFD-feeding. Letter E indicates end point after HFD. D: Euglycemia was maintained at ∼150 mg/dL during the clamp. E: GIR to maintain euglycemia. F: Estrogen treatment did not restore the insulin suppression of EndoRa after OVX in LKO mice (−, baseline period of clamp; clamp period). G: Plasma ¹⁴C-triglyceride (TG)-specific activity (SA) was not significantly different by E2 treatment after OVX in LKO mice. *P < 0.05. Differences from baseline (B) or end point (E) were defined by Student t test.

FIG. 6.

Estrogen treatment fails to protect against fatty liver with HFD feeding in LKO mice after OVX. A–D: Liver lipids were extracted with the Folch method, and neutral lipids were separated by TLC as described in research design and methods. Triglyceride (TG) (A) and diacylglycerol (DAG) (C) were quantified using an enzymatic assay; ¹⁴C-triglyceride (B) and ¹⁴C-diacylglycerol (D) were quantified by scintillation counting. E: Livers from mice in a cohort that was fasted but not clamped (−) and after hyperinsulinemic clamp study (+) were used for protein extraction and Western blotting for pACC, total ACC, apoB100, MTP, and PLTP. Expression of actin and Panseau S staining were used as loading controls.

FIG. 7.

Schematic representation of liver ERα signaling with regard to the regulation of liver glucose and lipid metabolism: HFD feeding after OVX resulted in pathway-selective insulin resistance where insulin failed to suppress hepatic glucose production (HGP), yet insulin was able to promote liver triglyceride (TG) storage by reducing hepatic apoB content, dephosphorylating ACC, and suppressing tracer incorporation into serum triglyceride. Estrogen treatment after OVX improved insulin suppression of hepatic glucose production and blocked insulin-mediated liver triglyceride storage.