Conditional ablation of mediator subunit MED1 (MED1/PPARBP) gene in mouse liver attenuates glucocorticoid receptor agonist dexamethasone-induced hepatic steatosis

Abstract

Glucocorticoid receptor (GR) agonist dexamethasone (Dex) induces hepatic steatosis and enhances constitutive androstane receptor (CAR) expression in the liver. CAR is known to worsen hepatic injury in nonalcoholic hepatic steatosis. Because transcription coactivator MED1/PPARBP gene is required for GR- and CAR-mediated transcriptional activation, we hypothesized that disruption of MED1/PPARBP gene in liver cells would result in the attenuation of Dex-induced hepatic steatosis. Here we show that liver-specific disruption of MED1 gene (MED1(delta Liv)) improves Dex-induced steatotic phenotype in the liver. In wild-type mice Dex induced severe hepatic steatosis and caused reduction in medium- and short-chain acyl-CoA dehydrogenases that are responsible for mitochondrial beta-oxidation. In contrast, Dex did not induce hepatic steatosis in mice conditionally null for hepatic MED1, as it failed to inhibit fatty acid oxidation enzymes in the liver. MED1(delta Liv) livers had lower levels of GR-regulated CAR mRNA compared to wild-type mouse livers. Microarray gene expression profiling showed that absence of MED1 affects the expression of the GR-regulated genes responsible for energy metabolism in the liver. These results establish that absence of MED1 in the liver diminishes Dex-induced hepatic steatosis by altering the GR- and CAR-dependent gene functions.

Figure 1

Histological analysis of liver from wild-type (MED1^+/+) and MED1^ΔLiv mice given Dex (50 mg/kg body weight, IP) or corn oil as vehicle once daily for 3 days. The liver sections of mice treated for 0 (A, E), and 3 days (B, F) were stained with hematoxylin and eosin. Liver sections from 3-day Dex-treated mice were processed for immunohistochemical localization of MED1 (C, G). Dex-treated MED1^+/+ mouse liver reveals prominent macrovesicular fatty change (B) and shows MED1 nuclear staining (C). Note the peripheral location of MED1-positive nucleus (arrows) due to displacement by large fat vacuole (C). Fatty change is minimal in MED1^ΔLiv mouse liver even at 3-day Dex treatment. An occasional MED1-positive nucleus is seen (G, arrow) due to escape from Cre-mediated deletion. Oil red O (D, H) staining of liver sections obtained from MED1^+/+ (D) and MED1^ΔLiv (H) mice treated with Dex for 3 days. Hepatic steatosis seen in Dex-treated MED1^+/+ mice with hematoxylin and eosin staining (B) is confirmed by Oil red O staining (D). In the livers of 3-day Dex-treated liver conditional knockout mouse an occasional large hepatocyte that escaped Cre-mediated deletion is present (F, arrows), which reveal MED1-positive nucleus in MED1^ΔLiv mouse liver (G, arrow).

Figure 2

(A) Immunoblot analysis of liver proteins for changes in some enzymes responsible for fatty acid oxidation. LCAD, MCAD, and SCAD represent mitochondrial β-oxidation system enzymes, while ACOX, L-PBE, and PTL are members of peroxisomal β-oxidation pathway. CYP4A1 is a microsomal fatty acid β-oxidation system enzyme. Also included are EFT, COT, and CTL. (B) The histogram is the densitometric analysis of the Western blot signals. Black bars refer to wild-type (con) and white bars to Dex treatment (Dex) in MED1^+/+ and MED1^ΔLiv mice. All data are presented as the mean ± SD of three independent measurements.

Figure 3

(A) Northern blot analysis for CAR and GR mRNA levels in MED1^+/+ and MED1^ΔLiv mouse livers without and with Dex treatment for 3 days. GAPDH is used as RNA loading control. (B) Quantitative real-time PCR analysis of CAR mRNA level in MED1^+/+ and MED1^ΔLiv mouse livers following treatment with Dex for 0 (con), 1 (1D), 2 (2D), and 3 (3D) days. Dark bars refer to wild-type (MED1^+/+) and white bars represent MED1^ΔLiv mice. All data are presented as the mean ± SD of three independent experiments.

Figure 4

Comparative expression of genes in liver selected from microarray profile data in the livers of MED1^+/+ and MED1^ΔLiv mice after 1, 2, and 3 days of treatment with Dex. Increases in CYP2b10, IGFBP2, glucokinase, thyroid responsive SPOT14, tissue inhibitor of metalloproteinase 4, Saa1, and ATP-binding cassette C3 are seen in Dex-treated wild-type (black bars) compared to MED1^ΔLiv mice (white bars). The microarray predicted increase in phophatidylinositol 3 kinase (p55) in Dex-treated MED1^ΔLiv mouse liver is confirmed by quantitative PCR. The specific amplification of genes was normalized with 18S RNA signal and the arbitrary values are shown. All data are presented as the mean ± SD of three independent experiments.

Figure 5

Further validation of microarray findings by quantitative PCR of hepatic RNA of MED1^+/+ and MED1^ΔLiv mice after 1, 2, and 3 days of treatment with Dex. Increases in CYP3A11, acyl-CoA synthetase, glucose transporter, glucose-6-phosphatase (G6pc) are seen in Dex-treated wild-type (black bars) compared to MED1^ΔLiv mice (white bars). On the other hand, increases in acyl-CoA thioesterase 1 and cyclin-dependent kinase inhibitor (p21) are observed in Dex-treated MED1^ΔLiv mouse liver. The specific amplification of genes was normalized with 18S RNA signal and the arbitrary values are shown. Data are shown as the mean ± SD of three independent experiments.