Nrf2 deletion from adipocytes, but not hepatocytes, potentiates systemic metabolic dysfunction after long-term high-fat diet-induced obesity in mice

Abstract

Nuclear factor erythroid 2-related factor 2 (Nrf2) is a canonical regulator of cytoprotective gene expression, but evidence of its cross talk with other pathways, including metabolic ones, is ever increasing. Pharmacologic or systemic genetic activation of the Nrf2 pathway partially protects from obesity in mice and ameliorates fasting hyperglycemia in mice and humans. However, systemic Nrf2 deletion also protects from diet-induced obesity and insulin resistance in mice. To further investigate the effect of the disruption of Nrf2 on obesity in a tissue-specific manner, we focused on adipocytes and hepatocytes with targeted deletion of Nrf2. To this end, mice with cell-specific deletion of Nrf2 in adipocytes (ANKO) or hepatocytes (HeNKO) were fed a high-fat diet (HFD) for 6 mo and showed similar increases in body weight and body fat content. ANKO mice showed a partially deteriorated glucose tolerance, higher fasting glucose levels, and higher levels of cholesterol and nonesterified fatty acids compared with their Control counterparts. The HeNKO mice, though, had lower insulin levels and trended toward improved insulin sensitivity without having any difference in liver triglyceride accumulation. This study compared for the first time two conditional Nrf2 knockout models in adipocytes and in hepatocytes during HFD-induced obesity. None of these models could completely recapitulate the unexpected protection against obesity observed in the whole body Nrf2 knockout mice, but this study points out the differential roles that Nrf2 may play, beyond cytoprotection, in different target tissues and rather suggests systemic activation of the Nrf2 pathway as an effective means of prevention and treatment of obesity and type 2 diabetes.

Keywords: Keap1; Nrf2; diabetes; fat; liver.

Fig. 1.

Validation of the tissue-specific nuclear factor erythroid 2-related factor 2 (Nrf2) knockout models. A: schematic representation of the Nrf2^flox locus before and after tissue-specific exon V deletion. N2, N3, and Ex-V-32 represent the primers used for genotyping and detection of the recombined Nrf2^flox allele. B: PCR-based Nrf2^flox genotyping and detection of the recombined Nrf2^flox allele using the primer pairs (N2, N3) for detection of the recombined allele and (N2, N3, Ex-V-32) for genotyping. Primer pair (Cre1, Cre2) was used for detection of Cre gene. Two percent agarose gel stained with ethidium bromide. #Band that corresponds to the presence of the recombined Nrf2^flox allele. Lanes are as follows: L, 1-kb ladder; N, nontemplate control; 1, wild-type liver; 2, Nrf2^flox/flox epididymal white adipose tissue (eWAT); 3, Nrf2^flox/flox::Adipoq-Cre eWAT; 4, Nrf2^flox/flox liver; 5, Nrf2^flox/flox::Albumin-Cre liver; 6, Nrf2^flox/flox::Adipoq-Cre liver. C: mRNA levels of Nrf2 in eWAT of Control and Nrf2^flox/flox::Adipoq-Cre (ANKO) 2-mo-old male mice on regular diet; n = 4 per genotype. Data are means ± SE. *P < 0.05. D: mRNA levels of Nrf2 in liver of Control and Nrf2^flox/flox::Albumin-Cre (HeNKO) 2-mo-old male mice on regular diet; n = 4 per genotype. Data are means ± SE. *P < 0.05. E: mRNA levels of NAD(P)H quinone oxidoreductase 1 (Nqo1) in eWAT of wild-type (WT), Control, and ANKO male mice on regular diet treated with vehicle (V) or 1[2-cyano-3,12-dioxooleana-1,9(11)-dien-28-oyl]imidazole (CDDO-Im; CD); n = 4–7 per genotype per treatment. Data are means ± SE. *P < 0.05, #P < 0.05 compared with Control-V, $P < 0.05 compared with vehicle treatment of the same genotype. F: mRNA levels of Nqo1 in liver of wild-type, Control, and HeNKO male mice on regular diet treated with vehicle or CDDO-Im; n = 4–7 per genotype per treatment. Data are means ± SE. *P < 0.05, #P < 0.05 compared with Control-V, $P < 0.05 compared with vehicle treatment of the same genotype.

Fig. 2.

Schematic representation of the experimental design. CLAMS, Comprehensive Laboratory Animal Monitoring System for indirect calorimetry; HFD, high-fat diet (60% kcal fat); IPGTT, intraperitoneal glucose tolerance test; IPIST, intraperitoneal insulin sensitivity test; StD; standard diet (10% kcal fat); ¶, body weight measurement.

Fig. 3.

Body weights, body fat content, and energy expenditure in the tissue-specific nuclear factor erythroid 2-related factor 2 (Nrf2) knockout mice. Α: body weights of Control (Nrf2^flox/flox), HeNKO (Nrf2^flox/flox::Albumin-Cre), and ANKO (Nrf2^flox/flox::Adipoq-Cre) mice fed either a standard (StD) or a high-fat diet (HFD) are plotted against time; n = 6–12 mice per diet type per genotype. Symbols: ×, Control on StD; □, HeNKO on StD; ▼, ANKO on StD; ○, Control on HFD; ●, HeNKO on HFD; △, ANKO on HFD. *P < 0.05 HFD-fed mice compared with the StD-fed mice after day 41 onward. B: body fat content assessed by EchoMRI of Control, HeNKO, and ANKO mice fed HFD is plotted against time; n = 9–12 mice per genotype. C and D: lean body mass (LBM)-normalized energy expenditure (EE) values in Control, HeNKO, and ANKO mice before HFD initiation and after 84 days on HFD; n = 8 for Control, n = 4 for HeNKO, and n = 4 for ANKO before HFD and n = 10 for Control, n = 7 for HeNKO, and n = 7 for ANKO after 84 days on HFD. E and F: respiratory exchange ratio (RER) in Control, HeNKO, and ANKO mice before HFD initiation and after 84 days on HFD; n = 8 for Control, n = 4 for HeNKO, and n = 4 for ANKO before HFD and n = 10 for Control, n = 7 for HeNKO, and n = 7 for ANKO after 84 days on HFD. Data are means ± SE.

Fig. 4.

Glucose tolerance and insulin sensitivity in the tissue-specific nuclear factor erythroid 2-related factor 2 (Nrf2) knockout mice. A–C: intraperitoneal glucose tolerance test (IPGTT) in Control, Nrf2^flox/flox::Albumin-Cre (HeNKO), and Nrf2^flox/flox::Adipoq-Cre (ANKO) mice after 2 mo on standard diet (StD, A), 2 mo on high-fat diet (HFD, B), and 4 mo on HFD (C); n = 6–12 mice per genotype. D: IPGTT in Control, HeNKO, and ANKO mice after 6 mo on HFD; n = 9–11 per genotype. *P < 0.05 ANKO vs. Control. E: area under the curve (AUC) calculations for the IPGTT shown in D. F: homeostatic model assessment (HOMA) index for insulin resistance (HOMA-IR) in Control, HeNKO, and ANKO mice after 6 mo on HFD; n = 9–11 per genotype. G: HOMA index for β-cell function (HOMA-%B) in Control, HeNKO, and ANKO mice after 6 mo on HFD; n = 9–11 per genotype. H: intraperitoneal insulin sensitivity test (IPIST) in Control, HeNKO, and ANKO mice after 6 mo on HFD; n = 9–11 per genotype. *P < 0.05 HeNKO vs. Control. I: IPIST as in H with the glucose values plotted as percentage of basal (time 0) glucose; n = 9–11 per genotype. Data are means ± SE.

Fig. 5.

A–G: serum metabolic parameters of Control, Nrf2^flox/flox::Albumin-Cre (HeNKO), and Nrf2^flox/flox::Adipoq-Cre (ANKO) mice after 6 mo on high-fat diet; n = 9–11 per genotype. Data are means ± SE. Glucose levels in A represent the glucose levels at time 0 in Fig. 4D. Fgf21, fibroblast growth factor 21; NEFA, nonesterified fatty acids. *P < 0.05.

Fig. 6.

Liver glycogen, triglyceride contents, and histology in the hepatocyte-specific nuclear factor erythroid 2-related factor 2 (Nrf2) knockout mice on high-fat diet (HFD). A: hepatic triglyceride levels normalized to total protein levels in Control, Nrf2^flox/flox::Albumin-Cre (HeNKO), and Nrf2^flox/flox::Adipoq-Cre (ANKO) mice after 6 mo on HFD; n = 9–10 per genotype. Data are means ± SE. B: liver glycogen content normalized to total protein levels in Control and HeNKO mice; n = 9–10 per genotype. Data are means ± SE. C: hepatic triglyceride species measured by HPLC-mass spectrometry in Control and HeNKO mice after 6 mo on HFD; n = 8 per genotype. Data are means (SD). D: changes of fatty acid content in liver of Control and HeNKO mice after 6 mo on HFD determined by HPLC-electrospray ionization-tandem mass spectrometry; n = 8 per genotype. Data are means (SD). E: representative hematoxylin-eosin (H&E) stain of liver sections of Control and HeNKO mice after 6 mo on HFD. Scale bar: 100 μm. F: representative trichrome stain of liver sections of Control and HeNKO mice after 6 mo on HFD. Scale bar: 100 μm. G: relative fibrosis index (%) as pathologically assessed by trichrome-stained liver sections in Control and HeNKO mice after 6 mo on HFD; n = 8 per genotype. Data are means ± SE. H: representative images of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay on liver sections of Control and HeNKO mice after 6 mo on HFD. Scale bar: 100 μm. Arrows indicate positive staining. I: serum alanine transaminase (ALT) concentration in serum of Control and HeNKO mice after 6 mo on HFD; n = 8 per genotype. Data are means ± SE.

Fig. 7.

Gene expression analysis of liver and white adipose tissue in hepatocyte- and adipocyte-specific nuclear factor erythroid 2-related factor 2 (Nrf2) knockout mice, respectively. A, B, and D: mRNA levels of genes assessed by quantitative real-time PCR in livers of Control and Nrf2^flox/flox::Albumin-Cre (HeNKO) mice (A) and in epididymal white adipose tissue (eWAT, B) and inguinal white adipose tissue (iWAT, D) in Control and Nrf2^flox/flox::Adipoq-Cre (ANKO) mice after 6 mo on HFD; n = 11 for Control, n = 10 for HeNKO, and n = 9 for ANKO. Data are means ± SE. *P < 0.05. Tg, triglyceride synthesis. C and E: representative hematoxylin-eosin-stained sections (from a total of 6 checked per genotype) of eWAT (C) and iWAT (E) of Control and ANKO mice after 6 mo on HFD. Scale bar: 50 μm.

Fig. 8.

Schematic summary of the main phenotypic changes of nuclear factor erythroid 2-related factor 2 knockout (Nrf2^−/−) mice compared with wild-type mice (based on references) and of Nrf2^flox/flox::Albumin-Cre (HeNKO) and Nrf2^flox/flox::Adipoq-Cre (ANKO) mice versus Control after long-term exposure to high-fat diet. Arrows pointing upward indicate increase, arrows pointing downward indicate decrease, horizontal left-right arrows mean no change, and arrows with two intersecting lines indicate trend. BW/F, body weight and body fat; chol, serum cholesterol; EE, energy expenditure; IR, insulin resistance; GT, glucose tolerance; NEFA, serum nonesterified fatty acids; TG, serum triglycerides.