The Keap1-Nrf2 system prevents onset of diabetes mellitus

Abstract

Transcription factor Nrf2 (NF-E2-related factor 2) regulates a broad cytoprotective response to environmental stresses. Keap1 (Kelch-like ECH-associated protein 1) is an adaptor protein for cullin3-based ubiquitin E3 ligase and negatively regulates Nrf2. Whereas the Keap1-Nrf2 system plays important roles in oxidative stress response and metabolism, the roles Nrf2 plays in the prevention of diabetes mellitus remain elusive. Here we show that genetic activation of Nrf2 signaling by Keap1 gene hypomorphic knockdown (Keap1flox/-) markedly suppresses the onset of diabetes. When Keap1flox/- mice were crossed with diabetic db/db mice, blood glucose levels became lower through improvement of both insulin secretion and insulin resistance. Keap1flox/- also prevented high-calorie-diet-induced diabetes. Oral administration of the Nrf2 inducer CDDO-Im {oleanolic acid 1-[2-cyano-3,12-dioxooleana-1,9(11)-dien-28-oyl] imidazole} also attenuated diabetes in db/db mice. Nrf2 induction altered antioxidant-, energy consumption-, and gluconeogenesis-related gene expression in metabolic tissues. Thus, the Keap1-Nrf2 system is a critical target for preventing the onset of diabetes mellitus.

Figures

Fig 1

Nrf2 prevents diabetes in db/db mice. (A and B) Blood glucose levels (A) and body weights (B) of db/db::Keap1flox/+ (control) or db/db::Keap1flox/− mice crossed with Nrf2+/+ or Nrf2−/− mice (n = 9 to 20) and fed ad libitum. Also shown are blood glucose (C) and plasma insulin (D) levels in the OGTT. Glucose (2 g/kg of BW) was orally administered to 13- to 18-week-old mice after 14 h of fasting (n = 9 to 20). The lower graphs demonstrate AUC. (E) IHC of insulin in pancreatic sections (low, 5× objective; high, 20× objective) of 18-week-old control or db/db::Keap1flox/− mice (left) and the quantified insulin-positive area in these sections (right; n = 9 to 12). Scale bars, 500 μm (low) and 100 μm (high). P < 0.001 (***), P < 0.01 (**), and P < 0.05 (*) versus db/db::Keap1flox/+. Error bars, SEM.

Fig 2

Nrf2 enhances insulin sensitivity and prevents obesity. (A and B) Blood glucose (n = 5) (A) and plasma insulin levels (n = 10) (B) of male lean Keap1flox/+ (control) or Keap1flox/− mice fed ad libitum. ***, P < 0.001 versus control. (C) Blood glucose levels of male lean control or Keap1flox/− mice crossed with Nrf2+/+ or Nrf2−/− mice fed ad libitum (10 weeks old; n = 9 to 19). (D) Body weights of male lean control or Keap1flox/− mice crossed with Nrf2+/+ (n = 5) or Nrf2−/− (n = 4 or 5) mice. ***, P < 0.001 versus control. (E) Blood glucose levels of male lean control or Keap1flox/− mice crossed with Nrf2+/+ (n = 10) or Nrf2−/− (n = 9 to 11) mice in the ipGTT. Glucose (2 g/kg of BW) was administered after 14 h of fasting (mice were 8 to 9 weeks old). ***, P < 0.001 versus control. (F) Blood glucose levels of male lean control or Keap1flox/− mice crossed with Nrf2+/+ (n = 9 or 10) or Nrf2−/− (n = 6 or 7) mice in the ITT. Regular insulin (0.75 U/kg of BW) was administered after 0.5 h of fasting (mice were 9 to 10 weeks old). ***, P < 0.001 versus control. (G and H) Blood glucose levels of male lean Keap1flox/flox (control) or tissue-specific Keap1 knockout mice in SkM (G) (Keap1 MuKO, n = 10 to 12) and liver (H) (Keap1 LKO, n = 8) in the ITT. Regular insulin (0.75 U/kg of BW) was administered after 0.5 h of fasting (mice were 8 to 9 weeks old). P < 0.001 (***) and P < 0.05 (*) versus control. (I) Blood glucose levels of male lean control, Keap1 MuKO, and Keap1 LKO mice fed ad libitum. ***, P < 0.001 versus control. (J to L) Body weight changes (J and K) and blood glucose levels (L) of male Keap1flox/+ mice fed the SD (Keap1 F/+ SD), Keap1flox/+ mice fed the HCD (HFD with 20% sucrose in drinking water; Keap1 F/+ HCD), Keap1flox/− mice fed the SD (Keap1 F/− SD), and Keap1flox/− mice fed the HCD (Keap1 F/− HCD) crossed with Nrf2+/+ (J and L) (n = 5) or Nrf2−/− (K) (n = 4 to 6) mice. P < 0.001 (***) and P < 0.01 (**) versus Keap1 F/− SD. Error bars, SEM.

Fig 3

Nrf2 increases energy consumption-related genes in SkM. (A) Energy consumption-related gene expression in liver, SkM, white adipose tissue (WAT), and BAT of male 8-week-old lean Keap1flox/+ (control) or Keap1flox/− mice. Data are normalized with Hprt (n = 5 or 6). **, P < 0.01 versus Keap1flox/+. (B and C) Plasma NEFA (B) and TAG (C) levels in male lean 20-week-old control or Keap1flox/− mice crossed with Nrf2+/+ or Nrf2−/− mice (n = 5 or 6). ***, P < 0.001; *, P < 0.05. (D) Cpt1b expression in SkM of 20-week-old male control or Keap1flox/− mice fed the HCD for 8 weeks crossed with Nrf2+/+ or Nrf2−/− (n = 5 or 6) mice. Data are normalized with Hprt. ***, P < 0.001; *, P < 0.05. (E) Immunoblot analysis for CPTIB in SkM of 20-week-old male control or Keap1flox/− mice fed the HCD and quantified protein expression level normalized with α-tubulin (n = 4). **, P < 0.01. (F) Oxygen consumption levels in SkM of male (n = 5) or female (n = 6) 8-week-old control or Keap1flox/− mice fed the HCD for 2 weeks. **, P < 0.01 versus control. Error bars, SEM.

Fig 4

Nrf2 induction in SkM prevents HFD-induced obesity. (A) Body weight changes of male Keap1flox/flox (control) or Keap1 MuKO mice fed the SD (n = 4) or the HFD (n = 11). (B) Blood glucose levels of male control or Keap1 MuKO mice 8 weeks after being fed the SD (n = 4) or HFD (n = 11) ad libitum. (C) Plasma insulin levels of male control or Keap1 MuKO mice 8 weeks after being fed the HFD (n = 5 each) ad libitum. (D and E) Locomotor activity in the open-field study of male control or Keap1 MuKO mice 8 weeks after being fed the HFD (n = 5). Moved distance and rated crossing zone number were quantified every 5 min (0 to 5, 5 to 10, 10 to 15, and 15 to 20 min) or through the whole testing period (0 to 20 min). Data are represented moved distance (D) and rated crossing zone number (E) per minute. (F and G) Correlation between locomotor activity and body weight. Locomotor activities are represented as moved distance (F) and rated crossing zone number (G) in the 15- to 20-min period of the open-field test. P < 0.001 (***) and P < 0.05 (*) versus control. Error bars, SEM.

Fig 5

Nrf2 induction decreases gluconeogenesis-related genes in liver. (A) Expression levels of gluconeogenesis-related genes in livers of 14- to 18-week-old db/db::Keap1flox/+ (control) or db/db::Keap1flox/− mice crossed with Nrf2+/+ or Nrf2−/− mice (n = 13 to 18). Data are normalized with Hprt. **, P < 0.01; *, P < 0.05. (B) Blood glucose levels in the PTT. Pyruvate (1.5 g/kg of BW) was intraperitoneally administered after 24 h of fasting to 8- to 10-week-old control or db/db::Keap1flox/− mice (n = 12 to 21). ***, P < 0.001 versus control. (C) Effect of CDDO-Im on G6pc mRNA expression in AML12 cells. After a 24-h incubation with 10 nmol/liter of CDDO-Im, cells were stimulated with 1 μmol/liter of Dex or 1 mmol/liter of dbcAMP for 6 h. Data are represented as G6pc expression levels and normalized with β-actin (n = 4). ***, P < 0.001. (D) Effect of CDDO-Im on G6pc regulatory activities in AML12 cells stably transfected with G6pc-Luc2P. After a 24-h incubation with 10 nmol/liter of CDDO-Im, cells were stimulated with 1 mmol/liter of dbcAMP for 9 h. Data are represented as luciferase activities after normalization to the total cellular protein amounts (n = 5). **, P < 0.01. (E) Role of Nrf2 in G6pc regulatory activities in AML12 cells transiently transfected with G6pc-Luc2P. Data are represented as firefly luciferase activities after normalization to Renilla luciferase activities (n = 6). P < 0.001 (***) and P < 0.01 (**) versus CREB. (F) Immunoblot analysis for CREB in liver and the quantified protein expression levels normalized with α-tubulin of 18-week-old control or db/db::Keap1flox/− mice (n = 5). ***, P < 0.001. (G) Plasma glucagon concentrations of male 18-week-old Keap1flox/− or Keap1flox/+ lean or db/db mice fed ad libitum (n = 3). **, P < 0.01 versus lean mice. Error bars, SEM.

Fig 6

Nrf2 contributes to β-cell protection. (A) Nrf2 mRNA expression in isolated islets of Nrf2+/+ or Nrf2−/− mice normalized with 18S rRNA (n = 3). ***, P < 0.001 versus Nrf2+/+ liver. (B) IHC of insulin in pancreatic sections (low, 5× objective; high, 20× objective) of 10-week-old male Nrf2+/+ or Nrf2−/− mice (left) and the quantified islet size in these sections (right; n = 368 to 406; islets from 10 mice in each group). Scale bars, 500 μm (low) and 100 μm (high). ***, P < 0.001 versus Nrf2+/+. (C) Insulin contents of isolated islets from 10-week-old male Nrf2+/+ or Nrf2−/− mice normalized with total protein contents (n = 8). ***, P < 0.001 versus Nrf2+/+. (D) GSIS from islets isolated from 10-week-old male Nrf2+/+ or Nrf2−/− mice (n = 8). ***, P < 0.001. (E) Insulin and antioxidant gene mRNA expression in islets from Nrf2+/+ and Nrf2−/− mice normalized with 18S rRNA (n = 3). ***, P < 0.001. (F) Antioxidant gene expressions in islets compared with livers of male 10-week-old wild-type mice. Data are normalized with 18S rRNA (n = 4). ***, P < 0.001 versus liver. Error bars, SEM.

Fig 7

Administration of CDDO-Im protects β-cell function and improves insulin resistance in db/db mice. (A to C) Blood glucose levels (A), body weight (B), and plasma insulin levels (C) of vehicle- and CDDO-Im-treated db/db mice fed ad libitum (male, n = 8). P < 0.001 (***), P < 0.01 (**), and P < 0.05 (*) versus vehicle. (D) HOMA-IR of 16-week-old male db/db mice (n = 8). **, P < 0.01 versus vehicle. (E and F) Blood glucose (E) and plasma insulin (F) levels in the OGTT of db/db mice. Glucose (0.5 g/kg of BW) was orally administered to 16-week-old male db/db mice after 14 h of fasting (n = 8). P < 0.001 (***) and P < 0.01 (**) versus vehicle; P < 0.05 (*) versus time zero. (G) IHC of insulin in pancreatic sections (low, 5× objective; high, 20× objective) of 18-week-old db/db mice (left) and quantified insulin-positive areas in these sections (right) (n = 8). Scale bars, 500 μm (low) and 100 μm (high). *, P < 0.05 versus vehicle. Error bars, SEM.

Fig 8

Suggested mechanisms of prevention of diabetes by the Keap1-Nrf2 system. Induction of Nrf2 activity increases expression of antioxidant enzyme genes in pancreatic β cells and energy consumption-related genes in SkM but decreases expression of gluconeogenesis-related genes in the liver. The Keap1-Nrf2 system suppresses diabetes onset via both restoration of sufficient insulin secretion and insulin resistance.