Diabetic Downregulation of Nrf2 Activity via ERK Contributes to Oxidative Stress–Induced Insulin Resistance in Cardiac Cells In Vitro and In Vivo

HL-1 cells were obtained from Dr. Claycomb (Louisiana State University Health Science Center, New Orleans, LA), who first established and characterized the cell line derived from adult murine atrial cardiomyocyte tumor lineage, and cultured in Claycomb Medium (Sigma-Aldrich) (16). Cells were infected with adenovirus of β-galactosidase (Ad-βGal) or murine Nrf2 (Ad-Nrf2) as previously described (15).

Cells were incubated in Krebs-Ringer-HEPES buffer (15 mmol/L HEPES [pH 7.4], 105 mmol/L NaCl, 5 mmol/L KCl, 1.4 mmol/L CaCl₂, 1 mmol/L KH₂PO₄, 1.4 mmol/L MgSO₄, and 10 mmol/L NaHCO₃) for 30 min. Cells next were incubated with insulin (100 nmol/L) in Krebs-Ringer-HEPES buffer for 30 min and then added to 2-deoxy-d-glucose (0.2 mmol/L) and 2-deoxy-D-[³H]glucose (2-DG) (1 μCi/mL, MP Biomedicals) for an additional 10 min. The cells were washed three times with ice-cold PBS, solubilized in 0.5 mol/L NaOH, and neutralized with 0.5 mol/L HCl, and the radioactivity in aliquots from cell extracts was measured using a liquid scintillation counter. Protein concentration of the cell extracts was measured using a protein assay kit (Bio-Rad). Deoxyglucose uptake was expressed as counts per minute per microgram of protein.

Cells were transfected with Nrf2 transcription reporter gene ARE-luc (firefly luciferase) and internal control pRL-TK-luc (renilla luciferase) plasmids, and Nrf2 transcriptional activity was quantified by measuring luciferase activities as previously described (17).

MT-overexpressing transgenic (MT-TG) mice were produced from the FVB mice that have been well characterized (12–14). Both 8- to 10-week-old MT-TG positive mice (heterozygotes) and negative littermates (wild type [WT]) were kept in the same cages with free access to rodent chow and tap water. All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of Louisville, which is certified by the American Association for Accreditation of Laboratory Animal Care. STZ (Sigma, St. Louis, MO) was dissolved in sodium citrate buffer (pH 4.5). Male mice were given an intraperitoneal injection of STZ at 40 mg/kg body wt daily for five days. Whole blood glucose obtained from the murine tail vein was detected using a SureStep complete blood glucose monitor (LifeScan, Milpitas, CA) five days after the last STZ injection. STZ-injected mice with glucose levels >12 mmol/L were considered diabetic, and mice serving as controls were given the same volume of sodium citrate (12–14).

Two animal studies were performed. The first study was to investigate the diabetic effect on cardiac Nrf2 expression along with nitrosative damage, measured by 3-nitrotyrosine (3-NT) and ERK expression, using MT-TG and WT diabetic mice at 2 and 5 months after diabetes onset. The second study was to investigate the preventive effect of Nrf2 activator Dh404 (provided by Reata Pharmaceuticals) on cardiac Nrf2 expression and 3-NT as well as insulin signaling targets including phosphorylation of Akt and glycogen synthase kinase-3β (GSK-3β) at 2 weeks after diabetes onset. For Dh404 treatment, diabetic and age-matched nondiabetic mice were administrated with Dh404 (10 mg/kg body wt) or vehicle (sesame oil; Sigma) every other day by gavage. Two weeks later, 6 h after the last dose administration, mice were killed and heart tissues were collected for the study.

Staining and Western blot analysis for Nrf2, total and phosphor-ERK1/2, total and phosphor-Akt, total and phosphor-GSK-3β, and 3-NT were performed as previously described (9,12–14). Briefly, heart tissues were homogenized in lysis buffer using homogenizer. Tissue proteins were collected by centrifuging at 12,000g at 4°C in a Beckman GS-6R centrifuge for 15 min. The protein concentration was measured by Bradford assay. The sample, diluted in loading buffer and heated at 95°C for 5 min, was then subjected to electrophoresis on 10% SDS-PAGE gel at 120 V. After electrophoresis of the gel and transfer of the proteins to nitrocellulose membrane, the membranes were rinsed briefly in Tris-buffered saline, blocked in blocking buffer (5% milk and 0.5% BSA) for 1 h, and washed three times with Tris-buffered saline containing 0.05% Tween 20. The membranes were incubated with different primary antibodies at a dilution of 1:1,000 for 2 h and then washed and reacted with secondary horseradish peroxidase–conjugated antibody for 1 h. Antigen-antibody complexes were then visualized using an ECL kit (Amersham, Piscataway, NJ).

Tissue sections of left ventricles were prepared from autopsy heart specimens of humans with or without a disease history or diabetes. All diabetic males had histories of hypertension and cardiac dysfunction (Supplementary Table 1). This study was approved by the Institutional Review Board for human subject research at the University of South Carolina School of Medicine and the Memorial Medical Center of Johnstown.

Values are expressed as means ± SD. The data were analyzed using ANOVA with the Newman-Keuls test. Values of P < 0.05 were considered to be statistically significant.

Chronic treatment with H₂O₂ has been demonstrated to induce oxidative stress leading to insulin resistance in various cell types (18); therefore, we tested whether such a treatment induces insulin resistance in HL-1 adult cardiomyocytes. Consistent with a previous report (19), insulin stimulated 20–30% increases in glucose uptake in HL-1 cells (Fig. 1A), and treatment with H₂O₂ at a nontoxic dose of 100 μmol/L for 16 h suppressed not only the basal but also the insulin-induced glucose uptake (Fig. 1A). Pretreatment of N-acetyl-cysteine (NAC), a ROS scavenger, partially reversed H₂O₂-induced inhibition of the basal and insulin-induced glucose uptake (Fig. 1A), suggesting that oxidative stress directly causes insulin resistance in adult cardiomyocytes.

Potential link of ERK1/2 with oxidative stress and insulin resistance in cardiomyocytes has been documented (19). Accordingly, we examined the effect of chronic treatment of H₂O₂ on ERK1/2 phosphorylation in HL-1 with and without insulin stimulation. As shown in Fig. 1B, acute stimulation of insulin (30 min) or chronic treatment of H₂O₂ (16 h) increased ERK phosphorylation; however, the insulin-induced ERK activity was significantly enhanced by H₂O₂. NAC did not affect the insulin-induced ERK activity but suppressed ERK activity induced by H₂O₂ per se or H₂O₂ with insulin (Fig. 1B). The reservation of H₂O₂-induced insulin resistance (Fig. 1A) and suppression of H₂O₂-mediated ERK activity (Fig. 1B) by NAC reflects a possible role of ERK in oxidative stress–induced insulin resistance in cardiomyocytes.

We next examined the role of ERK and PI3K in regulating insulin-mediated glucose uptake in HL-1 cells. In control cells, addition of U0126, an ERK inhibitor, enhanced the basal (Fig. 2A) and insulin-induced glucose uptake (Fig. 2B), respectively; surprisingly, LY294002, a PI3K inhibitor, had no effect (Fig. 2A and B). In the cells with chronic H₂O₂ treatment, U0126 was able to partially reverse, but LY294002 worsened, the H₂O₂-mediated inhibition of glucose uptake (Fig. 2A) and insulin resistance (Fig. 2B). These results support the above hypothesis that ERK as a negative regulator of insulin-mediated glucose uptake plays a mediator role in oxidative stress–induced insulin resistance in adult cardiomyocytes.

Because we have demonstrated that Nrf2 negatively regulates oxidative stress in the heart (15), we then determined whether Nrf2 plays a role in regulating the oxidative stress–induced insulin resistance in HL-1 cells. Nrf2 transcription reporter assay showed that insulin activated Nrf2, which was enhanced by U0126, whereas blocked by LY294002 (Fig. 3A), suggesting that the PI3K-mediated Nrf2 activation is negatively regulated by ERK in insulin-stimulated HL-1 cells. To further test whether Nrf2 plays a critical role in regulating insulin-mediated glucose uptake, especially under an oxidative status, we examined effects of forced activation of Nrf2 on the oxidative stress–induced activation of ERK as well as the subsequent insulin resistance in adult cardiomyocytes. As shown in Fig. 3B, a forced activation of Nrf2 by adenoviral overexpression of Nrf2 significantly suppressed H₂O₂-induced ERK activity in HL-1 cells with the presence of insulin. Most importantly, the forced activation of Nrf2 enhanced the insulin-mediated glucose uptake in non–H₂O₂-treated cells, relative to the control of Ad-βGal overexpression and completely recovered insulin sensitivity to stimulate glucose uptake in H₂O₂-treated cells (Fig. 3C).

To explore a pathological relevance of Nrf2 in diabetic hearts, we examined cardiac Nrf2 expression in STZ-induced diabetic mice at 2 and 5 months after diabetes onset. The glucose levels of these diabetic mice were 25.76 ± 2.14 mmol/L at 2 months and 26.61 ± 1.70 mmol/L at 5 months, respectively. As shown in Fig. 4, cardiac Nrf2 expression was slightly increased at 2 months (Fig. 4A) but significantly decreased at 5 months (Fig. 4B) after diabetes onset. The decreased cardiac Nrf2 expression at the late stage was accompanied with oxidative and nitrosative damage, shown by significantly increased cardiac 3-NT accumulation (Fig. 4C) and overactivation of ERK function (Fig. 4D), as observed in the in vitro study (Fig. 3). Importantly, all these pathogenic changes were not observed in the hearts of the MT-TG diabetic mice (Fig. 4B–D). These findings suggest that diabetic oxidative and nitrosative stress may induce ERK activation that downregulates Nrf2 expression.

Consistent with the finding from diabetic mice, immunochemical staining of human normal and diabetic hearts also showed that Nrf2 protein expression was dramatically suppressed in the diabetic failing hearts of both males (Fig. 5A) and females (Fig. 5B), as compared with nondiabetic hearts.

We recently demonstrated that a novel synthetic triterpenoid derivative, Dh404, is a potent Nrf2 activator and suppresses oxidative stress in cardiomyocytes (20). We next determined whether Dh404 could mimic the inhibitory effect of Nrf2 overexpression on H₂O₂-induced insulin resistance in HL-1 cells. As expected, Dh404 strongly activated Nrf2 transcriptional activity and additively increased the insulin-induced Nrf2 transcriptional activity (Fig. 6A). Dh404 treatment did not change insulin-induced ERK phosphorylation (Fig. 6B) and glucose uptake (Fig. 6C) but inhibited H₂O₂-induced ERK phosphorylation and partially reversed H₂O₂-inhibited basal glucose uptake or insulin-stimulated glucose uptake (Fig. 6B and C). These results further support the notion that Nrf2 is a critical regulator of insulin-mediated glucose uptake in cardiomyocytes in vitro.

To validate the in vitro finding, diabetic mice, induced by multiple-low doses of STZ, were treated with Dh404 at 10 mg/kg every other day for 2 weeks. Cardiac tissues were collected from these diabetic mice with and without Dh404 treatment and subject to Western blotting analysis for Nrf2 and 3-NT. Dh404 treatment did not change glucose levels of these mice (control 8.40 ± 1.20; diabetes: 25.40 ± 2.89; Dh404: 7.99 ± 0.56; diabetes/Dh404: 27.12 ± 1.22 mmol/L). Cardiac Nrf2 expression was increased in the hearts of either diabetic or Dh404-treated nondiabetic mice and synergistically increased in the hearts of Dh404-treated diabetic mice (Fig. 7A). Diabetes significantly induces 3-NT accumulation (Fig. 7B) and ERK1/2 phosphorylation (Fig. 7C), which were prevented by Dh404 treatment, suggesting that activation of Nrf2 expression by Dh404 in the heart plays an important role in the prevention of diabetes-induced cardiac oxidative and/or nitrosative damage and ERK1/2 activation. As expected, diabetes-induced oxidative damage and ERK1/2 activation is accompanied with downregulation of insulin signaling, shown by decreased phosphorylation of Akt (Fig. 7D) and GSK-3β (Fig. 7E). Activation of Nrf2 by Dh404 preserved the normal levels of Akt and GSK-3β phosphorylation in the heart of diabetic mice (Fig. 7D and E).

Oxidative stress is causative of insulin resistance in various cell types (2,6,7,18), which was confirmed by our present study using HL-1 adult cardiomyocytes treated with H₂O₂. Reportedly the stimulating effect of acute treatment with H₂O₂ on glucose uptake has been recognized for years in adipocytes (21). In fact, the cellular generation of ROS such as H₂O₂ by insulin has been considered as an integral part of the insulin signaling (21). These results emphasize the biological significance of ROS as signaling molecules, e.g., for the insulin-mediated glucose metabolism in cardiomyocytes; however, chronic accumulation of ROS causes oxidative stress, thereby contributing to the pathogenesis of various diseases such as cardiac dysfunction (9). Cardiac oxidative stress is often observed coincident with insulin resistance, and there is accumulating evidence that ROS and/or RNS mediates deleterious effects in the insulin-resistant heart although the mechanisms by which oxidative stress induces insulin resistance remain not well understood.

Increasing evidence indicates that ERK activation seems a causal factor for insulin resistance caused by oxidative stress in cardiomyocytes. For instance, through activation of ERK1/2, high levels of glucose or angiotensin II induces insulin resistance in several conditions (19,22,23). A recent study that showed that deficiency in the ERK1 protects leptin-deficient mice from insulin resistance without affecting obesity further confirms the causative of ERK activation in the insulin resistance (24). In the current study, we demonstrated for the first time that chronic treatment of HL-1 cells with H₂O₂ increased ERK phosphorylation along with a significant inhibition of glucose uptake. More importantly we also demonstrated in the diabetic hearts that upregulation of ERK1/2 along with nitrosative damage is associated with downregulation of phosphorylation of both Akt and GSK-3β, two key components of insulin signaling cascade. In the cells with chronic H₂O₂ treatment, inhibition of ERK1/2 activation by U0126 was able to partially reverse glucose uptake and insulin resistance (Fig. 2B). Findings that NAC prevents H₂O₂-mediated activation of ERK1/2 in vitro (Fig. 1B) and MT prevents diabetes-induced activation of ERK1/2 in vivo (Fig. 4D) also suggest that the activation of ERK1/2 is associated with oxidative stress induced either by H₂O₂ in vitro or diabetes in vivo.

Nrf2 has been shown to negatively regulate oxidative stress in the heart (15). To support this notion, our recent study showed that high levels of glucose induced ROS production in both primary neonatal and adult cardiomyocytes from the heart of Nrf2 WT (C57BL/6 background) murine heart, whereas, in Nrf2 knockout cells, ROS was significantly higher under basal conditions and further enhanced under high glucose exposure conditions than that in WT cells (11). Concomitantly high glucose induced significantly higher levels of apoptosis at lower concentrations and in shorter time in Nrf2 knockout cells than in Nrf2 WT cells (11). To explore a pathological relevance of Nrf2 in diabetic hearts, we examined cardiac Nrf2 expression in STZ-induced diabetic mice. As shown in Figs. 4 and 7, cardiac Nrf2 expression was slightly increased in the heart of diabetic mice at 2 weeks (Fig. 7A) and 2 months (Fig. 4A) but significantly decreased in the hearts of diabetic mice at 5 months (Fig. 4B) after diabetes onset. The slight increase in cardiac Nrf2 expression at the early stages of diabetes was consistent with our previous study (11) in which Nrf2 was increased in the heart of diabetic mice with C57BL/6 background at 2 weeks after STZ-induced hyperglycemia. These results suggest that Nrf2 is adaptively trying to remain functional to overcome diabetic damage at the early stage of diabetes. At the late stage of diabetes, however, cardiac antioxidant function is further impaired, leading to a decrease in cardiac Nrf2 expression. Importantly, the decreased cardiac Nrf2 expression in the diabetic hearts was accompanied with oxidative and nitrosative damage, shown by significantly increased 3-NT accumulation (Fig. 4C) and overactivation of ERK function (Fig. 4D) as observed in the in vitro study (Fig. 3). Peroxynitrite interacts with lipids, DNA, and proteins via direct nitration or via indirect, radical-mediated mechanisms. These reactions trigger cellular responses ranging from subtle modulations of cell signaling to overwhelming oxidative and nitrosative injury. In vivo, peroxynitrite generation represents a crucial pathogenic mechanism in conditions such as diabetic cardiomyopathy (12,25).

We have demonstrated that MT-TG mice are resistant to diabetes-induced oxidative and/or nitrosative damage and development of cardiomyopathy (13,26). Here we further showed that MT-TG mice are also significantly resistant to diabetes-induced decrease in cardiac Nrf2 expression (Fig. 4B) and increases in 3-NT accumulation (Fig. 4C) and ERK activation (Fig. 4D) at the late stage of diabetes (5 months after diabetes onset). Our finding is consistent with a recent study in which MT-TG mice are resistant to lipopolysaccharide (LPS)-induced cardiac oxidative stress, ERK1/2 activation, and cardiac dysfunction as compared with WT mice (27). Therefore, like NAC that protects cardiomyocytes from H₂O₂ downregulation of Nrf2 in the in vitro study (Fig. 3), MT as a potent antioxidant prevents diabetic oxidative and nitrosative stress, leading to a suppression of ERK activation and a preservation of cardiac Nrf2 function, as illustrated in Fig. 7F, to protect the heart from diabetes- and LPS-induced damage. We recently also showed that inactivation of GSK-3β by overexpression of cardiac MT gene results in an improvement of cardiac glucose metabolism in the diabetic hearts along with a significant suppression of cardiac nitrosative damage, inflammation, and remodeling (14).

In the current study, we demonstrated for the first time that Nrf2 plays a critical role in insulin-mediated glucose uptake, especially under an oxidative status. By Nrf2 transcription reporter assay we found that insulin activated Nrf2 (Fig. 3A), which could be enhanced by inhibition of ERK1/2 with U0126, whereas it is blocked by PI3K inhibition with LY294002 (Fig. 3A), indicating that insulin activates Nrf2 via PI3K-mediated signaling that is negatively regulated by ERK in HL-1 cells, as illustrated in Fig. 7F. Furthermore, forced activation of Nrf2 by adenoviral overexpression of Nrf2 significantly suppressed insulin-induced ERK activity in HL-1 cells with and without H₂O_2. This finding suggests that Nrf2 activation could prevent oxidative stress–induced ERK activity in adult cardiomyocytes (Fig. 7F). Most importantly, the forced activation of Nrf2 further enhances insulin-mediated glucose uptake in non–H₂O₂-treated cells and also completely recovered insulin-mediated glucose uptake in H₂O₂-treated cells (Fig. 3C). The in vitro finding was further supported by in vivo results showing that diabetes inhibits cardiac Akt and GSK-3β phosphorylation, effects that were prevented by enhanced expression of Nrf2 with Dh404 (Fig. 7D and E). Therefore, these results not only establish a key role of Nrf2 in the maintaining of insulin sensitivity but also raise a possibility that Nrf2 is a crucial player in the regulation cardiac insulin signaling in the cardiomyocytes and the heart, as illustrated in Fig. 7F. These findings provide further support for other recent findings implying the link of Nrf2 deficiency to oxidative stress and insulin resistance in other cells and tissues (28,29). Therefore, diabetes-induced oxidative stress suppresses Nrf2 expression (Fig. 5) (30), and lack of Nrf2 expression in the diabetic hearts and kidney contributes to an insulin resistance and eventually cardiac dysfunction (11) or renal dysfunction (31,32).

In summary, our data provide several novel findings of redox signaling in the regulation of insulin sensitivity in the cardiac cells in vitro and in vivo, as illustrated by Fig. 7F. First, oxidative stress plays a causative role in the development of insulin resistance. Second, ERK is a negative regulator of glucose uptake and mediates oxidative stress–induced insulin resistance. Third, ERK-mediated suppression of Nrf2 activity is linked to the oxidative stress–induced insulin resistance. Fourth, Nrf2 expression is upregulated at early stage of diabetic heart but significantly downregulated at the late stage in diabetic hearts. Fifth, Nrf2 activation suppresses the oxidative stress–induced ERK activity and reverses the oxidative stress–induced insulin resistance. Finally, antioxidants such as NAC and MT can prevent oxidative stress–induced ERK activation and Nrf2 downregulation. These results suggest that Nrf2 plays a critical role in regulating insulin sensitivity in the heart, and targeting Nrf2 might provide a novel therapeutic approach for the treatment of cardiac insulin resistance and diabetic cardiomyopathy. Importantly a close structural analog of Dh404 that also potently induces Nrf2, bardoxolone methyl (CDDO-Me; RTA 402), is in a pivotal clinical study in patients with type 2 diabetes mellitus with advanced kidney disease to assess changes in renal function and glucose control (33,34).

This study was supported in part by the American Heart Association as a Beginning Grant in Aid (BGIA) (0865101E to T.C.), the American Diabetes Association as a Career Development Award (05-07-CD-02 to L.C.), a Starting-Up Fund for Chinese-American Research Institute for Diabetic Complications from Wenzhou Medical College (to X.L. and L.C.), and a Zhejiang Provincial Extremely Key Subject Building Project “Pharmacology and Biochemical Pharmaceutics 2009” (to X.L. and L.C.).

C.J.M. is employed by Reata Pharmaceuticals. No other potential conflicts of interest relevant to this article were reported.

Y.T. and T.I. researched data and reviewed and edited the article. J.L., Q.S., H.Y., X.C., C.S.G., and C.J.M. researched data. X.L. contributed initial discussion of the project and reviewed the article. L.C. and T.C. wrote the article, contributed to discussion, and reviewed and edited the article.