Cardiac-Specific Overexpression of CYP2J2 Attenuates Diabetic Cardiomyopathy in Male Streptozotocin-Induced Diabetic Mice

Abstract

Cytochrome P450 (CYP) epoxygenases metabolize arachidonic acid to biologically active cis-epoxyeicosatrienoic acids, which have potent vasodilatory, antiinflammatory, antiapoptotic, and antidiabetes properties. Here, we showed the effects of cardiac-specific overexpression of CYP epoxygenase 2J2 (CYP2J2) on diabetic cardiomyopathy and insulin resistance in high-fat (HF) diet fed, low-dose streptozotocin-treated mice. Diabetic cardiomyopathy was induced by HF and streptozotocin in cardiac-specific CYP2J2 transgenic mice. Physiological parameters and systemic metabolic parameters were monitored using ELISA kits. Intraperitoneal injection glucose tolerance test and hyperinsulinemic-euglycemic clamp study were implied to indicate insulin resistance. Cardiac function was assessed by echocardiography and Millar catheter system. Real-time PCR and Western blotting were used in signal pathway detection. αMHC-CYP2J2 transgenic mice showed significantly lower plasma glucose and insulin levels, improved glucose tolerance, and increased cardiac glucose uptake. Furthermore, αMHC-CYP2J2 transgenic mice were significantly protected from HF-streptozotocin-induced diabetic cardiomyopathy. Strikingly, CYP2J2 overexpression attenuated myocardial hypertrophy induced by diabetes. We conclude that cardiac-specific overexpression of CYP2J2 significantly protects against diabetic cardiomyopathy, which may be due to improved cardiac insulin resistance, glucose uptake, and reversal of cardiac hypertrophy. Relevant mechanisms may include up-regulation of peroxisome proliferator-activated receptor γ, activation of insulin receptor and AMP-activated protein kinase signaling pathways, and inhibition of nuclear factor of activated T cells c3 signal by enhanced atrial natriuretic peptide production. These results suggest that CYP2J2 epoxygenase metabolites likely play an important role in plasma glucose homeostasis, and enhancement of epoxyeicosatrienoic acids activation may serve as an effective therapeutic strategy to prevent diabetic cardiomyopathy.

Diabetes mellitus is associated with a specific cardiomyopathy, independent of hypertension, coronary artery disease, or hyperlipidemia, as evidenced by clinical studies and experimental animal models (1, 2). Prominent early features of diabetic cardiomyopathy include impaired myocardial relaxation and increased diastolic stiffness (ie, diastolic dysfunction), abnormalities which are exacerbated after ischemia (1, 2). More advanced cardiomyopathy is characterized by systolic and autonomic dysfunction. However, the underlying mechanisms of diabetic cardiomyopathy are not completely understood.

Arachidonic acid is converted to eicosanoid mediators by cyclooxygenase, lipoxygenase, and cytochrome P450 (CYP) monooxygenase pathways (3). The CYP pathway produces 2 types of eicosanoid products, the epoxyeicosatrienoic acids (EETs), formed by CYP epoxygenases, and the hydroxyeicosatetraenoic acids, formed by CYP ω-hydroxylases (4). Over the last decade, accumulating evidences have suggested that EETs play crucial and diverse roles in cardiovascular homeostasis (5, 6). Exogenous application of EETs inhibits vascular smooth muscle cell migration, platelet aggregation, nuclear factor κB activation and vascular cell adhesion molecule-1 expression (5, 7–9), and protects endothelial cells from apoptosis (10). EETs also have strong vasodilator and angiogenic effects both in vivo and in vitro (11), suggesting an overall beneficial role for EETs within the vasculature. Recent study demonstrated that epoxygenase overexpression and EETs attenuate diabetes and insulin resistance (12).

CYP epoxygenase 2J2 (CYP2J2) is abundantly expressed in cardiomyocytes and coronary vascular endothelial cells. However, the roles of CYP2J2 and its EET products in the diseased heart are not fully understood. Diabetic cardiomyopathy is associated with significant endothelial cell dysfunction, attenuation of endothelial nitric oxide synthase (eNOS) activity, abnormal cardiac ion transport (13, 14), and reduced activation of AMP-activated protein kinase (AMPK) and Akt, also known as protein kinase B (PKB), which play important roles in energy metabolism and surviving of myocardium (15, 16). Recent studies have demonstrated that EETs can activate cardiac and vascular ATP-sensitive K⁺ channels and enhance cardiac L-type Ca²⁺ currents (17). EETs also activate Akt (18) and up-regulate eNOS (19), an effect which may lead to attenuated insulin resistance via activation of Akt and AMPK (20). EETs can protect heart via antiinflammation (21).

Together, these observations suggest that elevated cardiac EETs may protect against diabetic cardiomyopathy. Therefore, in the present study, we used cardiac-specific CYP2J2 transgenic mice to investigate this hypothesis and confirm the role of CYP-derived EETs in diabetic cardiomyopathy and relevant mechanisms.

Materials and Methods

Generation of cardiac-specific CYP2J2 transgenic mice

Cardiac-specific CYP2J2 transgenic mice driven by αMHC promoter on a pure C57BL/6 genetic background were generated in Dr D. Zeldin's laboratory, and transgenic mice were identified as described previously (22). All studies used heterozygous CYP2J2 transgenic progeny and age/sex-matched wild-type littermate control mice. All experimental procedures were approved by the Experimental Animal Research Committee of Tongji Medical College, Huazhong University of Science and Technology, and in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Operations were under sodium pentobarbital (50 mg/kg, ip) anesthesia, and efforts were made to minimize suffering. Mice were anesthetized preformed with ip injections of a xylazine (5 mg/kg) and ketamine (80 mg/kg) mixture and placed in a supine position before killing. To assess the adequacy of anesthesia during echocardiographic and hemodynamic examinations, parameters such as responsiveness, blood pressure, respiratory rate, and heart rate were monitored. Then, mice were killed by CO₂ inhalation.

Induction of diabetes

Noninsulin-dependent diabetes mellitus was induced as previously reported (23). In brief, male αMHC-CYP2J2 transgenic and wild-type mice weaned at 4 weeks of age were randomly assigned to either a high-fat (HF) diet (35.5% wt/wt; Bioserv, Frenchtown, New Jersey) or regular chow. At 7 weeks of age, mice were injected ip with streptozotocin (STZ) (Sigma, St Louis, Missouri). Some αMHC-CYP2J2 transgenic and wild-type mice were fed normal chow and injected ip with vehicle as controls. Animal feed (HF diet catalog D12451, chow diet catalog 1025) was supplied by Beijing HFK Bioscience (Beijing, China). All groups were maintained on the same diet until age 19 weeks and then killed.

Clinical blood tests and metabolic assessments

Under nonfasting conditions, blood samples were collected at 0, 1, 2, 4, 8, and 12 weeks after STZ administration for determination of plasma glucose, insulin, total cholesterol, triglycerides, and high density lipoprotein (HDL) Diagnostic kit (Sigma) and Insulin ELISA kit (Linco Research, St Charles, Missouri). Body weight was measured immediately before blood collection. Food intake and drinking water per 24 hours were recorded in individual metabolic cages weekly.

Intraperitoneal injection glucose tolerance test

Animals were fasted overnight and then injected ip with D-glucose (20% solution; 2 g per kg body weight). Blood samples were taken at −30 (“pre”), 15, 30, 60, and 120 minutes after the glucose administration on week 4 and 12 after induction of diabetes. Plasma glucose and insulin levels were determined as described above.

Hyperinsulinemic-euglycemic clamp study

The clamp studies were performed exactly as previously published protocol (24, 25). Indwelling catheters were inserted into the superior vena cava 7 days before the clamp studies. Then, the basal rates of glucose turnover were measured by continuous infusion of [3-³H] glucose (high-pressure liquid chromatography purified; PerkinElmer, Boston, Massachusetts) at a rate of 0.05 μCi/min for 2 hours, after an overnight fast. Then, a 120-minute hyperinsulinemic-euglycemic clamp was conducted. At the end of basal and clamp periods, additional blood samples were obtained for the measurement of plasma insulin and free fatty acid concentrations.

Glucose metabolic studies

To determine plasma ³H-glucose level, plasma was deproteinized with ZnSO₄ and Ba(OH)₂, dried to remove ³H₂O, resuspended in water, and then counted in scintillation fluid (Ultima Gold; PerkinElmer) on a scintillation counter (Beckman, Fullerton, California). Rates of basal and insulin-stimulated whole-body glucose turnover were determined as the ratio of the [3-³H] glucose infusion rate (GIR) (disintegrations per minute) to the specific activity of plasma glucose (disintegrations per minute per milligram) at the end of the basal period and during the final 30 minutes of the clamp experiment, respectively. Hepatic glucose production (HGP) was determined by subtracting the GIR from the total glucose appearance rate. The plasma concentration of ³H₂O was determined by the difference between ³H counts without and with drying. Whole-body glycolysis, whole-body glycogen synthesis, and heart glucose uptake were calculated as described previously, respectively (26–28). HGP and glucose disposal rate (GDR) were calculated in the hyperinsulinemic-euglycemic clamp studies using Steele's equation. At steady state, the rate of glucose disappearance (ie, total GDR) is equal to the sum of the rate of endogenous glucose production or HGP plus the exogenous GIR. The insulin-stimulated GDR is equal to the total GDR minus the basal glucose turnover rate.

Cell culture

H9c2 cells obtained from the American Type Culture Collection (Manassas, Virginia) were routinely cultured in DMEM with low glucose (5.5mM), L-glutamine, and 10% fetal bovine serum at 37°C in an atmosphere of 5% CO₂. 14,15-EET (1μM) (Cayman Chemical Co, Ann Arbor, Michigan), 14,15-epoxyeicosa-5(Z)-enoic acid (EEZE) (1μM) (Cayman Chemical Co), high glucose (25mM) (Sigma), angiotensin II (AngII) (1 μM) (Sigma, Shanghai, China), atrial natriuretic peptide (ANP) receptor antagonist A71915 (500nM) (Sigma), and protein kinase G inhibitor KT5823 (2 μM) (Sigma) were added to H9c2 cells for 24 hours as indicated. Transfections of small interfering RNA (siRNA) (100nM) (RiboBio, Guangzhou, China) were with Lipofectamine 2000 reagent (Invitrogen, Life Technologies Corp, Carlsbad, California) for 48 hours, according to the manufacturer's instructions. Conditions used for hyperglycemia (HG)-induced insulin resistance were described previously (29). Glucose uptake in H9c2 cells was measured by the method of Dyntar et al with few modifications (30). At the end of the stimulation period, cells were rapidly washed and equilibrated for 30 minutes at 37°C in Krebs-Ringer HEPES buffer (pH 7.4) containing the following constituents: 123mM NaCl, 5mM KCl, 1.3mM CaCl₂, 100mM HEPES, 5mM D-glucose, and 1.5% bovine serum albumin free fatty acid. This was followed by incubations in media containing 5.5mM glucose and 2-deoxy-d-[1-¹⁴C] glucose (final specific activity 0.45 mCi/mmol). Glucose uptake was terminated by rapid washes with ice-cold PBS. One aliquot of the cell lysate was used for quantification of the accumulated radioactivity and another for protein quantification.

Evaluation of cardiac function and ventricular wall via echocardiography

After anesthetization, echocardiographic examination was performed before killing, using a high-resolution imaging system with a 30-MHz high frequency scanhead (VisualSonics Vevo770; VisualSonics, Inc, Toronto, Canada) as previously described (31).

Hemodynamic measurements of left ventricular (LV) function

Measurements of LV function were performed using Millar catheter system as described before (32).

Histological analysis

Hearts were excised from mice weighted and then fixed in 10% neutral buffered formalin. Sections were cut 5 mm thick and stained with hematoxylin and eosin. Under microscope, cell surface areas were measured to assess effect of CYP2J2 on diabetic cardiac hypertrophy.

RNA extraction and quantitative RT-PCR

Total RNA was isolated and reverse transcribed using TRIzol (Invitrogen, Carlsbad, California) and MultiScribe (ABI, Foster City, California). Real-time PCR analysis was performed on a 7900HT Fast Real-Time PCR System (ABI). The relative amounts of specific transcripts were calculated by using the comparative threshold cycle method. Glyceraldehyde-3-phosphate dehydrogenase was used as endogenous control.

Protein extraction and Western blotting

The protein concentration was determined using the Bradford method. Lysates were resolved by SDS-PAGE, transferred to nitrocellulose membranes, and blocked with 5% nonfat dry milk in Tris-buffered saline-Tween. The membranes were then incubated with the primary antibody overnight at 4°C, followed by peroxidase-conjugated secondary antibody for 2 hours. The enhanced chemiluminescence system was used to visualize the separated proteins. β-Actin was probed as loading control.

Statistical analysis

The results obtained are expressed as the mean ± SEM. The data were analyzed using 1- or 2-way ANOVA. In all cases, P < .05 was considered significant.

Results

Changes in physiological parameters after STZ administration

Nonfasting plasma glucose, total cholesterol, triglycerides, HDL, and insulin concentrations before and 1, 2, 4, 8, and 12 weeks after STZ administration in wild-type and αMHC-CYP2J2 transgenic mice fed normal chow or HF diet are shown in Supplemental Figure 1, published on The Endocrine Society's Journals Online web site at http://endo.endojournals.org. Plasma glucose was significantly higher in all HF mice compared with chow-fed mice. There was no significant change in the mean plasma glucose concentration at any time in the chow-fed group, whereas a significant increase in plasma glucose was present 4 weeks after STZ in both wild-type and αMHC-CYP2J2 transgenic mice fed HF. Compared with wild-type, CYP2J2 overexpression resulted in a small, but statistically significant, reduction in plasma glucose 8–12 weeks after STZ injection (Supplemental Figure 1A). In the chow-fed group, glucose levels were within a normal range throughout the observation period.

In HF-fed mice, plasma total cholesterol and triglyceride levels were markedly higher, but HDL was significantly lower than in mice fed normal chow (Supplemental Figure 1, B–D). Interestingly, we observed that the plasma triglyceride concentration in the αMHC-CYP2J2 transgenic HF-STZ group was significantly lower than in the wild-type HF-STZ group for the entire observation period (Supplemental Figure 1C).

The plasma insulin concentrations are shown in Supplemental Figure 1E. Plasma insulin concentrations were significantly elevated, before STZ injection, in HF-fed mice. There was a decline in insulin levels in HF-STZ-treated mice over time, which suggests that diabetes was progressively induced. However, insulin levels remained significantly higher than normal at 12 weeks. Interestingly, αMHC-CYP2J2 transgenic HF-STZ mice had significantly lower insulin concentrations compared with wild-type HF-STZ mice, and this reached statistical significance at 8–12 weeks after STZ. This finding is consistent with the lower glucose levels observed in αMHC-CYP2J2 transgenic mice.

Changes in systemic metabolic parameters

The body weight of wild-type and αMHC-CYP2J2 transgenic mice increased gradually during the 15-week observation period. Body weight did not differ between genotype groups in either chow-fed mice or HF-STZ-treated mice at any of the time points examined. However, HF-STZ-treated mice had significantly higher body weights than mice with the same genotype fed normal chow (Supplemental Figure 2A). Drinking water and food consumption per 24 hours in the HF mice was also significantly greater than in mice fed normal chow (Supplemental Figure 2, B and C). Drinking water and food consumption in the HF mice steadily increased for 4–19 weeks, but there were no significant differences between genotypes in any of the examined time points.

Effect of cardiac-specific CYP2J2 overexpression on glucose tolerance and insulin resistance

Figure 1A shows results of glucose tolerance tests in HF diet mice compared with mice fed normal chow 4 weeks after STZ administration. The fasting plasma glucose level in the HF mice at 0.5 hours before glucose loading was significantly higher than in the respective normal control groups. In both the HF and normal chow-fed mice, plasma glucose levels reached a peak at 15 minutes after glucose loading and then gradually decreased. After glucose loading, plasma glucose levels in the HF mice were significantly higher than in normal chow mice at all time points. No differences were observed between genotypes.

Figure 1.

Changes in glucose tolerance. Plasma glucose concentrations after ip glucose administration in wild-type and cardiac-specific CYP2J2 transgenic mice fed standard chow or HF-STZ for (A) 4 weeks or (B) 12 weeks. (C) Insulin tolerance test assays were performed at 12 weeks. (D) Plasma insulin concentrations during ip glucose administration in wild-type and cardiac-specific CYP2J2 transgenic mice fed standard chow or HF-STZ for 12 weeks (n = 10 per group; *, P < .05 vs HF-STZ-treated αMHC-CYP2J2 transgenic mice). (E–G) Hyperinsulinemic-euglycemic clamp study in wild-type and cardiac-specific CYP2J2 transgenic mice fed standard chow or HF-STZ for 12 weeks (insulin-stimulated GDR). (H) Heart glucose uptake test in wild-type and cardiac-specific CYP2J2 transgenic mice fed standard chow or HF-STZ for 12 weeks (n = 10 per group; *, P < .05 vs wild-type chow; #, P < .05 vs wild-type HF-STZ). (I) Glucose uptake determination in H9c2 cells with different treatments (results are represented as mean ± SEM; n = 5 per group; *, P < .05 vs control; #, P < .05 vs HG; &P < .05 vs HG+EET).

Glucose tolerance testing was also assessed 12 weeks after HF-STZ administration. Fasting plasma glucose levels in HF-STZ mice at 0.5 hours before glucose loading was significantly higher than in mice fed normal chow (Figure 1B). These differences persisted after glucose loading was initiated and lasted throughout the entire observation period. Although there were no significant differences in glucose tolerance between genotypes in mice fed a normal chow diet, αMHC-CYP2J2 transgenic mice had significantly lower plasma glucose levels after glucose loading than wild-type controls after 12 weeks of HF-STZ administration (Figure 1B). Insulin tolerance test results showed that insulin sensitivity was improved in HF-STZ-treated αMHC-CYP2J2 transgenic mice compared with wild-type mice (Figure 1C). Likewise, plasma insulin levels were higher in HF-STZ-treated mice before glucose loading, and these differences persisted after glucose was administered (Figure 1D). Although there were no significant differences between genotypes in mice fed a normal chow diet, αMHC-CYP2J2 transgenic mice had significantly lower plasma insulin levels after glucose loading than wild-type mice after 12 weeks of HF-STZ (Figure 1D). These results suggest that cardiac-specific overexpression of CYP2J2 protects against HF-STZ-induced insulin resistance moderately.

Next, we assessed the effects of CYP2J2 overexpression on insulin resistance by calculating [3-³H] glucose tracer-derived GIR, insulin-stimulated GDR, and HGP. Hyperinsulinemic-euglycemic clamp study showed that wild-type mice plus 12 weeks HF-STZ treatment had significantly reduced GIR than normal chow control and αMHC-CYP2J2 transgenic mice plus HF-STZ treatment (Figure 1, E–G). αMHC-CYP2J2 overexpression in heart improved diabetes-induced insulin resistance.

Furthermore, the heart glucose uptake test representing tissue-specific glucose uptake also showed that diabetic mice had significantly reduced heart-specific glucose uptake. However, αMHC-CYP2J2 transgenic mice had significantly improved heart glucose uptake (Figure 1H). Glucose uptake improvements in other insulin-sensitive tissues (such as skeletal muscle and adipose tissue) were also observed (Supplemental Figure 3).

To determine the effects of cardiac overexpression of CYP2J2 on gluconeogenesis, the expression of rate-limiting enzymes in gluconeogenesis was measured. Quantitative real-time PCR assays revealed that hepatic expression of glucose-6-phosphatase, fructose-1,6-bisphosphatase, and phosphoenolpyruvate carboxykinase was significantly higher in the mice subjected to HF-STZ administration than in normal chow-fed animals. CYP2J2 overexpression significantly reduced the expression of these enzymes (Supplemental Figure 4).

In cultured cardiomyocytes, similarly, addition of 14,15-EET enhanced glucose uptake in HG stimulation (Figure 1I). All these data suggest that myocardial-specific CYP2J2 overexpression markedly enhances heart glucose uptake and attenuates systemic insulin resistance.

Overexpression of CYP2J2 in the heart protects against impaired cardiac function after HF-STZ administration

Cardiac performance was determined 12 weeks after HF-STZ administration using Millar catheter system, and results were summarized in Table 1. Cardiac performance was impaired both in wild-type and αMHC-CYP2J2 transgenic mice treated with HF-STZ. End-diastolic pressure and LV isovolumic relaxation rate were significantly increased in HF-STZ-treated mice compared with normal chow-fed mice of both genotypes. Likewise, heart rate, maximum pressure, end-systolic pressure, cardiac output, maximal rate of rise of left ventricular pressure, and minimum rate of rise of left ventricular pressure were significantly reduced in HF-STZ-treated mice compared with normal chow-fed mice of both genotypes. However, the response to 12 weeks of HF-STZ in all of the aforementioned parameters was significantly less pronounced in αMHC-CYP2J2 transgenic mice than in wild-type controls. These data show that cardiac-specific overexpression of CYP2J2 ameliorates both systolic and diastolic dysfunction of diabetic cardiomyopathy. Furthermore, echocardiography tests showed that FS of LV was significantly lower, and left ventricular internal dimension and LV volume were higher in HF-STZ-treated mice. In contrast, in αMHC-CYP2J2 transgenic mice, these cardiac function parameters were normal (Figure 2, A–D).

Table 1.

Effects of Overexpression of CYP2J2 on Ventricular Function

	Chow wild type	Chow αMHC-CYP2J2	HF-STZ wild type	HF-STZ αMHC-CYP2J2
Heart rate (bpm)	312 ± 33	335 ± 37	220 ± 33a	230 ± 30b,c
Maximum pressure (mmHg)	123 ± 7	122 ± 7	93 ± 3a	102 ± 2b,c
End-systolic pressure (mm Hg)	121 ± 4	121 ± 5	89 ± 3a	100 ± 3b,c
End-diastolic pressure (mm Hg)	5.85 ± 0.79	5.82 ± 0.59	9.57 ± 0.64a	7.53 ± 0.79b,c
Cardiac output (μL/min)	1482 ± 77	1481 ± 83	1033 ± 61a	1107 ± 69b,c
dP/dtmax (mm Hg/sec)	8197 ± 759	8482 ± 954	6139 ± 514a	7209 ± 373b,c
dP/dtmin (mm Hg/sec)	5549 ± 786	5384 ± 546	3611 ± 289a	4413 ± 238b,c
Tau (msec)	15.0 ± 0.7	14.8 ± 0.7	24.4 ± 0.9a	18.9 ± 0.9b,c

The data were analyzed using 1-way ANOVA with post hoc testing performed by Student-Newman-Keuls method (n = 10 mice per group; bpm, beats per minute; Tau, LV isovolumic relaxation rate).

^aP < .05 vs wild-type chow.

^bP < .05 vs αMHC-CYP2J2 chow.

^cP < .05 vs wild-type HF-STZ.

Figure 2.

Cardiac function after HF-STZ administration. (A) Measurement of M-mode echocardiography in various groups. (B–D) Echocardiographic parameters (fractional shortening [FS], left ventricular internal dimension [LVID], and left ventricular volume [LV Vol]) in various groups (results are represented as mean ± SEM; n = 10 per group; *, P < .05 vs wild-type chow; #, P < .05 vs wild-type HF-STZ).

Overall, these data further suggest that heart-specific CYP2J2 overexpression ameliorates cardiac dysfunction induced by diabetes.

Effect of cardiac CYP2J2 overexpression on expression of substrate metabolism and energy utilization regulating genes

Glucose transporter type 4 (GLUT4) is an insulin-regulated glucose transporter found in cardiac muscle responsible for insulin-regulated glucose disposal (33). Hearts from HF-STZ-treated wild-type and αMHC-CYP2J2 transgenic mice had significantly reduced GLUT4 mRNA levels compared with hearts from mice that were fed a normal chow diet (Figure 3A). The reduction in GLUT4 mRNA levels after HF-STZ treatment was significantly attenuated in αMHC-CYP2J2 transgenic mice (Figure 3A). In cultured cardiomyocytes, HG stimulation reduced membrane GLUT4 level, but 14,15-EET significantly reversed GLUT4 translocation to membrane, which was blocked by EET antagonist 14,15-EEZE (Figure 3B). Furthermore, we used RNA interference technique to silence GLUT4 expression in cultured cardiomyocytes. Results showed that the effects of EET on cardiac hypertrophy marker, ANP, and glucose-uptake all disappeared, which suggest that enhancing effect of EET is involved in GLUT4 (Figure 3, C and D). These results are consistent with that CYP2J2 overexpression and EET enhance the glucose uptake of heart and cardiomyocytes and reversal of insulin resistance.

Figure 3.

Expression of metabolism-related genes. (A) Cardiac mRNA levels of GLUT4 from wild-type and αMHC-CYP2J2 transgenic mice fed normal chow or treated with HF-STZ (n = 10 per group; *, P < .05 vs wild-type HF-STZ; #, P < .05 vs same genomic background mice with chow diet). (B) GLUT4 translocation detection by Western blotting in cultured cells (n = 5 per group; *, P < .05 vs control; #, P < .05 vs HG; &, P < .05 vs HG+EET). (C) ANP expression in H9c2 cells treated with GLUT4 siRNA (SiGlut4, siRNA specific for GLUT4; n = 5 per group; *, P < .05 vs control; #, P < .05 vs SiGlut4+HG; &, P < .05 vs SiGlut4+HG+EET). (D) Glucose uptake determination in H9c2 cells with different treatments (SiGlut4, siRNA specific for GLUT4; SiNC, siRNA negative control; n = 5 per group; *, P < .05 vs control; #, P < .05 vs HG; &, P < .05 vs HG+EET). (E and F) Cardiac mRNA levels of PDK4 and PPARγ from wild-type and αMHC-CYP2J2 transgenic mice fed normal chow or treated with HF-STZ (n = 10 per group; *, P < .05 vs wild-type HF-STZ; #, P < .05 vs same genomic background mice with chow diet). Real-time PCR data were obtained from 3 cardiac-specific CYP2J2 transgenic mice and 3 wild-type littermate controls. Each transcript was analyzed in triplicate. Data were shown as mean ± SEM. CON, control; DMSO, dimethyl sulfoxide.

Pyruvate dehydrogenase kinase 4 (PDK4) is an important negative regulator of pyruvate dehydrogenase that controls glucose oxidation in well-oxygenated tissues such as the heart (34). Real-time PCR analysis showed that PDK4 mRNA levels were significantly elevated in hearts from mice treated with HF-STZ compared with hearts from mice fed a normal chow diet (Figure 3E). This response was significantly attenuated in hearts from αMHC-CYP2J2 transgenic mice compared with wild-type mice. Peroxisome proliferator-activated receptor (PPAR)γ is a well-known transcriptional regulator that is involved in cardiac energy metabolism (35). PPARγ mRNA levels were reduced in hearts from mice treated with HF-STZ compared with mice fed normal chow (Figure 3F). The change in cardiac PPARγ gene expression in response to HF-STZ was significantly attenuated in αMHC-CYP2J2 transgenic mice compared with wild-type controls.

Effects of cardiac CYP2J2 overexpression on downstream signaling pathways

We next examined the effect of HF-STZ treatment on activation of cardiac signaling pathways that have been shown to be important in insulin resistance. Hearts of HF-STZ-treated mice showed suppression of tyrosine 989 phosphorylation of the insulin receptor substrate-1 (IRS-1), reduced phosphatidylinositol 3-kinase (PI3K) (P110 subunit) expression, and reduced phosphorylation of Akt at serine 308 (Figure 4, A–C). Each of these responses to HF-STZ treatment was significantly attenuated in hearts of αMHC-CYP2J2 transgenic mice compared with wild-type controls. In agreement with these findings, phosphorylation of serine 307 of IRS-1 was significantly increased in HF-STZ-treated mice (Figure 4A). However, there was no significant difference in serine 307 phosphorylation between wild-type and αMHC-CYP2J2 transgenic HF-STZ-treated mice. Because serine phosphorylation of IRS-1 is primarily mediated by c-Jun N-terminal kinase (JNK), we next examined the phosphorylation status of this protein in HF-STZ-treated mice. JNK phosphorylation was significantly elevated in HF-STZ-treated mice (Figure 4D). Importantly, JNK phosphorylation in αMHC-CYP2J2 transgenic HF-STZ-treated mice was significantly lower than in HF-STZ-treated wild-type controls.

Figure 4.

Effect on cardiac metabolism signaling pathways. (A) Immunoblot analysis of IRS-1-Tyr(P)989, IRS-1-Ser(P)307, and total IRS-1 in hearts. (B) Immunoblot analysis of PI3K content in hearts. (C) Immunoblot analysis of Akt-Thr(P)308 and total Akt in hearts. (D) Immunoblot analysis of AMPK-Thr(P)172 and total AMPK in hearts. (E) c-Jun-Ser(P)63/73 in hearts. (F) Immunoblot analysis of eNOS-Ser(P)1177, eNOS-Thr(P)495, and total eNOS in hearts. The upper panels show representative blots, and the lower panels show the densitometric analysis. Blots are representative of 3 separate experiments. Data were shown as mean ± SEM (n = 9–10 per group, *, P < .05 vs wild-type normal chow; #, P < .05 vs wild-type HF-STZ treated).

AMPK has been shown to facilitate energy expenditure and increase insulin sensitivity after treatment with antidiabetic agents (36). The cardiac expression of the active form of AMPK (phosphorylation at Thr172) was significantly reduced in HF-STZ-treated mice compared with mice fed a normal chow diet (Figure 4E). Interestingly, this effect was significantly attenuated in αMHC-CYP2J2 transgenic mice compared with wild-type controls.

Previous reports have shown that mice on a HF diet up-regulate phosphorylation of eNOS at threonine 495, down-regulate phosphorylation of eNOS at serine 1177, and have no alteration in total eNOS protein content. Together, these responses inhibit eNOS activity (19, 37). HF-STZ treatment significantly reduced eNOS serine 1117 phosphorylation and increased eNOS threonine 495 phosphorylation (Figure 4F). Although cardiac overexpression of CYP2J2 had no effect on eNOS threonine 495 phosphorylation, it significantly attenuated the reduction of eNOS serine 1117 phosphorylation observed in wild-type mice (Figure 4F).

Inhibitory effects of CYP2J2 overexpression and EET on cardiac hypertrophy

Echocardiography showed that HF-STZ treatment induced significantly increased thickness in ventricular septum compared with normal control mice. However, overexpression of CYP2J2 markedly prevented the diabetes-induced increase in thickness of ventricular septum in αMHC-CYP2J2 transgenic mice (Figure 5A). We compared heart weights in different treatments and found that HF-STZ treatment resulted in marked increases in both heart weights and ratio of heart/body weight, but in αMHC-CYP2J2 transgenic HF-STZ-treated mice, these parameters were normal (Figure 5B). And histological analysis showed that surface area of cardiomyocytes was significantly increased in diabetic mice compared with control, but heart-specific CYP2J2 transgene prevented the effect (Figure 5C). All these data suggest that in vivo heart-specific CYP2J2 transgene prevents diabetes-induced cardiac hypertrophy.

Figure 5.

Effects on cardiac hypertrophy. (A) Interventricular septum measurement by echocardiography. (B) Morphology and heart weight of mice with different treatment. (C) Histological analysis of surface area of cardiomyocytes by HE staining. (D) Immunoblot analysis of CYP2J2 and ANP in hearts of different treated mice (n = 10 per group, *, P < .05 vs wild-type normal chow; #, P < .05 vs wild-type HF-STZ treated). (E) Measurement of surface area of H9c2 cells with various treatment. (F) ANP detection by Western blotting in cultured cells (n = 5 per group; *, P < .05 vs control; #, P < .05 vs HG; &, P < .05 vs HG+EET). Data were shown as mean ± SEM, representative of 3 separate experiments. IVS, interventricular septum; WT, wild type; CON, control; DMSO, dimethyl sulfoxide.

In order to investigate possible mechanisms through which CYP2J2-EET inhibits diabetes-induced cardiac hypertrophy, we probed ANP expression and found that heart-specific CYP2J2 overexpression dramatically up-regulated ANP expression in heart although HF-STZ treatment enhanced ANP production (Figure 5D). In cultured cardiomyocytes, similarly, HG incubation induced cellular hypertrophy based on cell surface areas, but addition of EET inhibited the HG-induced hypertrophy, and the inhibitory effect was blocked by EEZE (Figure 5E). Also, EET promoted ANP expression in HG-treated H9c2 cells, and this effect was attenuated by EEZE as expected (Figure 5F). These in vitro data are consistent with above in vivo results.

Effects of CYP2J2 overexpression and EET on ANP expression and nuclear factor of activated T cells c3 (NFATc3) nuclear translocation

It was documented that calcineurin/NFAT pathway activation induces cardiac hypertrophy via stimulating GATA-binding protein 4 (GATA4) after nuclear translocation (38). We explored effect of CYP2J2 expression or EET incubation on calcineurin/NFAT expression and nuclear translocation in cultured cardiomyocytes. Results showed that HG incubation significantly enhanced nuclear translocation of NFATc3, but addition of EET dramatically attenuated the nuclear translocation effect, which was completely blocked by EEZE (Figure 6A). We also observed same results in cultured cardiomyocytes by fluorescence microscopy detection (Figure 6B). Consistently, increased nuclear translocation of cardiac NFATc3 induced by HF-STZ treatment was reversed by CYP2J2 overexpression in the hearts of αMHC-CYP2J2 mice (Figure 6C). Then, we used AngII in cultured cells, which showed similar antihypertrophy effects of EETs (Figure 6D). Further, we also detected calcineurin Aβ (CnAβ), which enhances NFATc3 translocation into nuclei, and found that AngII markedly increased CnAβ expression as NFATc3 translocation, EET incubation significantly reduced NFATc3 translocation and CnAβ level, which were all partially blocked by both ANP receptor antagonist (A71915) and protein kinase G inhibitor (KT5823) (Figure 6E).

Figure 6.

Effect on cardiac hypertrophy signaling pathways. (A) NFATc3 translocation detection by Western blotting in cultured cells. (B) NFATc3 translocation detection by fluorescence microscopy in cultured cells (n = 5 per group; *, P < .05 vs control; #, P < .05 vs HG; &, P < .05 vs HG+EET). (C) NFATc3 translocation detection by Western blotting in hearts of treated mice (n = 5 per group, *, P < .05 vs wild-type normal chow; #, P < .05 vs wild-type HF-STZ treated). (D) Measurement of surface area of H9c2 cells with AngII treatment. (E) CnAβ and NFATc3 translocation detection by Western blotting in cultured cells with AngII and/or different inhibitor (n = 5 per group; *, P < .05 vs control; #, P < .05 vs AngII; &, P < .05 vs AngII+EET). (F) Intracellular calcium level examination by fluorescence microscopy in cultured cells. (G) Intracellular calcium levels by ELISA in cultured cells (n = 5 per group; *, P < .05 vs control; #, P < .05 vs HG; &, P < .05 vs HG+EET). Data were shown as mean ± SEM, representative of 3 separate experiments. CON, control; DAPI, 4′6-diamidino-2-phenylindole; DMSO, dimethyl sulfoxide.

We also measured intracellular calcium level ([Ca²⁺]_i), which is related with cardiomyoctye hypertrophy, and found that HG stimulation significantly elevated [Ca²⁺]_i, and endogenous EET attenuated this effect, which was blocked by EEZE (Figure 6, F and G).

Discussion

Previous studies have shown that several animal models of diabetes lead to cardiomyopathy (39, 40). In the present study, we induced diabetic cardiomyopathy using HF diet and low-dose STZ in male wild-type and αMHC-CYP2J2 transgenic mice. The phenotypic characterization of this animal model was accompanied by in vitro mechanistic experiments performed on H9c2 cells. The results demonstrate that cardiac-specific CYP2J2 overexpression protects against diabetic cardiomyopathy as indicated by reversal of the diabetes-induced decrease in contractile performance and attenuation of cardiac hypertrophy, which were due to improved systemic and cardiac insulin resistance and glucose uptake (Figure 7). These findings were supported by results of glucose tolerance tests and hyperinsulinemic-euglycemic clamp studies at both the systemic and cardiac levels, as well as by changes in the expression of genes regulating cardiac metabolism by CYP2J2 expression. Possible molecular mechanisms for these effects include increased PPARγ expression, activated insulin receptor signaling pathway, enhanced AMPK phosphorylation, and increased eNOS activity. In vitro EET treatment ameliorated glucose uptake under high-glucose conditions by preserving the localization of GLUT4 at the plasma membrane in cultured cardiomyocytes. EETs also enhanced ANP expression in cells cultured with high-glucose conditions and limited cell hypertrophy as observed in vivo. The mechanism responsible for the decrease in cell hypertrophy is likely to involve decreased induction of calcineurin and reduced nuclear translocation of the transcription factor NFATc3.

Figure 7.

Signaling pathways showing mechanisms involved in the protective effects of cardiac-specific CYP2J2 transgenic mice subjected to diabetic stress. In case of high glucose and AngII, GLUT4 expression on the cell membrane is deceased, reduces the ability of glucose transport, and finally increases intracellular calcium, promotes NFTA into the nucleus, enhances transcription, leading to cardiac hypertrophy. CYP2J2/EETs directly increase the transfer of GLUT4 to the cell membrane and may also via IRS-1-PI3K/AKT activation indirectly, thus reversing the process of cardiac hypertrophy. On the other hand, CYP2J2/EETs alleviate cardiac hypertrophy through activation of IRS-1-PI3K/AKT pathway, up-regulation of AMPK and eNOS, and down-regulation of c-Jun. Meanwhile, CYP2J2/EETs raise the expression amount of PPARγ, inhibit PDK4, reduce the toxicity of FFA, and relieve myocardial hypertrophy.

Most importantly, cardiac-specific overexpression of CYP2J2 not only improved cardiac function but also alleviated cardiac hypertrophy, through the inhibition of NFATc3 signaling by enhanced ANP production. LV hypertrophy might develop in patients with diabetes as a result of their insulin resistance, which independently stimulates LV growth (41). Previous data showed that dephosphorylated NFATc family members enter the nucleus and cooperatively bind to DNA with activator protein 1, musculoaponeurotic fibrosarcoma, GATA4, and other transcription factors to induce the expression of proteins that contribute to cardiac hypertrophy (42). Consistent with these data, our results suggest that CYP2J2 overexpression could relieve diabetes-induced cardiac hypertrophy through reduced nuclear translocation of NFATc3. In this study, we revealed, for the first time, that CYP2J2 comprehensively protects diabetic cardiomyopathy.

The key mechanisms that have been shown to contribute to diabetic cardiomyopathy are metabolic disturbances, myocardial fibrosis associated with increased AngII, induction of IGF-1 and inflammatory cytokines, microangiopathy and insulin resistance (43). Diabetes is associated with reduced levels of GLUT4 in the heart of spontaneously diabetic Zucker fa/fa rats (ZDF/drt) and viable yellow mice (Avy/a) (33). In our study, the loss of GLUT4 expression in response to HF-STZ treatment was significantly attenuated in the hearts of αMHC-CYP2J2 transgenic mice. Furthermore, the increase in PDK4 expression in response to HF-STZ treatment was significantly attenuated in αMHC-CYP2J2 transgenic mice. This is interesting, because diminished activity of the pyruvate dehydrogenase complex, mediated in part by increased expression of PDK4, is one of the mechanisms that has been implicated in diabetic cardiomyopathy (34).

CYP2J2 overexpression also had beneficial effects on PPARγ expression in diabetic cardiomyopathy. PPARγ is the master regulator of adipogenesis and the molecular target of thiazolidinedione. EETs were reported as PPARγ ligands (44), and we show here that CYP2J2 can significantly attenuate the loss of PPARγ expression. At a functional level, this may result in enhanced glucose uptake in the heart and could potentially represent a complimentary approach for treatment of insulin resistance in diabetic patients.

We also examined the insulin receptor-related signaling molecules in the heart. The cardiac levels of tyrosine phosphorylation of IRS-1, PI3K, and p-Akt were significantly reduced in the HF-STZ-treated group. Our results show that cardiac-specific overexpression of CYP2J2 enhances these molecules in diabetic mice. The αMHC-CYP2J2 transgenic mice also showed significant attenuation of the increased levels of phosphorylated c-Jun in the hearts of HF-STZ-treated mice. AMPK regulates cellular metabolism in response to the availability of energy and is therefore a target for type 2 diabetes treatment (45). Moreover, AMPK has a central role in controlling whole-body metabolism in response to diet and hormonal signals (46). AMPK is involved in the regulation of lipid metabolism (47), feeding and body weight (48), glucose homeostasis (49), and mitochondrial biogenesis (50). Moreover, a recent study suggests that AMPK activation is essential for the therapeutic effects of metformin in the liver (36). Interestingly, our results show that cardiac-specific CYP2J2 overexpression increases the level of the active form of AMPK (phosphorylation on Thr172) in the hearts of HF-STZ-treated mice, an effect which could have contributed to the attenuation of diabetic cardiomyopathy and insulin resistance.

Nitric oxide (NO) is an important signaling molecule that is involved in many physiological processes. Endothelium-derived NO modulates myocardial relaxation, diastolic tone, and oxygen consumption in preparations ranging from single myocytes and isolated hearts in vitro to intact hearts in vivo (51, 52). NO enhanced the uptake of glucose in skeletal muscle and attenuated the insulin resistance of peripheral organs (53). In this study, HF-STZ treatment enhanced phosphorylation of eNOS at Thr495 and inhibited phosphorylation of eNOS at Ser1177, an effect which lead to inhibited eNOS activity. Cardiac-specific overexpression of CYP2J2 prevented the loss of eNOS phosphorylation at Ser1177. Our previous studies demonstrated that CYP epoxygenase-derived EETs significantly increased eNOS expression and enhanced its activity (19, 37). Furthermore, overexpression of eNOS significantly attenuated insulin resistance (20). Therefore, up-regulation of eNOS by EETs may play an important role in CYP epoxygenase-mediated improvement of cardiac function in diabetic cardiomyopathy.

However, there are some limitations in our study. CYP2J2 epoxygenase is the only member of the human CYP2J subfamily known for its role in metabolizing arachidonic acid to physiologically important epoxides. Classically, the epoxygenase pathway generates EETs, which are then metabolized mainly by soluble epoxide hydrolase to the dihydroxyeicosatrienoic acid, which have traditionally been considered to be less active than EETs. A novel CYP epoxygenase in human atherosclerotic plaques reported recently, CYP2S1, can metabolize prostaglandin G2 and prostaglandin H2 to 12(S)-hydroxyheptadeca-5Z,8E,10E-trienoic acid, which is a novel regulator of macrophage function that is expressed in classical inflammatory macrophages (54). It should be noted that there may be new fatty acid substrates, such as linoleic acid, for CYP2J2 or other CYP epoxygenase, and the resulting metabolites may be involved in the cardiac protective effects of CYP epoxygenase overexpression. To verify this hypothesis, liquid chromatography-tandem mass spectrometry or lipid metabolomics should be employed.

In summary, we have shown that cardiac-specific CYP2J2 overexpression markedly improved cardiac function by attenuating insulin resistance and reducing plasma glucose levels in an HF-STZ-induced model of diabetic cardiomyopathy. These data suggest that CYP2J2 epoxygenase metabolites likely play an important role in plasma glucose homeostasis, and enhancement of EET activation may serve as an effective therapeutic strategy to prevent diabetic cardiomyopathy.