Increased CYP2J3 Expression Reduces Insulin Resistance in Fructose-Treated Rats and db/db Mice

Materials were obtained from the following suppliers: Antibodies involved in this study were from Santa Cruz Biotechnology (Santa Cruz, CA); the rabbit polyclonal anti-CYP2J3 antibody was developed in our laboratory as described (4); fructose, glucose, triglyceride, and cholesterol reagents were from Ningbo Cicheng Biocompany (Ningbo, China); Rat/Mouse Insulin ELISA Kit was from Linco Research (St. Charles, MO); 14,15-DHET ELISA Kit was from Detroit R&D (Detroit, MI); and guanosine 3′,5′-cyclic monophosphate (cGMP) and cAMP ELISA kit was from Cayman Chemical (Ann Arbor, MI). All other chemicals and reagents were purchased from Sigma-Aldrich unless otherwise specified. Full-length CYP2J3 cDNA was cloned from rat liver RNA and then subcloned into the pcDNA plasmid vector in sense (CYP2J3⁺) and antisense (CYP2J3⁻) orientations.

All animal experimental protocols complied with standards stated in the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and were approved by The Academy of Sciences of China. Male Sprague-Dawley rats weighing 200 ± 20 g and male C57BL/6 and db/db mice (8 weeks old) were obtained from the Experimental Animal Center of Shanghai (Shanghai, People's Republic of China). Animals were treated before experiments as described previously (19).

After a 1-week adaptation period (i.e., beginning at week 0), rats were fed normal rat chow and either normal water (n = 32) or water containing 10% fructose (n = 24) for a total of 5 weeks. Systolic blood pressure was measured weekly until week 6, and gene delivery protocols were undertaken at week 3 as described previously (19).

For the mouse study, C57BL/6 and db/db mice were anesthetized with diethyl ether and received a sublingual vein injection of plasmid at a dose of 5 mg/kg body wt. C57BL/6 mice received either empty pcDNA (C57BL/6 + pcDNA), CYP2J3⁺ (C57BL/6 + CYP2J3⁺), or 0.9% NaCl (C57BL/6 normal) (n = 8 per group), and db/db mice similarly received either pcDNA (db/db + pcDNA), CYP2J3⁺ (db/db + CYP2J3⁺), or 0.9% NaCl (db/db normal) (n = 8 per group).

Systolic blood pressure was measured weekly in conscious rats with a manometer-tachometer (Rat Tail NIBP System; ADI Instruments, Bella Vista, NSW, Australia) using the tail-cuff method as described previously (19).

Serum and urine samples from all rats were collected. Fasting serum levels of insulin, glucose, sodium, potassium, magnesium, cholesterol, triglycerides, LDL cholesterol, and HDL cholesterol as well as urine levels of sodium, potassium, and magnesium were assessed as described previously (19). Similarly, serum and urine samples from all mice were collected, and serum levels of insulin, glucose, cholesterol, triglycerides, LDL cholesterol, and HDL cholesterol were assessed. Serum levels of alanine aminotransferase, urea, and creatinine in rats were also measured on an AEROSET Clinical Chemistry System (Abbott Laboratories) to evaluate renal and liver function. Additional details are provided in the supplementary methods, available in an online appendix at http://diabetes.diabetesjournals.org/cgi/content/full/db09-1241/DC1.

At 2 weeks after gene delivery, C57BL/6 and db/db mice were fasted overnight (for 16 h) and then injected intraperitoneally with d-glucose (20% solution; 2 g/kg body wt). At 0, 30, 60, and 120 min after glucose administration, blood samples were taken from the cavernous sinus with a capillary while under ether anesthesia. Plasma glucose and insulin levels were determined using methods identical to those used for serum samples in rats (described in detail in the online supplementary methods).

To assess in vivo EET production, an ELISA kit (Detroit R&D) was used to determine concentrations of the stable EET metabolite 14,15-dihydroxyeicosatrienoic acid (14,15-DHET) in the urine of rats and mice. 14,15-DHET was quantified by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's instructions, as previously described (7).

Total RNA was extracted from frozen rat aortas using TRIZOL reagent (Invitrogen, Carlsbad, CA) and used to assess mRNA levels of ET-1 and ETA-R. The following oligonucleotide primers were used for amplification of the ET-1 and ETA-R cDNAs from reverse-transcribed aortic RNA: ET-1 (forward): 5′-AAGCGTTGCTCCTGCTCCTCC-3′; ET-1 (reverse): 5′-TTCCCTTGGTCTGGTCTTTGTG-3′; ETA-R (forward): 5′-TGCTCAACGCCACGACCAAGT-3′; ETA-R (reverse): 5′-GGTGTTCGCTGAGGGCAATCC-3′.

Amplification was performed on an ABI7500 PCR system (Applied Biosystems, Darmstadt, Germany) after incubation with Moloney murine leukemia virus reverse transcriptase at 42°C for 15 min. A preheating step of 5 min at 95°C was followed by 40 cycles consisting of 30 s at 95°C, 20 s at 60°C, and 20 s at 72°C. PCR products were electrophoresed on 1.5% agarose gels. The quantities of specific ET-1 and ETA-R transcripts were normalized to levels of glyceraldehyde-3-phosphate dehydrogenase transcripts to control for RNA quality and amount.

At 2 weeks after gene injection, rats and mice from each group were anesthetized with pentobarbital (100 mg/kg i.p.), and skeletal muscles, aortas, hearts, kidneys, and livers were excised, frozen in liquid nitrogen, and stored at −80°C. Western blotting was performed as described previously (19). Expression was quantified by densitometry and normalized to β-actin expression. All groups were then normalized to their respective controls, and bar graphs represent quantification of at least three independent experiments.

Urinary cAMP and cGMP levels were measured by ELISA as previously described (20,21).

Continuous data were expressed as means ± SEM. Comparisons between groups were performed by a one-way ANOVA. Two-way ANOVA was used to examine differences in response to treatments and between groups, with post hoc analyses performed using the Student-Newman-Keuls test. Statistical significance was defined as P < 0.05.

All rats in the study were assessed for a variety of physiological parameters 3 weeks after receiving either control or fructose-containing drinking water. As expected, consumption of fructose-containing water resulted in significantly increased systolic blood pressure, significantly increased levels of serum insulin, serum triglyceride, urine potassium, urine magnesium, and urine volume, and significantly decreased urine osmolarity (all P < 0.05) (Table 1, Fig. 1, and supplementary Table 1). Homeostasis model assessment insulin resistance (HOMA-IR) also was significantly increased in fructose-treated rats (P < 0.05) (Table 1). These data indicate that fructose administration induced hypertension, insulin resistance, and hypo-osmolar diuresis as described previously (19).

At 2 weeks after gene delivery (i.e., at week 5 of the study), CYP2J3 protein levels were increased in the aorta, heart, liver, and kidney of fructose-treated CYP2J3⁺ (F + CYP2J3⁺) rats compared with fructose-treated rats injected with CYP2J3⁻ or with the empty pcDNA vector (Fig. 1,A). Importantly, CYP2J3 functionality was demonstrated by the nearly fourfold increase in urinary 14,15-DHET levels in rats injected with CYP2J3⁺ compared with those injected with CYP2J3⁻ or with the empty pcDNA vector (Fig. 1 B).

Injection of CYP2J3⁺ to fructose-treated rats resulted in decreased systolic blood pressure 1, 2, and 3 weeks after injection (i.e., at weeks 4, 5, and 6 of the study, respectively) compared with that observed in fructose-treated rats injected with the control pcDNA or CYP2J3⁻ vectors (Fig. 1,C). The maximum reduction in blood pressure in F + CYP2J3⁺ rats was observed 2 weeks after injection (i.e., at week 5 of the study) when blood pressure reached a level similar to those observed in rats drinking normal water (Fig. 1,C). No changes in blood pressure compared with saline-injected rats were observed in normal drinking water–treated rats administered pcDNA, CYP2J3⁺, or CYP2J3⁻ (Fig. 1 C). These data indicate that CYP2J3 overexpression reduced hypertension in fructose-treated rats but had no effect on blood pressure in normal control rats.

Other physiological and biochemical parameters related to hypertension and hyperinsulinemia were assessed in eight rats per experimental group 2 weeks after gene delivery (week 5) (Table 2 and supplementary Table 2). Compared with values in the normal water–treated groups, serum insulin, insulin resistance (HOMA-IR), serum triglycerides, urine volume, urine potassium, and urine magnesium levels were all higher, whereas urine osmolarity was lower in fructose-treated rats injected with the empty pcDNA vector (F + pcDNA group; all P < 0.05) (Table 2 and supplementary Table 2). With the exception of serum triglyceride levels, all of these changes were prevented by injection of fructose-treated rats with CYP2J3⁺, but not with CYP2J3⁻ (Table 2 and supplementary Table 2). These data indicate that CYP2J3 overexpression markedly attenuated fructose-induced insulin resistance in rats but that it had no effect on these parameters in normal water–treated rats.

Compared with values in the normal water–treated groups, serum alanine aminotransferase, urea nitrogen, and creatinine showed no significant changes in fructose-treated rats injected with the empty pcDNA vector (F + pcDNA group; all P > 0.05). Compared with values in rats injected with empty pcDNA vector, serum alanine aminotransferase, urea nitrogen, and creatinine were not significantly changed in rats injected with CYP2J3⁺ (supplementary Table 3). These data indicate that CYP2J3 overexpression has no detrimental effects on rat renal and liver function.

PI 3-kinase is recruited to insulin receptor substrate (IRS) signaling complexes through binding of Src homology 2 domains in its 85-kDa regulatory subunit to specific phosphotyrosine residues in IRS-1, a process that leads to activation of the PI 3-kinase p110 catalytic subunit (22). To investigate the signaling mechanisms through which CYP2J3 attenuates fructose-induced insulin resistance, we evaluated the expression of IRS-1 associated with the insulin signaling cascade in rat liver and skeletal muscle. Compared with levels in normal control rats, fructose drinking resulted in significantly decreased IRS-1 levels in liver and skeletal muscle (Fig. 2,A and B). Phospho-Y989-IRS-1 levels were similarly decreased in liver and skeletal muscle, whereas phospho-S307-IRS-1 was significantly increased (supplementary Fig. 1A and B). Administration of CYP2J3⁺ significantly reversed the changes in IRS-1 and phospho–IRS-1 levels induced by fructose in both liver and skeletal muscle (Fig. 2 A and B; supplementary Fig. 1A and B).

Compared with expression in corresponding control animals, eNOS protein expression was downregulated in aorta, liver, and skeletal muscle in fructose-treated rats; this effect was not observed in fructose-treated rats injected with CYP2J3⁺ (Fig. 2 C; supplementary Fig. 1C and D). Similar changes in eNOS expression and effects of CYP2J3⁺ were observed in rat heart and kidney (data not shown). These data suggest that CYP2J3⁺ treatment restored eNOS activity and support previous observations of a beneficial effect of eNOS against fructose-induced hypertension and hyperinsulinemia (23).

The expression of ET-1 and ETA-R mRNA transcripts in rat aortas was determined by RT-PCR 2 weeks after gene delivery to examine the effects of fructose feeding and CYP2J3 gene delivery on the endothelin pathway, which has been shown to play a role in blood pressure homeostasis. Fructose administration resulted in significant increases in aortic ET-1 and ETA-R mRNA levels (Fig. 2 D and E). These changes were prevented in rats administered CYP2J3⁺, but not in rats administered CYP2J3⁻.

Similar to observations in fructose-fed rats, levels of CYP2J3 protein were increased in the aorta, heart, liver, and kidney in diabetic db/db mice 2 weeks after injection with CYP2J3⁺ compared with levels observed after injection with the empty pcDNA vector (Fig. 3,A). Functionality of the CYP2J3 protein in db/db mice was demonstrated by the nearly fivefold increase in urinary 14,15-DHET levels in both C57BL/6 and db/db mice injected with CYP2J3⁺ compared with those injected with the empty pcDNA vector (Fig. 3 B).

The diabetic phenotype of db/db mice (injected with the empty pcDNA vector) was confirmed by significantly higher levels of serum glucose, serum insulin, insulin resistance (HOMA-IR), serum triglycerides, serum cholesterol, and urine volume in these animals compared with those in C57BL/6 mice (all P < 0.05; Table 3 and supplementary Table 4). Injection of CYP2J3⁺ significantly reduced insulin resistance (HOMA-IR) and urine volume in db/db mice, whereas injection of the empty pcDNA vector did not (Table 3 and supplementary Table 4). Although not all parameters of the diabetic phenotype in db/db mice were reversed by CYP2J3⁺ treatment, these data provide further evidence that CYP2J3 can attenuate insulin resistance in this animal model.

A glucose tolerance test was performed in C57BL/6 and db/db mice 2 weeks after gene delivery. Fasting plasma glucose levels before glucose loading and plasma glucose levels after glucose loading were not altered by the various gene therapy treatments in C57BL/6 mice (Fig. 3,C). Compared with levels in C57BL/6 mice, fasting plasma glucose levels before glucose loading and plasma glucose levels after glucose loading were significantly higher in all db/db mouse groups. However, the glucose levels before and after glucose loading in db/db + CYP2J3⁺ mice were significantly lower than those in the db/db control and db/db + pcDNA groups (P < 0.05) (Fig. 3,C). A similar profile for fasting plasma insulin levels and for plasma insulin levels after glucose loading was observed, with db/db + CYP2J3⁺ mice having levels lower than those of the other db/db groups but higher than those of all of the C57BL/6 groups at all time points assessed (Fig. 3 D). These data indicate that CYP2J3 gene delivery attenuated insulin resistance and has antidiabetic effects in db/db mice.

Similar to the effects of fructose drinking on rats, S307 IRS-1 phosphorylation was significantly increased and Y989-IRS-1 phosphorylation was significantly decreased in livers of db/db control and db/db + pcDNA mice relative to C57BL/6 mice (Fig. 4,A). Similar effects were observed in skeletal muscle (supplementary Fig. 2A). Administration of CYP2J3⁺ to db/db mice prevented these changes in both tissues (Fig. 4 A; supplementary Fig. 2A).

The effects of CYP2J3⁺ treatment on the phosphorylation status of eNOS were examined in liver and skeletal muscle. The level of phospho-T495-eNOS was higher but phospho-S1177-eNOS was lower in livers of db/db mice compared with C57BL/6 mice (Fig. 4,B). CYP2J3 gene delivery prevented the downregulation of phospho-S1177-eNOS but had no impact on phospho-T495-eNOS (Fig. 4 B). Similar results were found in skeletal muscle (supplementary Fig. 3B), heart, and kidney (data not shown). These results suggest that CYP2J3⁺ treatment in db/db mice altered eNOS phosphorylation status.

To investigate potential mechanisms underlying the observed effects of CYP2J3⁺ treatment in rats, the protein expression levels of a variety of intracellular signaling pathway molecules were investigated in liver, skeletal muscle, heart, and kidney of rats and mice in all treatment groups 2 weeks after gene delivery. The signaling molecules that were assessed included PI 3-kinase, phosphorylated AKT (P-T308-AKT), phosphorylated AMP-activated protein kinases (P-T172-AMPKs), and phosphorylated p42/44 MAPK (P-MAPK). A similar pattern of expression was observed for all four molecules in all four tissues that were examined. Specifically, tissue protein levels of the specified signaling molecules were significantly decreased in fructose-treated rats and db/db mice compared with levels in normal water–fed rats and C57BL/6 mice, respectively. The only fructose-treated rats or db/db mice in which these decreases were not observed were those that had been injected with CYP2J3⁺. PI 3-kinase (P110) expression level data for rat and mouse liver are shown in Fig. 5 A and B, whereas data for PI 3-kinase in rat and mouse liver and skeletal muscle and for P-AKT, AMPK, and P-MAPK are shown in supplementary Figs. 4–6. CYP2J3⁺ injection reversed changes in protein expression and phosphorylation seen in db/db mice or induced by fructose in rats. The notable exception to this pattern was that of P-MAPK in rat tissues, in which fructose feeding did not result in a decreased level of P-MAPK; however, CYP2J3⁺ injection nonetheless increased expression of P-MAPK (supplementary Fig. 6A and B). Similar results were found for all of these signaling molecules in heart and kidney of rats and mice (data not shown). These data suggest that reversal of insulin resistance by CYP2J3⁺ treatment was associated with restoration of these important intracellular signaling molecules to normal levels.

To investigate potential mechanisms underlying the observed effects of CYP2J3 on the upregulation of AMPK phosphorylation, cAMP and cGMP levels were determined in urine of rats in all treatment groups 2 weeks after gene delivery. Compared with rats treated with normal water, the concentration of urinary cAMP and cGMP decreased significantly in rats with fructose water treatment. Interestingly, CYP2J3 gene delivery significantly increased urinary cAMP and cGMP level in rats treated with both normal water and fructose water (Fig. 6 A and B). These data indicate that CYP2J3 overexpression induced a significant increase in urine cAMP and cGMP secretion, which may result from G-protein–coupled receptor (GPCR) activation by EETs.

This study was undertaken to examine the effects of CYP2J3 gene delivery on insulin resistance and diabetes in fructose-induced insulin-resistant rats and db/db diabetic mice. Results showed that a single intravenous injection of CYP2J3 in the eukaryotic expression plasmid pcDNA reduced blood pressure in fructose-fed rats and improved sensitivity to insulin in peripheral tissues and organs in both animal models. CYP2J3 overexpression significantly reduced blood pressure with upregulated eNOS expression and downregulated ET-1 and ETA expression. Furthermore, CYP2J3 overexpression significantly improved insulin resistance in fructose-induced insulin-resistant rats and db/db diabetic mice, at least in part through eNOS, IRS-1, and PI 3-kinase/AKT signaling pathways, as well as AMPK signaling pathways in liver, muscle, heart, and kidney. These data provide direct evidence that CYP2J3-derived EETs may alleviate insulin resistance through a variety of beneficial effects on critical intracellular signaling pathways.

Previous studies have demonstrated that rats treated with high-fructose drinking water develop systemic hypertension, hyperinsulinemia, and hypertriglyceridemia (24). Although the pathophysiological mechanisms responsible for elevated blood pressure and hyperinsulinemia in fructose-treated animals are not completely understood, elevated sympathetic nervous system activity, impaired endothelium-dependent dilation, reduction of capillary permeability, elevated vascular expression of ET-1 and ETA receptor genes, decreased eNOS activity, and increased salt absorption by the intestine and kidney have all been implicated (12,23,25). Polymorphisms in CYP2J2 (the human homolog of CYP2J3) have been associated with essential hypertension (26), and reduced renal CYP-derived eicosanoid synthesis has been reported in rats with high-fat diet–induced hypertension (27). Furthermore, EETs have direct vasodilatory activity (28,29). In the present study, CYP2J3 overexpression significantly elevated urinary levels of 14,15-DHET in rats and mice and attenuated fructose-induced changes in eNOS, ET-1, and ETA-R expression in rats. These effects may underlie the beneficial effects of CYP2J3 overexpression on blood pressure and hyperinsulinemia that we observed.

Fructose induces inflammatory changes in human aortic endothelial cells and vessel walls in rats (30,31). Cytokines such as TNF-α induce insulin resistance in endothelial cells via a p38 mitogen–activated protein kinase–dependent pathway (32). This inflammatory signaling is relevant to diabetes, as TNF receptor 1 blockade protects Wistar rats from diet-induced obesity and insulin resistance (33). Physiological concentrations of EETs or overexpression of CYP2J2 decreases cytokine-induced endothelial cell adhesion molecule expression, indicating that EETs have anti-inflammatory properties independent of their membrane-hyperpolarizing effects (34). Furthermore, increased NO release improves insulin resistance in fructose-treated rats (23,35). Potassium depletion in rats exacerbates endothelial dysfunction and lowers the bioavailability of NO, which blocks insulin activity and causes insulin resistance (35). We observed that CYP2J3 expression resulted in a significant reduction in urine volume and urine potassium in fructose-treated rats. This suggests a potential ameliorative effect of CYP2J3 treatment against hypertension-related end organ (kidney) damage and attenuation of insulin resistance.

The precise molecular mechanisms for attenuation of insulin resistance in CYP2J3-injected, fructose-treated rats remain to be elucidated. Recent studies indicate that the ability of insulin to vasodilate skeletal muscle vasculature is mediated by endothelium-derived NO (36). These actions may partially explain our observation of improved insulin sensitivity after CYP2J3 gene delivery. In addition, we observed effects on insulin signaling in tissues directly involved in insulin sensitivity (37,,–40), including liver, muscle, heart, and kidney, as well as in an islet cell line. Our data show that insulin-dependent signaling was significantly inhibited in fructose-treated rats and db/db mice, but dramatically reversed by CYP2J3 overexpression. These results indicate that CYP2J3 overexpression potentiates insulin receptor signaling in liver, muscle, heart, and kidney and thus improves insulin sensitivity.

We also evaluated the phosphorylation status of AMPK. Small molecule–mediated activation of AMPK improves insulin resistance in ob/ob mice (41) and represents a promising approach for the treatment of type 2 diabetes and metabolic syndrome (42). We found that CYP2J3 overexpression resulted in a significant increase in AMPK phosphorylation, which may contribute to the EET-mediated alleviation of diabetes and insulin resistance we observed. CYP2J3 overexpression induced a significant increase in urine cAMP and cGMP secretion, which may result from GPCR activation by EETs. Thus, EET-induced GPCR activation may play an important role in EET-mediated AMPK activation and insulin sensitization. Taken together, these data suggest that the improvement of insulin resistance after CYP2J3 gene delivery was due to, at least in part, increased activation of IR/IRS-1/PI 3-kinase/AKT and AMPK signaling pathways (43), as well as the upregulation of eNOS.

In conclusion, we have demonstrated alleviation of insulin resistance and diabetes by CYP2J3 gene therapy in db/db diabetic mice and a model of fructose-induced insulin resistance in rats. These effects were associated with increased insulin sensitivity in peripheral tissues and organs via upregulation of systemic eNOS and activation of the IRS-1/PI 3-kinase/AKT and AMPK signaling pathways in muscle, liver, heart, kidney, and aorta. However, the more precise mechanisms need to be further investigated. The ability of CYP epoxygenase delivery to exert a broad spectrum of beneficial effects in these animal models warrants further investigation of this approach in the treatment of hypertension associated with insulin resistance and diabetes in humans.

This work was supported in part by funds from the National Education Ministration project, Nature Science Foundation Committee projects (Nos. 30930039 and 30700377), Wuhan project (200870834407), 973 program (2007CB512004), and the Intramural Research Program of the National Institutes of Health, National Institute of Environmental Health Sciences.

No potential conflicts of interest relevant to this article were reported.

We thank Dr. Jeffrey Card for assistance with the editing of this article. We also thank Dr. Stavros Garantziotis and Michael Fessler for helpful comments during the preparation of this article.