Angiotensin Receptor Blockade Recovers Hepatic UCP2 Expression and Aconitase and SDH Activities and Ameliorates Hepatic Oxidative Damage in Insulin Resistant Rats

Abstract

Metabolic syndrome (MetS) is commonly associated with elevated renin-angiotensin system, oxidative stress, and steatohepatitis with down-regulation of uncoupling proteins (UCPs). However, the mechanisms linking renin-angiotensin system, steatosis, and UCP2 to hepatic oxidative damage during insulin resistance are not described. To test the hypothesis that angiotensin receptor activation contributes to decreased hepatic UCP2 expression and aconitase activity and to increased oxidative damage after increased glucose intake in a model of MetS, lean and obese Long Evans rats (n = 10/group) were randomly assigned to the following groups: 1) untreated Long Evans Tokushima Otsuka (lean, strain control), 2) untreated Otsuka Long Evans Tokushima Fatty (OLETF) (MetS model), 3) OLETF + angiotensin receptor blocker (ARB) (10 mg olmesartan/kg·d × 6 wk), 4) OLETF + high glucose (HG) (5% in drinking water × 6 wk), and 5) OLETF + ARB + HG (ARB/HG × 6 wk). HG increased body mass (37%), plasma triglycerides (TGs) (35%), plasma glycerol (87%), plasma free fatty acids (28%), and hepatic nitrotyrosine (74%). ARB treatment in HG decreased body mass (12%), plasma TG (15%), plasma glycerol (23%), plasma free fatty acids (14%), and hepatic TG content (42%), suggesting that angiotensin receptor type 1 (AT1) activation and increased adiposity contribute to the development of obesity-related dyslipidemia. ARB in HG also decreased hepatic nitrotyrosine and increased hepatic UCP2 expression (59%) and aconitase activity (40%), as well as antioxidant enzyme activities (50-120%), suggesting that AT1 activation also contributes to protein oxidation, impaired lipid metabolism, and antioxidant metabolism in the liver. Thus, in addition to promoting obesity-related hypertension, AT1 activation may also impair lipid metabolism and antioxidant capacity, resulting in steatosis via decreased UCP2 and tricarboxylic acid cycle activity.

Obesity prevalence has increased to epidemic proportions in recent years owing primarily to hypercaloric diets and sedentary lifestyles (1). The onset of obesity increases the risk for the development of metabolic syndrome (MetS), an assemblage of conditions that includes elevated arterial pressure, fasting blood glucose, plasma triglyceride (TG) levels, and waist circumference, along with reduced high-density lipoprotein cholesterol (2). The development of nonalcoholic steatohepatitis (NASH) (steatosis) increases as a function of obesity and is associated with insulin resistance and dyslipidemia, suggesting that steatosis likely plays an instrumental role in the development of MetS (3–6).

Oxidative stress and inflammation are increased in individuals with central obesity, NASH, MetS, and type 2 diabetes (7–9). Hepatic mitochondrial oxidant production is one of the primary mechanisms that promote oxidative stress during these conditions (10–12). Uncoupling protein (UCP)2 is a mitochondrial inner membrane protein that is present in a variety of tissues, including the liver (hepatocytes and Kupffer cells) (13–18). Although the physiological role of UCP2 is not fully understood (19, 20), its main function appears to be the control of mitochondria-derived oxidant production (21, 22). Skeletal muscle mitochondria of UCP3 knockout mice (23) and inhibition of skeletal muscle UCP3 (24) are associated with decreased aconitase activity and increased oxidative damage (25), suggesting that UCP3 attenuates oxidative stress by increasing aconitase activity. UCP2 also attenuates oxidative stress in adipose (26) and macrophages (21), but how UCP2 changed in relation to aconitase activity in these studies was not ascertained. Furthermore, in steatotic/NASH livers, levels of reactive oxygen species and oxidative damage are chronically elevated and associated with higher expression of UCP2 (15–18, 27), whereas alloxan-induced diabetes is associated with reduced hepatic mitochondrial aconitase activity (28). However, the impact of diabetes or oxidative stress on mitochondrial aconitase appears to be incongruent. In hearts from streptozotocin-diabetic rats, mitochondrial aconitase activity in the forward direction (i.e. isocitrate production) was unchanged (29), whereas in skeletal muscle subjected to exercise, mitochondrial aconitase activity was actually unchanged or increased rather than decreased despite increased oxidative stress (30).

Because insulin resistance and MetS are inherently related to dysregulation of glucose metabolism and, ergo, of the tricarboxylic acid (TCA) cycle, impaired metabolism and dysregulation of TCA intermediates (i.e. isocitrate, succinate, etc.) may contribute to the manifestation of NASH commonly associated with MetS. Succinate can increase the expression of genes that lead to the progression of tumors (31), suggesting that elevated succinate may be detrimental. Conversely, vitamin E succinate supplementation decreased hepatic UCP2 expression and oxidative stress and improved mitochondrial function in mice with NASH, suggesting that succinate may play a protective role in the liver (32). However, the changes in UCP2, TCA cycle enzymes [aconitase, succinate dehydrogenase (SDH)], and oxidative damage in steatotic livers during MetS conditions have not been examined.

The stimulation of oxidative stress in a variety of tissues by increased activation of the renin-angiotensin system (RAS) and angiotensin II (Ang II) infusion is well established. However, the effect of increased Ang II and activation of its receptors, angiotensin receptor type 1 (AT1) and AT2, on UCPs is not well defined. Infusion of Ang II induced the accumulation of intracardiac lipids, associated with down-regulation of UCP2 and UCP3 and up-regulation of reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (Nox)1 and Nox4 gene expression in the heart of nondiabetic rats (33), suggesting that, at least in a nondiabetic heart, Ang II-mediated increases in Nox genes suppressed UCPs. Conversely, in the pancreas of young diabetic mice, AT1 activation generated progressive islet β-cell failure through UCP2-mediated oxidative damage (34). Blockade of the Ang II receptor (AT1) improves pancreatic β-cell function and oxidative stress in obese, hypertensive patients, and fasting blood glucose (35, 36) and pancreatic insulin secretion and peripheral insulin sensitivity in rats with MetS (37, 38). Collectively, these findings suggest that activation of AT1 contributes to the manifestation of MetS, potentially through UCP and aconitase-integrated pathways.

The associations among UCP2, TCA cycle activity (aconitase and SDH), AT1 activation, and steatosis during insulin resistance have not been examined simultaneously. The incongruence in the literature regarding the impact of insulin resistance and/or diabetes on UCPs and oxidative stress warrants further investigation. Furthermore, the contribution of AT1 activation to steatosis, the generation of mitochondrial oxidants, and oxidative stress have not yet been assessed during high glucose (HG) supplementation in animals with MetS. Therefore, using the Otsuka Long Evans Tokushima Fatty (OLETF) rat, a model characterized by diet-induced obesity, dyslipidemia, hypertension, late-onset hyperglycemia, increased RAS, and insulin resistance (39, 40), we tested the hypothesis that glucose supplementation exacerbates the AT1-mediated suppression of UCP2 expression and TCA cycle activity (aconitase and SDH), associated with increased hepatic lipid content and oxidative stress.

Materials and Methods

All experimental procedures were reviewed and approved by the institutional animal care and use committees of Kagawa Medical University and the University of California, Merced, and were conducted in compliance with the ethical guidelines for animal research. The present article describes a set of data that complements our previous study using tissues from the same animals (37).

Animals

Male Long Evans Tokushima Otsuka (LETO) (lean control strain) (265 ± 6 g) and obese OLETF (351 ± 2 g) rats (Otsuka Pharmaceutical Co. Ltd., Tokushima, Japan) were maintained at Kagawa Medical University. At the onset, 9-wk-old rats from both strains were randomly assigned to one of the following groups (n = 10/group): 1) lean, untreated LETO; 2) obese, untreated OLETF; 3) OLETF + angiotensin receptor blocker (ARB) (10 mg olmesartan/kg·d in chow); 4) OLETF + HG supplementation (5% glucose in drinking water); and 5) OLETF HG + ARB. The ARB (olmesartan; Daiichi-Sankyo, Tokyo, Japan) is an Ang II type 1 receptor (AT1) blocker and was mixed in the food and added according to initial food consumption rates to attain the desired dosage. All untreated animals were given free access to water and fed ad libitum standard laboratory rat chow (MF; Oriental Yeast Corp., Tokyo, Japan). Animals were housed in groups of three or four rats per cage in a specific pathogen-free facility under controlled temperature (23 C) and humidity (55%) with a 12-h light, 12-h dark cycle (37).

Blood pressure

Systolic blood pressure (SBP) was measured after 6 wk of treatment in conscious rats by tailcuff plethysmography (BP-98A; Softron Co., Tokyo, Japan) (37).

Plasma and tissue collection

At the end of the study (6 wk), animals were fasted overnight, body mass (BM) was measured, and plasma and liver were collected. Blood was immediately collected in chilled tubes containing 50 mm EDTA and protease inhibitor cocktail (PIC) (Sigma-Aldrich, St. Louis, MO). Blood samples were centrifuged at 3000 × g for 15 min at 4 C. Plasma was transferred to cryo-vials, frozen by immersion in liquid nitrogen, and stored at −80 C until analyzed. Livers were harvested, weighed, and a piece (∼300 mg) of tissue was snap frozen in liquid nitrogen and stored at −80 C until analyzed.

Plasma analyses

Nonesterified free fatty acids (NEFA) and glycerol concentrations were measured using colorimetric kits (Wako Chemicals, Richmond, VA; Cayman Chemical, Ann Arbor, MI). TG concentrations were measured using an autoanalyzer (Hitachi-Roche 912; Roche, Indianapolis, IN). An aliquot of plasma (0.8 ml) was extracted for measurement of Ang II as previously described (41) and measured by RIA using a commercially available kit (Phoenix Pharmaceuticals, Burlingame, CA). Corticosterone (B) was also measured by RIA using a commercially available kit (MP Biomedicals, Solon, OH).

Hepatic lipid content

Lipids were extracted from the liver pieces to measure NEFA and TG content. Liver pieces (∼25 mg) were homogenized in 1 ml of 50 mm potassium phosphate buffer containing 1% Triton X-100. Homogenates were centrifuged (3000 × g 15 min at 4 C) and the supernatants collected for specific extraction of either NEFA or TG. An aliquot (100 μl) of the supernatant was extracted with 1:1 petroleum ether:isopropanol, and liver NEFA concentration was measured from this extract using a commercial colorimetric kit (Wako Diagnostics, Richmond, VA). A separate aliquot of the supernatant was extracted (dichloromethane, methanol, 0.1% sulfuric acid, and isopropanol) and liver TG concentration measured using a commercial colorimetric kit (Cayman Chemical). Hepatic lipid content was calculated as a function of tissue mass.

Hepatic enzyme activities, succinate, and oxidative damage

Hepatic aconitase, catalase, glutathione (GSH) peroxidase (GPx), GSH S-transferase (GST), and superoxide dismutase (SOD) activities were measured using commercial kits (Cayman Chemical). Frozen liver samples were homogenized in two volumes of 50 mm potassium phosphate buffer containing 1 mm EDTA, 1% Triton X-100, 1% phenylmethylsulfonylfluoride, and 1% PIC. Supernatants (crude extracts) were immediately used to measure catalase, GPx, GST, and SOD activities. Aconitase activity was measured in mitochondrial protein fractions prepared by differential centrifugation using a homogenization buffer containing sodium citrate to avoid aconitase ex vivo oxidation. Succinate-coenzyme Q reductase (complex II; SDH) activity was measured in liver mitochondria prepared using a mitochondria isolation kit (Pierce Biotechnology, Rockford, IL) supplemented with PIC. SDH was first immuno-captured from liver mitochondria using MaxiSorp plates coated with a monoclonal antibody specific for rat SDH. Immuno-capture and activity assay procedures were performed using a commercial kit (Abcam, Cambridge, MA). Hepatic succinate content was measured as previously described (42) from crude extracts. Total protein content in crude extracts and mitochondrial fractions was measured using the Bio-Rad Bradford protein assay (Bio-Rad Laboratories, Hercules, CA). Frozen liver samples were homogenized in two volumes of 50 mm potassium phosphate buffer, centrifuged, and diluted in PBS to 10 μg/ml. Subsequent supernatants were used to measure hepatic 4-hydroxy-2-nonenal (4-HNE), malondialdehyde (MDA), and nitrotyrosine (NT) concentrations as markers of oxidative damage (43). Commercial EIA kits (Cell BioLabs, San Diego, CA) were used to measure NT and 4-HNE. Twenty micrograms of total protein were loaded onto 0.45-μm nitrocellulose membranes using a Bio-Dot SF microfiltration apparatus (Bio-Rad Laboratories). Membranes were blocked, probed with an MDA antibody (diluted 1:1000, catalog no. sc-130087; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), and developed similar to Western blot (described below). Uniform protein loading was confirmed with Ponceau S staining.

Hepatic UCP2, AT1, AT2, and AO expressions

Protein expression of AT1 and AT2, angiotensinogen (AO), and UCP2 were measured by Western blotting using standard techniques. For UCP2 measures, mitochondrial protein fractions were prepared using a commercial kit (Pierce Biotechnology) supplemented with PIC as described above. Membrane fractions were obtained for measurements of AT1 and AT2, and cytosolic fractions were used for measurement of AO. Western blottings were performed as previously described (37) using a primary anti-UCP2 (diluted 1:5000, catalog no. NB 100-59742; Novus Biologicals, Littleton, CO), anti-AT1 (diluted 1:200, catalog no. sc-1173; Santa Cruz Biotechnology, Inc.), monoclonal anti-AT2 (diluted 1:500, catalog no. ab92445; Abcam), and anti-AO (diluted 1:1000; a kind gift from C. Sernia, University of Queensland, St. Lucia, Australia). For all primary antibodies, an horseradish peroxidase -conjugated secondary antibody (Pierce Biotechnology) was used. Percent change from LETO was calculated after band densities were normalized against cytochrome oxidase IV (diluted 1:5000; Abcam) or β-actin (diluted 1:1000; Santa Cruz Biotechnology, Inc.), which served as loading controls.

Statistics

Means (±se) were compared by ANOVA using a Fisher's protected least significant difference post hoc test and considered significant at P < 0.05. Statistical analyses were performed with the SYSTAT 11.0 software (SPSS, Richmond, CA).

Results

Blood pressure and insulin resistance

Data demonstrating the development of hypertension and insulin resistance in OLETF rats and improvement with ARB treatment have been previously published (37) but are briefly included here to confirm these effects. SBP increased (P < 0.001) 33% in OLETF compared with LETO (Table 1). Treatment with ARB decreased (P < 0.001) SBP by 27 and 29% compared with OLETF and OLETF HG, respectively (Table 1). Insulin resistance index (IRI) was calculated as the product of area under curve (AUC) for glucose and insulin (AUC_glucose × AUC_insulin) as previously described (44) after an oral glucose tolerance test. IRI increased (P < 0.0001) 97% in OLETF (12.9 ± 0.9) compared with LETO (6.5 ± 0.7), and ARB treatment decreased IRI 39% (P < 0.01) (7.9 ± 0.6). Glucose supplementation increased (P < 0.01) IRI an additional 28% (18.0 ± 1.1), and again ARB treatment decreased (P < 0.0001) IRI (11.6 ± 1.2).

Table 1.

Mean (±se) SBP, BM, absolute liver and relative liver masses (LM) from lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment

	LETO	OLETF	OLETF + ARB	OLETF HG	OLETF HG + ARB
SBP (mm Hg)	113 ± 4	146 ± 4a	107 ± 2b	151 ± 2	107 ± 2c
BM (g)	385 ± 8	489 ± 6a	437 ± 6b	536 ± 9b	474 ± 6c
Absolute LM (g)	10.5 ± 0.3	13.8 ± 0.2a	11.9 ± 0.3b	15.5 ± 0.4b	13.3 ± 0.2c
Relative LM (g/100 g BM)	2.72 ± 0.04	2.85 ± 0.03a	2.75 ± 0.03b	2.88 ± 0.03	2.83 ± 0.02

SBP taken from Ref. 37.

^aP < 0.05 vs. LETO.

^bP < 0.05 vs. OLETF.

^cP < 0.05 vs. OLETF HG.

Body and liver masses

BM data have been previously published (37) but presented here for completeness. Although a strain effect on food intake was observed, ARB treatment did not significantly change intake (37). A strain difference was associated with a 27% increase (P < 0.05) in mean BM. ARB treatment decreased (P < 0.05) BM by 11% in OLETF + ARB compared with OLETF (Table 1). HG exacerbated the increase (P < 0.05) an additional 10%, and again, ARB treatment decreased it by 12% (P < 0.05) (Table 1). Strain was also associated with a 31% increase (P < 0.0001) in mean absolute liver mass, whereas ARB treatment reduced it by 4% (P < 0.0001) but was still 13% greater (P < 0.01) than in LETO (Table 1). HG exacerbated the increase (P < 0.0001) in mean absolute liver mass an additional 12%, and again, ARB treatment was associated with a decrease (14%; P < 0.0001) (Table 1). The changes in mean absolute liver mass among LETO, OLETF, and OLETF + ARB were also consistent with the changes in mean relative liver mass (P < 0.05) (Table 1). However, those same changes among the OLETF, OLETF HG, and OLETF HG + ARB groups were not observed with mean relative liver mass (Table 1).

Plasma lipid concentrations

Plasma NEFA, TG, and glycerol concentrations were measured to assess the effects of glucose supplementation and ARB treatment on lipid content in a model of MetS. Plasma NEFA and TG concentrations have been published (37) but are reported here to provide a more thorough interpretation. Mean plasma NEFA concentration was 25% higher (P < 0.01) in OLETF than in LETO (Fig. 1A). HG increased (P < 0.001) mean plasma NEFA concentration an additional 28% compared with OLETF (Fig. 1A). Regardless of HG, ARB treatment had no effect on plasma NEFA concentrations (Fig. 1A). Mean plasma TG concentrations were 3-fold higher (P < 0.0001) in OLETF than in LETO, and HG increased (P < 0.0001) concentrations an additional 35% compared with OLETF (Fig. 1B). ARB treatment decreased (P < 0.05) mean plasma TG 19 and 15%, respectively, compared with OLETF and OLETF HG (Fig. 1B). No differences in mean plasma glycerol concentrations were detected among LETO, OLETF, and OLETF + ARB groups. However, HG increased (P < 0.0001) mean plasma glycerol concentrations 87% compared with OLETF, and ARB treatment decreased (P < 0.01) mean concentrations 23% compared with OLETF HG (Fig. 1C).

Fig. 1.

Mean (±se) plasma concentrations of NEFA (A), TGs (B), and glycerol (C) in lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF; §, P < 0.05 vs. OLETF HG.

Hepatic lipid content

Neither strain nor HG was associated with increased hepatic content of NEFA, but ARB treatment increased (P < 0.05) mean content 17 and 11%, respectively (Fig. 2A). Mean hepatic TG content was increased (P < 0.0001) approximately 3-fold in OLETF compared with LETO, and ARB treatment reduced (P < 0.05) the mean content 25% (Fig. 2B). HG exacerbated the increase (P < 0.01) in mean hepatic TG content by an additional 25%, and ARB treatment reduced (P < 0.0001) the mean content 55% (Fig. 2B).

Fig. 2.

Mean (±se) hepatic content of NEFA (A) and TGs (B) in lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF; §, P < 0.05 vs. OLETF HG.

Hepatic mitochondrial UCP2 protein expression, aconitase and SDH activities, and succinate content

Hepatic mitochondrial UCP2 protein expression, aconitase and SDH activities, and succinate content (a potential marker of impaired function and damage) were measured to assess the contributions of glucose supplementation and ARB treatment to mitochondrial oxidant production, TCA cycle activity, and substrate content in the liver of rats with MetS. The 18% decrease in mean UCP2 expression in OLETF compared with LETO did not reach significance (P < 0.10). However, ARB increased (P < 0.001) mean expression 44% (Fig. 3A). HG suppressed (P < 0.05) the expression of mean UCP2 31% compared with LETO, and ARB recovered (P < 0.001) mean protein expression to LETO levels (Fig. 3A). Mean aconitase activity decreased (P < 0.01) 64% in OLETF compared with LETO, and ARB almost completely recovered (P < 0.01) mean activity (Fig. 3C). HG also suppressed (P < 0.01) mean aconitase activity compared with LETO but not to levels that were significantly different from OLETF, but ARB increased (P < 0.05) mean activity 40% (Fig. 3C). Mean SDH activity was reduced by 56% in OLETF compared with LETO (P < 0.05), and ARB treatment completely recovered the activity by increasing it 3-fold (P < 0.05) (Fig. 3D). Mean SDH activity was 75% lower in OLETF HG compared with LETO and 50% lower compared with OLETF (P < 0.05) (Fig. 3D). ARB treatment increased mean SDH activity by 3-fold during HG (Fig. 3D). Mean hepatic succinate content was over 4-fold greater (P < 0.01) in OLETF compared with LETO, and ARB treatment completely ameliorated (P < 0.001) this increase (Fig. 3E). Mean succinate content was over 2-fold greater in HG compared with LETO but was not significant (P < 0.10), and the mean levels were lower (P < 0.05) than OLETF (Fig. 3E). ARB treatment reduced (P < 0.05) mean succinate content with HG (Fig. 3E).

Fig. 3.

A, Mean (±se) liver UCP2 protein expression (normalized using cytochrome oxidase IV) in LETO, OLETF, OLETF + ARB, OLETF HG, and OLETF HG + ARB after 6 wk of treatment. B, Representative Western blottings showing expression levels in each group and validation of expression. Lanes: 1, UCP2 lysate (UCP2-transfected HEK293T; catalog no. NBL-17582; Novus Biologicals); 2, empty vector; and 3, LETO mitochondrial protein fraction. Mitochondrial aconitase activity (C), mitochondrial SDH activity (D), and succinate content (E) in lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF; §, P < 0.05 vs. OLETF HG.

Hepatic 4-HNE, MDA, and NT content

Hepatic 4-HNE, MDA, and NT content were measured to assess the effects of consumption and the benefits of ARB treatment on oxidative damage in the liver of rats with MetS. Mean 4-HNE content was 3-fold greater (P < 0.001) in OLETF than in LETO, and ARB treatment completely ameliorated (P < 0.001) this increase (Fig. 4A). Interestingly, HG suppressed mean 4-HNE content to nearly half (P < 0.01) of normal (LETO) and almost to a quarter (P < 0.0001) of OLETF, and ARB treatment in the presence of HG normalized (P < 0.05) mean content (Fig. 4A). Mean MDA content increased approximately 50% (P < 0.05) in OLETF compared with LETO, but the 25% decrease in ARB did not reach significance (P < 0.10) (Fig. 4B). Similar to the pattern observed with 4-HNE, HG suppressed mean MDA content to LETO levels, and ARB treatment in the presence of HG returned levels to that of OLETF (P < 0.05) (Fig. 4B). Mean hepatic NT content was 56% greater (P < 0.05) in OLETF compared with LETO, and ARB treatment completely ameliorated (P < 0.01) this increase (Fig. 4C). Although the 21% increase in mean NT content with HG was not significant (P < 0.10) compared with OLETF, ARB treatment in the presence of HG normalized (P < 0.05) these levels (Fig. 4C).

Fig. 4.

Mean (±se) hepatic 4-HNE (A), MDA (B), and NT levels (C) in lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF; §, P < 0.05 vs. OLETF HG.

Hepatic antioxidant enzyme activities

Hepatic SOD, catalase, GPx, and GST activities were measured to evaluate the effects of HG consumption and AT1 activation on antioxidant enzymes during HG supplementation in a model of MetS (Fig. 5). Mean SOD activity decreased (P < 0.01) 57% in OLETF compared with LETO, and ARB treatment completely recovered (P < 0.01) mean activity (Fig. 5A). HG did not have an additional impact on mean SOD activity with levels decreased (P < 0.05) from LETO and similar levels to OLETF (P > 0.10), and ARB treatment again completely recovered (P < 0.01) mean activity (Fig. 5A). Mean catalase activity decreased (P < 0.05) 33% between LETO and OLETF, and ARB completely recovered (P < 0.001) the activity (Fig. 5B). Interestingly, HG inhibited the strain effect in catalase activity with activity levels remaining similar (P > 0.10) to LETO, but ARB in the presence of HG did not induce any further effect (Fig. 5B). Mean GPx activity decreased (P < 0.001) 58% in OLETF compared with LETO, and ARB treatment completely recovered (P < 0.05) activity levels (Fig. 5C). The approximately 33% decrease in mean activity with HG did not reach significance (P < 0.10) compared with OLETF, but the decrease was sufficient enough to prohibit ARB treatment from completely recovering activity, even though levels were doubled (Fig. 5C). Nonetheless, mean GPx activity levels were only 35% of OLETF + ARB group in the presence of HG (Fig. 5C). Mean GST activity decreased (P < 0.05) 44% in OLETF compared with LETO, and ARB treatment completely recovered (P < 0.05) activity levels (Fig. 5D). Interestingly, HG had no impact on mean activity levels, and ARB treatment in the presence of HG decreased (P < 0.05) activity levels 52% (Fig. 5D).

Fig. 5.

Mean (±se) activity of hepatic catalase (A), GPx (B), GST (C), and SOD (D) in lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF; §, P < 0.05 vs. OLETF HG.

Hepatic angiotensin receptors and AO protein expressions

The protein expressions of hepatic Ang II receptors (AT1 and AT2) and AO were measured to assess the contribution of hepatic RAS to the observed changes. No significant changes in hepatic AT1 (Fig. 6A) and AO (data not shown) content were detected among the groups despite observed trends for increases with OLETF and recovery with ARB. Mean AT2 content increased (P < 0.01) over 3-fold in OLETF compared with LETO, and ARB reduced (P < 0.05) levels approximately 50% (Fig. 6B). HG did not significantly change AT2 levels compared with OLETF but remained elevated compared with LETO, and ARB had no further effect on levels (Fig. 6B).

Fig. 6.

Mean (±se) hepatic AT1 (A) and AT2 (B) levels in lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. Insets, Representative Western blottings showing expression levels in each group. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF.

Plasma Ang II and corticosterone

To confirm the activation of RAS and elevated glucocorticoids, both of which are commonly associated with MetS, plasma Ang II and B were measured. Insulin resistant OLETF rats exhibited a nearly 3-fold greater (P < 0.05) mean plasma Ang II concentration. ARB effectively blocked Ang II receptors, resulting in a nearly 6-fold exacerbation in mean plasma Ang II as expected (Fig. 7A). HG induced a nearly 6-fold decrease (P < 0.01) in mean Ang II, and ARB treatment again was associated with an exacerbation in mean levels (P < 0.001), but these levels were still only about 50% of ARB-treated animals without HG (Fig. 7A). Similar to plasma Ang II concentrations, insulin resistance (OLETF) was associated with a 30% increase (P < 0.05) in mean plasma B concentrations, and ARB treatment completely normalized (P < 0.05) this mean concentration (Fig. 7B). HG suppressed (P < 0.01) mean B by 67%. However, treatment with ARB in the presence of HG did not result in an additive suppression in mean concentration but rather in a 87% increase (P < 0.05) that represented normalization of mean concentrations (Fig. 7B).

Fig. 7.

Mean (±se) plasma concentrations of Ang II (A) and corticosterone (B) from lean, LETO, obese, insulin resistant OLETF, OLETF + ARB, OLETF + HG supplementation, and OLETF HG + ARB rats after 6 wk of treatment. *, P < 0.05 vs. LETO; †, P < 0.05 vs. OLETF; §, P < 0.05 vs. OLETF HG.

Discussion

The consumption of calorically sweetened beverages increases central obesity, NASH, hepatic inflammation, oxidative stress, and the onset of the MetS (7–9, 45). In addition, MetS is commonly associated with activation of RAS. However, the contribution of AT1 activation to the manifestation of steatosis via oxidative stress associated with impaired lipid metabolism and uncoupling of the mitochondrial membrane is not well described. The present study demonstrates that insulin resistance is characterized by impaired lipid metabolism, increased accumulation of liver TGs, lipid peroxidation and protein oxidation, and reduced hepatic TCA cycle activity, UCP2 protein expression, and antioxidant enzyme activities. The recovery of all of these factors with ARB treatment and the associated lower BM indicates that these consequences of insulin resistance are AT1 mediated in the presence of reduced BM. Although the benefits of ARB treatment on systemic oxidative stress (46–48) by reducing superoxide production by limiting the assemblage of the membrane Nox complex (49, 50) are well described, the most important and novel finding of the present study is the discovery that hepatic UCP2 protein expression and activities of aconitase and SDH are AT1 mediated and that their suppression is associated with an increase in hepatic oxidative damage (increased 4-HNE, MDA, and NT), decreased antioxidant enzyme activities, and a reduction in TCA cycle activity during insulin resistance. Although these changes are not associated with changes in hepatic AT1 content, they appear to be associated with compensatory changes in AT2. Furthermore, the physiological role of succinate during pathological conditions (tumorigenesis, diabetic nephropathy, NASH, etc.) (31, 42) is not well defined, but the scarce, available data, including this study, suggest that it contributes in some capacity, potentially through its recently discovered receptor, GPR91 (G protein-coupled metabolic receptor 91) (42, 51).

The physiological role of UCP2, especially during insulin resistant or diabetic conditions, is not fully understood (19, 20), although it appears that the main function of UCP2 is the control of mitochondria-derived oxidant production (21, 22). The present study suggests that UCP2, in conjunction with activities of aconitase, SDH, and antioxidant enzymes, plays a critical role in ameliorating mitochondrial oxidant production and hepatic oxidative damage. Furthermore, and more importantly, we demonstrate that hepatic UCP2 protein expression and aconitase and SDH activities are partially regulated by AT1 activation, because ARB treatment completely recovered the suppressed levels associated with the insulin resistant phenotype. Additionally, because ARB treatment was consistently associated with lower BM, the effects of reduced mass on UCP2 and enzyme activities may have also contributed. Superoxide radical (12, 26, 52) and 4-HNE (53) have been reported to be potent activators of UCP1, UCP2, and UCP3. However, in the present study, UCP2 protein expression was suppressed in the presence of increased 4-HNE, suggesting that UCPs are differentially regulated. Some important distinctions need to be considered when comparing these studies. Activation of UCPs (including liver) were assessed by quantifying respiration rate, and not UCP protein expression directly, from isolated mitochondria in an ex vivo setting from nondiabetic/noninsulin resistant rats (53). Skeletal muscle mitochondria of UCP3 knockout mice (23) and inhibition of skeletal muscle UCP3 (24) are associated with decreased aconitase activity and increased markers of oxidative damage (25), consistent with the UCP2-aconitase activity relationship in the present study, suggesting that increased UCP2 content/activity may contribute to increased aconitase activity to ameliorate hepatic oxidative damage. Attenuation of oxidative stress by UCP2 in adipose (26) and macrophages (21) provides further evidence of its potential contribution to such a role in the liver in the present study. Furthermore, the reduction of hepatic mitochondrial aconitase activity in alloxan-induced diabetic mice (28) is consistent with the observed decrease in insulin resistant, OLETF rats in the present study. Conversely, in mice with various forms of hepatic steatosis (i.e. alcohol or high-fat diet induced), gene expression (15–17) and protein content (18) of UCP2 are increased. Although discrepancies in the response of hepatic UCP2 to an obese phenotype between mice and rats exist, we would argue that these data are not necessarily contradictory but enhance our understanding of hepatic UCP2 regulation and its role in mediating oxidative stress. Disregarding the potential for interspecies differences, a collective assessment of these datasets suggests that hepatic UCP2 may be sensitive to the status of insulin resistance and oxidative stress, because obesity and some degree of steatosis are common to both animal models. Furthermore, no difference in hepatic lipid peroxidation (assessed by thiobarbituric acid reactive substances) (18) was observed between lean and obese wild-type and UCP2 knockout mice, suggesting that some, yet undefined, obesity-related factor is responsible for up-regulating hepatic UCP2 in obese mice. Unfortunately, the insulin resistant status and level of RAS of these mice were not measured or reported. Thus, we propose that, independent of the obese phenotype, the increased activation of RAS suppresses UCP2 protein expression and enzyme activities (aconitase, antioxidants, and SDH), resulting in increased hepatic oxidative damage (4-HNE, MDA, and NT) during insulin resistance.

Another important aspect of the present study is the simultaneous measurements of hepatic UCP2 with aconitase, SDH, and antioxidant enzyme activities and markers of oxidative damage. Aconitase is a redox-sensitive enzyme that can be inactivated by superoxide coming from complex III in the electron transport chain (54, 55), and the loss of aconitase activity has been used as an indicator of mitochondrial-derived oxidative stress (56). Additionally, because aconitase activity is so tightly integrated with the TCA cycle, its activity may serve as an indicator of mitochondrial metabolism (29). The reduction of hepatic aconitase activity in the presence of elevated markers of oxidative damage is representative of the redox-sensitive nature of aconitase previously described (54, 55), suggesting that this reduced activity is indicative of oxidative stress. Although glucose supplementation did not suppress aconitase activity beyond that associated with the insulin resistance phenotype characteristic of OLETF rats, ARB treatment was equally effective at recovering the lost activity, suggesting that mitochondrial superoxide production and aconitase activity are mediated by AT1 and compensatory changes in AT2.

Elevated plasma Ang II is a consequence of insulin resistance in this model, likely contributing to many of the deleterious effects observed. However, the suppression of plasma Ang II in the presence of glucose supplementation is difficult to reconcile here and may partially explain why HG did not exacerbate many of these deleterious effects. Thus, reduced circulating Ang II in the HG animals may lessen the compensatory changes in the hepatic AT2 receptor. Although we recognize the concerns with AT receptor antibodies, thus limiting the interpretation of these results, these data are intriguing and provide the basis for future studies in which an AT2 antagonist is provided alone and in combination with olmesartan along with quantification of AT1/2 genetic expressions to better assess the contributions of these receptors to steatosis and hepatic oxidative stress during insulin resistance. Again, the lower BM associated with ARB treatment cannot be discredited as a contributing factor to the observed benefits of ARB treatment in the present study.

The decrease in aconitase activity with and without glucose supplementation also suggests a decrease in hepatic TCA cycle activity that may have resulted from the increased substrate availability (i.e. citrate) (57). Increased substrate availability may partially explain the increased plasma concentrations and liver contents of free fatty acids and TG, because increased citrate is shuttled into fatty acid synthesis. Although the exact mechanisms linking aconitase activity with UCP2 have not been elucidated, the expression of hepatic UCP2 and antioxidant enzyme activities may have increased in response to the ARB treatment-associated increase in aconitase activity to prevent oxidative damage in hepatocytes, resulting from increased TCA cycle activity and normalization of substrate utilization.

Additionally, the decrease in hepatic SDH activity associated with the MetS phenotype further suggests that insulin resistance is associated with a general reduction of TCA cycle activity, resulting in impaired glucose metabolism and accumulation of specific TCA cycle intermediates, such as succinate. The antiparallel changes in hepatic SDH activity and succinate content indicate that reduced SDH activity results in the accumulation of succinate as expected. However, the subsequent consequence of this succinate accumulation needs to be evaluated further. Nonetheless, because hepatic succinate content changed in parallel with hepatic markers of oxidative damage (4-HNE, MDA, and NT), increased succinate content may serve as a viable marker for assessing impaired TCA cycle activity in addition to oxidative damage. Furthermore, the contribution of the succinate receptor, GPR91, to mediating the observed hepatic oxidative damage cannot be dismissed (42, 51). The changes in hepatic UCP2 in the present study are consistent with those observed in vitamin E succinate supplemented, steatotic mice (32). However, in mice where succinate supplementation increased hepatic GSH content after 7 d in the presence of reduced UCP2 expression, OLETF rats exhibited reduced antioxidant enzyme activities (i.e. GPx) in the presence of elevated hepatic succinate content and a 20% reduction (not significant) in hepatic UCP2 content, suggesting that: 1) GSH may play a more formative role in alleviating the oxidative damage associated with NASH than antioxidant enzymes; and 2) succinate, either by impaired TCA cycle activity or dietary intake, decreases UCP2 content.

Because studies elucidating the contribution of AT1 activation and AT2 content to mitochondrial oxidative stress are scarce, the findings from the present study are particularly novel and enhance our understanding of the mechanisms contributing to the regulation of oxidative stress in the liver. Blockade of AT1 attenuates mitochondrial oxidative stress during aging and diabetic retinal neurodegeneration (58, 59). Mitochondrial oxidant generation increases with high sucrose and fructose consumption (60, 61) due to an increase in the production of electron donors from the TCA cycle (62, 63). The present study demonstrates that ARB treatment consistently reduced hepatic 4-HNE, MDA, and NT, surrogate measures of oxidative stress, in the absence, but not the presence, of glucose supplementation. Interestingly, ARB treatment in the presence of glucose increased hepatic 4-HNE and MDA levels beyond those of HG alone. These paradoxical effects in the HG animals may be consequence of the inability of ARB treatment to completely recover GPx and GST activities in these animals (compared with ARB animals without glucose), suggesting that appropriate GSH regulation is especially important to the maintenance of antioxidant capacity during insulin resistance. Although activation of AT1 appears to be a contributing factor in the mechanism responsible of mitochondrial oxidant generation in the liver, the redox-sensitive mitochondrial adenosine triphosphate-dependent potassium channels (64–67) also likely play a significant role. Because Nox stimulate mitochondrial oxidant generation by activating these channels, blockade of AT1 may impair the assemblage of the Nox, ultimately resulting in reduced mitochondrial oxidant formation and subsequently preventing hepatic oxidative damage in the liver of animals with MetS.

The contribution of antioxidant enzyme activities is another important component of the mechanisms regulating oxidative stress. Olmesartan (the same ARB used here) has been shown to improve extracellular SOD activity, whereas telmisartan increases catalase, GPx, and SOD in whole blood and peripheral mononuclear cells of hypertensive patients (68–70). Treatment with ARB has also been shown to restore the expression of antioxidant enzymes during renal failure in a rat model of spontaneous focal glomerulosclerosis and up-regulate SOD in diabetic mice (71, 72). In the present study, ARB treatment was associated with reduced hepatic oxidative damage and recovery of the activities of catalase, GST, GPx, and SOD, elucidating the importance of antioxidant capacity to ameliorating the AT1-mediated oxidative damage associated with the development of insulin resistance. Interestingly, glucose supplementation did not exacerbate the suppression of antioxidant activities or the increase in oxidative damage despite the exacerbation in hepatic TG accumulation and plasma levels. However, the protein expression of UCP2 and the activity of aconitase were also not suppressed to any greater extent with glucose supplementation, suggesting that AT1-mediated oxidative damage is more closely related to UCP2 content and the activities of the enzymes than the accumulation of hepatic TGs and dyslipidemia at this early stage in the manifestation of steatosis-associated insulin resistance.

The lack of change in both plasma glycerol and plasma NEFA with ARB treatment suggest that there was no meaningful change in the rate of adipose lipolysis during the study period (73). The observed decreases in hepatic and plasma TG along with increased hepatic NEFA in the ARB-treated animals suggest that hepatic reesterification of NEFA decreased over time, possibly leading to a decrease in the packaging of very low-density lipoprotein (74), subsequently decreasing the concentration of TG in circulation. Activation of AT1 has been reported to cause fibrosis by “activating” the hepatic stellate cells (75), which are responsible for lipid storage. Thus, the observed reduction in hepatic lipid content with ARB treatment may simply have been the result of inhibition of stellate cell activation rather than a direct effect of the ARB on lipid metabolism. The lower BM, and thus adiposity, associated with ARB treatment must also be considered a contributing factor to the changes in hepatic lipid content.

Hepatic steatosis is closely related to hypertension (76) and recently suggested that hypertension increases hepatic TG content and oxidative stress in spontaneously hypertensive rats (77). However, this conclusion is difficult to reconcile given that blood pressure was not measured (77), casting doubt on the contribution of hypertension to the manifestation of steatosis. In the present study, SBP in OLETF and OLETF HG animals were elevated but not different between the two groups (Table 1). However, liver TG content was 25% greater in OLETF HG animals, suggesting that the degree of steatosis is not associated with exacerbation of SBP (that is the effect is not linear). Additionally, liver TG content in ARB-treated animals remained approximately 2-fold greater than in lean LETO rats, even though SBP was similar among each of these groups, suggesting that steatosis is not related to arterial blood pressure in insulin resistant OLETF rats. Furthermore, ARB reduces SBP within 24 h and keeps pressure suppressed (78), whereas hepatic steatosis develops over 6 wk in these animals, including in ARB-treated animals in which SBP is normalized, suggesting that hypertension does not induce the manifestation of steatosis in insulin resistant rats. Nonetheless, the contribution of hypertension to the development of NASH remains elusive and warrants further investigation.

In summary, the present study demonstrates that hepatic UCP2 expression and TCA cycle activity (aconitase and SDH) are associated, at least in part, by activation of AT1 and the compensatory changes in AT2, and increased plasma Ang II and corticosterone. Additionally, the higher BM observed in the OLETF and OLETF + HG animals must also be considered as a contributing factor as many of the reported changes (i.e. UCP2 and enzyme activities) with ARB treatment included lower BM. The AT1-mediated suppression of this hepatic UCP2 content and TCA cycle activity are associated with increased hepatic content of succinate, 4-HNE, MDA, and NT, indicating that increased lipid peroxidation and protein oxidative damage are the consequence of insulin resistance in OLETF rats. These data also demonstrate that activation of RAS decreases hepatic antioxidant enzyme activities and that this suppression of activity is independent of increased glucose intake. Moreover, the benefits of AT1 blockade are not impaired by exacerbated plasma TGs associated with increased glucose intake, and thus, complete recovery of aconitase, SDH, and antioxidant enzyme activities is accomplished with ARB treatment, independent of the level of glucose intake. Therefore, besides the established antihypertensive effects of ARB, inhibition of RAS by blocking AT1 in addition to lower BM improves overall hepatic health, independent of glucose consumption, during insulin resistance in a model of MetS.