Brown adipose tissue redox status in response to dietary-induced obesity-associated oxidative stress in male and female rats

Abstract

Obesity is linked to systemic oxidative stress and, although brown adipose tissue (BAT) plays a crucial role in energy balance, BAT redox status effects on obesity have not been studied previously. Female rats exhibit a greater BAT thermogenic capacity, attributed to enhanced mitochondrial differentiation, than males. The aim of this study was to investigate whether the mitochondrial sexual dimorphism is related to differences in BAT redox status and to assess its role in the regulation of body weight gain in response to chronic high fat diet (HFD) feeding. Ten-week-old Wistar rats of both genders were fed a pelleted control diet or HFD for 26 weeks. Although mitochondria of female rats produced higher levels of hydrogen peroxide than those of males, females exhibited lower oxidative damage, attributed to greater glutathione peroxidase activity and higher glutathione content. In response to HFD, body weight increased markedly in females, but oxidative capacity increased only in males, thus maintaining improved BAT redox status compared with females. In conclusion, the sexual dimorphism in BAT redox status found in control animals is attenuated by the HFD. The enhanced oxidative capacity of HFD males can be related to their greater resistance to body weight gain.

Keywords::

• Brown adipose tissue
• gender differences
• mitochondria
• obesity
• oxidative stress
• redox status

Introduction

White and brown adipose tissues (WAT and BAT, respectively) play opposite roles in the regulation of whole-body energy balance: WAT is an energy storage tissue, whereas BAT is an energy dissipating tissue. BAT thermogenic potential is activated in response to several physiological stimuli, such as cold exposure or excessive calorie intake (Lowell and Spiegelman 2000). However, the role of other tissues, such as the liver, in diet-induced thermogenesis, has also been reported (Ma et al. 1988; Ma and Foster 1989; Bachman et al. 2002). BAT thermogenic capacity relies on the presence of a large number of mitochondria that are naturally uncoupled due to the function of uncoupling protein 1 (UCP1) (Heaton et al. 1978). When energy gain exceeds energy expenditure, the whole-body energy balance is altered, resulting in an increase of adipose tissue mass defined as obesity (Himms-Hagen 1984). Recent studies describe the presence of BAT in adult humans (Saito et al. 2009; Virtanen et al. 2009; Zingaretti et al. 2009); however, the role of BAT in energy homeostasis in humans remains to be elucidated.

In dietary-induced obesity, both the amount of substrate available for oxidation and the electron flux through the mitochondrial respiratory chain increase, thus enhancing free-radical production. Reactive oxygen species (ROS) produced during both control and pathological conditions can be counteracted by endogenous antioxidant defenses (Halliwell et al. 1992; Johnson 2002). The main antioxidant defenses include antioxidant enzymes, such as superoxide dismutase (SOD), glutathione peroxidase (GPx) or glutathione reductase, and non-enzymatic antioxidant defenses, such as glutathione. Hence, most of the superoxide ion produced by mitochondria is transformed into hydrogen peroxide (H₂O₂) owing to SOD activity; GPx catalyzes the reduction of H₂O₂ and other hydroperoxides by oxidizing glutathione, which is, in turn, reduced by the enzyme glutathione reductase (Navarro and Boveris 2007). When ROS production overwhelms the antioxidant defenses, an imbalance in the redox status takes place, which is designated as oxidative stress. In this context, ROS can cause oxidative damage in mitochondrial components in the form of lipid peroxidation, protein modifications, and DNA mutation, leading to altered mitochondrial functionality, which has been related to a wide range of human pathologies, including obesity-related pathologies (James and Murphy 2002). Indeed, systemic oxidative stress correlates with body mass index (Olusi 2002; Keaney et al. 2003) and obesity prevalence is linked to decreased concentrations of plasma antioxidants (Reitman et al. 2002). This association between obesity and pro-oxidative environment leads to the emergent view that characterizes obesity as a chronic inflammatory situation (Dandona et al. 2004).

The redox status of adipose tissues could play an important role in obesity and the metabolic syndrome. In the case of WAT, studies using several obesity models have shown disparity in the results. Some studies showed that obesity is characterized by a reduced redox state in WAT (Galinier et al. 2006a), which could, in turn, encourage preadipocyte proliferation and differentiation (Carriere et al. 2004), promoted by overexpression of antioxidant defenses (McClung et al. 2004). However, other studies revealed that WAT oxidative stress increased in obesity, which could act as an early instigator of the metabolic syndrome (Furukawa et al. 2004; Rebolledo et al. 2008). In the case of BAT, some previous studies suggest that cafeteria diet feeding enhances BAT thermogenic activity (Rothwell and Stock 1979; Brooks et al. 1980), whereas others show that a similar dietary treatment does not alter BAT oxygen consumption (Ma et al. 1988; Ma and Foster 1989). Either way, the effects of BAT redox status on obesity to our knowledge are as yet unknown.

BAT thermogenic capacity and mitochondrial function have been found to be gender dependent. In comparison with male rats, females show an enhanced thermogenic potential (Quevedo et al. 1998) which could be attributed to their more differentiated mitochondria, because females exhibit elevated UCP1 content, greater multilocular arrangement, as well as more cristae dense mitochondria (Rodriguez-Cuenca et al. 2002; Justo et al. 2005). Although the exact causes of this sexual dimorphism are not clear, the hormonal environment, such as norepinephrine, thyroid hormones, and sex steroid levels, seems to play a crucial role (Pedersen et al. 2001; Rodriguez-Cuenca et al. 2002; Rodriguez-Cuenca et al. 2007; Villena et al. 2007; Valle et al. 2008).

From this background, the aim of this study was to investigate whether the aforementioned mitochondrial sexual dimorphism is related to differences in BAT redox status and its role in the regulation of body weight in response to chronic high fat diet (HFD) feeding. To address this question, we analyzed the effects of oxidative stress associated with dietary-induced obesity on both body weight gain and BAT redox status in male and female rats. We determined, in BAT, oxidative capacity, H₂O₂ production, levels of enzymatic and non-enzymatic antioxidant defenses, and levels of oxidative damage markers. Finally, UCP1 protein expression was measured as a marker of mitochondrial thermogenic capacity.

Methods

Animals and diets

Animal experiments were performed in accordance with general guidelines approved by our institutional ethics committee and EU regulations (2003/65/CE and 86/609/CEE). Thirty male Wistar rats (10 weeks old, body weight mean 332 ± 4 (s.e.m.)g) and 30 female Wistar rats (10 weeks old, body weight mean 217 ± 2 g) purchased from Charles River (Barcelona, Spain) were used. The rats were housed (two per cage) in controlled environmental conditions (22°C; 12 h light/dark cycle, lights on at 08:00 h and lights off at 20:00 h) with free access to food and water. Both male and female rats were divided into the following experimental groups: control males (n = 15), HFD males (n = 15), control females (n = 15), and HFD females (n = 15). The control groups were fed a control pelleted diet (3,385 kcal/kg; 2.9% (w/w) fat) and the HFD groups were fed a HFD (3,876 kcal/kg diet; 26% (w/w) fat) for 26 weeks. The HFD components and the amount consumed of each high fat food are given in Table I. This highly palatable diet has been shown to be very effective in inducing dietary obesity in rats and mimics the development of human obesity (Sclafani and Springer 1976). The energy composition of the HFD was 13% protein, 33% carbohydrate, and 54% lipid, whereas that of the control pelleted diet was 19% protein, 73% carbohydrate, and 8% lipid. Rat body weights were assessed monthly. Food intake (HFD and control diet) was analyzed at the end of the dietary treatment by introducing rats into metabolic cages (one per cage) (Panlab, Barcelona) for 24 h. All the components of the HFD were presented in several small pieces and in gross excess so as to allow the recovery the following day of at least part of all the components offered. The amount of each component consumed by each rat was calculated from the difference between the amount offered and the amount recovered the next day.

Table I.  Food and nutrient composition of the HFD.

	Nutrient composition (g/100 g food)
High fat food	Total food amount (g/100 g diet)	Protein	Carbohydrate	Lipid
Fresh bacon	46.1 ± 2.7	17.3	–	29.9
Cookies	2.80 ± 0.89	5.82	68.0	21.3
Pork liver paté	26.1 ± 2.1	11.9	2.70	29.5
Ensaïmada (typical Majorcan pastry)	15.2 ± 1.9	8.10	50.6	29.1
Chocolate	4.35 ± 0.97	6.70	60.0	30.0
Pelleted standard diet	5.09 ± 1.67	18.7	73.3	8.00

Total food amount is the amount of each high fat food consumed expressed per 100 g of diet consumed. Given that no gender differences were found in the high fat food consumption, the values of total food amount per 100 g of diet are the means ± SEM taking into account the rats of both genders (n = 12).

Rats were killed in the morning (between 08:00 and 09:00 h) by decapitation without anesthesia after a 12-h period of fasting. Trunk blood was collected and centrifuged at 900 g for 30 min at 4°C to separate serum. Inguinal, gonadal, mesenteric, and lumbar WAT depots as well as the interscapular BAT depot were dissected and weighed. Mitochondrial isolation was performed in fresh samples of BAT. The rest of the tissue and serum samples were frozen in liquid N₂ and stored at − 80°C until analysis. Since the mitochondria remain functional for a short period of time after the mitochondrial isolation (about 2 to 3 h), some rats from each group (n = 5–7) were assigned to half of the determinations performed from the mitochondrial fraction, whereas the other rats (n = 5–7) were used in the rest of the analyses of the mitochondrial fraction. In this way, we ensured optimal conditions for mitochondrial functionality.

Serum parameters

Serum glucose and triglyceride levels were measured by using the Accutrend® system and serum levels of insulin were measured by an enzyme immunoassay kit (sensitivity: 0.15 μg/l; intra-assay variability: 3.7%; inter-assay variability: 5.9%).

Sample preparation

Fresh BAT was homogenized with a teflon/glass homogenizer in isolation buffer (250 mM sucrose, 5 mM Tris–HCl, and 2 mM EGTA, pH 7.4) in 10 ml of buffer per gram of tissue and was filtered through a layer of gauze. One aliquot of homogenate was frozen at − 20°C with protease and phosphatase inhibitors (leupeptin 10 μM, pepstatin 10 μM, PMSF 0.2 mM, and ortovanadate 0.2 mM) for Western blotting analysis. Another aliquot was treated with metaphosphoric acid 5% (w/v) to remove proteins for determination of glutathione concentration and was stored at − 80°C.

Mitochondrial isolation

Approximately, 5 ml of fresh BAT homogenate was centrifuged at 8500 g for 10 min at 4°C. After removing the fat, the pellet was resuspended and the nuclei and cell debris were removed by centrifugation at 500 g for 10 min. The supernatant was centrifuged at 8000 g to yield the mitochondrial pellet, which was washed once by resuspension and centrifugation (8000 g). The final pellet was resuspended with 300 μl of the isolation buffer. To calculate mitochondrial recovery cytochrome c oxidase (COX) activity was measured in both BAT homogenate and mitochondrial fraction, setting COX activity in the homogenate as 100%. The percentage of mitochondrial recovery was applied to the parameters determined in the mitochondrial fraction when expressed per gram of tissue or milligram of DNA.

BAT composition

Total protein and mitochondrial protein were determined in homogenates and in mitochondrial fractions, respectively, as previously described (Bradford 1976). Total DNA was measured in BAT homogenates (Thomas and Farquhar 1978).

Mitochondrial oxygen consumption

BAT mitochondrial oxygen consumption was measured polarographically in isolated mitochondria as previously described (Lopez-Torres et al. 2002), with minor modifications. Isolated mitochondria (0.3 mg protein/ml) were incubated in a water-thermostatically regulated chamber with a computer-controlled oxygen Clark-type electrode Oxygraph (Hansatech, Norfolk, UK) in respiration buffer (145 mM KCl, 30 mM HEPES, 5 mM KH₂PO₄, 3 mM MgCl₂, 0.1 mM EGTA, and 0.1% BSA, pH 7.4 at 37°C). Non-phosphorylating mitochondrial respiration rate was measured with glyceraldehyde 3-phosphate (G3P) (10 mM) or palmitoyl carnitine-malate (PCM) (50 μM/5 mM) as substrates.

Mitochondrial H₂O₂ production

H₂O₂ production was assayed in isolated BAT mitochondria by measuring the increase in fluorescence (530 nm excitation, 590 nm emission) due to the reaction of Amplex Red with H₂O₂ in the presence of horseradish peroxidase (Muller and Liu 2004). Assays were performed at 37°C in a 96-well microplate fluorometer FLx800 (BioTek instruments, Winooski, VT, USA). Mitochondria (0.10 mg protein/ml) were added to the respiration buffer (145 mM KCl, 30 mM HEPES, 5 mM KH₂PO₄, 3 mM MgCl₂, 0.1 mM EGTA, pH 7.4) supplemented with horseradish peroxidase 0.1 U/ml and 50 μM Amplex Red reagent. The assays were performed in the presence of G3P (10 mM) or PCM (50 μM/5 mM) as substrates and antimycin A (2 μM). The rate of H₂O₂ production was calculated using a standard curve of H₂O₂-stabilized solution.

Enzymatic activities

In isolated BAT mitochondria, COX (Chrzanowska-Lightowlers et al. 1993), citrate synthase (CS) (Nakano et al. 2005), Mn-SOD (Quick et al. 2000), and GPx (Paglia and Valentine 1967) activities were assayed by spectrophotometric methods.

Measurement of glutathione content

Glutathione content was measured in diluted BAT homogenates (1 g tissue/20 ml isolation buffer) using the HT Glutathione Assay Kit according to the manufacturer's protocol. The samples were treated with 4-vinylpyridine to determine the oxidized glutathione concentration. To obtain the levels of reduced glutathione, the oxidized glutathione was subtracted from the total glutathione.

Measurement of thiobarbituric acid-reactive substances (TBARS) and protein carbonyl groups

Levels of TBARS, as an index of lipid peroxides, were measured spectrophotometrically in BAT homogenates as previously described (Slater and Sawyer 1971). Protein carbonyl groups, as an index of protein oxidation, were measured in BAT homogenates by immunoblot detection by using the OxyBlot™ Protein Oxidation Detection Kit according to the manufacturer's protocol with minor modifications (Guevara et al. 2009).

Western blot analysis of UCP1 protein levels

Ten micrograms of BAT mitochondrial protein for UCP1 (32 kDa) were fractionated on SDS-PAGE gels and electrotransferred onto a nitrocellulose filter. Rabbit polyclonal antibody against rat UCP1 was used as primary antibody. Development of immunoblots was performed using an enhanced chemiluminescence kit. Bands were visualized with the ChemiDoc XRS system (Bio-Rad, CA, USA) and analyzed with the image analysis program Quantity One© (Bio-Rad, CA, USA). Bands revealed an apparent molecular mass of 32 kDa for UCP1.

Materials

The Accutrend® GCT-meter and glucose and triglyceride test strips were supplied by Roche Diagnostics (Basel, Switzerland). The rat insulin enzyme immunoassay kit was purchased from Mercodia (Uppsala, Sweden). The Colorimetric HT Glutathione Assay Kit was supplied by Trevigen (Gaithersburg, MD, USA). Amplex Red and Oxyblot™ protein oxidation detection kits were obtained from Invitrogen (Carlsbad, CA, USA) and from Chemicon International (Temecula, CA, USA), respectively. Rabbit polyclonal antibody against rat UCP1 (Cat. num. UCP12-A) was from Alpha Diagnostic International (San Antonio, TX, USA). The enhanced chemiluminescence Western blotting analysis reagents were supplied by Amersham (Little Chaffont, UK). Routine chemicals used were purchased from Sigma-Aldrich (St Louis, MO, USA) and Panreac (Barcelona, Spain). Control pelleted diet (A04) was obtained from Panlab (Barcelona, Spain).

Statistical analysis

Serum parameters, biometric measurements, and measures of BAT composition are expressed as mean values ± SEM of 15 rats per group. BAT mitochondrial oxygen consumption, oxidative enzyme activities, antioxidant defenses, mitochondrial H₂O₂ production, oxidative damage, and UCP1 protein content are expressed as mean values ± SEM of 5–7 rats per group. Each determination was repeated twice per sample. Statistical differences between experimental groups were analyzed by two-way analysis of variance (ANOVA). Student's t-test, as post hoc comparison, was performed when an interactive effect of gender and diet was shown. A p-value of less than 0.05 was considered statistically significant. All statistical analyses were performed using a statistical software package (SPSS 17.0 for Windows, Inc., Chicago, IL, USA).

Results

Serum parameters, energy intake, and body weight gain

Serum glucose concentrations were similar for both control and HFD rats. Insulin concentrations were higher in male rats than is females [F(1,52) = 12.16; p = 0.001], but they were not significantly affected by the dietary treatment, (Table II). Control male rats showed lower concentrations of triglyceride than females [interaction, F(1,52) = 8.458; p < 0.01; post hoc, p < 0.01], but in response to the dietary treatment serum triglyceride increased only in females [post hoc, p < 0.001], reaching the levels of males.

Table II.  Effects of HFD on serum parameters.

	Males		Females
	Control	HFD	Control	HFD	ANOVA
Glucose (mM)	6.76 ± 0.23	6.34 ± 0.20	5.95 ± 0.37	6.37 ± 0.36	NS
Insulin (μg/L)	1.79 ± 0.36	1.92 ± 0.47	0.442 ± 0.124	0.976 ± 0.151	G
Triglyceride (mM)	2.58 ± 0.25	2.43 ± 0.20	1.67 ± 0.14^†	2.68 ± 0.21*	D,G*D

HFD, high fat diet. Values are means ± SEM (n = 15). ANOVA (p < 0.05): G indicates gender effect, D indicates diet effect, G*D indicates gender and diet interactive effect, and NS stands for non-significant. Student's t-test (p < 0.05); * HFD vs. control; ^† females vs. males.

Energy intake increased significantly with the consumption of the HFD in both genders [F(1,23) = 11.41; p < 0.01] (Table III). HFD feeding induced an increase in body weight in both genders [F(1,57) = 57.04; p < 0.001], although male rats showed greater resistance to becoming obese than females (the excess of body weight was 51.8% for females and 22.8% for males compared with their control counterparts). Moreover, in response to the dietary treatment, females showed a greater increase in white fat depot weight than males [F(1,55) = 67.15; p < 0.001], although this parameter remained lower in females than in males [F(1,55) = 15.71; p < 0.001]. By contrast, relative white fat depot weight also increased in both genders with the dietary treatment [F(1,54) = 96.69; p < 0.001], but reached higher levels in females than in males [interaction, F(1,54) = 11.12; p < 0.01; post hoc, p < 0.05]. BAT weight was greater in control males rats than in females [F(1,58) = 9.748; p < 0.01], but when this parameter was adjusted to body weight, it was higher in females than in males [F(1,57) = 16.09; p < 0.001]. In response to HFD feeding, BAT weight increased in a similar way in both genders [F(1,57) = 73.46; p < 0.001].

Table III.  Effects of HFD on body weight gain, adipose tissue weight, and energy intake.

	Males		Females
	Control	HFD	Control	HFD	ANOVA
Energy intake (kJ/kg BW^0.75· day)	603 ± 24	1104 ± 101	972 ± 90	1121 ± 129	D
BW (g)	555 ± 12	671 ± 23	296 ± 6	450 ± 23	G,D
BW gain (g) WAT weight (g)	225 ± 11 57.2 ± 5.3	336 ± 22 106 ± 0.8	83.9 ± 5.0 23.0 ± 1.8	230 ± 21 86.2 ± 9.5	G,D G,D
(g/100g BW) BAT weight (g)	10.1 ± 0.8 0.714 ± 0.047	15.6 ± 0.6^* 1.37 ± 0.11	7.61 ± 0.52^† 0.466 ± 0.025	18.6 ± 1.2^*^,† 1.15 ± 0.08	D,G*D G,D
(g/100g BW)	0.128 ± 0.007	0.201 ± 0.010	0.155 ± 0.007	0.256 ± 0.014	G,D

HFD, high fat diet; BW, body weight; WAT, white adipose tissue; BAT, brown adipose tissue. WAT weight is the sum of inguinal, gonadal, mesenteric, and lumbar depot weights. Energy intake values are means ± SEM (n = 5–7) and other values are means ± SEM (n = 15). ANOVA (p < 0.05): G indicates gender effect, D indicates diet effect, and G*D indicates gender and diet interactive effect. Student's t-test (p < 0.05); * HFD vs. control; ^† females vs. males.

BAT composition

Interactive effects between gender and diet were found in DNA levels [interaction, F(1,49) = 7.494; p < 0.01], total protein content [interaction, F(1,48) = 15.09; p < 0.001], and mitochondrial protein content [interaction, F(1,58) = 9.282; p < 0.05]. DNA [post hoc, p < 0.01] and both total [post hoc, p < 0.001] and mitochondrial protein levels [post hoc, p < 0.001] were higher in the BAT of control female rats than in that of males (Table IV). In response to HFD feeding, DNA levels decreased only in female rats [post hoc, p < 0.01], whereas both total and mitochondrial protein contents increased in males [post hoc, p < 0.05 and p < 0.001, respectively]. Moreover, total protein levels decreased in female rats [post hoc, p = 0.001]. Mitochondrial protein relative to DNA (i.e. per cell) increased in response to the dietary treatment in both genders [F(1,55) = 13.29; p = 0.001], but total protein content per cell did not show diet effects.

Table IV.  Effects of HFD on BAT composition.

	Males		Females
	Control	HFD	Control	HFD	ANOVA
DNA
(mg/g tissue)	4.70 ± 0.30	4.78 ± 0.28	6.52 ± 0.46^†	4.65 ± 0.30^*	G,D,G*D
Protein
(mg/g tissue)	41.5 ± 2.6	50.6 ± 3.8^*	77.8 ± 4.1^†	58.8 ± 3.8^*	G,D,G*D
(mg/mg DNA)	8.99 ± 0.79	11.0 ± 1.0	12.1 ± 0.6	13.5 ± 0.9	G
mtProtein
(mg/g tissue)	11.0 ± 1.5	21.9 ± 2.5^*	27.0 ± 1.6^†	25.9 ± 1.9	G,D,G*D
(mg/mg DNA)	2.38 ± 0.44	4.55 ± 0.57	4.39 ± 0.38	5.63 ± 0.39	G,D

HFD, high fat diet; mtProtein, mitochondrial protein. Values are means ± SEM (n = 15). ANOVA (p < 0.05): G indicates gender effect, D indicates diet effect, and G*D indicates gender and diet interactive effect. Student's t-test (p < 0.05); * HFD vs. control; ^† females vs. males.

BAT mitochondrial oxygen consumption and oxidative enzyme activities

In comparison with control male rats, control females showed higher levels of COX activity [interaction, F(1,24) = 4.715; p < 0.05; post hoc, p < 0.01] and CS activity [interaction, F(1,20) = 6.728; p < 0.05; post hoc, p < 0.001] (Table V). In response to HFD consumption, CS activity increased only in male rats [post hoc, p < 0.05]. Mitochondrial oxygen consumption measured in the presence of G3P as substrate decreased with obesity in both genders [F(1,24) = 10.74; p < 0.01] (Figure 1). In the presence of PCM, mitochondrial oxygen consumption showed a gender effect, because it was lower in male rats than in females [F(1,25) = 9.877; p < 0.01].

Table V.  Effects of HFD on BAT COX and CS activities.

	Males		Females
	Control	HFD	Control	HFD	ANOVA
COX (A.U./g tissue)	273 ± 46	378 ± 46	460 ± 40^†	373 ± 43	G*D
CS (I.U./g tissue)	34.9 ± 7.7	62.3 ± 9.7^*	104 ± 9^†	79.9 ± 11.7	G,G*D

HFD, high fat diet; COX, cytochrome c oxidase; CS, citrate synthase; A.U., arbitrary units; I.U., international units. Enzymatic activities were measured in isolated mitochondria. Values are means ± SEM (n = 5–7). ANOVA (p < 0.05): G indicates gender effect, and G*D indicates gender and diet interactive effect. Student's t-test (p < 0.05); * HFD vs. control; ^† females vs. males.

Figure 1.  Effects of HFD on BAT mitochondrial oxygen consumption. HFD, high fat diet. O₂ consumption rates were measured in isolated mitochondria using glyceraldehyde 3-phosphate (G3P) or palmitoyl carnitine-malate (PCM) as substrates. Values are means ± SEM (n = 5–7). ANOVA (p < 0.05): G indicates gender effect, and D indicates diet effect.

BAT antioxidant enzyme activities and glutathione levels

Control female rats showed both higher levels of total glutathione [interaction, F(1,15) = 28.162; p < 0.001; post hoc, p < 0.001] and reduced glutathione [interaction, F(1,15) = 36.351; p < 0.001; post hoc, p < 0.001], which decreased with the HFD feeding only in females [post hoc, p = 0.001 for total glutathione and post hoc, p < 0.001 for reduced glutathione], reaching the levels of males (Table VI). GPx activity was also higher in control female rats compared to control male rats [interaction, F(1,22) = 4.773; p < 0.05; post hoc, p = 0.001] and augmented in both genders with the HFD feeding, but this increase was 235% in obese male rats [post hoc, p < 0.01] and only 44.6% in obese females [post hoc, p < 0.05], thus leading to similar levels in HFD male and female rats. With the dietary treatment, Mn-SOD activity increased only in females [interaction, F(1,22) = 6.000; p < 0.05; post hoc, p < 0.05]. Actually, in HFD females, Mn-SOD activity was almost double that in HFD males [post hoc, p = 0.001]. In response to the HFD feeding, the ratio of GPx/SOD activities, which is an index of cellular H₂O₂ metabolism, increased only in male rats (threefold compared to their control counterparts) [interaction, F(1,16) = 7.171; p < 0.05; post hoc, p < 0.01].

Table VI.  Effects of HFD on BAT GPx and SOD activities and glutathione levels.

	Males		Females
	Control	HFD	Control	HFD	ANOVA
GSH+GSSG (nmol/g tissue)	478 ± 37	585 ± 47	1082 ± 37^†	689 ± 62^*	G, D,G*D
GSH (nmol/g tissue)	409 ± 45	503 ± 51	990 ± 42^†	569 ± 32^*	G, D,G*D
GPx (I.U./g tissue)	0.755 ± 0.116	2.53 ± 0.44^*	1.39 ± 0.08^†	2.01 ± 0.21^*	D,G*D
Mn-SOD (mI.U./g tissue)	926 ± 163	764 ± 112	870 ± 162	1370 ± 38^*^,†	G*D
GPx/Mn-SOD	1.00 ± 0.11	3.15 ± 0.74^*	1.72 ± 0.33	1.68 ± 0.27	D,G*D

HFD, high fat diet; GSH, glutathione; GSSG, glutathione disulfide; GPx, glutathione peroxidase; Mn-SOD, manganese superoxide dismutase; I.U., international units. Enzymatic activities were measured in isolated mitochondria and glutathione levels were measured in homogenate. Values are means ± SEM (n = 5–7). ANOVA (p < 0.05): G indicates gender effect, D indicates diet effect, and G*D indicates gender and diet interactive effect. Student's t-test (p < 0.05); * HFD vs. control; ^† females vs. males. GPx/Mn-SOD levels of control male rats were set as 1.

BAT mitochondrial H₂O₂ production and oxidative damage

Mitochondrial H₂O₂ production was higher in control female rats than in control male rats, in the presence of G3P [interaction, F(1,25) = 12.51; p < 0.01; post hoc, p < 0.001] and in the presence of PCM [interaction, F(1,24) = 4.881; p < 0.05; post hoc, p < 0.01] (Table VII). With the consumption of the HFD, H₂O₂ production increased in male rats [post hoc, p < 0.01] but decreased in females [post hoc, p < 0.05], using G3P as substrate. Control female rats showed lower protein carbonyl groups than control male rats [interaction, F(1,22) = 17.46; p = 0.001; post hoc, p < 0.05]. In response to obesity induced by HFD feeding, protein oxidation decreased in males [post hoc, p < 0.01] and increased in females [post hoc, p < 0.05], reaching higher levels than HFD males [post hoc, p < 0.001]. Lipid peroxide levels were lower in control female rats than in males [interaction, F(1,20) = 8.986; p < 0.01; post hoc, p < 0.001], and the HFD intake increased these levels only in females [post hoc, p < 0.01].

Table VII.  Effects of HFD on BAT mitochondrial hydrogen peroxide production, lipid peroxide levels, and protein carbonyl groups.

	Males		Females
	Control	HFD	Control	HFD	ANOVA
H2O2 production (nmol/g tissue·min)
G3P	39.4 ± 4.9	75.2 ± 10.7^*	133 ± 15^†	86.4 ± 14.1^*	G,G*D
PCM	10.1 ± 1.2	15.1 ± 3.4	30.8 ± 5.6^†	18.6 ± 2.5	G,G*D
TBARS (nmol/g tissue)	358 ± 10	348 ± 9	262 ± 7^†	323 ± 16^*	G,D,G*D
Protein carbonyl groups (%)	100 ± 12	54.7 ± 2.6^*	64.9 ± 5.0^†	79.1 ± 4.3^*^,†	G,G*D

HFD, high fat diet; H₂O₂, hydrogen peroxide; G3P, glyceraldehyde 3-phosphate; PCM, palmitoyl carnitine-malate. Values are means ± SEM (n = 5–7). ANOVA (p < 0.05): G indicates gender effect, D indicates diet effect, NS stands for non-significant, and G*D indicates gender and diet interactive effect. Student's t-test (p < 0.05); * HFD vs. control; ^† females vs. males.

BAT UCP1 protein content

UCP1 protein content relative to mitochondrial protein increased in response to HFD feeding in both genders [F(1,19) = 23.31; p < 0.001] (Figure 2). UCP1 protein concentrations per gram of BAT were higher in control females than in control males [interaction, F(1,19) = 17.27; p = 0.001; post hoc, p < 0.01] and increased with the HFD feeding only in males (353% compared with their control counterparts) [post hoc, p < 0.001].

Figure 2.  Effects of HFD on BAT UCP1 protein levels. HFD, high fat diet; UCP1, uncoupling protein 1; mtProtein, mitochondrial protein. Values are means ± SEM (n = 5–7). ANOVA (p < 0.05): D indicates diet effect, and G*D indicates gender and diet interactive effect. Student's t-test (p < 0.05): ^aHFD vs. control; ^bfemales vs. males. Levels of control male rats were set as 100%.

Discussion

In this study, we determined whether there is a relationship between the redox status of BAT and the predisposition to becoming obese in male and female rats. In the control situation, female rats showed lower insulinemia and triglyceridemia, greater BAT oxidative and antioxidant capacity, and lower oxidative damage than males. This sexual dimorphism was attenuated by the effect of obesity, given that the insulin and lipid profile was impaired only in females, whereas BAT oxidative capacity tended to decrease in female rats and increase in males in response to HFD feeding. In addition, BAT of male rats seems better protected against oxidative damage induced by HFD feeding than that of females, as indicated by the markers of lipid and protein oxidation and the content of reduced glutathione.

Gender effect

BAT of control female rats showed a higher H₂O₂ production in comparison with that of control males (more than 200%, independent of the substrate used), although no significant gender differences were found in specific H₂O₂ production (nmol H₂O₂/mg mtProtein·min) under any of the conditions used in the experiment (data not shown). These results indicate that respiratory chain efficiency in BAT is similar in males and females, but female rats contain more mitochondrial machinery, consume more oxygen, and produce elevated mitochondrial H₂O₂. Moreover, the greater amount of mitochondrial protein and the elevated levels of both COX and CS activities shown by female rats can be related to the enhanced mitochondrial biogenesis process previously reported in BAT in females (Justo et al. 2005).

It is interesting to note that the greater H₂O₂ production shown by BAT of females did not lead to proportional levels of oxidative damage; indeed, this was lower in female rats than in males. These results could be explained by the enhanced antioxidant capacity found in female rats, which agrees with other studies performed in different tissues (Colom et al. 2007; Valle et al. 2007; Catala-Niell et al. 2008; Gomez-Perez et al. 2008).

In female rats, the higher mitochondrial oxidative capacity was accompanied by increased levels of UCP1 protein, which point to enhanced BAT recruitment in the sense of increased thermogenic potential which could be attributed to the greater sensitivity to norepinephrine signals found in this gender (Rodriguez-Cuenca et al. 2002). Accordingly, the threshold temperature for cold-induced thermogenesis is higher in female rats than in males (Quevedo et al. 1998). Taking into account of previous studies pointing to mitochondria as an important target for the actions of estrogens (Chen et al. 2009), sex hormones could play an important role in gender differences in BAT functionality. Indeed, ovariectomy in rats leads to an impairment of BAT UCP1 expression and to an increase of body weight gain, which is normalized by estrogen treatment (Richard 1986; Pedersen et al. 2001).

Diet effect

The gender-related differences in BAT redox status shown by control rats were attenuated in obese animals. In response to the chronic dietary treatment, mitochondrial H₂O₂ production increased in male rats and decreased in females but, surprisingly, oxidative damage showed the opposite profile. This apparent contradiction could be explained by the gender differences observed in the ratio of GPx/SOD activities, which can be taken as an index of cellular H₂O₂ metabolism (Galinier et al. 2006b). The high ratio of GPx/SOD activities shown by HFD male rats suggests that the H₂O₂ generated by SOD could be destroyed by GPx. In contrast, this ratio was not altered by the dietary treatment in female rats, and the increased BAT oxidative damage could result from the imbalance between SOD and GPx activities. In female rats, GPx activity would not be sufficient to eliminate the levels of the H₂O₂ potentially produced by SOD activity, causing oxidative damage. From the present results, it seems that, despite chronic HFD feeding, BAT from male rats is better preserved from oxidative stress than that from females, owing to its greater capacity to eliminate the H₂O₂ produced by the oxidative metabolism.

The sexual dimorphism in the effect of BAT redox status on obesity could be related to the different fate of BAT hypertrophy under HFD feeding in each gender. On the one hand, BAT of male rats maintains its specific key role in energy homeostasis. Both total and mitochondrial protein content and COX activity increased in HFD male rats, pointing to an enhanced mitochondrial function in response to the dietary treatment. Thus, male rats would increase H₂O₂ production as a consequence of their increased oxidative metabolism. On the other hand, BAT hypertrophy in HFD female rats occurred with the opposite profile. Total protein content decreased, and mitochondrial protein content as well as COX and CS activities also showed a trend to decrease, indicating a diminished functionality of BAT in energy homeostasis regulation. Moreover, this loss of oxidative capacity exhibited by HFD females could be related to their lower levels of H₂O₂ production. Moreover, our results do not indicate an involvement of UCP1 in the prevention of ROS production. In male rats, the increased ROS production occurred together with higher levels of UCP1, whereas ROS production decreased in females, and UCP1 protein levels were not altered. Indeed, although other UCPs, such as UCP2 and UCP3, seem to be important in the redox status (Echtay 2007), the role of UCP1 as an antioxidant defense dissipating the high proton gradient is still under debate (Cannon et al. 2006; Fisler and Warden 2006; Shabalina et al. 2006).

In this study, the sexual dimorphism found in oxidative capacity in response to HFD feeding is in agreement with previous reports using shorter dietary treatments (Roca et al. 1999); however, it differs from other studies performed in female mice, which indicated that HFD feeding increases energy expenditure to defend against weight gain (Schmid et al. 2004). The reason for the discrepancy is unclear, although both the genuine interspecies difference and the substrate used in the experiments could be relevant. Indeed, our results show that mitochondrial oxygen consumption decreases when the substrate used is a carbohydrate (G3P), especially in female rats (48.7% in HFD females and 18.4% in HFD males), an effect that was not found in the presence of the lipid substrate (PCM). These differences could be a consequence of oxidative metabolism adaptations to the obese state. Furthermore, fat-rich diets lead to an impairment of glucose metabolism in rats (Pichon et al. 2006; Terauchi et al. 2007).

Energy expenditure in BAT has been reported to be closely linked to UCP1 protein levels, because this protein can dissipate energy as heat by uncoupling oxidative phosphorylation (Porter 2008). In our study, the greater resistance to becoming obese shown by male rats (body weight excess was 51.8% for females and only 22.8% for males compared with their control counterparts) could be related, at least in part, to the marked increase of BAT UCP1 protein content. In fact, UCP1 activity has been suggested to contribute to the regulation of body weight (Matamala et al. 1996; Inokuma et al. 2006; Catala-Niell et al. 2008) and may be decisive for obesity development in thermoneutral conditions (Feldmann et al. 2009). These findings, however, are difficult to reconcile with some studies showing that transgenic mice with a complete absence of UCP1 are not obese (Enerback et al. 1997; Liu et al. 2003). The existence of UCP1-independent, diet-inducible thermogenic mechanisms in BAT or target tissues other than BAT in mediating diet-induced thermogenesis has been proposed (Bachman et al. 2002). Thus, in addition to an enhancement of UCP2 expression, which has been found in UCP1 deficient mice (Enerback et al. 1997), the stimulation of UCP1-independent thermogenesis in WAT (Granneman et al. 2003) and in liver (Ma and Foster 1989) could play a relevant role.

Altogether, our results provide support for one of the main features of BAT mitochondria: a strong oxidative potential accompanied by a low phosphorylative capacity (Lindberg et al. 1967), which is more evident in male rats in response to the HFD. These gender-associated differences in the response to excess food can be associated with the wider margin for environmental adaptation of females. The reduced capacity of female rats to increase BAT UCP1 protein content would be related to their greater white adipose tissue expandability, which could be an advantage to handle female biological functions.

Recent studies describing the presence of BAT in adult humans (Saito et al. 2009; Virtanen et al. 2009; Zingaretti et al. 2009) could suggest that BAT functionality may have some relevance to humans. However, understanding the process that keeps brown adipocytes in a proliferative state is a key factor to extrapolate the results to humans. However, the metabolic significance of this tissue in adult humans is still questioned (Nedergaard and Cannon 2010).

In summary, chronic HFD feeding attenuated gender differences in BAT redox status. Only obese male rats showed an increase in BAT oxidative capacity and improved protection against oxidative stress. This gender dimorphism in BAT function could be related to the gender differences observed in body weight gain under chronic HFD feeding. The enhanced oxidative capacity shown by male rats resulted in a greater resistance to increasing their body weight, whereas the wider margin for HFD feeding adaptation of females is in accordance with their reproductive function and decreased BAT functionality, which would favor the increase of body weight in this gender. In future experiments, it would be interesting to study whether the sex-dependent effects of obesity on BAT redox status could be related to differences between genders in the mitochondrial biogenesis process, and the role the hormonal status could play in this sexual dimorphism.

Acknowledgements

This work was supported by Fondo de Investigaciones Sanitarias of the Spanish Government (PI060293) and by Conselleria d'Innovació i Energia of the Comunitat Autónoma de les Illes Balears (PROGECIB-1C). Nadal-Casellas was funded by a grant from the Comunitat Autònoma de les Illes Balears.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.