The prevalence of obesity, metabolic syndrome and type 2 diabetes mellitus is increasing throughout the world (Ford, Reference Ford2004; Grundy, Reference Grundy2004). There is increasing interest in dietary supplementation with conjugated linoleic acid (CLA), because it prevents obesity and improves insulin resistance (Tsuboyama-Kasaoka et al. Reference Tsuboyama-Kasaoka, Takahashi, Tanemura, Kim, Tange, Okuyama, Kasai, Ikemoto and Ezaki2000; Plum et al. Reference Plum, Giesen, Kluge, Junger, Linnartz, Schurmann, Becker and Joost2002; Larsen et al. Reference Larsen, Toubro and Astrup2003; Rainer & Heiss, Reference Rainer and Heiss2004). CLA is a group of positional and geometric conjugated dienoic isomers of linoleic acid (18 : 2n-6) that are present in dairy products and meat (Wang & Jones, Reference Wang and Jones2004). Although the cis-9, trans-11 (c9, t11)-CLA isomer is the principal dietary form of CLA, commercial CLA supplements usually contain large quantities of the trans-10, cis-12 (t10, c12)-CLA isomer (Wang & Jones, Reference Wang and Jones2004).
The results of studies in which rodents were fed various ratios of c9, t11-CLA and t10, c12-CLA isomers indicate that t10, c12-CLA has a greater effect on weight gain and fat deposition than c9, t11-CLA does (Tsuboyama-Kasaoka et al. Reference Tsuboyama-Kasaoka, Takahashi, Tanemura, Kim, Tange, Okuyama, Kasai, Ikemoto and Ezaki2000; Nagao et al. Reference Nagao, Wang, Inoue, Han, Buang, Noda, Kouda, Okamatsu and Yanagita2003b). Some human studies have shown that both CLA isomers do not improve insulin sensitivity in obese men with the metabolic syndrome and diabetic patients (Riserus et al. Reference Riserus, Arner, Brismar and Vessby2002; Moloney et al. Reference Moloney, Yeow, Mullen, Nolan and Roche2004). In contrast, a mixture of CLA improved insulin sensitivity in young, sedentary subjects (Eyjolfson et al. Reference Eyjolfson, Spriet and Dyck2004). Although human consumption of t10, c12-CLA is not advocated because of the possibility of insulin resistance, this adverse effect was not observed when CLA was included in diets containing high levels of fat (Ryder et al. Reference Ryder, Portocarrero and Song2001; Ealey et al. Reference Ealey, El-Sohemy and Archer2002). More studies are required to fully understand the mechanisms involved in the beneficial or deleterious effects of CLA isomers on insulin resistance.
Studies using normal rats and diets with 30–40 % energy from fat, which mimic the meal patterns of individuals in developed countries, have given much information to further understand the mechanisms by which high-fat (HF) feeding induces insulin resistance, and to develop preventive measures. Consumption of diets high in fat reduces the expression of insulin receptor substrate (IRS), activity of phosphatidyl inositol (PI) 3-kinase, and phosphorylation of Akt in the liver (Shulman, Reference Shulman2000; Lowell & Shulman, Reference Lowell and Shulman2005). To investigate the effects of CLA on insulin resistance of rats fed diets with HF contents, it might be necessary to measure the expression or activation of molecules related to insulin signalling events.
Recently, it was suggested that receptor binding of adiponectin regulates insulin sensitivity (Yamauchi et al. Reference Yamauchi, Kamon and Ito2003). Decreased expression of the adiponectin receptor (AdipoR) in insulin-resistant animals is correlated with decreased AMP kinase activation, PPARα activation, fatty acid oxidation and glucose uptake (You & Crabb, Reference You and Crabb2004; Kadowaki & Yamaguchi, Reference Kadowaki and Yamaguchi2005). However, there are few studies on the effects of CLA on AdipoR expression and serum adiponectin concentrations. Impaired mitochondrial function by oxidative stress may play a role in the pathogenesis of insulin resistance and type 2 diabetes (Shulman, Reference Shulman2000). Because CLA decreases blood lipid concentrations and has antioxidant properties (Palacios et al. Reference Palacios, Piergiacomi and Catala2003; Su et al. Reference Su, Liu, Kim, Jeong and Sok2003), it may protect mitochondria from oxidative stress, especially when insulin resistance is induced by diets with HF contents.
The objectives of the present study were to investigate the effects of an HF diet supplemented with CLA preparations on insulin resistance and fatty acid oxidation, and mitochondrial function in normal rats. We monitored proteins involved in energy metabolism, the antioxidant defence system and the respiratory chain complex in mitochondria to identify the mechanism by which CLA improves insulin resistance.
Materials and methods
Animals and diets
Male Sprague–Dawley rats were obtained from the Experimental Animal Resources Laboratories of the Korean Food and Drug Agency. They were housed in plastic cages in a room with a 12 h light–12 h dark cycle and maintained at 22 ± 1°C. After adaptation for 1 week, rats were randomly assigned to four groups of seven animals each.
For 8 weeks, the rats received either an unsupplemented HF diet or the HF diet supplemented with one of three CLA preparations. The HF diet contained 23 % fat (20·5 % beef tallow and 2·5 % maize oil), which accounted for about 45 % of the energy content of the diet. The CLA diets contained about 1 % of the respective CLA preparation, 19·5 % beef tallow and 2·5 % maize oil. Maize oil was included to provide essential fatty acids. The first CLA preparation, termed as CLA-mix, consisted of 31·3 % c9, t11-CLA, 36·7 % t10, c12-CLA, 17·3 % other isomers of CLA, and 14·8 % of other fatty acids. Thus, the diet fed to the CLA-mix group contained a 30:40 mix of the c9, t11-CLA isomer and t10, c12-CLA isomer, an approximately equal ratio of the two major CLA isomers. The second CLA preparation, referred to as c9, t11-CLA-mix, was an 80:20 mix of the two CLA isomers, and consisted of 76·5 % c9, t11-CLA, 17·2 % t10, c12-CLA and 6·3 % other isomers of CLA (c9, t11-CLA-mix group). The third CLA preparation, referred to as t10, c12-CLA-mix, was a 10:90 mix of the two CLA isomers, and consisted of 89·6 % t10, c12-CLA and 10·4 % c9, t11-CLA (t10, c12-CLA-mix). Supplying the pure form of each isomer would enable us interpret the results much simpler than supplying CLA mixtures. However, we used the CLA mixtures for two reasons. One is that we expect the CLA mix, one of the CLA mixtures, to give the synergic effects by both major CLA isomers. The other reason is availability of CLA preparations. The pure form of c9, t11-CLA or t10, c12-CLA cannot be easily obtained due to economical and technical reasons. Although we did not use pure forms of each isomer, the c9, t11-CLA mix or t10, c12-CLA mix may represent c9, t11-CLA and t10, c12-CLA respectively since each preparation contains a high percentage of either c9, t11-CLA or t10, c12-CLA. All four diets also contained 23 % casein, 40·7 % sucrose, 3 % Brewer's yeast, 4·8 % celluose, 4 % mineral mix, 1·2 % vitamin mix and 0·3 % methionine (Choi et al. Reference Choi, Jung, Park and Song2004). Feed and water were available ad libitum throughout the experiment. Food intake was recorded at the same time each day and body weight was recorded weekly.
Blood and tissue collection for biochemical analysis
After 8 weeks, the rats were fasted overnight, anaesthetised, and blood was obtained by heart puncture. Serum was collected and stored at − 20°C. Liver and leg skeletal muscle were collected, weighed, frozen in liquid N2 and stored at − 70°C until further processing. Mitochondrial and cytosolic fractions were prepared from liver and muscle by ultracentrifugation and stored at − 70°C. Serum glucose concentrations were determined using an enzymic assay (Asan Pharmaceutical, Yongin, Gyeonggido, Korea). ELISA were used for measurement of serum insulin concentrations (Rat Insulin Kit RPN2567, Amersham, Bucks, UK) and adiponectin concentrations (Mouse/Rat Adiponectin ELISA Kit, K1002–1; B-Bridge International, Inc., Sunnyvale, CA, USA). Fasting serum insulin and glucose concentrations were used to calculate insulin resistance from the homeostasis model assessment (HOMA) for insulin resistance: insulin (μU/ml) × glucose concentration (mmol/l)/22·5 (Matthews et al. Reference Matthews, Hosker and Rudenski1985). A high HOMA index denotes low insulin sensitivity, although it should be acknowledged that the HOMA model has not been validated for use in animal models (Wallace et al. Reference Wallace, Levy and Matthews2004). To assess insulin sensitivity, another derived index of insulin resistance was suggested, i.e. the revised quantitative insulin sensitivity check index (R-QUICKI) (1/log insulin (μU/ml)+log glucose (mg/dl)+log NEFA (mmol/l)) (Perseghin et al. Reference Perseghin, Caumo, Caloni, Testolin and Luzi2001; de Roos et al. Reference de Roos, Rucklidge and Reid2005).
Glucose metabolising enzyme activities and glycogen concentrations
The activity of glucose-6-phosphatase was assayed according to a previously described method (Baginski et al. Reference Baginski, Foa, Zak and Bergmeyer1974) with a slight modification. After the reaction with glucose-6-phosphatase, the liberated inorganic phosphate in a sample of supernatant fraction was determined using a reaction based on the molybdenum blue method (Phosphor B-Test Wako; Wako Pure Chemical Industries Ltd, Osaka, Japan). The activity of phosphoenolpyruvate carboxykinase was measured as previously described (Chang & Lane, Reference Chang and Lane1966) with a slight modification. One unit phosphoenolpyruvate carboxykinase was defined as the enzyme activity resulting in the formation of 1 μmol NAD+/min per mg protein. Hepatic glycogen content was analysed using a modification of a previously described method (Heidelberger et al. Reference Heidelberger, Msenberg and Hassid1954).
Activities of mitochondrial electron transport chain enzymes
We assessed mitochondrial function from the activities of succinate dehydrogenase (SDH) (complex II) and cytochrome c oxidase (COX) (complex IV). SDH activity was measured by kinetic analysis (Owen & Freer, Reference Owen and Freer1970). The activity of SDH was estimated colorimetrically at 600 nm from reduction of 2,6-dichloroindolphenol during oxidation of succinate. Activity was expressed as μmol 2,6-dichloroindolphenol reduced/min per mg protein. The activity of COX was determined by a previously described method (Wharton & Tzogaloff, Reference Wharton, Tzogaloff, Estabrook and Pullman1967). One unit of enzyme activity was defined as that which resulted in the oxidation of 1 μmol ferrocytochrome c (the reduced form of cytochrome c)/min per mg protein.
Activities of hepatic mitochondrial antioxidant enzymes and lipid peroxidation product
Activities of manganese-superoxide dismutase, glutathione peroxidase (GPx) and glutathione reductase (GR) and concentrations of GSH and malondialdehyde were measured using the following kits from Oxis Research (Portland, OR, USA): Bioxytech SOD-525, Bioxytech GPx-340, Bioxytech GR-340, Bioxytech GSH-340, and Bioxytech MDA-586, respectively.
Ribonucleic acid extraction and analysis of messenger ribonucleic acid expression
Total RNA was extracted from tissues using TRI reagent (Molecular Research, Cincinnati, OH, USA) according to the manufacturer's instructions. RNA expression of uncoupling protein (UCP) 2, UCP3 and acyl-CoA oxidase was quantified by using specific primer sets and RT-PCR methods as described previously (Choi et al. Reference Choi, Jung, Park and Song2004). Gel electrophoresis and ethidium bromide staining were used for quantification of PCR products. SDH and COXIII mRNA levels were quantified by real-time quantitative RT-PCR using SYBR green PCR reagents (Applied Biosystems, Foster City, CA, USA), and the ABI PRISM 7900 HT sequence-detection system (Applied Biosystems). The Ct values obtained were the threshold cycles at which a statistically significant increase in SYBR green emission intensity occurred. Data are expressed as 2− ΔΔCT values obtained by normalising to 18S ribosomal RNA and then the mean ΔCT values to the HF group. The PCR primers used were as follows: SDH, forward, 5′-GGAGGGGTCTCTCTTTTTGG-3′, and reverse, 5′-GACAGGCCTTTCCCTAGGTC-3′; COXIII, forward, 5′-AAGGCCACCACACCCTATT-3′, and reverse, 5′-AAATGCTCAGAAAAATCCGGC-3′; 18S rRNA, forward, 5′-GTCGTACCACTGGCATTGTG-3′, and reverse, 5′-CTCTCAGCTGTGGTGGTGAA-3′.
Immunoblot analysis
Equal amounts of whole lysate protein were separated by 5–10 % SDS-PAGE, transferred to polyvinylidene difluoride membranes, incubated in blocking buffer and treated with primary antibodies. Rabbit polyclonal antibodies against IRS-1, PPARα, PPARγ, PPARγ coactivator 1α and actin were obtained from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). Rabbit polyclonal antibodies against phospho-IRS-1-ser307, Akt, and phospho-Akt-ser473 were obtained from Cell Signaling Technology Inc. (Danvers, MA, USA). Rabbit polyclonal antibodies against AdipoR1 and AdipoR2 were purchased from α-Diagnostic Inc. (San Antonio, TX, USA). Appropriate secondary antibodies were used, and the bands were visualised using ECL Western blotting detection reagents (RPN2106; Amersham, Bucks, UK) and X-ray film (AGFA, Mortsel, Belgium). Tina 2.0 software (Silk Scientific Inc., Orem, UT, USA) was used for densitometric analysis of immunoreactive bands. Actin was determined for each blot to verify equal protein loading.
Statistical analysis
All data are expressed as mean values with their standard errors. Differences between the group means were analysed by one-way ANOVA using the SAS statistical analysis program (SAS Institute, Cary, NC, USA). Differences between means were considered statistically significant at P < 0·05. Duncan's multiple-range tests were used to determine the significance of differences in group means.
Results
Dietary conjugated linoleic acid improved insulin resistance
Although serum glucose concentrations were not significantly affected by CLA supplementation (Fig. 1 (A)), serum insulin levels were significantly (P < 0·05) decreased by the c9, t11-CLA-mix (Fig. 1 (B)). Furthermore, the degree of whole-body insulin resistance, assessed by serum insulin concentration, HOMA and R-QUICKI, was decreased in general by all three CLA preparations (P < 0·05) (Fig. 1 (C) and Fig. 1 (D)). All three CLA preparations reduced the value of HOMA, the index of insulin resistance. We also calculated another index for insulin sensitivity, R-QUICKI, by adding another parameter, NEFA. We previously reported modest decreases in the levels of serum NEFA and TG (Choi et al. Reference Choi, Jung, Park and Song2004). R-QUICKI values were significantly higher in all three CLA groups than the HF group (Fig. 1 (D)). CLA did not affect serum adiponectin levels significantly. The group mean of serum adiponectin for the HF, CLA-mix, c9, t11-CLA-mix and t10, c12-CLA-mix group rats was 7·18 (se 0·49), 9·49 (se 1·81), 8·06 (se 1·08) and 7·93 (se 0·83) ng/ml respectively. We previously reported that consumption of CLA for 8 weeks did not affect concentrations of TG and cholesterol in liver or skeletal muscle (Choi et al. Reference Choi, Jung, Park and Song2004).
Fig. 1 Effect of three different conjugated linoleic acid (CLA) preparations on (A) serum glucose, (B) serum insulin, (C) homeostasis model assessment (HOMA) and (D) revised quantitative insulin-sensitivity check index (R-QUICKI) in rats fed a diet containing a high level of fat. (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (
), cis-9, trans-11-CLA-mix diet; (
), trans-10, cis-12-CLA-mix diet. Data are expressed as mean values with their standard errors represented by vertical bars (n 4–5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA). For details of diets, see Materials and methods.
Dietary conjugated linoleic acid activated insulin signalling pathways and inhibited gluconeogenesis and glycogenolysis
In response to insulin, IRS-1 becomes tyrosine phosphorylated and then generates the major docking sites for the PI3-kinase. If the phosphorylation of IRS-1 on Ser307 is increased, the interaction between IRS-1 and the insulin receptor is markedly reduced. And the activity of PI3-kinase and then Akt phosphorylation is reduced (Gual et al. Reference Gual, Le Marchand-Brustel and Tanti2005; Lowell & Shulman, Reference Lowell and Shulman2005; Taniguchi et al. Reference Taniguchi, Kondo and Sajan2005). In the liver, all three CLA mixtures decreased IRS-1 serine phosphorylation (Fig. 2 (A) and Fig. 2 (B)). The decreases in IRS-1 serine phosphorylation led to the improvement of insulin signalling. The t10, c12-CLA-mix increased pAkt activation in the liver. In the muscle, all three CLA mixtures tended to decrease IRS-1 serine phosphorylation. Only the t10, c12-CLA-mix significantly decreased serine phosphorylation of IRS-1. It could be that liver is a more sensitive tissue than skeletal muscle in respect to the effect of CLA on insulin signalling.
Fig. 2 Effect of three different conjugated linoleic acid (CLA) preparations (CLA-mix; cis-9, trans-11 (c 9, t 11)-CLA-mix; trans-10, cis-12 (t 10, c 12)-CLA-mix; for details of diets, see Materials and Methods) on insulin signalling proteins and glycogen content in livers of rats fed a diet containing a high level of fat. (A) Western analysis of liver. Representative immunoblots of p-insulin receptor substrate (IRS)-1 (Ser 307), total IRS-1, pAkt (Ser 473) and total Akt in liver. HF, unsupplemented high-fat diet. (B) Densitometry. Relative densities of pIRS-1 (Ser 307), total IRS-1, pAkt (Ser 473) and total Akt. (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (), c 9, t 11-CLA-mix diet; (), t 10, c 12-CLA-mix diet. Data are expressed as mean values with their standard errors indicated by vertical bars (n 4–5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA). (C) Activities of glucose-6-phosphatase (G6Pase) and phosphoenolpyruvate carboxykinase (PEPCK). (D) Glycogen content. Mean values with unlike letters are significantly different (P < 0·05; ANOVA).
Although the activities of glucose-6-phosphatase and phosphoenolpyruvate carboxykinase did not differ between groups (Fig. 2 (C)), the content of glycogen in the liver was greater in all CLA-treated groups than in the HF group, and significantly (P < 0·05) greater in the CLA-mix and c9, t11-CLA-mix groups than in the HF group (Fig. 2 (D)). In muscle, serine phosphorylation of IRS-1 was significantly (P < 0·05) decreased in all CLA-supplemented groups but the phosphorylation of Akt was significantly activated by only the t10, c12-CLA-mix (Fig. 3).
Fig. 3 Effect of three different conjugated linoleic acid (CLA) preparations (CLA-mix; cis-9, trans-11 (c9, t11)-CLA-mix; trans-10, cis-12 (t10, c12)-CLA-mix; for details of diets, see Materials and Methods) on insulin signalling proteins in muscle of rats fed a diet containing a high level of fat. (A) Western analysis of muscle. Representative immunoblots of p-insulin receptor substrate (IRS)-1 (Ser 307), total IRS-1, pAkt (Ser 473) and total Akt in muscle. HF, unsupplemented high-fat diet. (B) Densitometry. Relative densities of pIRS-1 (Ser 307), total IRS-1, pAkt (Ser 473) and total Akt. (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (), c9, t11-CLA-mix diet; (), t10, c12-CLA-mix diet. Data are expressed as mean values with their standard errors indicated by vertical bars (n 4–5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA).
Dietary conjugated linoleic acid increased expression of adiponectin receptors
CLA supplementations did not change serum adiponectin concentrations, but we observed increases in AdipoR1 mRNA level by the c9, t11-CLA-mix and the t10, c12-CLA-mix and increases in AdipoR2 mRNA level by all three CLA mixtures (P < 0·05) (Fig. 4). The c9, t11-CLA-mix was the more potent of the two isomers for activation of AdipoR expression. We could not observe any modification of mRNA concentrations of AdipoR in muscle (Fig. 4).
Fig. 4 Effect of three different conjugated linoleic acid (CLA) preparations (CLA-mix; cis-9, trans-11 (c 9, t 11)-CLA-mix; trans-10, cis-12 (t 10, c 12)-CLA-mix; for details of diets, see Materials and Methods) on adiponectin receptor (AdipoR) 1 and AdipoR2 levels in liver and muscle of rats fed a diet containing a high level of fat. (A) Western analysis. Representative immunoblots of AdipoR1 and AdipoR2. HF, unsupplemented high-fat diet. (B) Densitometry. Relative densities of AdipoR1 and AdipoR2 in liver. (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (), c 9, t 11-CLA-mix diet; (), t 10, c 12-CLA-mix diet. Data are expressed as mean values with their standard errors indicated by vertical bars (n 5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA). (C) Densitometry. Relative densities of AdipoR1 and AdipoR2 in muscle.
Dietary conjugated linoleic acid influenced the activity of peroxisome proliferator-activated receptor-α
The effect of adiponectin on insulin sensitivity is mainly mediated by AdipoR and may result in increased activation of PPARα pathways and fatty acid oxidation (Kadowaki & Yamaguchi, Reference Kadowaki and Yamaguchi2005). Although expressions of PPARγ and PPARγ coactivator 1α in the liver and muscle were not changed by CLA treatments, PPARα expression in liver and muscle was significantly (P < 0·05) increased by the c9, t11-CLA-mix (Fig. 5 (A) and Fig. 5 (B)).
Fig. 5 Effect of three different conjugated linoleic acid (CLA) preparations (CLA-mix; cis-9, trans-11 (c 9, t 11)-CLA-mix; trans-10, cis-12 (t 10, c 12)-CLA-mix; for details of diets, see Materials and Methods) on PPARγ coactivator 1α (PGC-1α), PPARα and PPARγ protein levels and expression of downstream target genes in liver and muscle of rats fed a diet containing a high level of fat. (A) Western analysis. Representative immunoblots of PGC-1α, PPARα and PPARγ. HF, unsupplemented high-fat diet. (B) Densitometry. Relative densities of PGC-1α, PPARα and PPARγ. (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (), c 9, t 11-CLA-mix diet; (), t 10, c 12-CLA-mix diet. Data are expressed as mean values with their standard errors indicated by vertical bars (n 5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA). (C) RT-PCR. Relative changes in mRNA expression of acyl-CoA oxidase (ACO) and uncoupling protein (UCP) in liver and muscle. Mean values with unlike letters are significantly different (P < 0·05; ANOVA).
Dietary conjugated linoleic acid increased messenger ribonucleic acid of genes associated with fatty acid oxidation and energy dissipation
Expression of peroxisomal acyl-CoA oxidase and UCP2 and UCP3 are regulated by PPARα (Argyropoulos & Harper, Reference Argyropoulos and Harper2002). Among the three CLA mixtures, the c9, t11-CLA-mix was the most effective one to increase acyl-CoA oxidase mRNA level in the liver (Fig. 5 (C)).
Hepatic UCP2 mRNA level was increased by the CLA supplementations, but did not differ significantly between groups (Fig. 5 (C)). The muscle UCP2 mRNA level of the c9, t11-CLA-mix group (Fig. 5 (C)) was significantly (P < 0·05) greater than that of the HF group, and was greater than those of the t10, c12-CLA-mix and CLA-mix groups. The CLA-mix group and the c9, t11-CLA group had increased (P < 0·05) levels of UCP3 mRNA in the liver (Fig. 5 (C)). UCP3 mRNA levels in muscle tissue were not significantly different between groups (Fig. 5 (C)).
Dietary conjugated linoleic acid enhanced activities of mitochondrial respiratory chain enzymes
The c9, t11-CLA-mix and t10, c12-CLA-mix significantly (P < 0·05) increased hepatic expression of SDH mRNA relative to that of the HF diet. The c9, t11-CLA-mix also significantly (P < 0·05) increased SDH mRNA expression in muscle (Fig. 6 (A)). However, the enzymic activity of SDH in liver and muscle was not greatly increased by CLA isomers. All CLA mixtures tended to increase mRNA levels of COX III (P < 0·05) in the liver, but not in muscle (Fig. 6 (A)). COX activity in the liver was significantly (P < 0·05) greater in the c9, t11-CLA-mix group than in the HF group. All CLA preparations significantly (P < 0·05) increased COX activity in the muscle (Fig. 6 (B)).
Fig. 6 Effect of three different conjugated linoleic acid (CLA) preparations on enzymic activities of succinate dehydrogenase (SDH) and cytochrome c oxidase (COX) in liver and muscle of rats fed a diet containing a high level of fat. (A) Real-time RT-PCR. Relative changes in mRNA expression of SDH and COXIII. (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (), cis-9, trans-11-CLA-mix diet; (), trans-10, cis-12-CLA-mix diet. Data are expressed as mean values with their standard errors indicated by vertical bars (n 5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA). (B) Kinetic analysis. Enzyme activities of SDH and COX. Mean values with unlike letters are significantly different (P < 0·05; ANOVA). For details of diets, see Materials and Methods.
Dietary conjugated linoleic acid enhanced mitochondrial antioxidant capacities
Although all three CLA mixtures did not decrease the level of mitochondrial malondialdehyde, the c9, t11-CLA-mix significantly (P < 0·05) increased the level of GSH in the liver (Fig. 7 (A) and Fig. 7 (B)). The hepatic activities of manganese-superoxide dismutase, GPx, and GR were significantly greater in the c9, t11-CLA group than in the HF group (Fig. 7 (A)) (P < 0·05). The activity of manganese-superoxide dismutase in muscle mitochondria was also significantly (P < 0·05) greater in the c9, t11-CLA-mix group than in the HF group (Fig. 7 (B)). These data suggest that the c9, t11-CLA-mix improved mitochondrial function by increasing levels of GSH and activities of antioxidative enzymes.
Fig. 7 Effect of three different conjugated linoleic acid (CLA) preparations on mitochondrial antioxidant capacities in liver and muscle of rats fed a diet containing a high level of fat. Mitochondrial enzymic activities of manganese-superoxide dismutase (Mn-SOD), glutathione peroxidase (GPx) and glutathione reductase (GR) and levels of GSH and malondialdehyde (MDA) in liver (A) and muscle (B). (■), Unsupplemented high-fat diet; (□), CLA-mix diet; (), cis-9, trans-11-CLA-mix diet; (), trans-10, cis-12-CLA-mix diet. Data are expressed as mean values with their standard errors indicated by vertical bars (n 5). Mean values with unlike letters are significantly different (P < 0·05; ANOVA). For details of diets, see Materials and Methods.
Discussion
Dietary CLA has positive effects on body weight and fat deposition (Sisk et al. 2001). There is much interest in the potential use of CLA to alleviate insulin resistance. We observed that inclusion of CLA in the HF diet reduced insulin resistance by improving one or several aspects of insulin signalling, fatty acid oxidation and mitochondrial antioxidant capacity. Because the effects of CLA on insulin resistance may differ between isomers, we investigated the role of three different CLA preparations on insulin resistance in rats fed a diet with an HF content. Three commercially available CLA supplements contained different ratios of the c9, t11-CLA isomer and t10, c12-CLA isomer. CLA-mix, c9, t11-CLA-mix and t10, c12-CLA-mix contained the c9, t11-CLA and t10, c12-CLA isomers at the ratios of 30:40, 80:20 and 10:90 respectively. All our three CLA mixtures significantly improved insulin resistance based on the values of HOMA and R-QUICKI. There are several contradictory results of studies on the effects of CLA on insulin resistance. While many studies have shown harmful effects of CLA on insulin sensitivity in mice models, some beneficial effects can be observed in rat models (Houseknecht et al. Reference Houseknecht, Vanden Heuvel, Moya-Camarena, Portocarrero, Peck, Nickel and Belury1998; Ryder et al. Reference Ryder, Portocarrero and Song2001; Nagao et al. Reference Nagao, Inoue, Wang and Yanagita2003a; Wargent et al. Reference Wargent, Sennitt and Stocker2005; Poirier et al. Reference Poirier, Shapiro, Kim and Lazar2006). Since mice rapidly lost fat mass together with hepatomegaly, differential effects of CLA on insulin sensitivity could be explained partly by animal differences (Tsuboyama-Kasaoka et al. Reference Tsuboyama-Kasaoka, Takahashi, Tanemura, Kim, Tange, Okuyama, Kasai, Ikemoto and Ezaki2000). Based on the size of organisms, the effect of CLA on insulin resistance in the rat model might reflect the changes in man better than in the mouse model. The rat model has shown beneficial effects of CLA on insulin resistance. For example, in a Zucker diabetic fatty (ZDF) rat model for obesity and diabetes, providing a 50:50 CLA mix at 1·5 % diet improved glucose tolerance and insulin sensitivity, and decreased fasting glucose and insulin levels (Houseknecht et al. Reference Houseknecht, Vanden Heuvel, Moya-Camarena, Portocarrero, Peck, Nickel and Belury1998, Ryder et al. Reference Ryder, Portocarrero and Song2001). Nagao et al. (Reference Nagao, Inoue, Wang and Yanagita2003a) showed that CLA attenuated plasma glucose and insulin and prevented hyperinsulinaemia by enhancing plasma adiponectin levels and mRNA expression in white adipose tissue from Zucker diabetic fatty rats. However, in mouse models, the intake of t10, c12-CLA in ob/ob mice elevated serum glucose and insulin levels and induced insulin resistance (Roche et al. Reference Roche, Noone, Sewter, McBennett, Savage, Gibney, O'Rahilly and Vidal-Puig2002; Poirier et al. Reference Poirier, Shapiro, Kim and Lazar2006). In a recent study (Wargent et al. Reference Wargent, Sennitt and Stocker2005), supplementation of 1·5 % CLA mix or t10, c12-CLA-enriched CLA for 2 weeks elevated fasting glucose and insulin levels of genetically obese C57BL/6 lepob/lepob mice. However, when supplementation was continued for 10 weeks, CLA had beneficial effects on both glucose and insulin levels. The study suggests that although initially CLA may have negative effects on insulin resistance, long-term treatment with CLA could improve insulin sensitivity and glucose tolerance. Furthermore, the effects of CLA may be dependent on the fat content of the diet. In human studies, CLA may decrease insulin sensitivity in obese men (Riserus et al. Reference Riserus, Arner, Brismar and Vessby2002), but CLA improves insulin sensitivity in sedentary human subjects (Eyjolfson et al. Reference Eyjolfson, Spriet and Dyck2004). Therefore, experimental conditions such as the proportion of CLA in the diets, strain of animal, metabolic state of the subjects (normal v. obese and diabetic) and duration of CLA feeding could contribute to the different results between studies. More studies are required to fully understand the mechanisms involved in the beneficial or deleterious effects of CLA and the purified isomers on insulin resistance.
The effect of CLA on insulin signalling events has not been subject to extensive study. In a cancer cell, CLA inhibits the activation of PI3-kinase, Akt, cell growth and tumour growth (Cho et al. Reference Cho, Kim, Kim, Jung, Park, Lee, Tyner and Park2003). Chung et al. (Reference Chung, Brown, Provo, Hopkins and McIntosh2005) reported that t10, c12-CLA decreased the IRS-1 and Glut-4 content of adipose cells but had no effect on the phosphorylation of IRS-1 (Ser 307 or Tyr 891) or Akt (Ser 473), suggesting that CLA does not affect insulin signal transduction per se. However, Chung's results were derived from cell cultures under in vitro conditions. In the present in vivo experiment, all three CLA mixtures decreased serine phosphorylation of IRS-1 and tended to increase the phosphorylation of Akt in the liver. The t10, c12-CLA-mix was more effective than the c9, t11-CLA-mix in activating the insulin signalling pathway. An activation of Akt by phosphorylation led to the inactivation of glycogen synthase kinase-3. Inactivated glycogen synthase kinase-3 led to less phosphorylated glycogen synthase. Inhibiting the phosphorylation of glycogen synthase increases the synthesis of glycogen and decreases in hepatic glycogenolysis (Henriksen & Dokken, Reference Henriksen and Dokken2006).
Impaired hepatic glycogen storage and dysregulated glycogen synthesis is a critical feature of diabetes mellitus, as glycogen synthesis rates of diabetic patients are about 50 % of the values in healthy subjects (Bogardus et al. Reference Bogardus, Lillioja, Stone and Mott1984; Damsbo et al. 1991). In the present study, diets supplemented with the CLA-mix and the c9, t11-CLA-mix resulted in greater hepatic glycogen accumulation than the unsupplemented diet. This pattern, which mirrors that of Akt activation, suggests that CLA improves the insulin resistance induced by the HF diet. However, Akt in skeletal muscle was not activated by all three CLA mixtures. Consequently, we presume that CLA mixtures, especially the t10, c12-CLA-mix, activate IRS-1 in both liver and muscle. The downstream Akt activation was affected by the t10, c12-CLA-mix only in the liver.
Adiponectin promotes fatty acid oxidation in liver and muscle and inhibits hepatic glucose production (Yamauchi et al. Reference Yamauchi, Kamon and Ito2003; Kadowaki & Yamaguchi, Reference Kadowaki and Yamaguchi2005). Adiponectin may act through binding to its receptors in tissues, which may regulate insulin sensitivity. Two AdipoR, AdipoR1 and AdipoR2, were recently identified in many tissues, including liver and muscle (Kadowaki & Yamaguchi, Reference Kadowaki and Yamaguchi2005). In mice, AdipoR1 is ubiquitously expressed but is most abundant in skeletal muscle, while AdipoR2 is primarily expressed in the liver. In contrast, in rats and man, both receptors are highly expressed in muscle and liver. Expressions of AdipoR is decreased in insulin-resistant animals and is correlated with decreased adiponectin binding to membrane fractions, AMP kinase activation and acetyl CoA carboxylase phosphorylation in skeletal muscle and liver, activating fatty acid oxidation and glucose uptake in muscle and liver and inhibiting gluconeogenesis in liver. In addition, adiponectin increases the expression of PPARα and its target genes, resulting in reduced liver and muscle TG content (Yamauchi et al. Reference Yamauchi, Kamon and Ito2003; You & Crabb, Reference You and Crabb2004; Kadowaki & Yamaguchi, Reference Kadowaki and Yamaguchi2005). There are few studies of the effects of CLA on the expression and serum concentrations of adiponectin. Decreases in AdipoR expression impair the metabolic effects of adiponectin (Yamauchi et al. Reference Yamauchi, Kamon and Ito2003; Inukai et al. Reference Inukai, Nakashima, Watanabe, Takata, Sawa, Kurihara, Awata and Katayama2005). Beylot et al. (Reference Beylot, Pinteur and Peroni2006) also stated that along with the levels of adiponectin expression and plasma adiponectin, the levels of AdipoR expression could control insulin sensitivity. The expressions of AdipoR1 and R2 in ob/ob mice were significantly decreased in skeletal muscle and adipose tissue, which was correlated with decreased adiponectin binding to membrane fractions of skeletal muscle and decreased AMP kinase activation by adiponectin. This adiponectin resistance in turn may play a role in worsening insulin resistance in ob/ob mice (Tsuchida et al. Reference Tsuchida, Yamauchi and Ito2004).
In the present study, serum adiponectin concentrations were not affected by any CLA mixture, neither could we observe any modification of mRNA levels of AdipoR in muscle. Although the expression of AdipoR in rats is poorly responsive to changes in nutritional conditions contrary to what was reported in mice, HF diet feeding decreased AdipoR2 mRNA level in the liver of Wistar rats (Beylot et al. Reference Beylot, Pinteur and Peroni2006). We could also observe effects of CLA treatments on AdipoR mRNA levels in the liver. We observed the increases in AdipoR1 mRNA level by the c9, t11-CLA-mix and t10, c12-CLA-mix and the increases in AdipoR2 mRNA level by all three CLA mixtures. These changes suggest that the improved insulin sensitivity observed in the CLA groups may be related to the levels of hepatic AdipoR. Increased binding of AdipoR by CLA might affect the action of PPARα. CLA may increase fatty acid oxidation through increased hepatic AdipoR1 and AdipoR2 protein expression. Dietary c9, t11-CLA-mix increased protein and mRNA expression of PPARα in the liver and the expression of mRNA of acyl-CoA oxidase, and UCP-2 and UCP-3, enzymes that are involved in fat oxidation. It is probable that the divergent effects of CLA isomers on insulin and glucose metabolism reflect differences between the metabolic effects of c9, t11-CLA and t10, c12-CLA. We observed that the t10, c12-CLA-mix improved insulin signalling. c9, t11-CLA, in addition to improving glucose metabolism, improved lipid metabolism and resulted in less fat deposition through up regulation of hepatic PPARα, acyl-CoA oxidase and UCP. Up regulation of UCP is thought to prevent fat deposition by dissipating energy as heat (Argyropoulos & Harper, Reference Argyropoulos and Harper2002).
Dysfunctional mitochondria decrease energy production and contribute to insulin resistance (Wallace, Reference Wallace2001; Lossa et al. Reference Lossa, Lionetti, Mollica, Crescenzo, Botta, Barletta and Liverini2003; Petersen et al. Reference Petersen, Dufour, Befroy, Garcia and Shulman2004; Lowell & Shulman, Reference Lowell and Shulman2005; Wisloff et al. Reference Wisloff, Najjar and Ellingsen2005). We studied the effects of the HF diet and CLA supplementation on mitochondrial gene expression, respiratory chain complex activity and antioxidative status. Impairment of electron transport by the HF diet might have increased formation of reactive oxygen species (ROS) in mitochondria, depleted antioxidants, and impaired the flow of electrons, propagating an increasing rate of mitochondrial ROS formation (Lossa et al. Reference Lossa, Lionetti, Mollica, Crescenzo, Botta, Barletta and Liverini2003). CLA modulated HF-induced mitochondrial dysfunction in liver and muscle, resulting in greater energy production capacity with less production of ROS. The c9, t11-CLA-mix significantly increased both SDH and COXIII mRNA levels and increased the COX activity of liver and muscle compared with the CLA mix and t10, c12-CLA-mix. Among the three CLA mixtures, the c9, t11-CLA-mix resulted in the greatest improvement of COX activity in the liver. Mitochondria produce ROS during energy production, which makes them susceptible to oxidative damage. Because the c9, t11-CLA-mix increased mitochondrial activity, we examined how the three different CLA preparations affect overall antioxidant capacity. Results of studies on the effects of CLA on antioxidant status are contradictory (Palacios et al. Reference Palacios, Piergiacomi and Catala2003; Bergamo et al. Reference Bergamo, Luongo and Rossi2004; Yamasaki et al. Reference Yamasaki, Miyamoto, Chujo, Nishiyama, Tachibana and Yamada2005). However, CLA protected mitochondria isolated from rat liver from peroxidative damage and dietary CLA improved the antioxidant status of rats deficient in vitamin E (Palacios et al. 2004; Kim et al. Reference Kim, Kim, Ahn, Cho, Yoon and Ha2005). We found that the c9, t11-CLA-mix increased the activities of the mitochondrial antioxidant enzymes, manganese-superoxide dismutase, GPx, and GR, the level of GSH, and decreased malondialdehyde production in the mitochondria of liver and muscle. The increased expression of UCP induced by the c9, t11-CLA-mix in the present study may also promote the removal of ROS, because UCP are thought to reduce ROS accumulation (MacLellan et al. Reference MacLellan, Gerrits, Gowing, Smith, Wheeler and Harper2005).
In conclusion, the present findings suggest that the c9, t11-CLA-mix and t10, c12-CLA-mix activated the insulin signalling pathway in the liver by decreasing serine phosphorylation of IRS-1. The t10, c12-CLA-mix especially activated Akt. The c9, t11-CLA-mix also increased the sensitivity of AdipoR1 and AdipoR2 for adiponectin, resulting in an increase in PPARα ligand activity and fatty acid oxidation. The c9, t11-CLA-mix was the most effective in promoting energy production and thermogenesis because it activated the mitochondrial respiratory chain and fat oxidation, which resulted in less fat accumulation and improved whole-body insulin resistance. The c9, t11-CLA-mix also reduced the production of ROS by increasing antioxidant capacity.
Acknowledgements
The present study was supported in part by a Korea National Institute of Health intramural research grant (347-6111-211-207) and a grant from the Korea Health 21 R&D Project, Ministry of Health & Welfare, Republic of Korea (03-PJ1-PG3-22 000-0014).
