Conjugated linoleic acid (CLA) is a collective term for a group of isomers of linoleic acid that have conjugated double bonds. Depending on the position and geometry of the double bonds, several isomers of CLA have been reported (Eulitz et al. Reference Eulitz, Yurawecz, Sehat, Fritsche, Roach, Mossoba, Kramer, Adlof and Ku1999). Most of the published studies have used mixtures of CLA isomers, which comprised two major forms, cis 9, trans 11-CLA (c9, t11-CLA) and trans 10, cis 12-CLA (t10, c12-CLA), and a number of minor isomers. The major dietary sources of c9, t11-CLA are dairy products and ruminant meat, while that of t10, c12-CLA are partially hydrogenated vegetable oils from margarines and shortenings (McGuire et al. Reference McGuire, McGuire, Ritzenthaler, Scultz and Yurawecz1999).
Feeding a mixture of CLA isomers to animal models altered blood lipids, atherogenesis, diabetes, body composition, chemically induced carcinogenesis and immune cell functions (Belury, Reference Belury2002). Diets containing CLA reduced the amount of adipose fat in several species including rat, pig, hamster, chicken and mouse (Kelley & Erickson, Reference Kelley and Erickson2003; Tricon et al. Reference Tricon, Burdge, Williams, Calder and Yaqoob2005). The loss of adipose tissue in mice was associated with a several-fold increase in the amount of fat stored in the liver (Belury & Kempa-Steczko, Reference Belury and Kempa-Steczko1997; Park et al. Reference Park, Storkson, Albright, Liu and Pariza1999; Tsuboyama-Kasaoka et al. Reference Tsuboyama-Kasaoka, Takahashi, Tanemura, Kim, Tange, Okuyama, Kasai, Shinji and Ezaki2000; Clement et al. Reference Clement, Poirer and Noit2002; Degrace et al. Reference Degrace, Demizieux and Gresti2003; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004; Poirier et al. Reference Poirier, Rouault and Clement2005). We, as well as others, have previously reported that t10, c12-CLA is the isomer that is responsible for the development of fatty liver in mice (Park et al. Reference Park, Storkson, Albright, Liu and Pariza1999; Clement et al. Reference Clement, Poirer and Noit2002; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004). Decreased expression of adipocytokines and up regulation of the expression and activity of the lipogenic enzymes acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS), malic enzyme (ME), stearoyl CoA desaturase (SCD)-1 and δ 5 and 6 desaturases have been postulated to be the underlying mechanisms that lead to the development of fatty livers in mice fed diets containing CLA (Tsuboyama-Kasaoka et al. Reference Tsuboyama-Kasaoka, Takahashi, Tanemura, Kim, Tange, Okuyama, Kasai, Shinji and Ezaki2000; Degrace et al. Reference Degrace, Demizieux and Gresti2003, Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004; Takahashi et al. Reference Takahashi, Kushiro, Shinohara and Ide2003; Warren et al. Reference Warren, Simon, Bartolini, Erickson, Mackey and Kelley2003; Javadi et al. Reference Javadi, Beynen and Hovenier2004; Ide, Reference Ide2005; Poirier et al. Reference Poirier, Rouault and Clement2005). Expression of mice hepatic fatty acid oxidation genes also increased in three studies with a mixture of CLA isomers (Takahashi et al. Reference Takahashi, Kushiro, Shinohara and Ide2003; Javadi et al. Reference Javadi, Beynen and Hovenier2004; Ide, Reference Ide2005) and in one study with t10, c12-CLA (Degrace et al. Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004). Authors who supplemented the mouse diets with the purified t10, c12-CLA expected a reduction in hepatic fatty acid oxidation because CLA increased the hepatic concentration of malonyl CoA and the sensitivity of carnitine palmitoyl transferase (CPT)-1 to inhibition with malonyl CoA (Degrace et al. Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004). Development of fatty liver and increased hepatic fatty acid oxidation is a paradox that may be true, but normally we would expect hepatic fatty acid oxidation to be reduced if more fat is stored in the liver. None of the published reports has used the microarray technology to systematically examine all the hepatic genes involved in fatty acid metabolism, whose expression may be modulated by CLA-containing diets.
The purpose of the present study was to use a comprehensive approach to identify all the mouse liver genes involved in fatty acid oxidation (α,β and ω) and synthesis, whose expression may be altered by feeding diets containing CLA preparations enriched in one of the two common isomers (c9, t11-CLA and t10, c12-CLA). We used the Affymetrix Inc. (Santa Clara, CA, USA) microarray chips to identify the genes whose expression was altered by CLA. Changes in gene expressions detected by microarrays were confirmed by quantitative real-time PCR. The expressions of several genes involved in fatty acid oxidation and synthesis were altered by the feeding of t10, c12-CLA isomer. The largest reduction in gene expression was found for the two microsomal enzymes, cytochrome P450 (cyt P450) and flavin-containing monooxygenase (FMO)-3; these enzymes are involved in microsomal ω oxidation of fatty acids and the detoxification of a variety of xenobiotic compounds (White et al. Reference White, Handler, Smith, Hill and Lehman1978; Falls et al. Reference Falls, Cherrington and Clements1997; Orellana et al. Reference Orellana, Rodrigo and Valdes1998; Reddy & Hashimoto, Reference Reddy and Hashimoto2001; Krueger & Williams, Reference Krueger and Williams2005; Sanders et al. Reference Sanders, Ofman, Valianpour, Kemp and Wanders2005; Weng et al. Reference Weng, DiRusso, Reilly, Blacks and Ding2005). Under normal conditions microsomal fatty acid oxidation represents less than 10 % total fatty acid oxidation; however, during starvation and diabetes the contribution of this pathway in overall hepatic fatty acid oxidation is significantly increased (Orellana et al. Reference Orellana, Rodrigo and Valdes1998). Microsomal fatty acid oxidation may have a significant role in fatty acid oxidation in mice fed diets containing t10, c12-CLA, because this isomer produces diabetes-like symptoms of increased blood glucose and insulin resistance (Poirier et al. Reference Poirier, Rouault and Clement2005). Expression of cyt P450 and FMO3 is altered by several compounds that induce non-alcoholic fatty liver (Krueger & Williams, Reference Krueger and Williams2005) and FMO3 is the major isoform of FMO found in human liver (Falls et al. Reference Falls, Cherrington and Clements1997). The affect of CLA on the expression of these genes has not been previously published. We, therefore, also investigated the combined enzymatic activity of all the hepatic microsomal FMO and amount of the FMO3 specific protein. The present results show that diets containing the t10, c12-CLA reduced FMO activity, FMO3 specific protein and the transcription of several genes involved in fatty acid oxidation.
Materials and methods
Conjugated linoleic acid isomers and diets
Highly enriched c9, t11-CLA and t10, c12-CLA isomers in the form of NEFA were a kind gift from Natural ASA, Hovdebygda, Norway. The analytical data for these isomers were provided by the supplier and confirmed by GLC in our laboratory (Warren et al. Reference Warren, Simon, Bartolini, Erickson, Mackey and Kelley2003; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004). The preparation enriched in c9, t11-CLA contained (%): c9, t11-CLA 84·6; t10, c12-CLA 7·7; 18: 1n-9 3·8; t9, t11-CLA+t10, t12-CLA 2·0; other fatty acids 1·9. In the preparation enriched in t10, c12-CLA, this isomer was 88·1 %; c9, t11-CLA 6·6 %; t9, t11-CLA+t10, t12-CLA 2·5 %; 18 : 1n-9 1·1 %; other fatty acids 1·7 %.
The concentration of CLA used in the present study was 0·5 weight % of the diet, which is comparable to the concentrations used in previous studies with rodent models, which ranged from 0·1 to 1·5 weight % of a mixture of CLA isomers. AIN-93G, high carbohydrate, mouse diet was used as the basal diet. The nutrient and fatty acid composition of this diet has been previously reported (Warren et al. Reference Warren, Simon, Bartolini, Erickson, Mackey and Kelley2003; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004) and is shown in Table 1. For the two CLA-containing diets, CLA isomer-enriched oils were added by replacing 5 g/kg maize oil with an equivalent amount of the CLA source. Diets were constantly flushed with N gas while being gently mixed in a blender. Diets were packaged in 30 g aliquots, flushed with N gas and stored at − 20°C. Fresh dietary packets were served each day. The animal protocol was approved by the Animal Use Committee at the University of California, Davis.
Table 1 Composition of the basal diet*
c9,t11-CLA, cis 9, trans 11-conjugated linoleic acid; t10,c12-CLA, trans 10, cis 12-conjugated linoleic acid.
* For details of diets and procedures, see p. 59.
Animals, feeding and tissue collection
Eighteen, 8 week old, pathogen free C57BL/6N female mice were purchased from Charles River (Raleigh, NC, USA). Female mice were chosen because of their docility for housing in groups. They were maintained in a sterile air curtain isolator at the animal facility of the University of California, Medical School with controlled temperature (25°C) and light and dark cycle (12 h each). All animals were fed the laboratory chow diet for the first 7 d, then divided into three groups of six each and fed the experimental diets for 56 d. The dose of CLA and the duration of its feeding used in the present study are the same as we have used previously (Warren et al. Reference Warren, Simon, Bartolini, Erickson, Mackey and Kelley2003; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004), which are well within the ranges used by many other investigators. Details regarding animal handling, killing, tissue collection and storage have been published (Warren et al. Reference Warren, Simon, Bartolini, Erickson, Mackey and Kelley2003; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004).
Real-time PCR
Total RNA from approximately 100 mg liver slices was extracted by using Trizol reagent (Invitrogen, Carlsbad, CA, USA). This RNA (1 μg) was denatured and used to synthesize cDNA by using an Invitrogen pre-amplication kit. After the first strand cDNA synthesis the RNA templates were degraded by treatment with RNase H. Specific primers for different enzymes were designed based on published full-length cDNA sequences (Table 2). The PCR reactions were performed in a programmable thermal cycler (denaturation at 94°C for 3 min followed by 40 cycles, denaturation at 94°C for 30 s, annealing at 56°C for 30 s and extension at 72°C for 30 s followed by final extension of 72°C for 7 min). The PCR products were analysed by electrophoresis in 3 % agarose gels and stained with ethidium bromide. The amplicons were cloned into PCR2.1 plasmid vector (Invitrogen) and transformed into Escherichia coli competent cells by heat-shock. Cells were grown in Luria broth (LB) medium for 16 h. White colonies were picked and grown and plasmid DNA from these transformed colonies were isolated and analysed for presence of the genes inserted. The insertions were verified by digestion with EcoRI restriction enzyme. These plasmids were sequenced using M13 reverse and M13 forward primers. The purified plasmids were serially diluted and used to generate the standard curve.
Table 2 Sequence of primers used for quantitative real-time PCR analysis*
FMO3, flavin monooxygenase 3; Cyt P450, cytochrome P450; CPT1, carnitine palmitoyl transferase 1; ACOX, acetyl CoA oxidase; ACC, acetyl CoA carboxylase; FAS, fatty acid synthase; ME, malic enzyme; SCD1, steoryl CoA desaturase. Hprt, hypoxanthine guanine phosphoribosyl transferase.
* For details of procedures, see pp. 59–60.
Real-time quantitative PCR was performed using a LightCycler rapid thermal cycler system (Roche Diagnostics Ltd, Palo Alto, CA, USA) according to the manufacturer's instructions. β Actin and hypoxanthine-guanine phosphoribosyl transferase 1 was used as the housekeeping gene and results for gene expression determined by real-time PCR are expressed as ratios between the RNA for the gene of interest and that for β actin and hypoxanthine-guanine phosphoribosyl transferase 1. Reactions were performed in a 20 μl volume with 0·5 nm primers and 4 nm-MgCl2, dNTP, Taq DNA polymerase and buffer. The following programme was used to carry out the reaction: 30 s denaturation step (94°C) followed by 40 cycles with a 95°C denaturation for 1 s; annealing for 5 s at 56°C; extension at 72°C for 17 s. To confirm specific amplification, the PCR products from each primer pair were subjected to a melting curve. The melting curve was determined by holding the reaction at 55°C for 10 s and then heating slowly to 94°C with a linear rate of 0·2°C/s while the fluorescence emitted was measured. Melting curves were generated by plotting fluorescence against temperature. All assays were carried out in duplicate. Melting curve analysis demonstrated that each of the primer pair described amplified a single product with a distinct melting temperature. The predicted length of each product was confirmed by agarose gel electrophoresis.
DNA microarray analysis
Because of the high cost of the microarray chips, this analysis was performed only on two animals per group; however, the real-time PCR were performed on six animals per group. We performed DNA microarray analysis using Affymetrix Mouse GeneChips (430A 2.0 Array, representing 14 000 genes) to determine the liver genes whose expression was altered by the diets containing CLA. Total RNA was extracted from the livers of mice fed control and CLA-containing diets with the TRIzol reagent; its quality and integrity were determined utilizing an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) and absorbance at A260/A280. Only high quality RNA, having a 28S/18S rRNA ratio of 1·5–2·0 and an A260/280 absorbance ratio of 2, was utilized for further experimentation. It was further purified with RNAeasy silica columns (Qiagen, Valencia, CA, USA). RNA was converted to double-stranded cDNA, which was then converted to biotin-labelled cRNA by in vitro transcription labelling with a HighYieldTM BioArrayTM RNA Transcript Labeling Kit (Enzo Diagnostics, Farmingdale, NY, USA). The quality of in vitro transcription and fragmentation products was assessed using the Agilent 2100 Bioanalyzer. Fragmented, biotin-labelled cRNA (15 fg) was hybridized at 45°C overnight as defined in the Affymetrix7 expression analysis protocol. The hybridization buffer contained 100 mm-2-(4-morpholino) ethanesulfonic acid (MES), 1 m-NaCl, 20 mm-EDTA, 0·01 % Tween 20, four eukaryotic hybridization controls (1·5 pm-BioB; 5 pm-BioC; 25 pm-BioD; 100 pm-cre), 0·1 mg/ml herring sperm DNA (Promega, Madison, WI, USA) and 0·5 mg/ml acetylated bovine serum albumin. After hybridization, the arrays were washed and stained with an Affymetrix® fluidics station following the Antibody Amplification Washing and Staining Protocol (Affymetrix Inc.). Hybridization was detected with streptavidin-phycoerythrin and a confocal laser scanner (Affymetrix Inc.).
Microarray Suite 5·0 (Affymetrix Inc.) was used to determine the probe intensities and to compare expression amongst different arrays. Scaling to a target median intensity value of 125 normalized average intensity for each array. The gene expression values were log transformed (log base 2). Genes were ranked on t test scores, P-values (P < 0·05) and fold changes computed as actual expression values. A particular transcript was considered significantly differentially expressed between the control and CLA groups if it had a fold change >2, a P-value < 0·05, and this was observed in both independent experiments (described earlier). The cross-reference of the differentially expressed genes was performed using information from the Affymetrix and National Center for Biotechnology Information EntrezGene websites.
Determination of flavin-containing monooxygenase activity
Liver tissues of experimental mice were collected, frozen in liquid N and stored at − 80°C until used. The liver samples were homogenized in cold Tris/KCl buffer (0·05 m/0·15 m, pH 7·4), using glass/Teflon homogenizers, and microsomal fractions were purified (Chung & Buhler, Reference Chung and Buhler1994). Total catalytic activities of all the liver FMO were determined by oxidation of methyl p-tolyl sulfide to methyl p-tolyl sulfoxide (Cashman & Proudfoot, Reference Cashman and Proudfoot1988). Reaction mixture contained 0·05 m-glycine buffer (pH 9·5), 50 μg microsomal protein, 0·065 mm-NADP+, 3·3 mm-glucose-6-phosphate, 0·4 unit/ml glucose-6-phosphate dehydrogenase, 3·3 mm-MgCl2 and 2·0 mm-methyl p-tolyl sulfide in a total volume of 0·25 ml. After 10 min incubation at 37°C, the reaction was stopped with 75 μl acetonitrile and centrifuged (10 000 g) for 5 min. A 50-μl aliquot of the supernatant was analysed with HPLC. Methyl p-tolyl sulfoxide was detected by measuring absorbance at 237 nm and its concentration was calculated by comparing the absorbance to a standard curve based on known concentrations. Enzyme activity is expressed as p mol sulfoxide/mg microsomal protein per 10 min.
Immunoprecipitation and Western blotting of flavin-containing monooxygenase 3 specific protein
Microsomes from mice liver were passed through a Qiashredder (Qiagen) by centrifugation for 10 min at 20 800 g at 4°C. These extracts were immunoprecipitated by incubating with primary polyclonal antibody that was originally made against human FMO3 amino acid sequence position 259–279. Since the immunogenic regions of this peptide exactly matched the amino acid sequence for mouse FMO3, we used it to detect mouse FMO3. These microsome extracts (100 μg) were immunoprecipitated by incubating with 2 μl primary polyclonal antibody for 1 h at 4°C, followed by additional 1-h incubation with 20 μl protein A/G PLUS-agarose. The pellet was collected by centrifugation at 2500 rpm for 5 min at 4°C and washed four times with PBS. After the final wash, electrophoresis sample buffer was added and the sample was heated at 95°C for 2 min. These extracts were separated by SDS polyacrylamide (10 % acrylamide) electrophoresis and transferred to nitrocellulose. The membrane was blocked with a solution of 5 % powdered non-fat milk, 25 mm-Tris (pH 7·5) and 150 mm-NaCl and then incubated with HRP-conjugated goat anti-rabbit Ig-G. The bound antibody was detected using ECL chemiluminescence detection kit (Amersham, Pharmacia Biotech, Inc. Piscataway, NJ, USA).
Statistical analysis
The SAS proc glm was used for a one-way ANOVA between treatments and Levene's test was used for heterogeneity of variance (Littell et al. Reference Littell, Stroup and Freund2002). The two treatment means were compared with the control using Dunnett's adjustment for multiplicity. When there was evidence of heterogeneity of variance, the SAS proc mixed was used to incorporate the heterogeneity in the model. In cases for which the control data is entirely zero, t tests were used to test for treatment means being significantly different from zero. Differences were considered statistically significant for P < 0·05.
Results
Effect of conjugated linoleic acid isomers on body, liver and liver lipid weights
Body weights of animals in the three dietary groups did not differ at the start of the study. However, at the end of feeding experimental diets body weights of the t10, c12-CLA group was significantly lower than in the other two groups (control 25·4 (sem 0·3) g; c9, t11-CLA 26·7 (sem 0·5) g; t10, c12-CLA 23·2 (sem 0·3) g; P = 0·02). Weights of the livers in animals fed the diets containing t10, c12-CLA were significantly (P < 0·05) greater than those in the control and c9, t11-CLA groups (mean 2·54 (sem 0·07) g v. 1·28 (sem 0·03) g and 1·47 (sem 0·06) g respectively); similarly the weight of total liver lipids was approximately four times greater in the t10, c12-CLA than those in the control and c9, t11-CLA groups (775 (sem 119) and 147 (sem 18) and 175 (sem 13) mg respectively).
DNA microarray analysis
Gene microarray analyses were performed to get clues regarding the genes whose expression may be altered by the dietary treatments. Fig. 1 shows an overview of the variation in hierarchical clustering of gene expression across liver tissue of mice fed diets containing c9, t11-CLA or t10, c12-CLA. The expression level of each gene (relative to its mean expression across all samples) is represented by different colours, and the colour intensities represent deviations from the mean. Mean expression is shown by the black colour, red colour indicates gene expression increased, while green colour indicates gene expression decreased relative to the average. The observed pattern of gene expression identified two major clusters. The first cluster representing t10, c12-CLA-fed mice and the second cluster representing the control and c9, t11-CLA-fed mice. Names of individual genes whose expression was altered by 2-fold or more are listed in the online version in Supplementary Table 1a–m.
Fig. 1 Hierarchical clustering of gene expression profiles of liver tissue of mice fed diets with or without conjugated linoleic acid (CLA) isomers. Each column represents an individual mouse. Regional hierarchical clustering identified two major clusters; one representing trans 10, cis 12-CLA and the other control and cis 9, trans 11-CLA. Black colour represents the mean expression of all six animals, green represents lower expression than the mean and the red represents higher than the mean. The scale at the bottom represents 1·5 and 3 SD below and above the mean. Names of the genes altered are given online in Supplementary Table 1a–m). For details of diets and procedures, see pp. 59–61.
The focus of the current paper is only on the genes involved in fatty acid metabolism; hence we performed further studies only on the genes involved in fatty acid oxidation and synthesis. Feeding a diet containing c9, t11-CLA caused a 2-fold or greater change in the transcription of thirty-one hepatic genes (nine decreased and twenty-two increased) including nine genes involved in fatty acid metabolism (Supplementary Table 1a–m and Table 3). Expression of ME was significantly (P < 0·05) increased (2·9-fold), while change in the expression of other enzymes involved in fatty acid metabolism did not attain statistical significance.
Table 3 Effect of conjugated linoleic acid (CLA) isomers on the expression of genes involved in fatty acid oxidation and synthesis† (Values are means with their standard errors)
c9, t11-CLA, cis 9, trans 11-CLA; t10, c12-CLA, trans 10, cis 12-CLA; FMO3, flavin-containing monooxygenase; Cyt P450, cytochrome P450; CPT1a, carnitine palmitoyl transferase 1a; ACOX, acetyl CoA oxidase; ACC, acetyl CoA carboxylase; FAS, fatty acid synthase; ME, malic enzyme; SCD1, stearoyl CoA desaturase 1.
Mean values were significantly different from those in the control group: *P < 0·05 %.
† For details of diets and procedures, see pp. 59–60.
‡The top numbers for each gene listed were determined by microarrays (n 2); the bottom numbers (n 6) were determined by QRT-PCR.
Feeding the diet containing t10, c12-CLA caused a 2-fold or greater change in the transcription 399 genes (121 increased, 278 decreased, Supplementary Table 1a–m). This isomer significantly (P < 0·05) reduced the expression of FMO3 (93 %), cyt P450 (54 %), CPT1a (60 %) and PPARα (53 %) and increased the expression of ME by 6·3-fold (P < 0·05). Expression of other lipogenic enzymes ACC, FAS and SCD1 increased by 3·2-, 2·6- and 1·9-folds respectively; however, these did not attain statistical significance (Table 2).
Confirmation of altered gene expression by real-time PCR
The afore-mentioned changes in the expression of fatty acid metabolism genes were confirmed by real-time PCR. Again, c9, t11-CLA diet did not alter the expression of most of the genes involved in fatty acid metabolism when compared with the control diet, with the exception of three genes. Expression of FMO3 and cyt P450 was reduced by 61 % and 38 % respectively, while that of SCD1 increased 5·9-fold (Table 3; P < 0·05 for all three genes). Real-time PCR also confirmed that t10, c12-CLA reduced the expression of genes involved in fatty acid oxidation as indicated by the DNA microarray method (Table 3); it reduced the expression of FMO3 95 % (P < 0·0001), cyt P450 61 % (P = 0·002), CPT1a 77 % (P = 0·025), acetyl CoA oxidase 50 % (P = 0·08) and PPARα 65 % (P = 0·05) when compared with the corresponding values in the control group. It also increased expressions of ACC by 18-fold (P = 0·01) and FAS by 3·5-fold (P = 0·03). Expressions of SCD1 and ME were increased by greater than 2-fold, but those did not attain statistical significance. Overall, the results from the microarray and real-time PCR data suggest that the t10, c12-CLA increased the expression of genes involved in fatty acid synthesis and reduced those of the genes involved in mitochondrial and peroxisomal β oxidation and microsomal ω oxidation; c9, t11-CLA did not alter the expression of any of these genes significantly except FMO3 (P = 0·001), cyt P450 (P = 0·06) and SCD1 (P = 0·07) (Table 3).
Effect of conjugated linoleic acid isomers on the hepatic microsomal flavin-containing monooxygenase activity and flavin-containing monooxygenase 3 specific protein
We determined the effect of CLA isomers on the hepatic microsomal FMO activity because t10, c12-CLA reduced the expression of FMO3 by 95 % and c9, t11-CLA reduced it by 61 % as compared with the control group. We also determined the sum of enzymatic activity for all the hepatic isomers of FMO (FMO1, FMO3 and FMO5). Dietary c9, t11-CLA reduced the FMO activity by 15 % and FMO3 specific protein by 10 % (both non-significant; Fig. 2(A)(B). Diet containing t10, c12-CLA reduced the FMO activity by 40 % and FMO3 protein by 67 % (P < 0·05; Fig. 2(A)(B)). Thus, the changes caused by the two CLA isomers in total FMO activity and FMO3 specific protein and mRNA are consistent.
Fig. 2 Effect of dietary conjugated linoleic acid (CLA) isomers on mouse liver flavin-containing monooxygenase (FMO)-3 expression (A) and FMO activity (B). Data shown for FMO3 expression are representative of three experiments, while those for FMO activity are means with their standard errors represented by vertical bars (n 3). a,bMean values with unlike superscript letters were significantly different (P < 0·05). c9, t11-CLA, cis 9, trans 11-CLA; t10, c12-CLA, trans 10, cis 12-CLA. For details of diets and procedures, see pp. 59–61.
Discussion
We compared the effects of feeding two CLA preparations enriched in c9, t11-CLA and t10, c12-CLA isomers on the expression of hepatic enzymes involved in fatty acid metabolism. Transcription of CPT1a, FMO3, cyt P450 and PPARα genes was significantly reduced by the diet containing t10, c12-CLA as determined by both the microarray and real-time PCR methods (Table 3). CPT1 is the rate limiting enzyme for mitochondrial β oxidation, while PPARα regulates β oxidation in mitochondria and peroxisomes; cyt P450 and FMO3 are involved in the ω oxidation and drug detoxification in the microsomes. Expression of acetyl CoA oxidase, the rate limiting enzyme for peroxisomal fatty acid oxidation, was reduced by 50 % by the diet containing t10, c12-CLA, although it was not statistically significant (P = 0·08; Table 3). The reduced expression of these five genes suggests that t10, c12-CLA may reduce fatty acid oxidation in the mitochondria, peroxisomes and the microsomes. Thus, reduced fatty oxidation in all these three organelles may contribute to the development of fatty liver in mice fed diets containing t10, c12-CLA. This diet also increased the expression of four lipogenic genes (ACC, FAS, ME and SCD1) by more than 2-fold; however, statistical significance was attained only for the expressions of ACC and FAS (P < 0·05; Table 3). Neither of the CLA isomers significantly altered the expression of nuclear factors, liver X receptor α and regulatory element-binding protein-1c, that regulate fatty acid synthesis (Supplementary Table 1a–m). We cannot determine the specific contributions of reduced fatty acid oxidation and increased fatty acid synthesis to the development of fatty liver in animals fed a diet containing t10, c12-CLA, but it appears that both reduced fatty acid oxidation and increased fatty acid synthesis may contribute to the development of fatty liver. Other factors such as reduced transport of lipids from liver may also contribute to the development of the fatty liver; results from a published report (Poirier et al. Reference Poirier, Rouault and Clement2005) and our own microarray data do not support this notion. Dietary c9, t11-CLA increased the liver lipid by only 15 %, which was not statistically significant. It did not reduce the expression of CPT1, acetyl CoA oxidase and PPARα genes and caused modest reductions in the expressions of FMO3 and cyt P450. These results suggest that c9, t11-CLA did not reduce the mitochondrial and peroxisomal fatty acid oxidation. The present study protocol cannot distinguish if the reductions in the expression of FMO3 and cyt P450 caused by c9, t11-CLA are specific to this isomer or because of its contamination with t10, c12-CLA. Completely pure CLA isomer preparations are needed to address this issue.
Supplementary Table 1a Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1b Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1c Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1d Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1e Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1f Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1g Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1h Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1i Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1j Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1k Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1l Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
Supplementary Table 1m Effect of CLA isomers on mouse liver gene expression
a All means are from 2 animals per group.
The present results showing down regulation of the genes involved in fatty acid oxidation by the t10, c12-CLA are at variance with those showing an increased expression of the CPT1a in mice fed a mixture of CLA isomers (Takahashi et al. Reference Takahashi, Kushiro, Shinohara and Ide2003; Javadi et al. Reference Javadi, Beynen and Hovenier2004; Ide, Reference Ide2005) or a purified t10, c12-CLA isomer (Degrace et al. Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004). We are not sure of the reasons for this discrepancy, but the design and methods used were considerably different between the present and other studies. We used female mice and fully quantitative real-time RT-PCR techniques, while those investigators used male mice and semi-quantitative RT-PCR methods. Furthermore, it is inappropriate to compare our results with those obtained with a mixture of CLA isomers (Takahashi et al. Reference Takahashi, Kushiro, Shinohara and Ide2003; Javadi et al. Reference Javadi, Beynen and Hovenier2004; Ide, Reference Ide2005), since the two isomers have contrasting effects on fatty acid metabolism (Roche et al. Reference Roche, Noone and Sewter2003; Kelley et al. Reference Kelley, Bartolini, Warren, Simon, Erickson and Mackey2004). The amount of CLA and the duration of its feeding were also different between the present and other studies. Most significantly, the basal diet in these studies with a mixture of CLA isomers was a high-fat diet (total fat 15–19 weight %) rich in SFA (palm oil in Takahashi et al. Reference Takahashi, Kushiro, Shinohara and Ide2003; Ide, Reference Ide2005) or SFA and MUFA (coconut, olive, palm and high oleic sunflower oils in Javadi et al. Reference Javadi, Beynen and Hovenier2004). In the present study, maize oil (5 weight %) was the source of fat and 0·5 % CLA was incorporated by replacing an equivalent amount of the maize oil. It is possible that CLA may increase hepatic fatty acid oxidation when fed with high-fat diets that are rich in SFA or MUFA, since the amount and type of dietary fatty acids regulate the expression of hepatic genes and the development of non-alcoholic steatosis and steatohepatitis (Demizieux et al. Reference Demizieux, Degrace and Gresti2002). In the study with the purified t10, c12-CLA, the basal diet contained sunflower and linseed oils and 1 weight % oleic acid or CLA were added to the control and test diets, respectively (Degrace et al. Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004). In this study both the in vitro CPT1 activity and mRNA expression were significantly increased by t10, c12-CLA. Also increased were the liver malonyl CoA that inhibits CPT1 and the sensitivity of CPTI to malonyl CoA (50 % inhibition, 2 v. 12 μmol/l). These authors recognize the inconsistency between their results and propose that in vivo hepatic fatty acid oxidation may actually be suppressed by this isomer. We found reduction in mRNA not only for CPT1, but also for four other genes involved in fatty acid oxidation. The present results are consistent with the proposal put forward by the investigators, which was discussed earlier (Degrace et al. Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004), the in vitro reduction of fatty acid oxidation by isolated mitochondria treated with CLA (Clarke, Reference Clarke2001) and the development of fatty liver observed in many studies with CLA. Increased expression of the lipogenic genes, ACC, FAS, ME and SCD1, found in the present study are consistent with those of published reports (Takahashi et al. Reference Takahashi, Kushiro, Shinohara and Ide2003; Degrace et al. Reference Degrace, Demizieux, Gresti, Chardigny, Sebedio and Clouet2004; Javadi et al. Reference Javadi, Beynen and Hovenier2004; Ide, Reference Ide2005).
The most dramatic effect of t10, c12-CLA in the present study was the down regulation of the genes for cyt P450 and FMO3, which are involved in the ω hydroxylation of the fatty acids and the production of dicarboxylic fatty acids (White et al. Reference White, Handler, Smith, Hill and Lehman1978; Krueger & Williams, Reference Krueger and Williams2005; Sanders et al. Reference Sanders, Ofman, Valianpour, Kemp and Wanders2005; Weng et al. Reference Weng, DiRusso, Reilly, Blacks and Ding2005). Once formed, dicarboxylic fatty acids can be shortened from either end of the molecule by β oxidation. This pathway plays a significant role in overall fatty acid oxidation during starvation and diabetes (Orellana et al. Reference Orellana, Rodrigo and Valdes1998). It may play a significant role in the overall hepatic fatty acid oxidation in mice fed diet containing t10, c12-CLA, since these animals do develop symptoms of diabetes (Poirier et al. Reference Poirier, Rouault and Clement2005). This interpretation is also consistent with a recent finding that mice deleted of the cyt P450 gene develop a fatty liver (Weng et al. Reference Weng, DiRusso, Reilly, Blacks and Ding2005). In addition to their role in fatty acid metabolism, both these enzymes detoxify numerous foreign compounds and also limit the length of time during which different drugs may be effective (Krueger et al. Reference Krueger, Vandyke, Williams and Hines2006). Thus, the suppression of these detoxifying enzyme systems may have additional health risks in addition to the development of fatty liver. FMO3 is the major hepatic isomer in man and a mutation in this gene causes trimethyaminuria, a condition wherein individuals excrete trimethylamine rather than trimethylamine oxide; trimethylamine produces a fishy odour in urine, sweat, breath and other bodily excretions (Seibel & Walsh, Reference Seibel and Walsh2002). It is difficult to extrapolate the results of this and many other mice studies to man because the amount of CLA used in most of the mice studies is equivalent to 30–60 g/60 kg person per d (Kelley & Erickson, Reference Kelley and Erickson2003); however, the long-term consumption of even lesser amounts of t10, c12-CLA by man may have serious health consequences.
In summary, the results of the present study indicate that the development of the fatty liver in mice fed diets containing t10, c12-CLA may be due both to reduced fatty acid oxidation and increased fatty acid synthesis. Results in addition to FMO3 need to be confirmed at the level of enzyme specific proteins and activities. To the best of our knowledge this is the first in vivo report that shows reduced expression of fatty acid oxidation and drug detoxification genes by t10, c12-CLA. Further studies are needed to determine the health consequences of the reduced expression of these genes by t10, c12-CLA.
Note
Supplementary information accompanies this paper on the journal's website (http://www.nutritionsociety.org).
Acknowledgements
We are most appreciative of the help provided by Terry Neidlinger in the HPLC analysis of FMO activity, Dr Colin Baron with the analysis of the microarray data and Dr John Newman in the preparation of tables and figures.
