Fish oil prevents sucrose-induced fatty liver but exacerbates high-safflower oil-induced fatty liver in ddy mice

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

Diets high in sucrose/fructose or fat can result in hepatic steatosis (fatty liver). We analyzed the effects of dietary fish oil on fatty liver induced by sucrose, safflower oil, and butter in ddY mice. In experiment I, mice were fed a high-starch diet [70 energy% (en%) starch] plus 20% (wt/wt) sucrose in the drinking water or fed a high-safflower oil diet (60 en%) for 11 weeks. As a control, mice were fed a high-starch diet with drinking water. Fish oil (10 en%) was either supplemented or not. Mice supplemented with sucrose or fed safflower oil showed a 1.7-fold or 2.2-fold increased liver triglyceride content, respectively, compared with that of control mice. Fish oil completely prevented sucrose-induced fatty liver, whereas it exacerbated safflower oil-induced fatty liver. Sucrose increased SREBP-1c and target gene messenger RNAs (mRNAs), and fish oil completely inhibited these increases. In experiment II, mice were fed a high-safflower oil or a high-butter diet, with or without fish oil supplementation. Fish oil exacerbated safflower oil–induced fatty liver but did not affect butter-induced fatty liver. Fish oil increased expression of peroxisome proliferator-activated receptor gamma (PPARγ) and target CD36 mRNA in safflower oil-fed mice. These increases were not observed in sucrose-supplemented or butter-fed mice. Conclusion: The effects of dietary fish oil on fatty liver differ according to the cause of fatty liver; fish oil prevents sucrose-induced fatty liver but exacerbates safflower oil-induced fatty liver. The exacerbation of fatty liver may be due, at least in part, to increased expression of liver PPARγ. (HEPATOLOGY 2007.)

The prevalence of obesity in Western societies has increased dramatically, due in large part to high-fat (HF) and high-sucrose diets. Among the consequences of obesity are the emerging epidemics of hepatic steatosis and nonalcoholic fatty liver disease (NAFLD). NAFLD occurs in patients who consume little or no alcohol.1 Fatty liver has been thought to be benign. However, it is now understood that fatty liver is a precursor of nonalcoholic steatohepatitis, which progresses to cirrhosis in up to 25% of patients.2

The accumulated hepatic lipids in patients with NAFLD include plasma nonesterified fatty acids (NEFAs) from adipose tissue, fatty acids produced in the liver via de novo lipogenesis, and dietary fatty acids, which enter the liver via spillover into the plasma NEFA pool and via hepatic uptake of intestinally derived chylomicron remnants. Analysis of multiple stable isotopes in patients with NAFLD has revealed that, of the triglyceride (TG) in the liver, 59% is derived from NEFAs, 26% from de novo lipogenesis, and 15% from dietary fatty acids.3 Roughly one fourth of fatty acids in the liver is produced from de novo 2-carbon precursors derived from glucose, fructose, and amino acids. Dietary fat also contributes significantly to liver TG storage pools.

Fish oil contains n-3 fatty acids such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), which, when included in the diet, decrease blood TG concentrations in hypertriglycemic patients and are considered to have protective effects against fatty liver.4 This effect is attributable mainly to the combined effects of inhibition of lipogenesis and stimulation of fatty acid oxidation in the liver.5, 6 Fish oil supplementation may be effective in the prevention of NAFLD; however, it is not clear whether fish oil is effective against all types of fatty liver.

The C57BL/6J inbred mouse strain has been used for studies of obesity and diabetes because of its susceptibility to these diseases in response to an HF diet.6, 7 C57BL/6J mice also develop fatty liver in response to an HF diet.8–10 However, they are resistant to sucrose/fructose–induced fatty liver because they possess adenine −468 bp from the putative 5′ end of the sterol regulatory element-binding protein (SREBP)-1c gene.11Mice with guanine at this site show increased liver SREBP-1c messenger RNA (mRNA) in response to a high-fructose diet, whereas mice with adenine do not.11 It has been reported that ddY mice show increased body weight and liver weight in response to 20% sucrose supplementation.12 We found that ddY mice possess guanine −468 bp in the SREBP-1c promoter and show hepatic steatosis when fed either sucrose supplementation or an HF diet (Yamazaki et al., unpublished observation).

In this study, we used ddY mice to examine whether fish oil can prevent fatty liver induced by sucrose supplementation or 2 types of HF diet (safflower oil or butter), and the mechanisms involved were elucidated.

Abbreviations

ACC, acetyl-CoA carboxylase; ACO, acyl-CoA oxidase; ANOVA, analysis of variance; CD36, fatty acid translocase; ChREBP, carbohydrate response element-binding protein; CPT, carnitine palmitoyltransferase; CREB, cAMP response element-binding protein; DGAT, acyl-CoA:diacylglycerol acyltransferase; DHA, docosahexaenoic acid; en%, energy percent; EPA, eicosapentaenoic acid; FAS, fatty acid synthase; GPAT, acyl-CoA:glycerol-3-phosphate acyltransferase; HF, high-fat; LPK, liver-type pyruvate kinase; LXR, liver X receptor; MCAD, medium-chain acyl-CoA dehydrogenase; mRNA, messenger RNA; NAFLD, nonalcoholic fatty liver disease; NEFA, nonesterified fatty acid; PCR, polymerase chain reaction; PGC, peroxisome proliferator-activated receptor r coactivator; PPAR, peroxisome proliferator-activated receptor; QUICKI, quantitive insulin sensitivity check index; SCD, stearoyl-CoA desaturase; SREBP, sterol regulatory element-binding protein; WAT, white adipose tissue; TG, triglyceride.

Materials and Methods

Animals.

Six-week-old male ddY mice were obtained from Japan SLC Inc. (Hamamatsu, Japan) and fed a normal laboratory diet (CE2; Clea, Tokyo, Japan) for 1 week to stabilize metabolic conditions. Mice were exposed to a 12-hour light/12-hour dark cycle, and the room was maintained at a constant temperature of 22°C. All animal procedures were in accordance with institutional guidelines.

Diet.

At 7 weeks of age, ddY mice were assigned to 1 of 6 groups (n = 5-9 in each group). In experiment I, to investigate fatty liver induced by sucrose or high-safflower oil, 3 groups were created: control mice were fed a high-starch diet containing 70 energy% (en%) starch and 10 en% safflower oil, sucrose-supplemented mice were fed a high-starch diet plus 20% sucrose (wt/wt) in the drinking water, and high-safflower oil-fed mice were given 20 en% sucrose plus 60 en% safflower oil. In the HF diet, to promote fat consumption, sucrose rather than starch was used as the carbohydrate source. Control and HF diet-fed mice were given distilled water. In experiment II, to investigate fatty liver induced by high-safflower oil and high-butter, 3 groups were created: control mice were fed a high-starch diet, high-safflower oil-fed mice were given 60 en% safflower oil, and high-butter-fed mice were given 60 en% butter. To examine the effects of fish oil, 10 en% safflower oil or butter was replaced with 10 en% fish oil in experiments I and II. Duration of the dietary manipulation was 11 weeks. The detailed compositions of the experimental diets are listed in Table 1. Fatty acid compositions of dietary oils were measured by gas–liquid chromatography. Safflower oil (high-oleic type) contained 45% oleic acid (18:1n-9) and 46% linoleic acid (18:2n-6), fish oil from tuna contained 7% EPA (20:5n-3) and 24% DHA (22:6n-3), and butter contained 71% (wt:wt) saturated fatty acid including 45% palmitic acid (16:0), 11% stearic acid (18:0), and 22% oleic acid. Diet preparations were similar to those of our previous studies.5, 6 Fish oil was provided by NOF Corp. (Tokyo, Japan). Butter was purchased from Snow Brand Milk Corp. (Hokkaido, Japan).

Table 1. Dietary Composition in Experiments I and II
IngredientsStSt+FOSucSuc+FOSafSaf+FOButterButter+FO
  1. Abbreviations: St, high-starch diet; St+FO, high-starch diet plus fish oil supplementation; Suc, high-starch diet with 20% (w/w) sucrose drink; Suc+FO, high-starch diet with 20% (w/w) sucrose drink plus fish oil supplementation; Saf, high-safflower oil, Saf+FO, high safflower oil plus fish oil supplementation; Butter, high-butter; Butter+FO, high butter plus fish oil supplementation.

g/100g
Safflower oil4.04.033.529.5
Butter33.529.5
Fish oil4.04.04.04.0
Casein20.020.020.020.029.029.029.029.0
Sucrose23.323.323.323.3
α-Starch66.266.266.266.2
Vitamin mix (AIN-93)1.01.01.01.01.51.51.51.5
Mineral mix (AIN-93)3.53.53.53.55.15.15.15.1
Cellulose powder5.05.05.05.07.37.37.37.3
L-cysteine0.30.30.30.30.40.40.40.4

Consumption of food was measured daily. Drinking of sucrose water was measured weekly. The mean food intake per day was estimated by subtracting the food weight of that day from the initial food weight of the previous day and dividing by the number of mice housed in the cage. With the use of these data, average energy intakes were estimated.

Quantitative Reverse Transcription Polymerase Chain Reaction.

Mice were killed, and livers were isolated for RNA preparation in the morning from 3-hour fasted animals to avoid acute effects of food intake. RNA was extracted with TRIzol Reagent (Invitrogen Corp., Carlsbad, CA) according to the manufacturer's instructions. Total RNA isolated from liver was reverse transcribed with ReverTra Ace (Toyobo Co. Ltd., Osaka, Japan) with random hexamers. The resulting complementary DNA was polymerase chain reaction (PCR) amplified in the 96-well format with SYBR Green PCR Master Mix and a 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA). Expression levels of test genes were normalized to those of an endogenous control, acidic ribosomal phosphoprotein P0 (36B4). The primers used for quantitative reverse transcription PCR are listed in Table 2.

Table 2. Primers Used for Quantitative PCR
GeneForward primer (5′ to 3′)Reverse primer (5′ to 3′)
ACC1GGACAGACTGATCGCAGAGAAAGTGGAGAGCCCCACACACA
ACOGCCCAACTGTGACTTCCATTGGCATGTAACCCGTAGCACT
CD36AATGGCACAGACGCAGCCTGGTTGTCTGGATTCTGGA
ChREBPGATGGTGCGAACAGCTCTTCTCTGGGCTGTGTCATGGTGAA
CPT IGCACTGCAGCTCGCACATTACAACTCAGACAGTACCTCCTTCAGGAAA
CREBGAAGAAGCAGCACGGAAGAGATCTCTTGCTGCCTCCCTGTT
DGAT1GTGCACAAGTGGTGCATCAGCAGTGGGATCTGAGCCATC
DGAT2AGTGGCAATGCTATCATCATCGTAAGGAATAAGTGGGAACCAGATCA
FASGCTGCGGAAACTTCAGGAAATAGAGACGTGTCACTCCTGGACTT
GPATCAACACCATCCCCGACATCGTGACCTTCGATTATGCGATCA
HES-1CACGACACCGGACAAACCATCCATGATAGGCTTTGATGACTTTC
LPKGAGTCGGAGGTGGAAATTGTCCGCACCACTAAGGAGATGA
LXRαGGGAGGAGTGTGTGCTGTCAGGAGCGCCTGTTACACTGTTGC
MCADGATCGCAATGGGTGCTTTTGATAGAAAGCTGATTGGCAATGTCTCCAGCAAA
PGC1βGGTCCCTGGCTGACATTCACGGCACATCGAGGGCAGAG
PPARαCCTGAACATCGAGTGTCGAATATGGTCTTCTTCTGAATCTTGCAGCT
PPARγ1GAGTGTGACGACAAGATTTGGGTGGGCCAGAATGGCATCT
PPARγ2TCTGGGAGATTCTCCTGTTGAGGTGGGCCAGAATGGCATCT
SCD1CCCCTGCGGATCTTCCTTATAGGGTCGGCGTGTGTTTCT
SREBP-1aGAGGCGGCTCTGGAACAGATGTCTTCGATGTCGTTCAAAACC
SREBP-1cGGAGCCATGGATTGCACATTCCTGTCTCACCCCCAGCATA

Liver Lipid Analysis.

Liver lipids were measured by enzymatic colorimetry as described previously.13

Serum Chemistries.

In experiment II, serum was obtained at 2 points, after 24 hours of fasting and 3 hours after refeeding (postprandial). Serum glucose was measured on an Ascensia autoanalyzer (Bayer Medical Ltd., Tokyo, Japan). Serum TG and NEFA levels were assayed by enzymatic colorimetry with TG E and NEFA C test kits (Wako Pure Chemical Industries Ltd., Osaka, Japan). Serum insulin was determined with a mouse insulin enzyme-linked immunosorbent assay kit (Morinaga, Kanagawa, Japan).

Hepatic Histology.

In experiment I, mouse livers were fixed in 4% neutral-buffered formalin, embedded in paraffin, cut into sections, and stained with hematoxylin-eosin. Frozen sections of formalin-fixed liver were stained with Oil red O with the use of standard techniques.

Statistical Analysis.

Two-way analysis of variance (ANOVA) was used to examine the 2 main effects of diet (starch, sucrose, and safflower oil) or fat (safflower oil and butter), fish oil supplementation, and their interaction. Statistical significance of the interaction of diet (or fat) and fish oil indicates that the effects of fish oil on the 3 types of diet (or 2 types of fat) were statistically different. When differences were significant with respect to main or interaction effects, each group was compared with the others by Fisher's protected least significant difference test (StatView 5.0; Abacus Concepts Inc., Berkeley, CA). Statistical significance was set at P < 0.05. Values are shown as mean ± standard error of the mean (SEM).

Results

Energy Intake and Body, White Adipose Tissue, and Liver Weights.

Average energy intakes for the 6 groups over 11 weeks in experiment I are shown in Fig. 1A. Energy intakes in sucrose-supplemented mice were 1.4-fold greater than those in control mice (starch), whereas those in HF diet-fed mice were similar to those of control mice. The increased total energy intake in sucrose-supplemented mice was due to the extra energy supply from the sucrose-supplemented drinking water. The macronutrient composition of en% in each group is also shown in Fig. 1B. Fish oil supplementation did not affect the intake of total energy or of other macronutrients.

Figure 1.

Daily energy intake and macronutrient intake in experiment I. St, high-starch diet; St+FO, high-starch diet plus fish oil; Suc, high-starch diet plus 20% (wt/wt) sucrose drink; Suc+FO, high-starch diet plus 20% (wt/wt) sucrose drink and fish oil; Saf, high-safflower oil; Saf+FO, high-safflower oil plus fish oil. (A) Daily energy intake. Intakes of food and drinking water were measured daily during the study and were averaged weekly. The data represent the mean daily energy intake. (B) Average energy percentage intakes of macronutrients throughout the study. Safflower oil, fish oil, starch, sucrose, and casein percentages of total energy intake were calculated.

Increases in body weight were observed in high-safflower oil–fed mice (relative to high-starch–fed mice) (Table 3), although the energy intake was similar in these 2 groups. This discrepancy in body weight increase between groups may be attributable to differences in physical activity level, dietary-induced thermogenesis, or basal metabolic rate. Feeding of 10 en% fish oil did not affect body weight or epididymal, retroperitoneal, mesenteric, or subcutaneous white adipose tissue (WAT) weight in the high-starch diet, sucrose-supplemented, or high-safflower oil-fed mice. However, 2-way ANOVA analysis showed that, among several types of WAT, epididymal WAT was sensitive to fish oil (fish oil effect, P = 0.010). Liver weight in sucrose-supplemented and in high-safflower oil-fed mice was 1.20-fold and 1.28-fold, respectively, that in high-starch-fed mice. The effects of fish oil on liver weight differed. Fish oil partially prevented the sucrose-induced increase in liver weight (not significant) but significantly exacerbated the high-safflower oil–induced increase in liver weight.

Table 3. Body and Tissue Weights of Mice After 11 Weeks on Study Diets in Experiment I
  StSucSafTwo-Way ANOVA P Value
DietFish oilDiet × Fish Oil
  1. Abbreviations: St, high-starch diet; Suc, high-starch diet with 20% (w/w) sucrose drink; Saf, high-safflower oil diet; BW, body weight; WAT, white adipose tissue. Values are mean ± SEM. Means without a common letter differ (P < 0.05).

 Fish oil      
n989   
 +588   
Weight (g)       
 BW at start31.7 ± 0.631.3 ± 0.731.4 ± 0.7   
 +31.2 ± 0.731.3 ± 0.731.3 ± 0.6   
 Final BW52.8 ± 2.6a56.2 ± 3.1ab61.6 ± 2.0bc   
 +51.9 ± 4.9a55.4 ± 1.3ac62.7 ± 1.5b0.0020.9440.904
 Liver1.94 ± 0.09a2.33 ± 0.18b2.48 ± 0.14b   
 +1.93 ± 0.17a2.17 ± 0.11ab3.07 ± 0.06c<0.0010.2030.011
 Epididymal WAT2.10 ± 0.29abc2.33 ± 0.27cde2.80 ± 0.19e   
 +1.52 ± 0.30a1.84 ± 0.14abd2.30 ± 0.14be0.0120.0100.983
 Retroperitoneal WAT0.43 ± 0.050.50 ± 0.070.48 ± 0.05   
 +0.36 ± 0.060.40 ± 0.030.43 ± 0.020.3850.0680.869
 Mesenteric WAT1.39 ± 0.461.46 ± 0.291.57 ± 0.20   
 +0.99 ± 0.271.02 ± 0.111.47 ± 0.090.4040.1650.789
 Subcutaneous WAT1.45 ± 0.21ab1.49 ± 0.26ab1.85 ± 0.16b   
 +0.95 ± 0.34a1.52 ± 0.11ab1.96 ± 0.21b0.0090.5030.344

Fish Oil Prevents Sucrose-Induced Liver TG Accumulation and Exacerbates High-Safflower Oil–Induced Liver TG Accumulation.

In parallel with alterations in liver weight, liver TG concentration in sucrose-supplemented mice and in high-safflower oil–fed mice was 1.7-fold and 2.2-fold, respectively, that in high-starch–fed mice (Fig. 2A). The effects of fish oil on liver TG accumulation differed. Fish oil prevented the sucrose-induced increase in liver TG but did not affect the high-safflower oil–induced increase in liver TG. On a whole liver basis, fish oil increased liver TG by 50% in high-safflower oil–fed mice (Fig. 2B). The tuna oil used in this study contained 7% EPA, 24% DHA, and other highly polyunsaturated fatty acids. Other fish oils were also examined. Sardine oil contains 28% EPA and 12% DHA, and a mixture of tuna and sardine oils contains 6% EPA and 13% DHA and is also effective in preventing sucrose-induced hepatic TG accumulation (data not shown).

Figure 2.

Fish oil prevents sucrose-induced hepatic TG accumulation (Suc) but not high-safflower oil–induced TG accumulation (Saf). (A) Hepatic lipids were extracted, and TG concentrations were measured. (B) TG content in a whole liver of mouse. Values are mean ± SEM (n = 5-9). Means without a common letter differ (P < 0.05). Two-way ANOVA P values are significant in effects of diet and interaction. St, high-starch diet.

Oil red O staining confirmed hepatic TG accumulation in both sucrose-supplemented mice and high-safflower oil–fed mice and also confirmed that fish oil prevented sucrose-induced hepatic TG accumulation but not high-safflower oil–induced hepatic TG accumulation (Fig. 3). This beneficial effect of fish oil was also observed in control mice; mild liver TG accumulation was observed in high-starch–fed mice, and this accumulation was prevented by fish oil. Hematoxylin-eosin staining more clearly depicted the microvesicular fatty change within hepatocytes in both sucrose-supplemented mice and high-safflower oil-fed mice (Fig. 4). These changes were not observed in fish oil–supplemented, sucrose-supplemented mice but were observed in fish oil–supplemented, high-safflower oil–fed mice.

Figure 3.

Oil red O staining of liver sections from mice fed high-starch (St) (A), high-starch plus fish oil (B), sucrose (Suc) (C), sucrose plus fish oil (D), high-safflower oil (Saf) (E), and high-safflower oil plus fish oil (F) for 11 weeks. Fish oil markedly decreased hepatic TG accumulation in sucrose-supplemented mice (D) but not in high-safflower oil–fed mice (F).

Figure 4.

Comparative liver morphology as revealed in sections stained with hematoxylin-eosin. Microvesicular fatty change in hepatocytes is observed in sucrose-supplemented (Suc) (C) and high-safflower oil-fed (Saf) mice (E). Note the absence of microvesicular fatty change in sucrose-supplemented plus fish oil-supplemented mice (D).

Effects of Sucrose Supplementation on Fatty Liver.

The mechanisms underlying the development of fatty liver in mice supplemented with sucrose were elucidated by hepatic gene expression profiling. SREBP-1c is a major regulator of lipogenesis in the liver.14 SREBP-1a is less abundant in the liver but is more potent in fatty acid synthesis than SREBP-1c.15 To examine whether the increased hepatic TG content in sucrose-supplemented mice was caused by activation of SREBP-1 (relative to that in high-starch–fed mice), mRNAs for SREPB-1 and the lipogenic enzymes fatty acid synthase (FAS), stearoyl-CoA desaturase 1 (SCD1), and acetyl-CoA carboxylase 1 (ACC1) were measured by quantitative reverse transcription PCR (Fig. 5). Sucrose-supplementation resulted in a 2.2-fold increase in SREBP-1c mRNA but did not affect SREBP-1a mRNA, compared with high-starch–fed mice. FAS and ACC1 mRNAs, but not SCD1 mRNA, were also increased in sucrose-supplemented mice. Acyl-CoA:diacylglycerol acyltransferase (DGAT) is responsible for the final esterification step of diacylglycerol to triacylglycerol,13 and acyl-CoA:glycerol-3-phosphate acyltransferase (GPAT) is responsible for the first esterification step of glycerol-3-phosphate to monoacylglycerol.16 Both are involved in TG synthesis but were not increased in sucrose-supplemented mice (Fig. 5). Carbohydrate response element-binding protein (ChREBP) binds to and activates the transcription of several lipogenic enzyme genes such as those for liver-type pyruvate kinase (LPK), a regulatory enzyme in the hepatic glycolysis pathway, ACC1, and FAS, leading to fatty liver.17 LPK mRNA was increased in sucrose-supplemented mice (Fig. 5). Activation of peroxisome proliferator-activated receptor (PPAR) γ causes fatty liver. Hepatic PPARγ expression is increased in several animal models of diabetes and obesity.18, 19 Adenovirus delivery of PPARγ in hepatocytes leads to fatty liver,20 and PPARγ RNA interference is reported to decrease hepatic TG levels.21 Therefore, expression levels of PPARγ mRNA and that of its target, fatty acid translocase (CD36), were measured. PPARγ consists of 2 forms, PPARγ1 and PPARγ2, and both forms are expressed in hepatocytes. However, mRNAs for both isoforms were not altered in sucrose-supplemented mice (Fig. 6). Sucrose supplementation may inhibit PPARα, which increases fatty acid β oxidation in the liver.22 However, mRNA levels of PPARα and its target genes, including acyl-CoA oxidase (ACO, a marker of peroxisomal β-oxidation), medium-chain acyl-CoA dehydrogenase (MCAD, a marker of mitochondrial β-oxidation), and carnitine palmitoyltransferase 1 (CPT1, a marker of fatty acid transport), were not altered in sucrose-supplemented mice (Fig. 6). These data suggested that sucrose supplementation activates the transcription factors SREBP-1c and ChREBP and promotes fatty acid synthesis.

Figure 5.

Hepatic gene expression of transcription factors related to lipogenesis, target genes of SREBP-1c, enzymes related to TG synthesis, and ChREBP and LPK in mice fed high-starch (St), sucrose supplementation (Suc), and high-safflower oil (Saf) with or without fish oil for 11 weeks in experiment I. Quantitative PCR analysis was performed with the specific primers listed in Table 2. Values are mean ± SEM (n = 5-9). Means without a common letter differ (P < 0.05). The percentage in mRNA levels relative to those of high-starch fed mice without fish oil are shown. Two-way ANOVA P values with respect to diet effect are significant for SREBP-1c, SREBP-1a, LXRα, FAS, DGAT1, GPAT, ChREBP, and LPK; with respect to fish oil effects are significant for FAS, SCD1, and DGAT1; those with respect to interaction are significant for SCD1, ACC1, and GPAT.

Figure 6.

Hepatic gene expression of PPARγ and PPARα and their target genes and of transcription factors regulating PPARγ in mice fed high-starch (St), sucrose supplementation (Suc), and high-safflower oil (Saf) with or without fish oil for 11 weeks in experiment I. Quantitative PCR analysis was performed with the specific primers listed in Table 2. Values are mean ± SEM (n = 5-9). Means without a common letter differ (P < 0.05). The percentage in mRNA levels relative to those of high-starch–fed mice without fish oil are shown. Two-way ANOVA P values with respect to diet effect are significant for PPARγ1, PPARγ2, CD36, ACO, CPT1, and MCAD; those with respect to fish oil effects are significant for PPARγ1, CD36, ACO, and MCAD; those with respect to interaction are significant for PPARγ1, PPARγ2, CD36, and ACO.

Effects of High-Safflower Oil on Fatty Liver.

The mechanisms underlying the development of fatty liver in mice fed a high-safflower oil diet (relative to high-starch–fed mice) were elucidated by hepatic gene expression profiling. High-safflower oil–fed mice did not show altered expression of SREBP-1c, SREBP-1a, FAS, ACC1, DGATs, GPAT, ChREBP, or LPK mRNA but did show decreased SCD1 mRNA (Fig. 5). High-safflower oil increased PPARγ1, PPARγ2, and CD36 mRNAs, although these increases were not significant (Fig. 6). High- safflower oil slightly increased ACO, CPT1, and MCAD mRNAs, but only the increase in ACO mRNA was significant (Fig. 6).

Effects of Fish Oil Supplementation on Control, High-Starch Diet-Fed Mice.

Fish oil decreased SCD-1 and LPK mRNAs and increased ACO mRNA significantly (Figs. 5 and 6). Fish oil–mediated suppression of SREBP-1c and ChREBP activity may lead to a decrease in liver TG content in high-starch–fed mice.

Effects of Fish Oil Supplementation on Sucrose-Supplemented Mice.

Fish oil significantly decreased FAS, SCD1, and ACC1 mRNAs (Fig. 5). The transcription factor liver X receptor (LXR)α binds to the SREBP-1c promoter and increases its expression23; however, fish oil did not significantly decrease the LXRα mRNA level. The coactivator peroxisome proliferator-activated receptor γ coactivator (PGC)-1 β binds to SREBP-1c and activates its transcriptional activity24; however, the PGC-1β mRNA levels were not altered. These data indicated that repression of sucrose-induced SREBP-1c activation may lead to a decrease in liver TG content.

Effects of Fish Oil Supplementation on High-Safflower Oil–Fed Mice.

Fish oil increased SREBP-1a, SCD1, and GPAT mRNAs (Fig. 5). However, these increases were not observed in experiment II. The reason for this discrepancy is unknown, but it suggests that increased expression of these genes is not important for the fish oil–induced liver TG increase. Surprisingly, fish oil markedly increased the expression of PPARγ1, PPARγ2, and CD36 mRNAs (Fig. 6). PPARγ expression is suppressed by the transcription factor hairy enhancer of split (HES-1), and HES-1 expression is increased by the cyclic adenosine monophosphate–responsive transcription factor-binding protein (CREB). This pathway is involved in the inhibition of hepatic TG synthesis observed in fasting.21 However, HES-1 and CREB mRNAs were not altered. Fish oil did not alter the PPARα mRNA level, but it did increase ACO, CPT1, and MCAD mRNAs, suggesting that PPARα activation may not be sufficient to decrease the TG accumulation in these mice. Therefore, the exacerbation of fatty liver may be attributable to increased PPARγ expression.

High-Butter-Fed Mice Show Increased Liver TG Concentration Compared with High-Safflower Oil–Fed Mice, But Fish Oil Does Not Exacerbate High-Butter–Induced Fatty Liver.

To examine the mechanism(s) underlying the exacerbation of fatty liver by fish oil in mice fed a high-safflower oil, we fed mice butter, saturated fatty acid–rich oil, with or without fish oil supplementation, in experiment II. As negative and positive controls, mice were fed a high-starch diet and a high-safflower oil diet, respectively. Energy intake was similar between high-safflower oil–fed and high-butter–fed mice and between fish oil–supplemented and nonsupplemented mice (data not shown). Because the 2 HF diets are hypercaloric and contain the same amount of other nutrients, we limited statistical comparisons to the effects of the 2 HF diets rather than comparing them with the effects of the high-starch diet (Tables 4–6). After 11 weeks, high-butter–fed mice showed increases in body weight and liver and epididymal WAT weights (relative to high-safflower oil-fed mice), but these increases were not significant. The liver TG concentration in high-butter–fed mice was twice that in high-safflower oil–fed mice (Table 4). As observed in experiment I, fish oil further increased the liver TG content in high-safflower oil–fed mice; surprisingly, this exacerbation was not observed in high-butter–fed mice (Table 4).

Table 4. Body and Tissue Weights and Liver TG of Mice After 11 Weeks on Study Diets in Experiment II
  StSafButterTwo-Way ANOVA P Value
FatFish OilFat × Fish Oil
  1. Abbreviations: St, high-starch diet; Saf, high-safflower oil diet; BW, body weight; WAT, white adipose tissue. Values are mean ± SEM. Means without a common letter differ (P < 0.05).

 Fish oil      
n888   
 +788   
Weight (g)       
 BW at start34.3 ± 0.534.3 ± 0.534.3 ± 0.3   
 +34.6 ± 0.434.4 ± 0.434.2 ± 0.40.8590.9760.882
 Final BW49.3 ± 2.055.9 ± 1.560.3 ± 1.0   
 +49.7 ± 1.458.2 ± 1.557.0 ± 2.30.3410.7570.097
 Liver1.73 ± 0.072.19 ± 0.112.63 ± 0.17   
 +1.82 ± 0.092.74 ± 0.212.46 ± 0.280.7110.3540.082
 Epididymal WAT2.01 ± 0.222.48 ± 0.293.05 ± 0.28   
 +1.69 ± 0.202.31 ± 0.252.48 ± 0.120.1350.1420.415
Liver TG (mg/g liver)124 ± 10240 ± 37a473 ± 59b   
 +103 ± 19357 ± 61ab468 ± 109b0.0230.4380.400
Liver TG (mg/liver)218 ± 23547 ± 81a1307 ± 180b   
 +194 ± 431037 ± 108b1304 ± 207b0.0020.1220.119
Table 5. Serum Analysis of Mice After 11 Weeks on Study Diets in Experiment II
  StSafButterTwo-Way ANOVA P Value
FatFish OilFat × Fish Oil
  1. Abbreviations: St, high-starch diet; Saf, high-safflower oil diet; 0h, serum obtained from 24-hour-fasted mice; 3h, serum obtained from animals 3 hours after food was presented; QUICKI, Quantitative Insulin Sensitivity Check Index, which is calculated by the following formula: 1/[loginsulin+logglucose], where insulin concentration is measured in microunit per milliliter and glucose concentration in milligram per deciliter. Values are mean ± SEM. Means without a common letter differ (P < 0.05).

 Fish oil      
n888   
 +788   
Serum analysis       
Glucose (mg/dL)
 0h92 ± 5112 ± 7a106 ± 6a   
 +82 ± 6142 ± 9b95 ± 8a0.0020.2110.012
 3h189 ± 11147 ± 9151 ± 12   
 +195 ± 15138 ± 8155 ± 110.3360.8270.537
NEFA (mEq/L)
 Oh1.07 ± 0.131.27 ± 0.11ab1.41 ± 0.15bc   
 +1.06 ± 0.131.02 ± 0.09a1.13 ± 0.08ab0.2570.0230.916
 3h0.85 ± 0.091.58 ± 0.15a1.97 ± 0.18a   
 +0.91 ± 0.111.57 ± 0.09a1.11 ± 0.12b0.7860.0040.006
TG (mg/dL) 
 0h140 ± 17189 ± 25189 ± 23   
 +197 ± 22153 ± 19167 ± 210.7360.1930.762
 3h184 ± 31380 ± 63ab519 ± 94bc   
 +229 ± 30274 ± 42a426 ± 49ab0.0340.1380.926
Insulin (ng/mL)
 0h0.28 ± 0.130.34 ± 0.08a0.38 ± 0.06a   
 +0.35 ± 0.151.28 ± 0.35b0.32 ± 0.08a0.0200.0270.013
 3h4.33 ± 0.765.54 ± 1.395.75 ± 0.75   
 +7.74 ± 1.186.86 ± 1.606.78 ± 1.530.9610.3960.913
 
QUICKI0.385 ± 0.0290.342 ± 0.013a0.333 ± 0.009ab   
 +0.374 ± 0.0240.290 ± 0.015b0.365 ± 0.021a0.0410.5270.011
Table 6. Hepatic Gene Expression Profile Related to TG Synthesis of Mice After 11 Weeks on Study Diets in Experiment II
  StSafButterTwo-Way ANOVA P Value
FatFish OilFat × Fish Oil
  1. Abbreviations: St, high-starch diet; Saf, high-safflower oil diet; BW, body weight; WAT, white adipose tissue.

  2. Values are mean ± SEM. Means without a common letter differ (P < 0.05). The percentage in mRNA levels relative to those of high-starch–fed mice without fish oil are shown.

 Fish oil      
n888   
 +788  
Transcription factors related to lipogenesis
SREBP-1c100 ± 18114 ± 24ab151 ± 25b   
 +51 ± 22123 ± 31b52 ± 6a0.4840.0630.029
SREBP-1a100 ± 8101 ± 7104 ± 12   
 +96 ± 21109 ± 1572 ± 70.1270.2740.078
LXRα100 ± 10103 ± 885 ± 10   
 +102 ± 18112 ± 1390 ± 80.0490.5130.818
PGC-1β100 ± 1792 ± 8111 ± 29   
 +147 ± 30118 ± 2865 ± 130.4350.6310.106
Target tenes or SREBP-1c
FAS100 ± 884 ± 24a136 ± 38a   
 +29 ± 474 ± 13ab14 ± 2b0.8670.0090.024
SCD1100 ± 1118 ± 6ac54 ± 11b   
 +34 ± 1724 ± 6a1 ± 1c0.3890.003<0.001
ACC1100 ± 469 ± 11a88 ± 14a   
 +77 ± 2967 ± 6a29 ± 4b0.3160.0040.007
Enzymes related to TG synthesis
DGAT1100 ± 987 ± 776 ± 7   
 +102 ± 1379 ± 669 ± 60.1360.2560.958
DGAT2100 ± 9108 ± 12104 ± 16   
 +119 ± 1992 ± 1390 ± 90.8210.2600.932
GPAT100 ± 2387 ± 9ab141 ± 14cd   
 +110 ± 38106 ± 19bd51 ± 5a0.9500.0390.002
ChREBP and target gene
ChREBP100 ± 18103 ± 12102 ± 20   
 +97 ± 19106 ± 1679 ± 100.3610.5110.388
LPK100 ± 1734 ± 6a61 ± 11b   
 +37 ± 839 ± 4a29 ± 8a0.2450.0870.022
PPARγ and target genes
PPARγ1100 ± 23405 ± 110a315 ± 149a   
 +113 ± 331043 ± 287b143 ± 60a0.0080.1900.027
PPARγ2100 ± 15587 ± 244ab952 ± 396ab   
 +32 ± 51030 ± 220b240 ± 53a0.4200.6080.034
CD36100 ± 8185 ± 57a161 ± 33a   
 +97 ± 10435 ± 92b123 ± 37a0.0090.0890.023
Transcription factors regulating PPARγ
CREB100 ± 14110 ± 11106 ± 11   
 +118 ± 24123 ± 1393 ± 140.1770.9800.305
HES-1100 ± 18103 ± 12102 ± 20   
 +97 ± 19106 ± 1679 ± 100.3610.5120.388
PPARα and their target genes
PPARα100 ± 10102 ± 1287 ± 11   
 +114 ± 1497 ± 1376 ± 60.1090.4580.787
ACO100 ± 8114 ± 18ab99 ± 17a   
 +159 ± 27159 ± 16b97 ± 14a0.0270.2140.158
CPTI100 ± 11107 ± 12ab83 ± 11ac   
 +121 ± 19119 ± 15b70 ± 5c0.0040.9840.306
MCAD100 ± 9132 ± 14ad88 ± 11bc   
 +155 ± 24169 ± 14a121 ± 15cd0.0020.0140.873

Postprandial TG Concentrations Are Increased in Mice Fed Safflower Oil or Butter, and Fish Oil Causes Insulin Resistance in Mice Fed High-Safflower Oil.

Increased plasma concentrations of glucose and NEFA may promote hepatic steatosis.25 In experiment II, we measured serum glucose, NEFA, TG, and insulin concentrations at 2 points, after 24 hours of fasting (fasting) and 3 hours after refeeding (postprandial). Fish oil increased fasting glucose and insulin concentrations in high-safflower oil–fed mice but not in high-butter–fed mice (Table 5). We estimated insulin sensitivity by the Quantitative Insulin Sensitivity Check Index26 (QUICKI; see Table 5 legend for formula). Fish oil decreased insulin sensitivity in high-safflower oil–fed mice but not in high-butter–fed mice, suggesting that the interaction of fish oil and safflower oil may lead to insulin resistance. Postprandial glucose and insulin concentrations were not altered. Fish oil decreased the postprandial NEFA concentration in high-butter–fed mice but not in high-safflower oil–fed mice, suggesting that relatively increased postprandial NEFA may contribute to fish oil–induced liver TG accumulation in high-safflower oil–fed mice. The postprandial TG concentration was increased both in high-safflower oil–fed and butter-fed mice (relative to high-starch–fed mice), and fish oil decreased the postprandial TG concentration by 28% and 18%, respectively, but these decreases were not significant.

Fish Oil Increases PPARγ Expression in Safflower Oil–Fed Mice But Not in Butter-Fed Mice.

Hepatic gene expression profiles are shown in Table 6. Fish oil markedly decreased SREBP-1c and target gene mRNAs in high-butter–fed mice but not in high-safflower oil–fed mice. Fish oil increased expression of PPARγ1 and CD36 mRNAs in high-safflower oil–fed mice, but decreased expression in high-butter–fed mice, although these decreases were not significant. However, fish oil did not decrease liver TG content in high-butter–fed mice, suggesting that increased postprandial TG, which supplies a large amount of fat to the liver, may overwhelm decreased fat synthesis in the liver.

Discussion

In this study, we found that the effects of fish oil on fatty liver differ according to the cause of fatty liver. In the dietary models of fatty liver, fish oil supplementation showed different effects on hepatic TG accumulation. Fish oil completely prevented sucrose-induced fatty liver, whereas it exacerbated high-safflower oil–induced fatty liver. Fish oil did not affect high-butter–induced fatty liver. Sucrose supplementation increased SREBP-1c expression and induced lipogenesis, and this process was effectively inhibited by fish oil supplementation. Fish oil increased the expression of hepatic PPARγ mRNA in high-safflower oil–fed mice, which may worsen fatty liver, whereas this increase was not observed in sucrose-supplemented mice or high-butter oil–fed mice.

The role of SREBP-1c in sucrose/fructose-induced fatty liver has been examined extensively. In mice with on a pure 129 SV background, fructose supplementation induces fatty liver.27 Fructose induces the hepatic expression of SCD1 and other lipogenic proteins in SREBP-1c-knockout mice, but not in SCD1-knockout mice,27 indicating that SCD1 expression is important for the development of fructose-induced fatty liver. However, in ddY mice, sucrose supplementation increased SREBP-1c mRNA and that of its target genes FAS and ACC1 but did not affect SCD1 mRNA expression (Fig. 5), suggesting that the induction of SCD1 is not always necessary for the development of fatty liver. In Wistar rats, increases in hepatic FAS and malic enzyme mRNAs were observed in response to a high-sucrose diet, and fish oil suppressed these increases.28 In ddY mice, fish oil decreased hepatic TG accumulation, along with a decrease in FAS and SCD1 mRNAs in response to sucrose supplementation. Therefore, fish oil may prevent sucrose-induced fatty liver by inhibiting the effects of lipogenic enzymes.

The mechanism of fatty liver in response to a HF diet has also been examined. In Fisher 344 rats, those fed a large amount of lard (45 en%) showed 3.7-fold greater hepatic TG concentration than those fed a large amount of fish oil (45 en%).26 The postprandial TG concentration in lard-fed rats was 221 mg/dL, whereas that in fish oil-fed rats was 75 mg/dL, suggesting that the increased amount of absorbed fat may be responsible for TG accumulation in the liver.26 In experiment II, we also observed increased postprandial TG concentration in high-butter–fed and high-safflower oil–fed mice (relative to high-starch–fed mice); however, fish oil slightly inhibited this increase. The difference in the decrease in postprandial TG by fish oil in these 2 studies may be attributable to a difference in the dose of fish oil (45 en% versus 10 en%). However, the fish oil–induced liver TG increase in high-safflower oil–fed mice was not explained by alterations in postprandial TG concentration.

With respect to hepatic gene expression profiles, HF diet (36-82 en%)-fed C57BL/6J mice show fatty liver without increases in SREBP-1c or target gene mRNAs8–10 but increased PPARγ and target gene mRNAs.9, 10 However, the dietary fatty acid compositions were not described in these studies. In the current study, fish oil further increased PPARγ and target gene mRNAs, when given to mice fed high-safflower oil diet but not to mice fed high-butter diet. This may be a major reason why fish oil exacerbates high-safflower oil–induced fatty liver, despite increased PPARα activation and decreased postprandial serum TG concentration, which may favor a decrease in liver TG. Insulin increases the expression of PPARγ2 in human adipocytes and in mouse hepatocytes.29, 30 An increased fasting blood insulin level may increase PPARγ expression. We speculate that unknown interactions between safflower oil and fish oil may lead to insulin resistance (lower QUICKI in Table 5), and fasting hyperinsulinemia attributable to insulin resistance may up-regulate PPARγ mRNA, which leads to the exacerbation of fatty liver. It is unlikely that fatty liver causes hyperinsulinemia because high-butter–fed mice did not show this condition. Sustained fatty acid synthesis in response to fish oil in high-safflower oil–fed mice could also be the consequence of fasting hyperinsulinemia; insulin is known to enhance both the transcription and activation of SREBP-1c.31

The increase in fatty liver in response to fish oil is probably attributable to the additive effects of several factors: increased PPARγ, high serum NEFA, and insulin resistance, as well as an absence of inhibition of fatty acid synthesis. Thus, the balance is toward fat accumulation.

In humans, overconsumption of sucrose/fructose or fat are 2 major causes of fatty liver. With the use of stable isotopes, it has been shown that in healthy men, a high-fructose diet increased the fractional de novo lipogenesis by 6-fold and that supplementation with fish oil (7.2 g/day) partially prevents this increase.32 It is likely that most humans are responders to sucrose/fructose overconsumption and that fish oil supplementation inhibits sucrose-induced fatty liver. With respect to HF diet–induced fatty liver, patients with fatty liver have been shown to consume significantly more saturated fat per day than control subjects matched for body mass index.33 The amount of hepatic fat appears to be related to the amount of fat in the diet rather than to endogenous fat deposits in obese women,34 suggesting that increased postprandial TG may favor fatty liver formation. In the current study, saturated fat–rich butter favored fatty liver formation more than did safflower oil, with greater postprandial TG concentration (Table 5).

The 10 en% fish oil (2.9 en% n-3 fatty acid) used in this study is comparable to levels consumed by humans. It is likely that fish oil supplementation is most effective in preventing fatty liver in those overfeeding sucrose, but not in those in overfeeding fat. To prevent fatty liver in a HF diet, it may be effective to inhibit intestinal absorption of dietary fat.

Ancillary