Metabolic disorders (obesity, hypertriacylglycerolaemia, hypercholesterolaemia, hyperinsulinaemia, insulin resistance and hypertension) are a risk factor for CHD and atherosclerosis. They are also likely to result in fatty liver and cholesterol gallstones(Reference Reaven1–Reference Portincasa, Moschetta and Palasciano3). Hypercholesterolaemia is considered a major risk factor for atherosclerosis; and was found to cause hepatic damage(Reference Buyssens, Kockx and Herman4, Reference Jeong, Jeong and Do5). Although excess of dietary cholesterol reportedly contributes to inflammatory diseases such as atherosclerosis(Reference Kleemann and Kooistra6, Reference Kleemann, Verschuren and Erk7), proof that liver fibrosis and focal nodular hyperplasia (FNH) are caused by the long-term consumption of a high-cholesterol (HC) diet has not yet been obtained.
Day & James(Reference Day and James8) proposed the ‘two-hits’ theory for the progression of steatohepatitis: the first hit being hepatic steatosis and the second being steatohepatitis with inflammation. A HC diet causes vascular inflammation including atherosclerosis, but the pathogenesis of non-alcoholic steatohepatitis and FNH caused by the chronic consumption of a HC diet is not clear. Therefore, here we examined the effects of a HC diet taken long term on the progression of hepatic damage.
Materials and methods
Materials
Maize starch, casein, the mineral mixture American Institute of Nutrition (AIN)-76 and the vitamin mixture AIN-76 were purchased from Oriental Yeast Co., Ltd (Tokyo, Japan). The Triglyceride E-Test, Triglyceride-Test, Total Cholesterol E-Test and NEFA C-Test were obtained from Wako Pure Chemical (Osaka, Japan). The Adiponectin and Leptin ELISA Kits were purchased from Otsuka Pharmaceutical Co., Ltd (Tokyo, Japan) and Morinaga Institute of Biological Science Inc. (Yokohama, Japan), respectively. The monocyte chemoattractant protein-1 (MCP-1), TNF-α, transforming growth factor (TGF)-β1 and IL-1β ELISA Kits were from R&D Systems, Inc. (Minneapolis, MN, USA). 3-Hydroxy-3-methyl [3-14C]glutaryl-coenzyme A and [1-14C]acetyl-coenzyme A were obtained from Amersham Biosciences UK Limited (Little Chalfont, Bucks, UK). Other chemicals were of reagent grade.
Animals and diet
Male C57BL/6J mice, aged 4 weeks, were obtained from Japan SLC (Shizuoka, Japan). They were housed in a room with a 12 h light–dark cycle and controlled temperature and humidity. They were fed a standard laboratory diet (8 g water, 51·3 g crude carbohydrate, 24·6 g crude protein, 5·6 g crude lipid, 3·1 g crude fibre, 6·4 g mineral mixture and 1 g vitamin mixture per 100 g diet; Oriental Yeast Co., Ltd, Osaka, Japan) for 1 week to adapt to the lighting conditions. After this period of adaptation, they were allowed free access to water and the experimental diets listed in Table 1 for 25 or 55 weeks. A low-fat and low-sucrose diet was used as the control diet in the present study. Animal experiments were performed according to the ethical guidelines of the Animal Experimentation, Ehime University and Japanese Pharmacological Society, and guide for the care and use of laboratory animals of the National Institutes of Health. The animal experiments were approved by the Ethics Committee on Animal Experimentation, Ehime University (approval no. YA-8-1).
Table 1 Composition of experimental diets (g/kg)
* Mineral and vitamin mixtures (American Institute of Nutrition (AIN)-76) were purchased from Oriental Yeast Co., Ltd (Tokyo, Japan).
Plasma analysis
At 25 or 55 weeks, after fasting overnight, blood was obtained from each mouse by venous puncture under anaesthesia with Nembutal® (Dainippon Pharmaceutical Co. Ltd, Osaka, Japan), and then the mice were killed with an overdose of Nembutal®. Blood samples were chilled in test-tubes containing heparin and centrifuged to give plasma. The plasma was frozen at − 20°C before analysis. TAG and total cholesterol concentrations in the plasma were determined using the TAG E-Test and Total Cholesterol E-Test kits. The concentrations of MCP-1, TNF-α, leptin and adiponectin in the plasma were measured using the respective ELISA kits.
Oral glucose tolerance test
The oral glucose tolerance test was carried out at weeks 25 and 55. After at least 4 h of food deprivation, glucose (100 mg/mouse) was administered orally to the mice. Blood samples were taken from the tail 0, 10, 20, 30 and 60 min after glucose administration. The blood glucose concentrations were measured using GLUCOCARD® (GT-1640; Arkray Co., Kyoto, Japan).
Liver lipid contents and cytokine levels, and cholesterol content in gallstones
TAG and total cholesterol levels in liver were measured by the methods of Fletcher(Reference Fletcher9) and Zak et al. (Reference Zak, Dickman and White10). A piece of the liver was homogenised in the protein extraction reagent (T-PER®; Pierce Biotechnology, Inc., Rockford, IL, USA), and centrifuged. The TGF-β1 and IL-1β levels in the supernatant fraction were measured using the respective ELISA kits. Gallstones were collected from gallbladder and extracted with an acetone–ethanol (1:1, v/v) mixture, and then cholesterol contents in gallstones were measured by the methods of Zak et al. (Reference Zak, Dickman and White10).
Histological examination
Liver was fixed in 10 % (v/v) buffered formalin, embedded in paraffin, and sectioned. The liver segments were stained with haematoxylin and eosin and with the Azan stain. The main artery was embedded in Tissue-Tek® (OTC compound; Sakura Finetech Co. Ltd, Tokyo, Japan) and frozen, and frozen sections were examined for macrophages at the vessel wall by an immunoperoxidase technique using an anti-macrophage antibody (clone MAC 387; Lab Vision Co., Fremont, CA, USA).
Enzyme activity
3-Hydroxymethylglutaryl-CoA reductase
The liver microsomal fraction was separated by the method of Daniele et al. (Reference Daniele, Rajas and Payrastre11). Briefly, the liver was cut and homogenised immediately on ice in 10 mm-HEPES buffer (pH 7·2) containing 0·25 m-sucrose and protease inhibitor, using a Teflon glass homogeniser. The homogenate was centrifuged at 9000 g for 20 min at 4°C. Then the supernatant fraction was further centrifuged at 105 000 g for 60 min at 4°C. The microsomal pellet was suspended in 0·1 m-phosphate buffer (pH 7·4), and added to a reaction mixture containing 0·128 mm-3-hydroxymethylglutaryl (HMG)-CoA ([14C]HMG-CoA, 144 MBq/mmol), 1 mm-NADPH, 10 mm-dithiothreitol and 10 mm-EDTA in 0·12 m-phosphate buffer (pH 7·4). The mixture was incubated at 37°C for 15 min, added to 2 m-HCl, and further incubated for 15 min to allow lactonisation of the mevalonic acid produced. This reaction mixture was spotted on a silica gel 60 F254 TLC plate (Merck, Germany), and developed with benzene–acetone (1:1, v/v)(Reference Shapiro, Imblum and Rodwell12). The radioactive spot corresponding to mevalonate lactone was measured using a BAS 1000 image analyser (Fuji Film, Tokyo, Japan). The HMG-CoA reductase activity was expressed as nmol mevalonic acid per mg protein.
Fatty acid synthase
Liver was homogenised on ice in 10 mm-HEPES buffer (pH 7·2) containing 0·25 m-sucrose and protease inhibitor, and centrifuged at 12 000 g for 10 min at 4°C. The activity was determined by measuring the incorporation of [14C]acetyl-CoA into fatty acids(Reference Tijburg, Maquedano and Bijleveld13). The supernatant fraction was added to a reaction mixture containing 0·062 mm-[14C]acetyl-CoA (148 MBq/mmol), 0·195 mm-malonyl-CoA, 0·5 mm-NADPH, 0·45 mm-dithiothreitol, 1 mm-EDTA and 0·93 % bovine serum albumin in 100 mm-HEPES buffer (pH 6·8), and then incubated at 37°C for 10 min. The reaction was stopped by the addition of chloroform–methanol (2:1, v/v), and mixed with a vortex. After centrifugation, the supernatant fraction was removed by aspiration, and the residue was washed with water in acidic conditions twice. The residual radioactivity was quantified with a liquid scintillation counter. The fatty acid synthase (FAS) activity was expressed per mg protein.
Immunoblotting
Liver was homogenised in a solubilising buffer (50 mm-HEPES buffer (pH 7·4) containing 5 mm-EDTA, 150 mm-NaCl, 1 % TritonX-100, 0·5 % sodium deoxycholate and protease inhibitor) and then centrifuged at 12 000 g for 15 min at 4°C. The amount of protein in the supernatant fraction was measured using a protein assay reagent (Bio-Rad, Hercules, CA, USA). After being denatured at 99°C, the protein was resolved by 12 % SDS-PAGE. The gel was transferred to a polyvinylidene fluoride membrane, and the membrane was blocked with 5 % skimmed milk. The membrane was incubated with an anti-platelet-derived growth factor (PDGF) antibody (Novus Biological, Inc., Littleton, CO, USA; 1:1000). Immunoreactivity was visualised with alkaline phosphatase-conjugated goat anti-rabbit IgG (ICN Pharmaceuticals, Cleveland, OH, USA) and BCIP/NBT solution (Sigma, St Louis, MO, USA). The blots were measured using the ImageJ program (National Institutes of Health, Bethesda, MD, USA). The ratio is the intensity of the PDGF blots normalised to the blot of glyceraldehyde-3-phosphate dehydrogenase.
Statistical analysis
All values are expressed as the mean values with their standard errors. The Dunnett test was used to compare the two groups. Differences were considered significant at P < 0·05.
Results
Plasma lipid, glucose, monocyte chemoattractant protein-1, TNF-α, leptin and adiponectin levels in mice fed the high-cholesterol diet
The body weights and epididymal adipose tissue weights of mice fed the control diet and HC diet were similar at weeks 25 and 55 (Table 2). The mean food consumption/mouse per d for 55 weeks was not different between mice fed the control diet and the HC diet: 54·5 (sem 0·53) kJ/mouse per d (control diet) and 55·2 (sem 0·99) kJ/mouse per d (HC diet). The fasting plasma glucose levels were not significantly different between the groups at weeks 25 and 55. To confirm glucose tolerance, oral glucose tolerance tests were performed at weeks 25 and 55. There were no differences in plasma glucose levels after the oral administration of glucose between the two groups (data not shown). At week 25, plasma levels of total cholesterol were significantly higher in the mice fed the HC diet. Levels of plasma, NEFA, MCP-1, leptin and adiponectin were not significantly different between the groups at week 25, but plasma TAG and NEFA levels were significantly elevated in the mice fed the HC diet at week 55. MCP-1 levels of mice fed the HC diet long term (55 weeks) were markedly increased compared with those of mice fed the control diet. On the other hand, plasma leptin and adiponectin levels did not differ between the two groups at week 55. Plasma TNF-α was not detected in the control diet- and HC diet-treated mice (data not shown).
Table 2 Body and adipose tissue weight, and plasma levels of lipids, monocyte chemoattractant protein-1 (MCP-1) and adiponectin in mice fed a control diet or high-cholesterol (HC) diet for 25 or 55 weeks
(Mean values with their standard errors for six or seven mice per group)
* Mean value was significantly different from that of the mice fed the control diet (P < 0·05).
Hepatic lipid accumulation and gallstone formation induced by the high-cholesterol diet
At week 25, liver hypertrophy was observed, and liver weight was significantly increased in the mice fed the HC diet (Table 3). The hepatic accumulation of lipids was also markedly enhanced in these mice. In addition, consumption of the HC diet resulted in cholesterol gallstones. Hepatic hypertrophy induced by the consumption of cholesterol may have contributed to the hepatic lipid accumulation. At week 55, hepatic TAG levels were greater in mice fed the control diet than in mice fed the HC diet (Table 3).
Table 3 Liver weight, hepatic lipid contents and cholesterol gallstones in mice fed a control diet or high-cholesterol (HC) diet for 25 or 55 weeks
(Mean values with their standard errors for six or seven mice per group)
* Mean value was significantly different from that of the mice fed the control diet (P < 0·05).
Hepatic lipogenesis
FAS activity in the liver was greater in mice fed the control diet than in those on the HC diet at weeks 25 and 55 (Fig. 1). The activity of HMG-CoA reductase, a key enzyme of cholesterol synthesis, was lower in the mice fed the HC diet at weeks 25 and 55 (Fig. 1).
Fig. 1 Fatty acid synthase activity (a) and 3-hydroxymethylglutaryl-CoA (HMG-CoA) reductase activity (b) in the liver of mice fed a high-cholesterol (HC) diet or a control diet. Values are means (n 6–7), with standard errors represented by vertical bars. * Mean value was significantly different from that of the mice fed the control diet (P < 0·05).
Liver focal nodular hyperplasia and fibrosis, and macrophage accumulation into the vessel walls
Hepatic hyperplastic pseudotumours were found in the mice fed the HC diet for 55 weeks (Fig. 2(a)). Based on the histological analysis, the HC diet caused FNH, and lipid accumulation and fibrosis were observed around the hyperplastic region (Fig. 2(b) and (c)). The chronic intake of the HC diet for 55 weeks caused the accumulation of macrophages into the vessel walls (Fig. 2(d)).
Fig. 2 Livers (a), liver sections (b, c), and vessel wall sections (d) of mice fed a control diet or high-cholesterol (HC) diet for 55 weeks. In mice fed the HC diet, liver focal nodular hyperplasia (a, b: haematoxylin and eosin stain) and mild fibrosis (c: azan stain) were detected, and macrophage accumulation ( ↓ ) into the vessel walls was detected by immunohistochemistry using an anti-macrophage antibody (d).
Effects of the high-cholesterol diet on hepatic IL-1β, transforming growth factor-β1 and platelet-derived growth factor protein levels
As shown in Fig. 3(a) and (b), TGF-β1 and IL-1β levels did not differ between the two groups at weeks 25 and 55. However, PDGF-B protein levels were significantly increased in the mice fed the HC at weeks 25 and 55 (Fig. 3(c)).
Fig. 3 Transforming growth factor (TGF)-β1 (a), IL-1β (b) and platelet-derived growth factor (PDGF)-B (c) levels in the liver of mice fed a high-cholesterol (HC) diet or a control (Co) diet. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. Values are means (n 6–7), with standard errors represented by vertical bars. * Mean value was significantly different from that of the mice fed the control diet (P < 0·05).
Discussion
The body and adipose tissue weights, and the food consumption of the mice fed the control diet were similar to those of mice fed the HC diet (Table 2). Chronic consumption of the HC diet resulted in the hepatic accumulation of cholesterol and crystallisation in the gallbladder (Table 3). Therefore, these findings suggest that the influence on liver function may be due to dietary cholesterol in the diet composition rather than the difference of energy intake of the two diets. Blood and liver total cholesterol levels were increased by feeding the HC diet, and the increases of blood and hepatic total cholesterol might have been due to dietary cholesterol rather than endogenous cholesterol synthesis in the liver, because these mice showed a reduction of HMG-CoA reductase (Fig. 1(a)). In fact, the HMG-CoA reductase in the control diet-treated mice was significantly greater than that in the HC diet-treated mice. HMG-CoA reductase was not significantly different between feeding for 25 and 55 weeks in the control diet-fed mice. On the other hand, HMG-CoA reductase in the HC diet-fed mice for 25 weeks tended to be greater than that in the HC diet-fed mice for 55 weeks. Therefore, it seems likely that the longer feeding HC diet results in the reduction of endogenous cholesterol synthesis by the supply of dietary cholesterol; consequently the differences for increasing rate in plasma total cholesterol may have occurred during week 25 to 55 in the control and the HC diet-fed mice. Thus, cholesterol synthesis is controlled by the absorption of dietary cholesterol. On the other hand, hepatic FAS activity was increased in mice fed the control diet, and hepatic TAG accumulation was observed at week 55. It has been reported that a high-carbohydrate (starch) diet is implicated in stimulating FAS activity and lipogenesis(Reference Kabir, Rizkalla and Quignard-Boulangé14, Reference Jentsch, Derno and Klein15). In fact, the control diet used in the present study is a high-carbohydrate diet including 46·5 % (w/w) carbohydrate. We found that hepatic FAS activity in the control diet-fed mice was significantly (P < 0·05) greater than that in the HC diet-fed mice, and hepatic TAG accumulation was significantly (P < 0·05) higher in mice fed the control diet than in mice fed the HC diet at week 55. These findings suggested that the higher levels of FAS activity in mice fed the control diet stimulated the hepatic TAG accumulation.
An excess of dietary cholesterol contributes to inflammatory diseases such as atherosclerosis(Reference Kleemann and Kooistra6). In the present study, chronic consumption of the HC diet caused hepatic fibrosis and FNH (Fig. 2(b)) and macrophages to accumulate in vessel walls (Fig. 2(d)). MCP-1 is an angiogenic factor(Reference Bernardini, Ribatti and Morbidelli16, Reference Kim, Byeon and Hong17) involved in the progression of atherosclerosis associated with inflammation in vascular disease(Reference Egashira, Zhao and Kataoka18–Reference Ni, Kitamoto and Ishibashi20). Furthermore, we reported that a high-fat diet caused obese mice to develop insulin resistance associated with higher plasma MCP-1 levels(Reference Sumiyoshi, Sakanaka and Kimura21). The progression of obesity is associated with the accumulation of macrophages and overexpression of MCP-1 in adipose tissue; which induces inflammation and insulin resistance(Reference Sartipy and Losktoff22, Reference Weisberg, McCann and Desai23). Kanda et al. (Reference Kanda, Tateya and Tamori24) reported that the overexpressing MCP-1 gene in adipose tissue was caused by a high-fat diet (33 weeks) in C57BL/6J mice, and that MCP-1 contributes to macrophage infiltration into adipose tissue, insulin resistance, and hepatic steatosis with the accumulation of TAG in obesity. Thus, it seems likely that an increase in blood MCP-1 is closely associated with the progression of obesity and glucose intolerance(Reference Sumiyoshi, Sakanaka and Kimura21, Reference Sartipy and Losktoff22). Furthermore, it has been reported that chronic infection such as periodontal diseases induces hypertriacylglycerolaemia and increases serum MCP-1 levels in a manner that increases the risk of atherosclerotic CVD(Reference Uchiumi, Kobayashi and Tachikawa25). However, we found that elevation of the plasma MCP-1 level was independent of adiposity in mice fed a high-sucrose diet for a long period(Reference Sumiyoshi, Sakanaka and Kimura21). In the present study, consumption long-term (for 55 weeks) of the HC diet markedly increased the plasma MCP-1 level without affecting obesity and glucose intolerance, compared with that in mice fed the control diet (Table 2). It has been reported that MCP-1(Reference Kanno, Tazuma and Nishioka26–Reference Desai, Mariscalco and Tawil29) and leptin(Reference Ikejima, Okumura and Kon30–Reference Bertolani and Marra32) play a significant role in the development of progressive liver fibrosis. Conversely, adiponectin inhibits the proliferation of hepatic stellate cells and hepatic fibrosis through the activation of adenosine monophosphate kinase-activated protein kinase(Reference Adachi and Brenner33). Thus, MCP-1, leptin and adiponectin were closely associated with the development of liver fibrosis. In the present study, we found that plasma leptin and adiponectin levels did not differ between the control diet-fed group and the HC diet-fed group at week 55, but the plasma MCP-1 level was greater in the mice fed the HC diet. Therefore, it seems likely that the fibrosis and macrophage accumulation into the vessel walls caused by long-term consumption of a HC diet is due to an elevation of the plasma MCP-1 level brought about by chronic inflammation, and that the elevated MCP-1 may be due to increases in plasma TAG, and the elevated TAG may be a causal factor for the liver fibrosis and FNH. Isoda et al. (Reference Isoda, Sawada and Ayaori34) reported that IL-1Ra− / − mice developed severe non-alcoholic fatty liver disease and portal fibrosis following 20 weeks on an atherogenic diet when compared with wild-type mice. Furthermore, they found that mRNA levels of IL-1β and TGF-β were significantly elevated in the livers of IL-1Ra− / − mice(Reference Isoda, Sawada and Ayaori34). Liver fibrosis is the excessive accumulation of extracellular matrix proteins including collagen that occurs in most types of chronic liver diseases. There is substantial evidence that hepatic stellate cells are the main producers of collagen and extracellular matrix during liver fibrosis. PDGF and TGF-β are activated in the hepatic stellate cells, and it is well documented that the two cytokines cause the development of liver fibrosis(Reference George, Roulot and Koteliansky35–Reference Borkham-Kamphorst, van Roeyen and Ostendorf42). Studies have identified PDGF as the main mediator for proliferation and TGF-β as the most important cytokine stimulating fibrogenesis in hepatic stellate cells(Reference George, Roulot and Koteliansky35–Reference Borkham-Kamphorst, van Roeyen and Ostendorf42). As expected, hepatic IL-1β and TGF-β1 levels were greater in mice fed the control diet or HC diet than in those given a laboratory chow diet (Oriental Co., Osaka, Japan) (data not shown). Feeding on the HC diet long-term (55 weeks) caused hepatic FNH and fibrosis, but the control diet caused neither hepatic FNH nor fibrosis. Hepatic IL-1β and TGF-β1 levels did not differ between the control diet-fed group and the HC diet-fed group. Therefore, the hepatic FNH and fibrosis caused by chronic consumption of the HC diet could not be explained by the increase in IL-1β and TGF-β1 levels in the liver. PDGF-B had been the most potent mitogen in cultures of activated hepatic stellate cells(Reference Pinzani, Gesualdo and Sabbah43). Czochra et al. (Reference Czochra, Klopcic and Meyer44) reported that hepatic PDGF-B overexpression causes liver fibrosis without significantly up-regulating TGF-β1 expression. Long-term consumption of the HC diet caused an increase in the hepatic PDGF-B protein level (Fig. 3). In addition, we found FNH of the liver together with liver fibrosis in the HC diet-fed mice (Fig. 2). Although the pathogenesis of FNH in the liver remains unclear, it has been reported that TGF-β1 protein was strongly expressed in FNH of human liver(Reference Vogelbruch, Wellmann and Maschek45), and that liver FNH was closely associated with vascular malformation and liver fibrosis through inflammation(Reference Kondo46–Reference Bioulac-Sage, Laumonier and Cubel49).
Chronic consumption of a HC diet causes non-obese fatty liver, FNH and fibrosis with an elevation of the plasma MCP-1 level and an increase in hepatic PDGF expression without affecting plasma leptin and adiponectin levels, or hepatic IL-1β and TGF-β1 levels. It is suggested that the incidence and/or development of liver FNH and fibrosis caused by the chronic intake of a HC diet may be partly due to the increases in blood MCP-1 and TAG levels and/or hepatic PDGF-B expression in the absence of increase in leptin in the present study. Further studies are needed to clarify the processes of hepatic inflammation and damage caused by a diet rich in cholesterol.
Acknowledgements
Animal experiments were performed according to the ethical guidelines of the Animal Experimentation, Ehime University and Japanese Pharmacological Society, and guide for the care and use of laboratory animals of the National Institutes of Health. The animal experiments were approved by the Ethics Committee on Animal Experimentation, Ehime University (approval number: YA-8-1).
M. S. designed the experiment, performed all the experimental analysis and wrote the manuscript. M. S. contributed to the discussion of the manuscript. Y. K. conducted all the experimental work, wrote the manuscript and discussed it with the whole group of authors. All authors read and approved the final manuscript.
There were no sources of funding.
The authors declare that there are no conflicts of interest in respect of the present study.
