The Effect of Fasting on Hepatic Lipid Accumulation and Transcriptional Regulation of Lipid Metabolism Differs between C57BL/6J and BALB/cA Mice Fed a High-fat Diet

Abstract

The effects of fasting on hepatic lipid metabolism in mice fed a high-fat diet (HFD) are still unclear. After fasting, the degree of hepatic lipid accumulation differs between HFD-fed C57BL/6J (B6) and BALB/cA (BALB/c) mice. It is not clear whether this difference is due to sensitivity to fasting or HFD. The aim of this study is to elucidate this difference among strains. After nine weeks of HFD feeding, both B6 and BALB/c mice showed moderate hepatic steatosis. However, after a subsequent twenty-hour fast, the hepatic lipid accumulation was markedly decreased in B6 but not in BALB/c mice. Moreover, the mRNA expression of a transcription factor, Srebp1(regulates hepatic lipid metabolism), and its target genes—malic enzyme, acetyl-CoA carboxylase, fatty acid synthase(regulate fatty acid synthesis), and glycerol-3-phosphate acyltransferase(regulates triacylglycerol synthesis)—were more markedly reduced in B6 than BALB/c mice. In conclusion, fasting may modify hepatic lipid accumulation in HFD-fed B6 and BALB/c mice differently. The difference may be partly owing to a marked downregulation of the expression of some lipid-metabolism–related genes in B6 mice. These results suggest that fasting per se has a significant effect on hepatic lipid accumulation in mouse strains. SREBP1 might play a role in this fasting effect.

Introduction

Hepatic lipid metabolism is influenced by the balance between the degradation and synthesis and/or import and export of triacylglycerol (TAG) and fatty acids. Liver incorporates plasma glucose and fatty acids. Glucose is degraded by the glycolytic system, and the acetyl CoA produced by this system is used for fatty acid synthesis. Fatty acids are degraded through β-oxidation, or esterified and then stored as TAG. TAG is degraded into fatty acids and glycerol in the liver, or directly exported to the extrahepatic circulation as very-low–density lipoprotein (VLDL) (Bender and Mayes 2006).

A high-fat diet (HFD) may induce hepatic TAG accumulation, owing to the import of excess amounts of fatty acids into the liver. The excess fatty acids are then esterified and stored as TAG. A methionine- or choline-deficient diet also induces hepatic TAG accumulation, which further induces a reduction in the synthesis of VLDL, and finally results in hepatic steatosis (Kirsch et al. 2003; Lombardi et al. 1968; Powell et al. 2005) with an accumulation of hepatic TAG.

Both satiety and fasting are thought to change the balance of hepatic lipid metabolism. In a satiated state, the concentrations of circulating glucose and insulin are high, and the hepatic import of plasma glucose and synthesis of fatty acids are increased. Then, hepatic TAG is increased. Moreover, β-oxidation of fatty acid is inhibited. As a result, a lipogenic metabolic state occurs in the liver. On the other hand, in a fasted state, the concentrations of circulating glucose and insulin are low, which stimulates the degradation of TAG into fatty acids in adipose tissue. Then, plasma fatty acid levels are increased and they are incorporated into the liver. In the liver, fatty acids are degraded through β-oxidation, and ketone bodies are synthesized to complement energy shortages. Finally, TAG synthesis is inhibited in the liver. The metabolic state of the liver condition is lipolytic (Botham and Mayes 2006). However, in cases in which the quantity of fatty acid taken up exceeds the hepatic lipolytic capability, excess fatty acids are stored in the liver as TAG. Actually, hepatic TAG increases with fasting in mice (Lin et al. 2005).

We previously showed that HFD-induced hepatic lipid accumulation is mild in C57BL/6J (B6) mice but still moderate in BALB/cA (BALB/c) mice after a twenty-hour fast (Nishikawa et al. 2007). This result is inconsistent with a report that B6 mice fed a HFD suffered from moderate hepatic steatosis (West et al. 1992). Therefore, the hepatic lipid accumulation in HFD-fed mice would be influenced not only by fasting but also by mouse strain. In many studies of obesity, modifications of liver steatosis in HFD-fed mice during fasting have not been fully considered, and the differences of lipid metabolism with fasting among strains have not yet been taken into full account. Therefore, in the present study, we investigate the effect of fasting on HFD-induced hepatic lipid accumulation in B6 and BALB/c mice.

Materials and Methods

Animals

Four-week-old male C57BL/6J Jcl (B6) mice and BALB/cA Jcl (BALB/c) mice were used. The animals, purchased from CLEA Japan (Tokyo, Japan), were housed individually in plastic mouse cages and maintained under a 14:10-hour light:dark cycle. They were fed a standard rodent chow (CE-2, CLEA Japan, Tokyo, Japan) and water ad libitum for one week. Then, the mice were fed a HFD (High Fat Diet 32, CLEA Japan, Tokyo, Japan) for nine weeks. The composition of each of the two diets is shown in Table 1. One half of the mice of each strain were fasted for twenty hours in the ninth week of the experiment (fasted groups). The other half were not fasted (nonfasted groups). Before and after the fasting, the mice were weighed. After the fasting, blood was collected from the tail vein for measuring the plasma glucose concentration. All mice were exsanguinated from the caudal vena cava under ether anesthesia, and the liver and abdominal and subcutaneous fat tissues were collected and weighed. The ratio of fat to body weight was obtained by dividing abdominal and subcutaneous fat tissues weight by body weight. The left lateral lobe of the liver was fixed in 10% neutral buffered formalin for histopathology. The right and caudate lobes were stored at –80°C until used for the examination of gene expression. All procedures involving mice were conducted according to the guidelines for the care and use of laboratory animals approved by the Graduate School of Agricultural and Life Sciences of Tokyo University.

Calculation of Energy Intake

Diet intake was obtained by subtracting the weight of diet remaining from that initially supplied. Energy intake was calculated on the basis of 5.068 kcal/g and 3.404 kcal/g for the HFD and CE-2 diets, respectively (Table 1).

Blood Biochemistry

Serum samples were prepared by centrifugation of total blood at 600 to 1400 × g for ten minutes. Using commercial kits, serum cholesterol, serum triacylglycerol, serum free fatty acid, and serum insulin levels (Cholesterol E-test Wako, Triglyceride E-test Wako, NEFA C-test Wako, Wako, Osaka, Japan and Mouse Insulin ELISA, Mercodia, Uppsala, Sweden, respectively) were measured according to the manufacturer’s instructions.

Histopathological Examination

The left lateral lobe of the liver was embedded in paraffin and 4-μm–thick paraffin sections were stained with hematoxylin and eosin (HE).

Semiquantitative RT-PCR

Total RNA was extracted from the liver tissue using an ISO-GEN kit (Nippon Gene Co. Ltd., Tokyo, Japan). Briefly, the liver tissue was lysed in ISOGEN solution on ice, and a water-soluble layer was obtained after centrifugation, followed by phenol-chloroform extraction and ethanol precipitation. For the first strand cDNA synthesis, a reverse transcriptase reaction was carried out using 5 μg of total RNA, the oligo (dT) primer, and the SUPERSCRIPT Preamplification System (Invitrogen, Carlsbad, CA, USA).

RT-PCR was performed with pairs of oligonucleotide primers corresponding to the cDNA sequence of the mouse mRNA (Table 2). PCR was carried out with 1 or 2 μl of RT product in a 50-μl reaction mixture containing 50 pM of the sense and antisense primers, 1.25 U of rTaq, 10×PCR buffer, and dNTP mixture (Takara, Siga, Japan). This procedure was immediately followed by preheating at 94°C for two minutes, denaturation at 94°C for thirty seconds, annealing at the melting temperature (Table 2) for thirty seconds, and extension at 72°C for one minute using a thermal cycler (TP-400, Takara, Siga, Japan). PCR products were identified by electrophoresis on 2% agarose gels, followed by ethidium bromide staining. Fluorescent gel imaging was carried out using the ultraviolet-CCD video system Fas-III (Toyobo, Osaka, Japan). The band density was digitized and normalized by using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression as a reference.

Statistical Analysis

Results were represented as the mean ± standard deviation (SD). Student’s ttest was used for comparing fasted and non-fasted groups.

Results

Histopathological Examination of the Liver

The severity of hepatic lipid accumulation after nine weeks on the HFD was moderate in nonfasting, but mild in fasting B6 mice (Figure 1). By contrast, it was moderate in BALB/c mice of both the fasted and nonfasted groups.

Energy Intake and Body Weight before Fasting

There were no significant differences in energy intake and body weight between the fasted and nonfasted groups in either strain (Table 3).

Body and Liver Weight, and the Ratio of Fat to Body Weight after Fasting

Body weight and the ratio of fat to body weight were significantly decreased in B6 mice after fasting, whereas they were unchanged in BALB/c mice (Table 4). The weight of the liver was significantly decreased after fasting in both strains (Table 4).

Blood Biochemistry

Serum levels of free fatty acid, triglyceride, total cholesterol, and insulin, and the plasma glucose level were significantly lower in the fasted group than the nonfasted group in both strains (Table 5).

RT-PCR

To examine the changes concerning lipid metabolism in the liver, we collected liver samples after fasting and performed semiquantitative RT-PCR.Table 6and Figure 2show the genes’ names and their roles in lipid metabolism, respectively. The mRNA expressions of these genes were investigated.

Srebp1, Pparγ, Pparα, and Foxa2are transcription factors, which regulate lipid metabolism. The expression of Srebp1was significantly reduced by fasting in B6 mice, but not in BALB/c mice (Figure 3A). The expression of Pparγ, Pparα,and Foxa2did not change in either strain (Figure 3Bto 3D). That of Gpam, which synthesizes TAG, was significantly reduced by fasting in B6 mice, but not in BALB/c mice (Figure 3E). The expression of Lipc,which degrades TAG in the blood, did not change (Figure 3F). That of Mttp,which is needed for the release of TAG from the liver as VLDL, significantly increased with fasting in B6 mice, but not in BALB/c mice (Figure 3G). The expression of Lipe, which degrades TAG, did not change (Figure 3H).

The expression of CD36,which imports circulating free fatty acid into the liver, did not change (Figure 3I). Levels of Acadsand Cpt1a,which are enzymes involved in β-oxidation, did not change (Figures 3Jand 3K). Levels of Acaca, malic enzyme,and FASN,which synthesize fatty acid, were reduced in fasted B6 and BALB/c mice compared to nonfasted animals (Figures 3Lto 3N). The reductions were greater in B6 mice than in BALB/c mice (Acaca: B6, 63%; BALB/c, 39%) (malic enzyme: B6, 53%; BALB/c, 38%) (FASN: B6, 58%; BALB/c, 46%).

Acyl-coA synthetases (ACSs) are needed for both anabolic and catabolic reactions. The expression of Acsl1significantly increased in fasted compared to nonfasted BALB/c mice, but did not change in B6 mice (Figure 3O). The expression of Acsl3significantly decreased in both strains with fasting (Figure 3P). That of Acsl5was not changed by fasting (Figure 3Q).

Discussion

In the present study, the hepatic lipid accumulation decreased with fasting in B6 mice, but not in BALB/c mice. Before fasting, energy uptake and body weight were the same between the fasted and nonfasted groups of both strains. Therefore, HFD-induced hepatic lipid accumulation may be affected by fasting, and the result differs between B6 and BALB/c mice.

Fasting induced decreases in circulating fatty acid, TAG, cholesterol, and glucose levels in both B6 and BALB/c mice. On the other hand, body weight and the ratio of fat weight to body weight were decreased in B6 mice, but not in BALB/c mice. These results suggest systemic differences in the reaction to fasting between HFD-fed B6 and BALB/c mice. Such strain differences of hepatic lipid accumulation are considered to be owing to differences in the expression of genes concerning lipid metabolism in the liver (Desvergne et al. 2006).

The mRNA expression of Gpam, which is involved in TAG synthesis, was significantly reduced in B6 mice, but not in BALB/c mice. This reduction may directly reflect the difference in hepatic lipid accumulation between B6 and BALB/c mice. It was reported that there was less hepatic TAG in Gpam-deficient mice than in wild-type mice (Hammond et al. 2002; Hammond et al. 2005). Moreover, it has also been shown that an increase in hepatic TAG was induced by overexpression of Gpamin mice (Linden et al. 2006). In rat hepatocytes, excess Gpamexpression leads to a decrease in β-oxidation and an increase in glycerolipid synthesis (Linden et al. 2004). Considering these findings, the fasting-induced reduction of GpammRNA expression may contribute to the difference in hepatic lipid accumulation between B6 and BALB/c mice (Figures 2and 4).

ACACA, malic enzyme, and FASNare enzymes related to fatty acid synthesis, using Acetyl-CoA as a substrate (Brownsey et al. 2006; Wakil 1989). In the present study, we demonstrated that the fasting-induced reduction in the mRNA expression of such enzymes was greater in B6 mice than BALB/c mice, suggesting that fatty acid synthesis in B6 mice decreases during fasting and induces a decrease of TAG synthesis, followed by less hepatic lipid accumulation (Figures 2and 4). Indeed, inactivating hepatic ACACA enzyme induces a decrease in the concentration of TAG in the liver (Mao et al. 2006). Additionally, inhibition of AcacamRNA expression improves dietary-induced hepatic steatosis (Savage et al. 2006).

ACS is needed for both anabolic and catabolic reactions, however details are not clear. In the present study, fasting-induced increases in Acsl1and Acsl3mRNA expression differed between B6 and BALB/c mice. This finding may reflect the differences in lipid metabolism between the two strains.

We showed that the expression of Srebp1mRNA was reduced by fasting in B6 mice. This finding is consistent with results reported by another group (Horton et al. 1998). Srebp1is a transcription factor associated with lipid metabolism, and it regulates the expression of many lipid-metabolism–related genes. Actually, Gpam, Acaca, FASN,and malic enzymeare the target genes of Srebp1(Horton et al. 2002). Therefore, the reduction in the mRNA expression of these genes in the present study may be partly caused by the reduction in Srebp1mRNA expression (Figures 2and 4).

In rats, the Srebp1mRNA level decreased with the injection of streptozotocin which abolished insulin secretion, and rose again with the injection of insulin (Shimomura et al. 1999). This finding suggests that the plasma insulin level would affect Srebp1mRNA expression. In the present study, plasma insulin concentrations decreased with fasting in both B6 and BALB/c mice. However, Srepb1mRNA expression was reduced by fasting in B6 mice, but not in BALB/c mice. This finding suggests that an insulin-associated reduction in Srebp1mRNA expression occurs only in B6 mice. Insulin resistance in B6 mice may be one possible cause of this impairment. However, we previously showed that glucose intolerance is similar between B6 and BALB/c mice (Nishikawa et al. 2007), suggesting that insulin sensitivity would not differ between B6 and BALB/c mice. Nevertheless, this result cannot completely exclude a difference in insulin sensitivity between B6 and BALB/c mice. Because previous glucose tolerance testing showed systemic glucose intolerance, insulin sensitivity and insulin signaling pathways under abnormal conditions (fasting or diabetes) differ between tissues, including skeletal muscle, adipose tissue, and liver (Heijboer et al. 2005; Kim et al. 1999; Rondinone et al. 1997), and lipid metabolism and glucose metabolism are regulated by different insulin-signaling pathways (Wolfrum et al. 2004). More investigations are needed to clarify the mechanism that leads to the strain differences.

PPARα regulates the expression of some genes related to β-oxidation (Aoyama et al. 1998; Leone et al. 1999). Hashimoto (2000)showed that PPARα-deficient mice developed severe fatty liver when fasting. This suggests that PPARα is crucial for lipid metabolism in the fasting condition. In the present study, however, the expression of PPARα mRNA did not differ between B6 and BALB/c mice, suggesting PPARα is not involved in the difference in hepatic lipid accumulation between B6 and BALB/c mice.

The activity of an enzyme is regulated not only by mRNA expression but also by reversible phosphorylation and allosteric activation of the protein by an allosteric modifier (Kennelly and Rodwell 2006). For example, the activity of Acaca, an allosteric enzyme, is stimulated by citric acid and inhibited by acetyl-CoA. Moreover, glucagon and epinephrine inhibit Acaca activity through phosphorylation of the enzyme, and insulin stimulates Acaca activity through dephosphorylation. Another example is CPT1, which is inhibited by malonyl-CoA produced by Acaca. In the present study, only the expression of lipid metabolism-related enzymes was examined, and the activities of the enzymes were not measured. The activities of each enzyme should be examined.

In conclusion, we showed that fasting induced differences in hepatic lipid accumulation between B6 and BALB/c HFD mice. This difference may be partly owing to the marked reduction in levels of Srebp1, Gpam, Acaca, malic enzymeand FASNcaused by fasting in B6 mice (Figure 4). The results suggest that fasting per se had a significant effect on HFD-induced hepatic lipid accumulation, which depends on the strain of mouse.