Single Nucleotide Polymorphism (–468 Gly to Ala) at the Promoter Region of Sterol Regulatory Element-binding Protein-1c Associates with Genetic Defect of Fructose-induced Hepatic Lipogenesis

To evaluate the genetic susceptibility to metabolic disorders induced by high fructose diet, we investigated the metabolic characteristics in 10 strains of inbred mice and found that they were separated into CBA and DBA groups according to the response to high fructose diet. The hepatic mRNA expression of the sterol regulatory element-binding protein-1 (SREBP-1) in CBA/JN was remarkably enhanced by high fructose diet but not in DBA/2N. Similar results were observed in primary hepatocytes after exposure to fructose. The nucleotide sequence at –468 bp from the putative starting point of the SREBP-1c gene was adenine in the DBA group while it was guanine in the CBA group. In hepatocytes from CBA/JN, the activity of CBA-SREBP-1c promoter was significantly increased by 2.4- and 2.2-fold, in response to 30 mm fructose or 10 nm insulin, respectively, whereas the activity of DBA-SREBP-1c promoter responded to insulin but not to fructose. In hepatocytes from DBA/2N, both types of SREBP-1c promoter activities in response to insulin were attenuated. Furthermore, electrophoretic mobility shift assay revealed an unidentified nuclear protein bound to the oligonucleotides made from the region between –453 to –480 bp of the SREBP-1c promoter of CBA/JN but not to the probe from DBA/2N. Thus, in DBA/2N, the reduced mRNA expression of SREBP-1 after fructose refeeding appeared to associate with two independent mechanisms, 1) loss of binding of unidentified proteins to the region between –453 to –480 bp of the SREBP-1c promoter and 2) impaired insulin stimulation of SREBP-1c promoter activity.

A high fructose diet in rats induces metabolic derangements similar to those that accompany with metabolic syndrome (4,5), and rats fed high fructose diet have been used as animal models for metabolic syndrome (6,7). We previously reported that the high fructose diet up-regulated hepatic expression of the sterol regulatory element-binding protein-1 (SREBP-1), 1 a key transcription factor for hepatic expression of lipogenic enzymes, but down-regulated the expression of peroxisome proliferator-activated receptor ␣ (PPAR␣), a ligand-activated nuclear receptor for expression of enzymes involved in fatty acid oxidation (7). These alterations in the expression of the transcription factors may play a central role in the pathogenesis of metabolic derangements in rats fed a high fructose diet. The molecular mechanisms, however, by which fructose induces the SREBP-1 gene expression and suppresses the PPAR␣ gene expression are not yet fully understood. Likewise, the genetic predisposition to metabolic disorders in response to a high fructose diet has not yet been determined. Therefore, we hypothesized that inducibility of SREBP-1 and PPAR␣ expressions in response to a high fructose diet could be genetically determined.
To investigate the genetic heterogeneity in the regulation of hepatic SREBP-1 and PPAR␣ gene expressions after consumption of high fructose diet, we selected two inbred mouse strains, one of which is highly responsive to a high fructose diet. We found that there were marked differences in CBA/JN mice and DBA/2N mice between the hepatic mRNA expressions of SREBP-1 in response to a high fructose diet. We also observed a single nucleotide mutation in DBA/2N mice from guanine to adenine at Ϫ468 bp from the putative starting point of the SREBP-1c gene, which caused impaired activation of SREBP-1c gene transcription in response to fructose. These results indicate that the loss of nuclear protein binding to the specific site at the promoter region of SREBP-1c may cause impaired hepatic lipogenesis in mice. zuoka, Japan). The mice were housed in an environmentally controlled room with a 12-h light/dark cycle and provided free access to a laboratory diet and water. The animals were divided into two groups (a control diet group and a high fructose diet group) and pair-fed for 8 weeks. The control diet (Oriental Yeast, Tokyo, Japan) consisted of 58% carbohydrate (no fructose), 12% fat, and 30% protein (energy percent of diet). The high fructose diet (oriental yeast) contained 67% carbohydrate (98% of which was fructose), 13% fat, and 20% protein. The day before the experiment we withdrew the food from all animals at 20:00. The next day half of the animals were fed either the control diet or the high fructose diet in the dark from 6:00 to 8:00 before the experiment, and the other half were kept in the fasting state. At 10:00, after 10 mg/kg intraperitoneal pentobarbital injection and under deep anesthetization, the liver and the epididymal fat were excised, immediately frozen in liquid nitrogen, and stored at Ϫ80°C. Blood samples were also taken for several blood tests such as triglyceride, total cholesterol, blood sugar, and insulin. RNA and DNA were extracted from the frozen samples. Shiga University of Medical Science Animal Care Committees approved all experiments.
Mouse Primary Hepatocytes-Mouse primary hepatocytes were isolated by the collagenase method with minor modifications (9). Under deep anesthetization the liver of each mouse was perfused in situ via the portal vein with 150 ml of Krebs-Ringer buffer followed by 100 ml of Krebs-Ringer buffer containing collagenase (Sigma-Aldrich). The cells were dispersed in an equal volume of ice-cold William's E medium (Sigma-Aldrich). The cells were precipitated and washed twice at 4°C with the same medium. Aliquots of 1 ϫ 10 6 cells in William's E medium supplemented with 5% (v/v) fetal calf serum, 1 nM insulin, 100 nM triiodothyronine, 100 nM dexamethasone, 100 units/ml penicillin, and 100 g/ml streptomycin were plated onto 6-well rat collagen I-coated dishes (Asahi Techno Glass, Chiba, Japan). After incubation for 2 h at 37°C in 9% CO 2 , the cells were incubated with William's E medium supplemented with 10% (v/v) fetal calf serum, 1 nM insulin, 1 nM triiodothyronine, 100 nM dexamethasone, 100 units/ml penicillin, and 100 g/ml streptomycin.
Northern Blot Analysis-Total hepatic RNA was isolated from the livers with TRIzol reagent (Invitrogen) after perfusion of ice-cold phosphate-buffered saline (Ϫ) in situ via the portal vein. In cases of primary hepatocytes they were starved in William's E medium supplemented with 0.75% bovine serum albumin, 100 units/ml penicillin, and 100 g/ml streptomycin followed by 6 h with insulin (100 nM), mannitol (30 mM), or fructose (30 mM). Total RNA was isolated with TRIzol reagent. Then 10 -30 g of RNA samples were run on a 1.2% agarose gel containing formaldehyde and transferred onto a nylon membrane (Nytran N; Schleicher & Schuell). The cDNA probes for Northern blot analyses were generated as previously described (7). The probes were labeled with [␣-32 P]dCTP (Amersham Biosciences) using a labeling kit (Takara, Shiga, Japan), hybridized to ultraviolet cross-linked blots overnight at 68°C in the hybridization buffer (Perfecthyb; Toyobo, Tokyo, Japan), and then washed at 68°C over 40 min with 1ϫ salinesodium citrate and 0.1% SDS. The blots were exposed to Kodak Biomax (Eastman Kodak Co.) film at Ϫ80°C. The signal was quantified with a densitometer, and loading differences were normalized to the signal generated with a probe for 18 S ribosomal RNA (8).
Sequencing of SREBP-1c Promoter Region-DNA sequences were analyzed by the dideoxynucleotide chain termination method.
Cell Transfection and Luciferase Assays-Primary hepatocytes were prepared as described. All transfections were performed using Superfect (Invitrogen) according to the manufacturer's instructions. Transfections were carried out with 1 g of pRSV-␤-gal expression plasmid and 2 g of each reporter plasmid. After the transfection for 18 h, the medium was replaced with William's E medium supplemented with 0.75% bovine serum albumin for starving followed by 8 h with either insulin (10 nM), mannitol (30 mM), or fructose (30 mM). The results were quantified with a luminometer and normalized to the ␤-galactosidase activity measured in the extract of the cells.
Preparation of Nuclear Protein Extracts-Nuclear protein of HeLa (HeLaScribe) was purchased from Promega (Madison, WI). Nuclear protein extract from the livers of mice and primary hepatocytes were isolated according to the procedure of Gorski et al. (9). The nuclear extract was suspended in 20 mM HEPES (pH 7.9), 330 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 25% glycerol, 0.5 mM dithiothreitol, and 0.2 mM phenylmethylsulfonyl fluoride, and aliquots were frozen in liquid nitrogen and stored at Ϫ80°C.
Western Blot Analysis-For Western blot analysis, whole cell lysates (50 g of protein per lane) were denatured by boiling in Laemmli sample buffer containing 100 mM dithiothreitol and resolved by SDS-PAGE. Gels were transferred to nitrocellulose by electroblotting in Towbin buffer containing 20% methanol. For immunoblotting, membranes were blocked and probed with the specified antibodies. Blots were then incubated with horseradish peroxidase-linked second antibody followed by chemiluminescence detection, according to the manufacturer's (PerkinElmer Life Sciences) instructions.
Statistical Analysis-The data are expressed as the means Ϯ S.E. unless otherwise stated. Tukey-Welsh's step-down multiple comparison test was used to determine the significance of any differences among four or more groups. p Ͻ 0.05 was considered significant.  Table I. Compared with the control mice fed a normal laboratory chew, C3H/He showed significantly increased body weights after feeding of the high fructose diet for 8 weeks. The epididymal fat weights for all strains except the DBA/2N and DBA/1JN strains a showed significant increase after the high fructose diet for 8 weeks compared with those of the control animals. The levels of blood glucose did not significantly differ in the control and the high fructose diet groups for any strains.

Characteristics of Experimental Animals-In
As shown in Table I, the levels of serum triglycerides in the postprandial state were significantly elevated after feeding the control diet in the C3H/He and CBA/JN strains and after feeding the high fructose diet in the C3H/He, CBA/JN, and BALB/c strains. In addition, these three strains showed higher postprandial plasma insulin levels after feeding the high fructose diet than those after feeding the control diet. In contrast, in C57BL/6N, DBA/1JN, and DBA/2N, the levels of serum triglycerides in the postprandial state were not significantly elevated after feeding control or high fructose diet. Although the postprandial plasma insulin levels in these strains after feeding of either control or high fructose diet were increased compared with the corresponding fasting level, the postprandial insulin levels did not differ between the control and the high fructose groups, respectively.
The mRNA Expression of SREBP-1 in the Liver-To explore the molecular mechanisms of the differences in response to the high fructose diet, we compared the hepatic mRNA expressions of SREBP-1, a key transcription factor regulating fatty acid synthesis in various strains of mice. As shown in Fig. 1, A and B, the hepatic mRNA expression of SREBP-1 in the CBA/JN mice was increased after feeding. In the control-postprandial group, it was stimulated by 3.9-fold (p Ͻ 0.001) as compared with that of the control-fasting group. The fructose-fasting group showed stimulation by 1.9-fold (p Ͻ 0.05) as compared with the control-fasting group. Furthermore, the fructose-postprandial group showed enhancement by 23-fold (p Ͻ 0.001) compared with the control-fasting group. Thus, the fructose diet stimulated the level of SREBP-1 mRNA expression by 11-fold (p Ͻ 0.001) compared with fructose-fasting levels and by 6-fold (p Ͻ 0.001) compared with the control-postprandial group.
Similar changes in the hepatic mRNA expression of SREBP-1 were observed in the C3H/He strain ( Fig. 1, C and D). The hepatic mRNA expression of SREBP-1 in the C3H/He control-postprandial group was stimulated by 3.8-fold (p Ͻ 0.05) compared with that in the control-fasting group. The level in the fructose postprandial group was enhanced by 5.8-fold (p Ͻ 0.001) compared with the control-fasting group. As a result, a fructose diet stimulated the level of SREBP-1 mRNA expression by 3.5-fold (p Ͻ 0.001) compared with the fructosefasting level and by 1.5-fold (p Ͻ 0.001) compared with the control-postprandial group. In addition, compared with that in the control-fasting group, the hepatic mRNA expression of SREBP-1 in the BALB/c strain was also significantly increased by 4.2-fold (p Ͻ 0.05) or by 6.1-fold (p Ͻ 0.001), respectively, after feeding the control or the high fructose diet.
On the other hand, the hepatic mRNA expressions of SREBP-1 in the DBA/2N mice (Fig. 1, E and F) and the DBA/ 1JN mice (Fig. 1, G and H) were not affected by either the control or the high fructose diet. The hepatic mRNA expression of SREBP-1 in the C57BL/6N mice was increased significantly but to a lesser degree, i.e. the control postprandial group showed a 2.1-fold (p Ͻ 0.05) increase, and the fructose-post-prandial group showed a 2.5-fold (p Ͻ 0.01) increase compared with the control fasting group, respectively. However, hepatic SREBP-1 mRNA expressions in either fasting or postprandial state were not different between the control and the high fructose diet groups.
The Hepatic mRNA Expression of Fatty Acid Synthase (FAS)-To investigate the effect of increased hepatic SREBP-1 expression on the downstream activation of the enzyme, we examined the mRNA expression of hepatic FAS, one of the target genes of SREBP-1, by Northern blot analysis. The hepatic mRNA expression of FAS from postprandial CBA/JN mice fed the high fructose diet was increased by 6-fold (p Ͻ 0.001) over that of the postprandial mice fed the control diet and by 4-fold (p Ͻ 0.001) when compared with that of the fructose-fasting mice (Fig. 2, A  and B). In contrast, the hepatic mRNA expression of FAS in the DBA/2N mice was not significantly affected by either control diet or high fructose diet (Fig. 2, C and D).
The mRNA Expression of PPAR␣ in the Livers from the CBA/JN and the DBA/2N Mice-As shown in Fig. 3, A and B, the mRNA expression of PPAR␣, a key transcription factor for fatty acid oxidation, in the livers from the CBA/JN mice was decreased after intake of either the control (p Ͻ 0.01) or the high fructose (p Ͻ 0.01) diet. Similar significant reductions of the hepatic mRNA expression of PPAR␣ were also found in DBA/2N mice after intake of either diet (Fig. 3, C and D).
The mRNA Expression of SREBP-1 in Primary Cultured Hepatocytes-To study the direct effect of fructose on the expression of SREBP-1 in the liver, we examined the mRNA contents of SREBP-1 in primary cultured hepatocytes isolated from the CBA/JN (Fig. 4A) or the DBA/2N (Fig. 4B) mice in the presence or absence of 30 mM fructose. In this experiment we used cells cultured with 30 mM mannitol as a control for fructose-treated or insulin-treated cells, since 30 mM mannitol did not affect the mRNA expression of SREBP-1 in either CBA/JN or DBA/2N primary hepatocytes; the values for hepatocytes cultured with 5 mM glucose and with 30 mM mannitol were    other hand, the mRNA expression of FAS in primary hepatocytes from DBA/2N mice was not affected by 30 mM fructose (Fig. 5B). However, a 32% (p Ͻ 0.01) increase in mRNA expression of FAS was induced by 100 nM insulin; the effect of insulin on the FAS mRNA expression was reduced significantly (p Ͻ 0.01) compared with that on the hepatocytes from CBA/JN mice.
Single Nucleotide Polymorphism in SREBP-1c Promoter Region-We have cloned a 1.2-kilobase pair fragment of the 5Ј upstream region of SREBP-1c gene from each inbred mouse strain and found a single nucleotide polymorphism at Ϫ468 bp from the putative starting point of the SREBP-1c gene. The nucleotide at Ϫ468 bp in the C3H/HeN, C3H/He, C3H/HeJ, BALB/c, and CBA/JN strains is guanine, whereas it is adenine in the C57BL/6, C57BL/6J, C57BL/6N, DBA/1JN, and DBA/2N strains.
Luciferase Activities of SREBP-1c Promoters-To investigate the significance of the single nucleotide mutation from guanine to adenine at Ϫ468 bp of the SREBP-1c promoter, we analyzed the promoter activity using the luciferase reporter carrying the 1.2-kilobase pair SREBP-1c promoter region of the CBA/JN or DBA/2N mice with 5 individual experiments in a total of 12 determinations. The activity of SREBP-1c promoter in primary hepatocytes isolated from CBA/JN mice was significantly increased by 2.4-fold (p Ͻ 0.01) by exposure of the cells to 30 mM fructose (Fig. 6A), but no increase was observed in primary hepatocytes isolated from the DBA/2N mice (Fig. 6B). The activity of the SREBP-1c promoter from DBA/2N mice was not induced by 30 mM fructose in primary hepatocytes isolated from either the CBA/JN or the DBA/2N mice.
We also investigated the effect of insulin with 5 individual experiments in a total of 12 determinations. As shown in Fig.  6C, the activity of SREBP-1c promoter from either the DBA/2N or CBA/JN strain in hepatocytes isolated from the CBA/JN mice was significantly increased by 2.4-fold (p Ͻ 0.01) or by 2.1-fold (p Ͻ 0.01), respectively, on the exposure of the cells to 10 nM insulin. These results indicate that a Ϫ468 Gly to Ala single nucleotide polymorphism did not affect the effect of insulin to SREBA-1c promoter. The activity of SREBP-1c promoter from either DBA/2N or CBA/JN strain in primary hepatocytes isolated from DBA/2N was increased only by 1.5-fold (p Ͻ 0.01) or by 1.4-fold (p Ͻ 0.01), respectively, with the exposure of the cells to 10 nM insulin. Thus, the effect of 10 nM insulin on the activity of SREBP-1c promoter from either DBA/2N or CBA/JN was significantly reduced by 27% (p Ͻ 0.05) and 36% (p Ͻ 0.01), respectively, in the hepatocytes isolated from DBA/2N mice compared with those in the cells from the CBA/JN mice.
The Effect of Insulin on Akt Phosphorylation and PPAR␣ mRNA Expression in Primary Cultured Hepatocytes-Because the effect of insulin on the SREBP-1c promoter activities of both CBA/JN and DBA/2N in hepatocytes isolated from DBA/2N was reduced, we analyzed the level of phosphorylation of Akt to compare the differences of insulin signal transduction. As shown in Fig. 7A, 100 nM insulin stimulated the level of phosphorylation of Akt in both DBA/2N and CBA/JN primary hepatocytes to the same degree. We also examine the mRNA expression of PPAR␣, known as an insulin-regulated gene (12). Insulin decreased the level of mRNA expression on both DBA/2N and CBA/JN primary hepatocytes (Fig. 7B) equally. The values in medium without or with insulin were 1.0 Ϯ 0.1 and 0.3 Ϯ 0.2 (p Ͻ 0.001, n ϭ 7) for hepatocytes from CBA/JN mice and 1. 2 Ϯ 0.2 and 0.3-0.1 (p Ͻ 0.001, n ϭ 7) for hepato-cytes from DBA/2N mice. There was no significant difference between the strains.
EMSA of Nuclear Protein Binding to the Region at the Single Nucleotide Polymorphism-As shown in Fig. 8, A and B, EMSA of nuclear protein extracted from the livers of the CBA/JN mice revealed a specific band recognizing the oligonucleotides from Ϫ453 to Ϫ480 bp of the upstream region of SREBP-1c gene of the CBA/JN mice. In contrast, we could not detect this band clearly when we used the oligonucleotides from Ϫ453 to Ϫ480 bp of the upstream region of the SREBP-1c gene of DBA/2N mice as a probe (Fig. 8B). The band was hardly detectable with an EMSA of the nuclear protein extracted from the livers of DBA/2N mice using either CBA/JN or DBA/2N probe (Fig. 8, A  and B). As shown in Fig. 8C, this band disappeared with the addition of an excess of unlabeled CBA/JN probe in dose-dependent manner. DBA/2N probe also competed with CBA/JN probe but less effectively. To assess the loading difference of the amount of the nuclear protein samples, we compared the content of OCT-1, a ubiquitously expressed transcriptional factor, between the samples from CBA/JN and DBA/2N. As shown in Fig. 8D, the binding activity of the hepatic nuclear protein isolated from DBA/2N mice to the oligonucleotides containing the consensus sequence of OCT-1 did not differ from that from CBA/JN mice.
Because the nucleotide sequences of the CBA/JN probe contain sequences similar to that of AP4, 5Ј-CAGCTG-3Ј, we examined whether or not the band detected with CBA/JN oligonucleotides was AP4. As shown in Fig. 9A, the band detected with CBA/JN probe was not competed by unlabeled AP4 consensus oligonucleotides or by antibody against AP4. Moreover, no band was detected when we used AP4 consensus oligonucleotides as the probe. In contrast, using a nuclear protein sample from HeLa cells we could detect the band with the AP4 consensus nucleotide as shown in Fig. 9B. 1JN, C57BL/6, C57BL/6J, and C57BL/6N strains showed a lesser or no expression of such metabolic derangements. Both C3H and CBA strains are developed from a cross between a Bagg albino female and a DBA male, so they are rather closely related. BALB/c is established from a Bagg albino. The DBA/1 and DBA/2 strains are established from DBA, but they are regarded as different strains. The C57BL/6 strains have a less direct relationship compared with other strains (Mouse Genome Informatics, www.informatics.jax.org).
We previously reported that the high fructose diet induced metabolic disorders as well as overexpression of SREBP-1 mRNA and suppression of PPAR␣ mRNA expression in the livers of Sprague-Dawley rats (7). We also found that the mRNA expression of SREBP-1 was greatly increased in the livers of the CBA/JN and C3H/He mice after feeding the high fructose diet but not in the livers of the DBA/2N and the DBA/1JN mice. The difference in SREBP-1 expressions induced by the high fructose diet in these mice was also suggested by the alteration in the mRNA expression of FAS, a target gene of transcriptional regulation by SREBP-1 (13), although the actual change in FAS mRNA seen at 4 h after feeding may be due to a separate mechanism because SREBP-1 mRNA but not protein may be elevated at this time point. Interestingly, inhibition of the hepatic mRNA expression of PPAR␣ was observed in both the CBA/JN and DBA/2N mice fed the high fructose diet. These results indicate that the difference in postprandial serum triglyceride levels in CBA/JN and DBA/2N mice fed a high fructose diet is more closely associated with the SREBP-1 expression in liver rather than with PPAR␣. In our study the mice were pair-fed, and they consumed equal amounts of food. Therefore, it is reasonable to conclude that the genetic background appears to associate with these character-istics of metabolic syndrome, and the variation in enhancement of the hepatic SREBP-1 mRNA expression induced by high fructose diet is possibly determined genetically in these mice.
SREBP-1 consists of two isoforms, i.e. SREBP-1a and SREBP-1c (14,15). It has already been reported that SREBP-1c gene expression in the liver is transcriptionally regulated by glucose (16,17), insulin (18), cyclic AMP (19), and polyunsaturated fatty acid (20). In the present study, although the plasma glucose levels did not differ in the two strains, the fasting and the postprandial plasma insulin levels of the CBA/JN mice were significantly higher than those of the DBA/2N mice. The elevated plasma insulin level may affect the mRNA expression of SREBP-1 in the liver (21). Comparable results were also obtained with the primary cultured hepatocytes, isolated from the CBA/JN and DBA/2N strains, and exposed to the high fructose media or insulin. The mRNA expression of SREBP-1 in primary hepatocytes from DBA/2N mice showed the absence or reduced response to the high fructose medium or insulin, respectively. Those results indicate that the difference in the SREBP-1c mRNA expression between the CBA/JN and DBA/2N mice is at least to some extent explained by the difference in the responses to fructose or insulin. Thus, genetic alterations of transcriptional regulation at the promoter region of the SREBP-1c gene may explain the difference in the SREBP-1 mRNA expression in response to the high fructose diet between these two strains. It has been shown that the promoter activity of SREBP-1c is regulated by SREBP-1 itself (22) or the liver X receptor through a cluster of putative binding sites for several transcription factors (NF-Y, SRE, Ebox, and Sp1 sites) and liver X receptor element binding sites (23). However, the mechanisms of the stimulation of SREBP-1c promoter activity by fructose or glucose are not fully under- FIG. 9. EMSA assay was performed using nuclear protein isolated from either the liver of CBA/JN mice or HeLa cells. A, the mobility band was removed by the addition of a 100-fold molar excess of unlabeled CBA/JN probe. Instead, the 100-fold molar excess of the unlabeled AP4 failed to compete. The mobility band was not removed after the addition of anti-AP4 antibody. B, CBA/JN nuclear protein and AP4 oligonucleotides failed to form a complex. With HeLa nuclear protein, AP4 successfully formed the protein-DNA complex that was competed by a 100-fold molar excess of unlabeled AP4 oligonucleotides, and that was removed after the addition of AP4 antibodies. The position of AP4-specific antibody supershift species is indicated by the asterisk (*). stood. In fact, a stimulatory effect of glucose on the transcription of SREBP1-c was observed using a hepatoma cell line (24). In primary cultured hepatocytes, however, such an effect of glucose was not observed (7, 25, and 26). On the other hand, a stimulatory effect of fructose on SREBP-1 expression was observed in primary cultured hepatocytes (7) but not in a hepatoma cell line (24).
To investigate the mechanism by which fructose is involved in the activation of SREBP-1c promoter and the variations of the response to fructose or insulin in mice strains, we compared the nucleotide sequences of proximal SREBP-1c promoter including SRE complex and liver X receptor element sites among the strains. We found a difference in the nucleotide sequence at Ϫ468 bp from the putative transcriptional starting site of SREBP-1c. The nucleotide from the CBA/JN, C3H/He, and BALB/c mice at Ϫ468 bp is the same as that previously reported for European house mice (23), whereas the nucleotide from the DBA/2N mice at this site is adenine instead of guanine. This change of nucleotide was also observed in the DBA/ 1JN and C57BL/6 strains. Because we compared only the 1.2kilobase pair 5Ј-flanking regions of the SREBP-1c promoter genes among these mouse strains, it is possible that the single nucleotide polymorphism at Ϫ468 bp is in linage disequilibrium with another functional mutation of the gene. However, the significance of this single nucleotide polymorphism in the promoter region of the SREBP-1c gene is also underscored by the results of the studies on the promoter activity of SREBP-1c. We showed that the incubation of hepatocytes from the CBA/JN mice with medium high in fructose increased the activity of SREBP-1c promoter from the CBA/JN mice but not of that from the DBA/2N mice. Interestingly, the activity of SREBP-1c promoter from DBA/2N mice was increased by insulin to a similar extent as CBA/JN when it was transfected to the primary hepatocytes isolated from CBA/JN mice. In addition, in the presence of insulin, the activity of SREBP-1c promoter from CBA/JN mice in the hepatocytes from DBA/2N mice was significantly less than that in the hepatocytes from CBA/JN mice. These results suggest that in contrast to a possible role in the response to fructose, the single nucleotide polymorphism at Ϫ468 bp is not involved in the stimulation of SREBP-1c promoter by insulin. However, it should be noted that even though a significant increase in insulin levels was observed in DBA/2N mice after feeding of either the control or high fructose diet, the hepatic expression of SREBP-1 in the liver was not increased in the postprandial state, indicating that DBA/2N showed the insufficient response of the hepatic SREBP-1 mRNA expression of SREBP-1c to insulin. Thus, a reduced response of hepatic SREBP-1 expression to feeding in DBA/2N mice was caused at least in part by an impaired response to insulin. Concerning insulin effects, we found that insulin-induced phosphorylation of Akt, an important marker of insulin metabolic effects, and the suppression of PPAR␣ mRNA expression by insulin were not different between CBA/JN and DBA/2N mice. Therefore, DBA/2A mice have some specific impairment in the transcription of SREBP-1c in response to insulin although further study is needed to identify the impaired sites. The variations in strains of SREBP-1 expression in response to insulin may explain part of the differences in strains such as DBA/2N and C57BL/6N, which have the same single nucleotide polymorphism but show different hepatic SREBP-1 expressions after feeding.
Concerning the importance of the single nucleotide polymorphism at Ϫ468 bp in SREBP-1c gene, we also found a specific protein that recognizes the sequence from Ϫ453 to Ϫ480 bp of CBA/JN SREBP-1c promoter but not that of DBA/2N SREBP-1c promoter. A search on the data base of MatInspector program (www.genomatix.de/cgi-bin/matinspector/matinspector.pl) for the sequence from Ϫ453 to Ϫ480 bp of SREBP-1c gene isolated from CBA/JN but not from DBA/2N predicted the existence of the consensus DNA sequence to which AP4 can bind. However, the band we recognized using the CBA/JN probe was not competed by unlabeled AP4 probe and not recognized by antibody against AP4, suggesting AP4 was not involved with the formation of the band. The cloning of the binding protein is now under way in our laboratory.
In addition to the alteration of the sequence of the proximal promoter site in the DBA/2N mice, we observed that the activity of intact SREBP-1c promoter of the CBA/JN mice was not stimulated in primary cultured hepatocytes from DBA/2N mice which were exposed to 30 mM fructose. Consistently, we also found a defect in a protein band that recognizes the intact sequence from CBA/JN mice in the nuclear protein isolated from the livers of DBA/2N mice after fructose feeding. These results suggest that in DBA/2N mice at least two independent causes of the defect in binding activities, a promoter sequence and a defect of a binding protein, which might result in the loss of response to fructose refeeding or insulin exposure of the SREBP-1 mRNA expression. Cloning of this binding protein may shed light on the defect in the binding protein that recognizes the region between Ϫ453 to Ϫ480 bp of the SREBP-1c promoter in DBA/2N mice. However, recent reports on DBA/2 mice revealed a genetic alteration in glycoconjugating enzyme (27,28) regulated by glucose or insulin (29) that affects certain transcription factor activities (30). Thus, abnormality of glycosylation observed in DBA/2N mice might relate to the defect of the binding protein to the SREBP-1c promoter.
In summary, in screening 10 strains of inbred mice we found that the diet high in fructose induced such metabolic disorders as postprandial hypertriglyceridemia, hyperinsulinemia, and visceral fat accumulation in the CBA/JN, C3H/He, and BALB/c mice. In contrast, the DBA/2N, DBA/1JN, and C57BL/6N mice showed lesser or no response to the high fructose diet. These variations in response to the high fructose diet accorded with the difference in the hepatic mRNA expression of SREBP-1 but not with that of PPAR␣. In DBA/2N mice, the reduced mRNA expression of SREBP-1 after fructose refeeding appeared to associate with two independent mechanisms, 1) loss of binding of an unidentified protein to the region between Ϫ453 to Ϫ480 bp of the SREBP-1c promoter and 2) impaired insulin stimulation of SREBP-1c promoter activity.