Xanthohumol Improves Diet-induced Obesity and Fatty Liver by Suppressing Sterol Regulatory Element-binding Protein (SREBP) Activation*

Background: Flavonoids are naturally occurring compounds and are known to regulate various transcription factors. Results: A prenylated flavonoid in hops, xanthohumol, improves diet-induced obesity and fatty liver by suppressing SREBP activation. Conclusion: Xanthohumol is a novel SREBP inactivator. Significance: Xanthohumol can be used as a nutraceutical food or pharmacological agent for improving metabolic syndrome. Sterol regulatory element-binding proteins (SREBPs) are key transcription factors that stimulate the expression of genes involved in fatty acid and cholesterol biosynthesis. Here, we demonstrate that a prenylated flavonoid in hops, xanthohumol (XN), is a novel SREBP inactivator that reduces the de novo synthesis of fatty acid and cholesterol. XN independently suppressed the maturation of SREBPs of insulin-induced genes in a manner different from sterols. Our results suggest that XN impairs the endoplasmic reticulum-to-Golgi translocation of the SREBP cleavage-activating protein (SCAP)-SREBP complex by binding to Sec23/24 and blocking SCAP/SREBP incorporation into common coated protein II vesicles. Furthermore, in diet-induced obese mice, dietary XN suppressed SREBP-1 target gene expression in the liver accompanied by a reduction of the mature form of hepatic SREBP-1, and it inhibited the development of obesity and hepatic steatosis. Altogether, our data suggest that XN attenuates the function of SREBP-1 by repressing its maturation and that it has the potential of becoming a nutraceutical food or pharmacological agent for improving metabolic syndrome.

mice (9,10), and liver-specific SREBP-1c transgenic mice have displayed hepatic steatosis, hypertriglyceridemia, and insulin resistance (11). These studies indicate that the dysregulation of SREBP-1c contributes to the pathogenesis of these diseases.
In this study, we screened for SREBP modulators in natural food constituents and identified xanthohumol (XN) as an inactivator. XN is the most abundant prenylated flavonoid in hops (12) and is known as a putative cancer chemopreventive agent because of its antioxidant (13) and anti-inflammatory properties (14). Moreover, XN has been shown to inhibit triglyceride synthesis and apolipoprotein B secretion associated with the low activity of diacylglycerol acyltransferase 1 and microsomal triglyceride transfer protein (15). Several reports have recently shown that XN exerts anti-metabolic syndrome effects in vivo (16 -19); however, detailed mechanisms contributing to these effects remain to be clarified.
In this study, we demonstrated that XN affects SREBP processing; it interacted with Sec23/24 and blocked the sorting of the SCAP/SREBP complex into COP II vesicles. XN reduced the mature forms of SREBPs and de novo biosynthesis of fatty acid and cholesterol in cultured cells. Furthermore, XN ameliorated obesity and fatty liver in mice fed a high fat diet (HFD), and this is associated with the down-regulation of hepatic SREBP processing.

cells were maintained in medium E.
Luciferase Assays-Huh-7/FAS-luc cells were plated in 12-well plates at a density of 1.0 ϫ 10 5 cells/well and cultured with medium A for 24 h. The cells were then switched to medium B for 16 h. After incubation for another 24 h in the absence or presence of 10 or 30 M XN, luciferase activity was measured as described previously (25). XN was dissolved in DMSO. The final DMSO concentration of the cultured medium was 0.1%. Normalized luciferase values were determined by dividing luciferase activity by the protein content in cell extracts quantified using the BCA protein assay (Pierce).
Immunoblotting-Cells and mouse liver were lysed in RIPA buffer (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% (v/v) Triton X-100, 0.5% (w/v) deoxycholate, and 0.1% (w/v) SDS) supplemented with a protease inhibitor mixture. The lysates were subjected to SDS-PAGE, transferred to a polyvinylidene difluoride membrane (Millipore, Billerica, MA), and probed with the antibodies indicated in the figure legends. The immunoreactive proteins were visualized using ECL (GE Healthcare) or Immobilon (Millipore, Billerica, MA) immunoblotting detection reagents. The signals on the membrane were detected by ImageQuant LAS 4000 mini (GE Healthcare).
Measurement of de Novo Fatty Acid and Cholesterol Synthesis-Huh-7 cells were plated in 12-well plates at a density of 2 ϫ 10 5 cells/well and cultured with medium A for 24 h. The cells were then switched to medium D for 16 h. After incubation for another 18 h in the absence or presence of 10 or 30 M XN, the cells were placed in fresh medium supplemented with 1.6 Ci/ml [ 14 C]acetate and incubated for 6 h. Furthermore, the cells were washed with phosphate-buffered saline (PBS), scraped, and collected by centrifugation. The cell pellets were mixed with 1 ml of 8 N potassium hydrate and 1 ml of ethanol. The mixture was heated at 100°C for 2 h and extracted two times with 2 ml of petroleum ether (cholesterol extract). Moreover, 2 ml of the lower aqueous layer was mixed with 1 ml of 12 N hydrochloric acid and then extracted two times with 3 ml of petroleum ether (fatty acid extract). The lipid extracts were then purified by thin layer chromatography on Silica Gel 60 (Merck) and quantified with a BAS2000 image analysis system (Fujifilm, Tokyo, Japan). Normalized fatty acid and cholesterol synthesis rates were determined by dividing the signals of 14 Clabeled fatty acid and [ 14 C]cholesterol by the amounts of total cellular protein quantified using the BCA protein assay.
Immunofluorescence Staining and Fluorescence Microscopy-CHO/pGFP-SCAP cells were plated on a poly-D-lysine Cellware 4-well culture slide (BD Biocoat) and cultured with medium E for 48 h. Moreover, the cells were switched to medium F for 3 h. After incubation for another 3 h in the absence or presence of 30 M XN, the cells were fixed with 3% (w/v) paraformaldehyde in PBS. The cells were permeabilized with 0.2% (v/v) Triton X-100 in PBS, blocked with 2% (w/v) bovine serum albumin in PBS, treated with primary antibody against GM130, and subjected to reaction with Cy3-conjugated secondary antibody. Then the cells were mounted on glass slides with Gel Mount aqueous mounting medium (Sigma). Images were obtained using an FV500 confocal laser scanning microscope (Olympus, Tokyo, Japan).
Cell Fractionation-Cell fractionation was performed according to a previously published procedure (27), with minor modifications. CHO-7 cells were plated in 100-mm dishes at a density of 2 ϫ 10 6 cells/dish and cultured with medium E for 24 h. The cells were then switched to medium F for 16 h. After incubation for another 3 h in the absence or presence of 30 M XN, the cells were washed with PBS, scraped, and collected by centrifugation. The cell pellets were resuspended in buffer B supplemented with 15% (w/v) sucrose and passed through a 25-gauge needle 20 times. The disrupted cell suspension was centrifuged at 3,000 ϫ g for 10 min, and the supernatant was diluted to 3 ml with buffer A supplemented with 15% sucrose. A discontinuous sucrose density gradient was prepared in a 13 PET tube (Hitachi Koki, Tokyo, Japan) by layering the following sucrose density solutions in buffer A: 3 ml of 45% sucrose, 5 ml of 30% sucrose, 3 ml of diluted supernatant on 15% sucrose, and 2 ml of 7.5% sucrose. The gradient solution was centrifuged at 24,000 rpm in a Himac CP85␤ (Hitachi Koki, Tokyo, Japan) for 1 h, and whole cells were separated into two fractions, light (14 -18%) and heavy (34 -38%). The light fraction contained the Golgi apparatus, plasma membrane, and endosome. The heavy fraction contained the ER and peroxisome. Aliquots of each fraction were subjected to immunoblotting to examine the protein levels of SCAP and SREBPs.
Detection of Binding Proteins for XN Beads-The preparation of XN beads and binding reactions were performed as described previously (28). CHO-7 cells were plated in 100-mm dishes at a density of 2 ϫ 10 6 cells/dish and cultured with medium E for 24 h. The cells were then switched to medium F for 16 h. After incubation for another 3 h in the absence or presence of 50 M XN, the cells were washed with PBS, scraped, and collected by centrifugation. The cell pellets were resuspended in binding buffer (10 mM Tris-HCl (pH 7.6), 50 mM KCl, 5 mM MgCl 2 , and 1 mM EDTA) supplemented with a protease inhibitor mixture and passed through a 25-gauge needle 20 times. The disrupted cell suspension was centrifuged at 15,000 ϫ g for 1 min, and the supernatant was collected. The cell lysate was precleared by incubation with control beads at 4°C for 1 h and then incubated with control or XN beads at 4°C for 12 h. The pelleted beads were washed three times with binding buffer, and the bound proteins were eluted with SDS sample buffer and were subjected to immunoblotting.
Immunoprecipitation-CHO-7 cells were plated in 100-mm dishes at a density of 2 ϫ 10 6 cells/dish and cultured with medium E for 24 h. The cells were then switched to medium F for 16 h. After incubation for another 3 h in the absence or presence of 30 M XN, the cells were harvested and lysed with Nonidet P-40 lysis buffer (50 mM Hepes-KOH (pH 7.6), 100 mM NaCl, 1.5 mM MgCl 2 , 1% (v/v) Nonidet P-40, 1 mM DTT) supplemented with a protease inhibitor mixture. The cell lysate was passed through a 25-gauge needle 20 times, rotated at 4°C for 1.5 h, and centrifuged at 20,000 ϫ g for 30 min. The supernatant was rotated with anti-SCAP (9D5) at 4°C for 1.5 h, and then protein A-Sepharose beads (GE Healthcare) were added. After rotation for 1.5 h, the pelleted Sepharose was washed three times with Nonidet P-40 lysis buffer, and the bound proteins were eluted with SDS sample buffer and were subjected to immunoblotting.
Mice and Diet-All experiments were performed according to the guidelines of the Animal Usage Committee of The University of Tokyo. Six-week-old C57BL/6J mice purchased from Clea Japan (Tokyo, Japan) were housed in the animal care facility under controlled conditions (12-h light/dark cycle, 23°C). The mice were provided free access to water and a standard chow diet (Labo MR Stock; Nosan Corp. Bio Department, Kanagawa, Japan) for the 1st week following their arrival.
Oral Administration of XN to Mice-Mice were randomized into three groups (n ϭ 5/group). Vehicle, 75 or 150 mg/kg body weight of XN, was administered to the mice by oral gavage once daily. After a 3-day treatment, livers were isolated, frozen in liquid nitrogen, and stored at Ϫ80°C until they were analyzed.
Administration of XN-supplemented HFD to Mice-HFD (60% energy supplied by fat; D12492) was purchased from Research Diet (New Brunswick, NJ). After feeding the HFD for 10 weeks, the mice were divided into three groups with similar average body weight and blood glucose levels (n ϭ 10 -11/ group). The mice were fed a HFD, 0.2% (w/w) or 0.4% (w/w) XN-supplemented HFD for 50 days. The body weight and food intake were measured every 2 days. Mice were then sacrificed, and their blood and liver were collected. Blood was collected in tubes with 0.1 M EDTA and centrifuged at 3,000 rpm for 30 min to obtain plasma. Livers were rapidly excised, frozen in liquid nitrogen, and stored at Ϫ80°C until further analysis.
Plasma Measurements-Plasma glucose, triglyceride (TG), and cholesterol concentrations were determined using kits purchased from Wako (Osaka, Japan). Plasma insulin and C-peptide concentrations were determined using an ELISA kit (Shibayagi, Gunma, Japan). Plasma lipoprotein profile was analyzed by LipoSEARCH (Skylight Biotech, Tokyo, Japan).
Measurement of TG and Cholesterol in Liver-Mouse livers (100 -150 mg) were homogenized in 4-ml of ice-cold chloroform/methanol (2:1, v/v) and incubated for 30 min at room temperature. Then 1 ml of 50 mM NaCl was added, and the mixture was centrifuged at 1,500 ϫ g for 30 min. The lower organic layer was collected, and 1-ml 0.36 M CaCl 2 /methanol (1:1, v/v) was added. This mixture was then centrifuged at 1,500 ϫ g for 10 min, and the upper aqueous layer was removed. The lower organic layer was washed again with 1-ml 0.36 M CaCl 2 /methanol (1:1, v/v) and diluted up to 5 ml with chloroform. Aliquots of the lipid extract (20 l for TG determination and 500 l for cholesterol) contained Triton X-100/chloroform (1:1, v/v) and were then air-dried. TG and cholesterol concentrations were determined using kits purchased from Wako (Osaka, Japan).
Measurement of TG in Feces-Dried mouse feces were homogenized and mixed with 25 ml of ethanol. The mixture was heated at 65°C for 2 h and filtered. The filtrate was diluted up to 25 ml with ethanol, and 5 ml of the fecal extract was evaporated to dryness under reduced pressure to remove ethanol. Isopropyl alcohol (2 ml) was added to the pellet (lipid extract), and TG concentration was determined using kits purchased from Wako.
Computed Tomography (CT) Scan Analysis-Mice were anesthetized with isoflurane and scanned with an x-ray CT scanner designed for experimental animals (LaTheta LCT-200; Hitachi Aloka Medical, Tokyo, Japan). Contiguous slice images were obtained at 2-mm intervals in the abdominal region (from the diaphragm to the bottom edge of the pubic bone) and at 0.1-mm intervals in the liver. The percentage of fat area in the abdomen and liver and the abdominal visceral and subcutaneous fat volumes were calculated using LaTheta software (version 3.00).
Oral Glucose Tolerance Test (OGTT)-After a 16-h fasting, glucose (2 g/kg body weight) was administered to mice by oral gavage. Blood glucose levels were measured from a nick in a tail vein before and at 15, 30, 60, 90, and 120 min after administration using a hand-held glucometer (Ascensia Breeze 2; Bayer Yakuhin, Osaka, Japan).
Statistical Analysis-All data are represented as mean Ϯ S.E. Statistical analysis was performed using the Ekuseru-Toukei Version 2.0 (Social Survey Research Information). One-way analysis of variance followed by the Bonferroni procedure was used to compare more than two groups. Differences were considered significant at p Ͻ 0.05.

Results
Identification of XN as a New SREBP Inactivator-To identify SREBP inactivators, we used a human hepatoma Huh-7 cell line that stably expressed a luciferase reporter gene driven by an SRE-containing FAS promoter (Huh-7/FAS-luc). To increase SREBP activity, the cells were depleted of sterols by incubating them with LPDS and fluvastatin, an agent that inhibits HMGCR and subsequently blocks cholesterol biosynthesis. The cells were then treated with ϳ100 naturally occurring food components for identifying potential SREBP inactivators. As a positive control, 25-HC, a well known inhibitor of SREBP processing, was tested and was shown to markedly lower luciferase activity (Fig. 1A). Using this screening system, we identified XN as a potent SREBP inactivator (Fig. 1, A and B).
Quantitative real time PCR analyses in Huh-7 cells revealed that XN treatment reduced mRNA levels of two SREBP-1 target genes, FAS and SCD1, involved in fatty acid synthesis, but not of ACC1 (Fig. 1C). Similarly, three SREBP-2 target genes, HMGCS, HMGCR, and SQS, involved in cholesterol synthesis, were down-regulated by XN (Fig. 1C). Consistent with these changes in gene expression, XN decreased the de novo synthesis of both fatty acid and cholesterol (Fig. 1D). Altogether, these results indicate that XN suppresses SREBP activity and reduces lipid synthesis.
XN Reduces Level of Mature Forms of SREBPs Independently of Insigs-SREBPs are synthesized as inactive precursors. When cellular cholesterol is low, the inactive precursors are proteolytically processed to the mature forms that function as active transcription factors. To determine whether XN affects cellular levels of mature SREBPs, Huh-7 cells were treated with XN for 3 h under low cholesterol conditions; the cell lysates were subjected to immunoblotting. Similar to 25-HC, XN reduced the level of the mature forms of both SREBP-1 and -2 ( Fig. 2A). We also examined the effect of three flavonoids with structures similar to XN, IXN, NG, and 8-PN, on cellular levels of mature SREBPs (Fig. 2, B-E). Although not to the extent of XN, IXN and 8-PN markedly decreased the level of the mature forms of SREBPs, whereas NG had little effect (Fig. 2F).
To further analyze this effect, we used a previously established Insig-1-and -2-deficient Chinese hamster ovary cell line (SRD-15). Previous studies have reported that 25-HC blocks SREBP processing by binding to Insigs, triggering their binding to SCAP and retaining the SCAP/SREBP complex in the ER (29). Consistent with these findings, we found that 25-HC did not affect the level of the mature forms of SREBPs in SRD-15 cells, whereas it suppressed SREBP processing in parental CHO-7 cells (Fig. 2G). XN reduced the level of the mature forms of SREBPs in both CHO-7 and SRD-15 cells (Fig. 2G). When endogenous Insig-1 and Insig-2 expressions in Huh7 cells were reduced with gene-specific siRNA (Fig. 2H), XNmediated suppression of mature forms of SREBPs was still observed, whereas 25-HC-mediated suppressions were partly suppressed (Fig. 2I). It should be noted that treatment with 25-HC and XN for 3 h decreased the expression of the Insig-1 gene possibly due to the suppression of SREBP activity. In addition, treatment with XN unexpectedly decreased the expression of the Insig-2 gene. These results indicate that the decrease in mature SREBPs by XN occurs without Insigs, acting in a manner different from 25-HC.
XN Suppresses SREBP Processing by Blocking ER-to-Golgi Transport of the SCAP/SREBP Complex-Next, we investigated the mechanism by which XN reduces the level of mature forms of SREBPs. One possibility is that XN promotes the protein degradation of mature SREBPs. To examine this, CHO-7 cells were transfected with an expression plasmid encoding a mature form of human SREBP-1a (amino acids 2-487; pCMV-3ϫFLAG-SREBP-1a(2-487), and a turnover of exogenous SREBP-1a was measured. In the presence of the translation inhibitor cycloheximide, the amount of mature SREBP-1a gradually reduced (Fig. 3A, lanes 1-4). 25-HC did not promote the degradation of mature SREBP-1a (Fig. 3A, compare lanes 1-4 and 5-8), consistent with a previous report (30). Moreover, XN did not promote the degradation of mature SREBP-1a (Fig. 3A,  compare lanes 1-4 and 9 -12). The quantification of these sig- nals is shown in Fig. 3B. Similar results were obtained when cells were transfected with an expression plasmid coding a FLAG-tagged mature form of human SREBP-1c (amino acids 2-463) (Fig. 3, C and D) or that of human SREBP-2 (amino acids 2-481) (Fig. 3, E and F). Next, we examined whether XN promotes the degradation of endogenous mature SREBPs. Sterol-depleted CHO-7 cells were preincubated with the serine protease inhibitor, AEBSF, that inhibits S1P, thereby inhibiting SREBP cleavage and maturation (31). In the presence of AEBSF, the level of mature SREBPs decreased in a time-dependent manner, and 25-HC and XN had no effect on the degradation (Fig. 3, G, compare lanes 1-4, 5-8, and 9 -12, and quantification in H and I). These data indicate that XN does not accelerate the degradation of mature SREBPs.
Another possibility in which XN may reduce the level of mature SREBPs is by inhibiting S1P and S2P activities. S1P and S2P are proteases that cleave the precursors of SREBPs in the Golgi apparatus, releasing the mature forms. To examine whether XN inhibits the proteolytic cleavage of SREBPs in the Golgi, we pretreated CHO-7 cells with BFA, a fungal metabolite that blocks the anterograde transport of proteins from the ER to the Golgi. BFA causes SREBPs to be proteolytically cleaved in the ER independently of sterols by translocating S1P and S2P from the Golgi back to the ER (32). Furthermore, the mature forms of SREBPs were present in cells treated with BFA, even in the presence of 25-HC or XN (Fig. 4). These data suggest that XN does not inhibit the activity of S1P or S2P. The precursors of SREBPs undergo two sequential cleavage events; however, before this, the SCAP/SREBP complex embedded in the ER must be transported to the Golgi. Therefore, we hypothesized that XN suppresses the ER-to-Golgi translocation of the SCAP/SREBP complex. To investigate this, we used two types of assays. First, we used immunofluorescence microscopy to examine SCAP-deficient cells stably expressing GFP-SCAP (CHO/pGFP-SCAP). When CHO/pGFP-SCAP cells were incubated under sterol-depleted conditions, GFP-SCAP exhibited bright juxtanuclear fluorescence (Fig. 5A). The colocalization of GFP-SCAP with the Golgi marker GM130 indicated the localization of GFP-SCAP in the Golgi (Fig. 5,   A-C). In contrast, when the cells were incubated in the presence of sterols, the GFP signal was diffuse and reticular, suggesting ER localization (Fig. 5D). There was little colocalization of GFP and GM130 staining under this condition (Fig. 5, D-F). Importantly, a similar diffuse, reticular signal was observed when the cells were treated with XN under sterol-depleted conditions (Fig. 5, G-I), indicating that XN blocks the transport of the SCAP/SREBP complex to the Golgi. Second, we performed cell fractionation to examine the protein levels of SCAP and SREBPs in both the ER and Golgi of the XN-treated cells. Before cell fractionation, aliquots of protein were divided, and the suppressive effects of XN and 25-HC on SREBP maturation were confirmed (Fig. 5J, lanes 1-3). The lysates from CHO-7 cells were then subjected to centrifugation through a discontinuous sucrose gradient, generating two discrete fractions, light and heavy. The light fraction contained the Golgi and expressed GM130 (Fig. 5J, lanes 4, 6, and 8), and the heavy fraction contained the ER and expressed the ER-resident protein Sec61␣ (Fig. 5J, lanes 5, 7, and 9). When the cells were cultured under sterol-depleted conditions, SCAP was localized to the light fraction (Fig. 5J, lanes 4 and 5), and SREBPs were not detected in either the light or heavy fractions. This may have occurred if SREBPs were proteolytically cleaved in the Golgi in response to sterol depletion and released from the Golgi. In contrast, in the presence of 25-HC, SCAP, and SREBP precursors were found in the heavy fraction (Fig. 5J, lanes 6 and 7). These results are consistent with previous reports, wherein the SCAP/SREBP complex was anchored to the ER in sterol-loaded cells but transported to the Golgi in sterol-depleted cells. Similar to 25-HC, XN treatment led to SCAP and SREBPs found in the heavy fraction (Fig. 5J, lanes 8 and 9). These data indicate that XN contributes to the tethering of the SCAP/SREBP complex to the ER membrane and suppresses the transport of the SCAP/SREBP complex to the Golgi.
XN Inhibits Incorporation of the SCAP/SREBP Complex into COP II Vesicles-Considering our findings from previous experiments, we hypothesized that XN interacts with proteins involved in the ER-to-Golgi transport of the SCAP/SREBP complex. When cells are depleted of sterols, Sec23/24 binds to SCAP, mediating the sorting of the SCAP/SREBP complex into COP II vesicles. Furthermore, SREBPs are transported via the vesicles to the Golgi apparatus. To examine whether XN interacts with SCAP, SREBPs, or COP II proteins, we prepared XNimmobilized agarose beads (XN beads). The lysates from steroldepleted CHO-7 cells were incubated with XN beads, and the coprecipitated proteins were subjected to immunoblotting. As shown in Fig. 6A, the proteins coprecipitated with XN beads contained Sec23 and Sec24 but not SCAP and SREBP-1, whereas the control beads contained trace levels of proteins (Fig. 6A, lanes 3 and 5). Moreover, when cells were pretreated with additional XN, the amount of Sec23 and Sec24 associated with XN beads reduced (Fig. 6A, lanes 5 and 6), indicating that the XN used in the pretreatment competed with the XN beads for associating with Sec23 and Sec24. These results suggest that XN binds to Sec23/24.
The association of XN with Sec23/24 and its inhibitory effect on the translocation of the SCAP/SREBP complex indicate that XN impairs sorting of the SCAP/SREBP complex into COP II vesicles. To examine this, we investigated whether the association between Sec23 and SCAP is attenuated by XN. CHO-7 cells were treated with XN under sterol-depleted conditions; the cell lysates were then coimmunoprecipitated with SCAP and subjected to immunoblotting with Sec23. Although SCAP associated with Sec23 proteins, this association was attenuated in cells treated with 25-HC (Fig. 6B, lanes 4 and 5). We also found that the association between SCAP and SREBP-1 precursor was promoted in cells treated with 25-HC (Fig. 6B, lanes 4  and 5); this is consistent with an increase in the amount of SREBP-1 precursor in the cell lysates before immunoprecipitation because of the suppression of proteolytic processing (Fig.  6B, lanes 1 and 2). Furthermore, these observations are consistent with a previous report that found that sterols cause a conformational change in SCAP, rendering it inaccessible to COP II proteins (29). XN also reduced the amount of Sec23 associated with SCAP and increased the amount of SREBP-1 associated with SCAP (Fig. 6B, lanes 4 and 6). These data suggest that XN blocks the incorporation of SCAP/SREBP into COP II vesicles by binding to Sec23/24. ATF6 is also transported from the ER to the Golgi by COP II vesicles (33) and is cleaved by S1P and S2P (34). Therefore, we examined whether XN suppresses ATF6 processing. Huh-7 cells were pretreated with Tg, which causes ER stress by blocking the ER Ca 2ϩ -ATPase and stimulates ATF6 processing; the cells were then treated with XN. ATF6 processing stimulated by Tg was inhibited in cells treated with AEBSF but not in those treated with XN (Fig. 7). In contrast, SREBP processing was suppressed in both AEBSF-and XN-treated cells (Fig. 7). Therefore, XN appears to block the COP II-dependent transport of specific cargo proteins.
XN Reduces the Mature Form of SREBP-1 in Mouse Liver-Next, we investigated the effect of XN in mice in vivo. First, to determine whether XN suppresses SREBP processing in mouse liver, 75 or 150 mg/kg body weight of XN was administered to C57BL/6J mice by oral gavage once daily for 3 days. The food intake and body weight were not affected by XN (Fig. 8, A and  B), whereas the level of hepatic mature SREBP-1 was reduced (Fig. 8C). SCD1 gene expression decreased in the livers of mice treated with XN; however, other SREBP target genes, such as ACC1, FAS, HMGCS, HMGCR, and SQS, were not affected by XN (Fig. 8, D and E). These results indicate that XN inhibits the proteolytic activation of SREBP in mouse liver as it does in cultured cells.
XN Ameliorates Diet-induced Obesity and Fatty Liver in Mice-Next, to verify the effect of XN on diet-induced obesity, C57BL/6J mice were fed a HFD or HFD supplemented with XN (0.2 or 0.4%) for 50 days. During the feeding experiment, the average daily food intake was similar in all groups (Fig. 9A). Considering daily food intake, dietary supplementation of 0.2% XN is estimated ϳ150 mg/kg/day XN. The mice fed XN-supplemented HFD exhibited lesser weight gain than those fed a HFD alone, in a concentration-dependent manner (Fig. 9B). Moreover, the weight of the liver, epididymal, mesenteric, subcutaneous, and inguinal white fat pads, as well as interscapular brown fat pads, was reduced in the mice fed a HFD supplemented with XN (Table 1). Consistent with this, CT scan analysis demonstrated that XN reduced the percentage of fat area in the abdomen and liver (Fig. 9C). In this analysis, the abdominal visceral and subcutaneous fat volumes were separately calculated, and XN decreased both volumes (Fig. 9D). These results suggest that XN can ameliorate diet-induced obesity. To determine whether dietary XN reduces fat absorption from the intestine, we analyzed lipid levels in feces. The triglyceride levels in feces showed a tendency to increase in the HFD ϩ 0.2% XN group and significantly increased in the HFD ϩ 0.4% XN group (Fig. 9E), suggesting that dietary XN reduced lipid absorption from the intestine.
As shown in Table 2, the plasma total TG levels were lower in the mice fed a HFD supplemented with XN than in mice fed a HFD alone, and there was a marked decrease in cholesterol levels in the mice fed a HFD supplemented with XN. Table 2 also summarizes the plasma lipoprotein profiles in each group, as measured by HPLC. TG and cholesterol levels in chylomicron and very low density lipoprotein (VLDL) fractions tended to decrease, and levels in low density lipoprotein (LDL) and high density lipoprotein (HDL) fractions markedly decreased in the mice fed a HFD supplemented with XN. Consistent with the decreases in liver weight and fat percentage in the mice fed a HFD supplemented with XN, the livers of these mice had lower TG and cholesterol accumulation than those of the mice fed a HFD alone.
To evaluate whether XN affects glucose metabolism, we performed the OGTT 44 days after the initiation of XN supplementation. There was no marked difference in the plasma glucose level between the groups during OGTT (Fig. 9, F and G). However, the average glucose levels did trend to lower values in the mice fed a HFD supplemented with XN (Fig. 9H). Moreover, there was a significant decrease in insulin and C-peptide levels in the mice fed a HFD supplemented with XN at the end of the 50-day experiment (Fig. 9, I and J). This suggests that XN improves insulin sensitivity in mice experiencing diet-induced obesity. Next, to assess the activity of hepatic proteins involved in insulin signaling, we analyzed their phosphorylation state. The phosphorylation of hepatic Akt was not altered by XN (Fig.  10A). Phosphorylated S6K increased in the liver of mice fed a HFD supplemented with XN; however, the amount of total S6K was also elevated by XN (Fig. 10A). In contrast, phosphorylated AMPK levels were increased by XN (Fig. 10A), indicating that XN activated AMPK in mouse liver. However, XN did not change the phosphorylation state of AMPK in CHO-7 cells (Fig.  10B).
XN reduced the mature form of hepatic SREBP-1 (Fig. 11A) and mRNA levels of its target genes involved in fatty acid synthesis, such as ACC1, FAS, and SCD1 (Fig. 11B). This suggests that XN lowered lipogenesis by inhibiting SREBP maturation in the mouse liver. In contrast, SREBP-2 target genes involved in cholesterol synthesis, such as HMGCS, HMGCR, and SQS, were not affected in the liver of mice fed a HFD supplemented with XN (Fig. 11C). However, mRNA levels of two other genes involved in cholesterol metabolism, ATP-binding cassette subfamily G member 8 (ABCG8) and CYP7A1, increased in mice fed a HFD supplemented with XN (Fig. 11D), suggesting that XN accelerated the catabolism of cholesterol. The expression of genes involved in fatty acid ␤-oxidation, such as ACO and CPT-1a, was not altered by XN (Fig. 11E). As shown in Fig. 12A, XN did not reduce the mature form of SREBP-1, although it reduced the precursor form of SREBP-1 in the subcutaneous white adipose tissue (WAT). FAS gene expression decreased in the subcutaneous WAT of mice treated with XN; however, other SREBP target genes, such as ACC1 and SREBP-1c, were not affected by XN (Fig. 12B). These results indicate that XN does not inhibit the proteolytic activation of SREBP in WAT. XN increased the expression of ATGL, adiponectin, and UCP1 genes, although it did not affect the expression of HSL, PPAR␥2, and TNF␣ gene in the subcutaneous WAT (Fig. 12B). XN did not alter the expression of UCP1, Dio2, and PGC1␣ gene in brown adipose tissue (BAT) (Fig. 12C). Altogether, these results suggest that although XN has different effects on several metabolic parameters, the suppression of hepatic SREBP activation by XN contributed to the amelioration of obesity and hepatic steatosis in HFD-fed mice.

Discussion
In this study, we identified XN as a new SREBP inactivator by screening numerous food components. XN suppresses SREBP processing by binding to Sec23/24 and attenuating the association between SCAP and Sec23/24, causing the SCAP/SREBP complex to remain anchored to the ER. Furthermore, XN reduces mRNA levels of SREBP target genes involved in fatty acid and cholesterol synthesis and inhibits de novo lipid synthesis.
It has been reported that several naturally derived small molecules inhibit SREBP processing. Betulin, a pentacyclic triterpene isolated from birch bark, impairs the ER-to-Golgi translocation of SREBPs by promoting the association of SCAP with Insig, a step also regulated by sterols (35). Polyunsaturated fatty acids suppress the proteasomal degradation of Insig-1 by triggering the dissociation of Insig-1 from UBX domain-containing protein 8, a protein that mediates the extraction of ubiquitinated Insig-1 from the ER membrane (36). It is noteworthy that these small molecules affect the function of Insig, inhibiting SREBP processing. In contrast, in this study, we analyzed the regulation of SREBPs in the absence of Insigs. We demonstrated that XN inhibits the transport of the SCAP/SREBP complex from the ER to the Golgi independently of Insigs and in a manner different from betulin, polyunsaturated fatty acids, and sterols (Fig. 2G); in contrast, XN acts on the COP II components Sec23/24. The Sec23/24 complex associates with cargo proteins that have sequences recognized by Sec24 and clusters these proteins into COP II transport vesicles. Because COP II vesicles transport various proteins, it is possible that the association between XN and Sec23/24 blocks the translocation of some proteins besides SREBPs. However, we found that XN does not affect the transport of ATF6, a cargo normally transported in COP II vesicles (Fig. 7). This indicates that XN may inhibit only a specific set of cargo proteins. Sec24 has multiple independent sites of cargo recognition (37); therefore, XN may bind near a specific site of Sec24, making it inaccessible to specific cargo proteins. It should be noted that because we used whole cell lysates to determine XN-binding proteins (Fig. 6A), but it is still unclear whether XN directly binds to Sec23/24. It is    Fig. 2, XN, IXN, and 8-PN reduced the mature forms of SREBPs in Huh-7 cells. XN has an open C-ring, whereas IXN and 8-PN lack the open C-ring, suggesting that this component is not essential for decreasing the observed mature SREBPs. However, the nonprenylated analog of 8-PN, NG, had little effect on the level of mature SREBPs. Therefore, the prenyl group at the A-ring in XN, IXN, and 8-PN may play an important role in reducing the level of mature forms of SREBPs. Prenylation has also been shown to enhance the membrane permeability and cellular uptake of compounds by introducing hydrophobic properties (38). This may assist XN and its derivatives in regulating SREBPs within cells. In vivo, XN is metabolized into IXN and 8-PN (39 -42). Therefore, the effects of these XN metabolites on mice fed a HFD supplemented with XN must also be considered. Because IXN and 8-PN decrease the mature forms of SREBPs in Huh-7 cells, when XN is consumed in the diet, its metabolites, such as IXN and 8-PN, may also exert an inhibitory effect on SREBP activity and metabolic disorders.
To evaluate and validate the effect of XN in vivo, we performed an experiment wherein XN was included in the diet of mice with diet-induced obesity. XN inhibited the development of obesity and fatty liver in a concentration-dependent manner Different superscript letters denote statistical significance (p Ͻ 0.05). The mice used here were the same as those used in Fig. 9.
(0.2 and 0.4%). Consistent with the results in cultured cells, XN decreased mature SREBP-1 and down-regulated the expression of its target genes in mouse liver. In contrast, mRNA levels of hepatic SREBP-2 target genes were not reduced (Fig. 11C), indicating that XN does not decrease mature SREBP-2 in the mouse liver. Although we could not detect SREBP-2 protein with immunoblotting in this experiment, it is important to analyze SREBP-2 processing in the livers of mice fed a HFD supplemented with XN. A previous study reported that in contrast to the marked increase in mature SREBP-1c, the amount of mature SREBP-2 changes very little in the livers of obese mice (9). Considering this, one possibility is that XN had little effect on SREBP-2 because of insufficient activation of SREBP-2 in the livers of mice fed a HFD. Alternatively, the down-regulation of SREBP-2 processing and its target genes by XN did occur; however, the subsequent decrease in cholesterol may have positively stimulated and restored the production of mature SREBP-2.
In the mice fed a HFD supplemented with XN, plasma insulin levels were markedly reduced (Fig. 9I). Because insulin activates the proteolytic processing of SREBP-1c, the predominant SREBP-1 isoform in the liver (43), an effect of decreased plasma insulin levels on hepatic SREBP-1c processing in the mice fed a HFD supplemented with XN may have occurred. Furthermore, the stimulation of SREBP-1c processing by insulin has been shown to require the activation of the Akt/mTOR/S6K pathway (44). However, the phosphorylation of Akt and S6K did not decrease in the livers of mice fed a HFD supplemented with XN (Fig. 10A). This suggests that the decrease in mature hepatic SREBP-1 by XN is not attributable to the reduction in plasma insulin levels. Despite the remarkable decline in plasma insulin levels, plasma glucose levels and glucose tolerance did not markedly improve by XN (Fig. 9, F-I). In the mice fed a HFD supplemented with XN, lower insulin levels produced nearly the same effect on glucose removal as higher insulin levels did in mice fed a HFD alone. This indicates that insulin sensitivity was improved by XN.
It has been previously demonstrated that AMPK phosphorylation is induced in the liver of ApoE Ϫ/Ϫ mice fed a westerntype diet supplemented with XN (45). Our results also demonstrate that dietary XN increases phosphorylated AMPK in the livers of mice fed a HFD (Fig. 10A). Because phosphorylation activates AMPK and inhibits the proteolytic processing of SREBPs (46), in this study the decrease in the mature forms of SREBPs caused by XN may be partly attributable to AMPK activation. However, XN treatment did not affect the phosphorylation of AMPK in CHO-7 cells (Fig. 10B), suggesting that AMPK activation is not required for the suppression of SREBP processing by XN.
Consistent with our result, several reports indicate that dietary or oral administration of XN exhibits an anti-obesity effect in obese model animals (16,18,19). Yui et al. (19) reported that a dietary XN-rich extract from the hop suppressed fat absorption from the intestine and reduced the activity of fatty acid synthesis enzyme, such as FAS, glucose-6-phosphate dehydrogenase, and malic enzyme, in the liver accompanied by the reduction of SREBP-1c mRNA expression. It has been reported that oral administration of XN reduced plasma levels of dicar-boxylic fatty acids and acylcarnitines, markers of dysfunctional ␤-oxidation, indicating that XN promotes fatty acid oxidation (17). Thus, XN exhibit anti-obesity effect through multiple functions. Given that SREBP inactivators, such as betulin and fatostatin, exhibit anti-obesity effect (35,47), the suppression of SREBP activation by XN is, at least in part, attributable to the anti-obesity effect.
SREBPs function as master regulators of fatty acid and cholesterol biosynthesis, and aberrant SREBP activity may be linked to metabolic disease states. The inhibition of SREBP activation is a promising therapeutic approach for the treatment of metabolic disorders. Here, we examined the regulation of SREBP activity and demonstrated that XN suppresses SREBP processing not only in cultured cells but also in the livers of mice. This contributed to the amelioration of diet-induced obesity and fatty liver. Although XN is the principal prenylated flavonoid in hops, drinks made from hops, beers, for example, contain very small amounts of XN (1-100 g/liter) (48). Thus, supplemental intake of XN is suitable to get its metabolic improvement. Altogether, our results indicate that XN is an effective nutritional and pharmacological agent that can be used for improving metabolic syndrome, including obesity and fatty liver.