Luteolin Attenuates Hepatic Steatosis and Insulin Resistance Through the Interplay Between the Liver and Adipose Tissue in Mice with Diet-Induced Obesity

Obesity is characterized by excessive fat accumulation and is associated with metabolic complications such as adiposity, dyslipidemia, insulin resistance, and steatohepatitis. The liver is one of the major organs responsible for metabolizing fats from the diet. Dysregulation of lipid metabolism in the liver induces abnormal accumulation of lipids and the subsequent formation of lipid droplets (LDs), known as hepatic steatosis. Hepatic steatosis is common in obese individuals and is strongly linked to insulin resistance (1,2). Adipose tissue is also a critical component of metabolic control and is an endocrine organ that secretes a number of adipokines known to mediate lipid metabolism, inflammation, and insulin sensitivity (3). Although the complex relationship between adiposity and hepatic steatosis has not been completely elucidated, dysregulation of lipid metabolism in the liver and adipose tissue may be involved in the pathogenesis of adiposity as well as its associated metabolic complications. One hypothesis for the explanation of a link between visceral adiposity and hepatic steatosis and insulin resistance is the “portal hypothesis,” which proposes that a high rate of lipolysis of the visceral adipose tissue can lead to an increased delivery of free fatty acids (FFAs) to the liver via the portal vein, and that chronic exposure of the liver to elevated FFAs can promote hepatic steatosis and hepatic insulin resistance (4,5). However, it is still unclear how visceral adiposity is associated with hepatic steatosis and insulin resistance, in particular at the level of transcriptional changes that occur in adipose tissue and the liver in obesity.

Luteolin (3′,4′,5,7-tetrahydroxyflavone), a crucial member of the flavones, widely occurs in vegetables, fruits, and natural herbal drugs such as parsley, thyme, peppermint, and celery. Luteolin possesses a variety of antitumor and anti-inflammatory properties by inhibiting NF-κB activation and inducing apoptosis (6,7). As other flavonoids, luteolin is most often found in plant materials in the form of glycosides, which are eventually metabolized and absorbed passively by intestine (8). However, little is known about the detailed mechanisms associated with the antiobesity action of luteolin based on the integration of the transcriptional profile and the phenotype biomarkers from the liver and adipose tissue. The current study was designed to elucidate the metabolic actions of luteolin in C57BL/6J mice with diet-induced obesity (DIO), in particular by focusing on its role in ameliorating obesity-associated hepatic steatosis and insulin resistance. Above all, it provides novel insights into the effect of luteolin on the interplay between the liver and adipose tissue.

Thirty-six 4-week-old male C57BL/6J mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All mice were individually housed under a constant temperature (24°C) and 12-h light/dark cycle, fed the AIN-76 semipurified diet for a 1-week acclimation period, and subsequently divided randomly into three groups and fed a normal diet (ND, n = 10), high-fat diet (HFD, n = 13, 20% fat and 1% cholesterol), or HFD with 0.005% (weight for weight) luteolin (LU, n = 13; Sigma-Aldrich, St. Louis, MO) for 16 weeks. The dose of luteolin was based on previous in vivo studies (9–12). All experimental diets were prepared every week and stored at −4°C. At the end of the experimental period, all mice were anesthetized with ether after a 12-h fast. Blood was taken from the inferior vena cava to determine the glucose, plasma lipid, and hormone concentrations. The liver and adipose tissue were removed, rinsed with physiological saline, weighed, immediately frozen in liquid nitrogen, and stored at −70°C until analysis. The animal study protocols were approved by the Ethics Committee at Kyungpook National University.

Every 2 weeks, the 12-h fasting blood glucose was measured in tail vein blood with a glucose analyzer (GlucoDr SuperSensor, Allmedicus, Korea). For the intraperitoneal glucose tolerance test (IPGTT) analysis performed in the 15th week after the start of the diet experiments, mice fasted for 12 h were injected intraperitoneally with glucose (0.5 g/kg body weight), and the blood glucose level was determined in tail vein blood at 0, 30, 60, and 120 min post–glucose injection. Radioimmunometric assays were used to measure the plasma insulin concentration (Milliplex MAP Mouse Endocrine Kit; Millipore, St. Charles, MO). The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated as [fasting insulin concentration (mU/L) × fasting glucose concentration (mg/dL) × 0.05551]/22.5.

The total plasma cholesterol (TC), HDL cholesterol (HDL-C), triglyceride (TG), glutamic oxaloacetic transaminase (GOT), and glutamic pyruvic transaminase (GPT) levels were measured using enzymatic assay kits (Asan Pharm, Seoul, Korea), as were those of plasma FFAs (Wako, Tokyo, Japan), apolipoprotein AI (apoA-I) (Eiken Chemical, Tokyo, Japan), apoB48 (MyBioSource, San Diego, CA), and apoB100 (Eiken Chemical). The non–HDL-C level was calculated as [(total cholesterol) − (HDL-C) − (TG/5)]. The hepatic and fecal lipid levels were determined using the same enzymatic kit used for the plasma analyses after extraction using the methods of Folch et al. (13).

The liver and epididymal adipose tissue were removed from each mouse. Samples were subsequently fixed in 10% (volume for volume) paraformaldehyde/PBS and embedded in paraffin for staining with hematoxylin and eosin. The stained area was visualized using a microscope set at a 200× magnification.

Hepatic mitochondrial, cytosolic, and microsomal fractions were prepared as previously described (14), and the protein concentrations were determined using the Bradford method. Glucose-6-phosphate dehydrogenase (G6PD) (15), fatty acid synthase (FAS) (16), malic enzyme (ME) (17), and phosphatidate phosphohydrolase (PAP) (18) activities were measured as previously described. Microsomal HMG-CoA reductase (HMGCR) (19) and acyl-CoA:cholesterol acyltransferase (ACAT) (20) activities were measured with [¹⁴C]HMG-CoA and [¹⁴C]oleoyl CoA as substrates, respectively. Spectrophotometric assays were used to determine the glucose-6-phosphatase (G6Pase) (21), glucokinase (22), and PEPCK (23) activities as previously published. The hepatic glycogen (24) and H₂O₂ (25) concentrations were determined as previously described.

Total RNA was extracted from adipose tissue and the liver using TRIZOL reagent (Invitrogen Life Technologies, Grand Island, NY) according to the manufacturer’s instructions. DNase digestion and subsequent reprecipitation in ethanol were performed to remove any DNA and phenol contamination, respectively. For quality control, the RNA purity and integrity were evaluated using the Agilent 2100 Bioanalyzer (Agilent Technologies, Wilmington, DE). For the liver and epididymal adipose tissue, three pooled RNA sample sets were constructed for the ND, HFD, and luteolin groups at each time point as previously described (26). RNA was stored at −70°C prior to further analysis by microarray and real-time quantitative PCR (RT-qPCR).

Total RNA (1 μg) was reverse transcribed into cDNA using the QuantiTect reverse transcription kit (Qiagen, Hilden, Germany), and the mRNA expression was quantified by RT-qPCR using the SYBR green PCR kit (Qiagen) and the CFX96TM real-time system (Bio-Rad, Hercules, CA). Gene-specific mouse primers were used as mentioned in Supplementary Table 1. Ct values were normalized to GAPDH, and the relative gene expression was calculated with the 2^−△△ Ct method (27).

Total RNA was amplified and purified using the Ambion Illumina RNA amplification kit (Ambion, Carlsbad, CA). Biotinylated cRNA (750 ng per sample) was hybridized to Illumina Mouse WG-6 v2 Expression BeadChips (Illumina, San Diego, CA) according to the manufacturer’s instructions. Hybridized arrays were washed and stained with Amersham Fluorolink streptavidin-Cy3 (GE Healthcare, Little Chalfont, U.K.) following the standard protocol provided in the bead array manual. Hybridization quality and overall chip performance were determined by visual inspection of both the internal quality controls and the raw scanned data in the Illumina BeadStudio software. Probe signal intensities were quantile normalized and log transformed. Microarray analysis was performed in the ArrayAssist software (Stratagene, La Jolla, CA) and with the R programming language.

The statistical differential gene expression analysis between groups was performed by the nonparametric Rank Prod approach. Those oligonucleotides that present changes between groups with a false discovery rate <0.05 were considered significant. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways mapper (www.genome.jp/kegg) was consulted for the analysis of gene functions involved in lipid metabolism and the tricarboxylic acid (TCA) cycle. These microarray data were deposited in the Gene Expression Omnibus (GEO) database (GEO accession number: GSE54189).

Liver and epididymal adipose tissue proteins were extracted with lysis buffer and quantified using the Bradford method. Total protein (80–100 g) was electrophoresed on 10% SDS polyacrylamide gels and transferred to polyvinylidene fluoride membranes (Millipore), blocked, and probed with rabbit anti-CIDEA (cell death–inducing DFFA-like effector A) (1:1,000; Abcam), rabbit anti-ACC (acetyl-CoA carboxylase) (1:1,000; Cell Signaling), rabbit anti-SREBP1 (sterol regulatory element-binding protein 1) (1 μg/mL; Abcam), rabbit anti-SREBP2 (4 μg/mL; Abcam), mouse anti-PPARγ (peroxisome proliferator–activated receptor γ) (1:1,000; Santa Cruz Biotechnology), rabbit anti-SIRT1 (sirtuin 1) (1:1,000; Cell Signaling), and mouse anti–β-actin (1:1,000; Cell Signaling), respectively. The immunoreactive antigen was then recognized with a horseradish peroxidase–labeled anti-rabbit or anti-mouse IgG (1:1,000; Cell Signaling) and an enhanced ECL kit from Pierce Biotechnology (Rockford, IL). The immunoreactive bands were quantified with a G-box (50S; BI System Co.).

¹H-nuclear magnetic resonance (¹H-NMR) spectra were phased and baseline corrected using AMIX software (version 3.9.4; Bruker Biospin GmbH). NMR spectral data were analyzed using Chenomx NMR Suite 7.1 (Chenomx, Edmonton, Canada), and concentrations were determined using the 600-MHz library from the Chenomx NMR Suite 7.1. The metabolite concentrations were determined based on the internal TSP/TMS references. The regions corresponding to water-acetonitrile (4.69–4.91 ppm) in the aqueous fraction, methanol-chloroform (3.30–3.87 ppm) in the lipophilic fraction, and the reference peaks (TMS and TSP, 0.0–0.5 ppm) were excluded from all spectra, and the remaining spectral regions were divided into 0.005-ppm bins. The spectral data were then normalized to the total spectral area.

The resulting data sets were imported into SIMCA-P version 12.0 (Umetrics AB, Umeå, Sweden) for multivariate analysis. Principal component analysis was performed to explain the “clustering” and trends within the multivariate data set. The quality of each model was determined using the goodness of fit parameter (R²) and the goodness of prediction parameter (Q²). The Mann-Whitney U test was applied to the concentrations of the hepatic metabolites using Prism version 5.0 for Windows (GraphPad Software, San Diego, CA). Statistical significance was set at P < 0.05.

The parameter values were expressed as the mean ± SEM. Because there is a difference of caloric intake between the ND and HFD, significant differences between the groups were determined by Student t test and Wilcoxon t test using the SPSS program (SPSS Inc., Chicago, IL). Results were considered statistically significant at P < 0.05.

Luteolin treatment significantly decreased body weight after the 12 weeks of high-fat feeding and food efficiency ratio (Fig. 1A and B). Luteolin also slightly decreased the liver weight and significantly reduced the weight of all white adipose tissue depots (epididymal, perirenal, retroperitoneal, mesenteric, subcutaneous, and interscapular) compared with the HFD group (Fig. 1C). The TC and non–HDL-C were significantly decreased in the LU group compared with their levels in the HFD group. The plasma TG, FFA, apoB48, and apoB100 levels were also significantly lower in the LU group compared with the HFD group (Fig. 1D).

Luteolin treatment markedly decreased the hepatic cholesterol, TG, and fatty acid (FA) contents (Fig. 2A), whereas it increased fecal TG and FA contents (Fig. 2B). Furthermore, the levels of the hepatic lipotoxicity markers plasma GOT, GPT, and hepatic mitochondrial H₂O₂ were significantly decreased by luteolin treatment compared with those in the HFD group (Fig. 2C). Tissue morphology analysis also revealed that the accumulation of hepatic LDs as well as the cell size of the LD was decreased in the LU group compared with the HFD group (Fig. 2D).

The activities of the hepatic enzymes involved in FA and TG synthesis (FAS, G6PD, ME, and PAP) were significantly decreased by luteolin treatment compared with the HFD group. The hepatic cholesterol-regulating enzymes, HMGCR, and ACAT activities were also significantly lower in the LU group compared with the HFD group (Fig. 2E). Western blot analysis revealed that the protein expression of the lipogenic factors PPARγ, ACC, SREBP1, SREBP2, and CIDEA was markedly decreased in the luteolin-fed mice when compared with their expression in the liver of the HFD mice. In contrast with the liver, PPARγ, SREBP1, and ACC protein expressions in epididymal adipose tissue were significantly increased by luteolin (Fig. 2F).

The fasting blood glucose and plasma insulin concentration were significantly lowered by luteolin treatment at weeks 4, 12, and 16 (Fig. 3A and B). The IPGTT revealed that luteolin significantly improved glucose tolerance (Fig. 3C). HOMA-IR was also significantly lowered by luteolin (Fig. 3D), indicating a decrease in insulin resistance. The hepatic glycogen level was lowered by luteolin (Fig. 3E), whereas glucokinase activity was significantly elevated in the LU group compared with its activity in the HFD group (Fig. 3F). Hepatic PEPCK and G6Pase activities were suppressed by luteolin treatment (Fig. 3F).

As shown in Fig. 4A and Supplementary Table 2, the hepatic microarray analysis showed that luteolin enhanced the hepatic expression of the chylomicron remnant receptor (Lrp4) and the LDL receptor (Ldlr, Pcsk9, and Stab1) genes when compared with their expression in the HFD group. Luteolin treatment resulted in downregulation of hepatic genes that are associated with FA synthesis (Acaa2, Acot1, and Decr2) and lipogenesis (Cidea, Acat3, Acsl1, Acsl5, Adrp, Apgat2, Apgat9, Gpam, Mttp, Ppap2a, and Ppapdc1b) while upregulating FA oxidation–associated genes (Acadl, Acox1, Ppargc1b, Hadh, and Scp2) and lipolysis-associated genes (Lpl and Mgll). Gene expressions responsible for cholesterol and bile acid synthesis were decreased by HFD, and among them Dhcr7 and Ndsh1 were further decreased by luteolin treatment. The expression of the Abcg5 and Abcg8 genes, which are responsible for biliary excretion of sterols, was also decreased by HFD and further downregulated by luteolin treatment.

Microarray analysis of the epididymal adipose tissue revealed that luteolin enhanced the expression of FA and lipid transporter (Cd36, Fabp4 [aP2], Ffar3, and Lpl). Luteolin reversed the expression pattern of genes involved in FA synthesis (Acacg, Acot4, Echdc2, Echdc3, Echs1, Fasn, Mcat, and Mecr2) and TG synthesis (Chrebp, Apgat6, and Dgat1), those genes downregulated by HFD, the cholesterol synthesis–associated genes (Dhcr24, Hsd17b7, Nsdhl, Sc4mol, and Sqle), and those upregulated by HFD (Fig. 4B and Supplementary Table 3). In particular, luteolin treatment lowered the expression of Adrp while increasing the expression of FA oxidation genes and their transcription regulators (Adrb3, Ppargc1a, Ppargc1b, Acadl, Acads, Acadsb, Acadvl, Cpt2, and Hadh) and lipolysis-associated genes (Lipe and Pnpla2). Moreover, luteolin increased TCA cycle–associated gene expressions, including Phdb, Pcx, Cs, Aco1, Aco2, Idh2, Idh3b, Idh3g, Ogdh, Dlst, Suclg2, Sdhb, Sdhc, Mdh1, and Mdh2, that were downregulated by HFD (Fig. 5 and Supplementary Table 4).

Seven genes, including Cidea, Ppargc1a (PGC1α), Ppargc1b (PGC1β), Cpt1a, Irs2, Adrb3, and Pck1, were selected for further validation by RT-qPCR (Supplementary Fig. 1). Both RT-qPCR and microarray analysis showed a similar transcriptional pattern for hepatic Cidea, Ppargc1a, and Irs2, as well as adipocyte Adrb3, Ppargc1a, Ppargc1b, and Pck1, genes for each group. No significant differences were observed in the expression of the Ppargc1a and Cpt1α genes by microarray analysis between the HFD and LU groups. However, those genes were significantly increased by luteolin, as shown by the RT-qPCR method.

Results from the quantitative analysis of hepatic metabolite concentrations are shown in Table 1 and Supplementary Fig. 2. Our principal component analysis models clearly differentiated the hepatic metabolite levels not only between the ND and HFD groups but also between the HFD and LU groups (Supplementary Fig. 2). In the liver, luteolin increased the levels of valine, alanine, glutamate, glutamine, phosphatidylcholine (PC), glutathione, and taurine and decreased the levels of glycogen and R-CH₃ (Table 1).

In the current study, we evaluated the effects of luteolin on and its potential mechanisms underlying the metabolic regulation of an obesogenic diet using “Omics” tools. We first showed that luteolin attenuated DIO, hepatic steatosis, dyslipidemia, and insulin resistance in DIO mice: 1) luteolin treatment resulted in an improvement in hepatic steatosis by suppressing lipogenesis and absorption of dietary fat; 2) the inverse relationship of adipocyte FA uptake gene (Cd36, Ffar3, Fabp4 [aP2], and Lpl) expressions modulated by PPARγ and dyslipidemia was observed in luteolin-treated obese mice, which is linked to an attenuation of hepatic lipotoxicity; 3) luteolin reduced adiposity via the restriction of TG availability by increasing lipolysis and the TCA cycle prior to LD formation in adipose tissue; and 4) furthermore, luteolin treatment improved the hepatic insulin sensitivity by suppressing gluconeogenesis while increasing insulin receptor substrate 2 (Irs2) gene expression.

HFD generally induces hepatic steatosis through multiple mechanisms, including an increased flux of dietary and liberated visceral FA, increased hepatic FA synthesis, and reduced hepatic FA oxidation. In this study, HFD decreased lipogenesis as well as FA oxidation due to the adaptation to overnutrition (28). Interestingly, luteolin treatment decreased further the reduced hepatic lipogenic protein expressions (PPARγ, SREBP1, SREBP2, and ACC) and enzyme activities (G6PD, FAS, ME, PAP, HMCGR, and ACAT) by adaptive response to HFD, suggesting that luteolin may limit hepatic lipid availability by inhibiting lipogenesis, thereby reducing hepatic lipotoxicity markers. Our NMR analysis result indicates that luteolin increased the expression of hepatic alanine, glutamine, and glutamate metabolites. In the liver, alanine, along with glutamate, carries amino groups into mitochondria for their conversion into urea. It is plausible that an increase in the hepatic alanine and glutamate level by luteolin treatment could activate the urea cycle to remove amino groups derived from the catabolism of proteins or amino acids.

Oligonucleotide microarray profiling also revealed that luteolin treatment combined with an obesogenic diet for 16 weeks downregulated the expression of hepatic genes involved in lipid synthesis compared with this expression in HFD, as is similar to the pattern of hepatic protein expression. Especially, luteolin dramatically lowered the mRNA expressions of Cidea and Adrp, which are important markers of lipid accumulation, with a simultaneous decrease in CIDEA protein expression, which in turn induced a significant decrease in hepatic LD formation and accumulation (29,30). In addition, luteolin-induced elevation of hepatic Abcg5 and Abcg8 mRNA expressions and fecal TG and FA levels could contribute to the inhibition of hepatic lipid load by promoting biliary sterol excretion and decreasing the absorption of dietary fat, thereby reducing the transport of exogenous fats in chylomicrons from the intestines to tissues, as suggested by a decrease in plasma apoB48 level. Furthermore, the preventive effect of luteolin on hepatic lipid accumulation via the suppression of lipogenesis and absorption of dietary fat can limit the TG availability for the assembly of VLDL, which also seemed to be reflected by a reduction of plasma apoB100. ApoB100 is synthesized exclusively by the liver and is required for VLDL secretion by Mttp (31). Huff et al. (32) reported that an ACAT inhibitor decreased apoB secretion in miniature pigs in vivo. Thus, the reduced Mttp mRNA expression and ACAT activity by luteolin we observed may be associated with a reduced VLDL assembly and secretion by limiting TG availability. Hepatic steatosis reflects hepatic oversupply of lipids, which is conducive to hyperlipidemia. The noticeable improvement in hepatic steatosis by luteolin treatment was associated with a significant reduction in the plasma lipid levels. Luteolin enhanced the hepatic gene expression of receptors for CR (Lrp4) and LDL (Ldlr, Stab1, and Pcsk9), resulting in reduced plasma TC and LDL-C levels. These observations indicate that luteolin has the potential to regulate hepatic lipid metabolism, thereby ameliorating hepatic steatosis and dyslipidemia in DIO mice.

In contrast to the liver, luteolin treatment in the HFD-fed mice elevated lipogenesis through increasing PPARγ protein expression in adipose tissue. Wu et al. (33) reported overexpression of PPARγ delivered by adenovirus-PPARγ attenuated steatohepatitis by redistribution of FA from liver to adipose tissue by enhancing expression of FA uptake genes and lipogenic genes in adipose tissue. Luteolin significantly reduced the plasma FFA level with a simultaneous increase in adipocyte FA uptake gene (Cd36, Ffar3, Fabp4, and Lpl) expressions, leading to the reduction of hepatic lipotoxicity via the increased FFA flux to the liver. In general, the excessive release of FAs from dysfunctional and insulin-resistant adipocytes results in lipotoxicity, causing the accumulation of TG-derived toxic metabolites in ectopic tissues (liver, muscle, and pancreatic β-cells) (4,5). Luteolin led to increased FA re-esterification into newly synthesized TG by enhancing the adipose FA uptake and Pck1 genes that can prevent hepatic lipotoxicity. Increased Pck1 expression in adipose tissue can modulate the conversion of oxaloacetate into phosphoenolpyruvate, which is further used in glycerol 3-phosphate (G3P) production for TG synthesis.

Interestingly, despite activating the pathway involved in TG synthesis in adipose tissue, luteolin markedly decreased the weights of all white adipose tissue depots. This observation seemed to be associated with a reduction in the mRNA expression of Adrp, which is involved in LD formation and in the enhancement of the mRNA expression of lipolysis-associated genes such as Pnpla2 and Lipe by luteolin treatment. It is plausible that luteolin limits TG availability by increasing lipolysis prior to LD formation as proposed in Fig. 6. Luteolin also elevated the expression of genes involved in both FA oxidation and the TCA cycle, such as Adrb3, Ppargc1α, and Ppargc1β, which are known to contribute to the activation of FA oxidation. These white adipose tissue microarray data suggest that although luteolin activates lipogenesis through the increase in PPARγ protein to prevent hepatic steatosis and lipotoxicity, it also simultaneously increases lipolysis, FA oxidation, and the TCA cycle, which may contribute to the reduced adiposity.

The striking improvement of hepatic steatosis coupled with the decreased adiposity in luteolin-treated mice was associated with a normalization of the plasma glucose and insulin levels, which was a reflection of improved hepatic insulin sensitivity as evidenced by a reduced HOMA-IR and the area under the curve of the IPGTT. It would be expected that a luteolin-induced increase in hepatic FA oxidation and decrease in hepatic lipogenesis improves hepatic insulin sensitivity. Insulin binds to receptors on the hepatocytes and results in the inhibition of enzymes involved in gluconeogenesis (34). Thus, the suppression of gluconeogenesis by decreased hepatic G6Pase and PEPCK activities and the activation of glucose use by an increased hepatic glucokinase activity seemed to be associated with the improved hepatic insulin sensitivity observed in luteolin-treated DIO mice. Luteolin also increased the formation of hepatic PC metabolites and the expression of Irs2 mRNA. When the membrane PC concentration is low, it activates the transcription of SREBP1 (35). In the liver, SREBP1 modulates Irs2 expression through its negative feedback to an insulin response element on the Irs2 promoter, and the decrease in Irs2 expression can lead to insulin resistance (34,36). Therefore, the increased PC concentration could result in a decrease in SREBP1 and an increase in Irs2 expression, thereby leading to an increase in insulin sensitivity. This observation suggests that the suppression of the hepatic lipogenesis pathway by luteolin is partially linked to glucose homeostasis, leading to the prevention of insulin resistance and glucose intolerance as proposed in Fig. 6.

In summary, the data obtained from our animal study indicate that luteolin can suppress DIO and modulate obesity-associated metabolic disorders. Luteolin decreased hepatic PPARγ protein expression and increased adipocyte PPARγ protein expression, resulting in the suppression of hepatic lipogenesis and an increase in uptake of circulating FFAs by adipocytes, which contributes to improvement of both hepatic steatosis and lipotoxicity. Luteolin also prevents adiposity by enhancing lipolysis and the TCA cycle prior to LD formation in adipose tissue. Taken together, the present findings suggest that luteolin ameliorates the deleterious effects of DIO and its metabolic complications such as adiposity, dyslipidemia, hepatic steatosis, and insulin resistance.

Funding. This work was supported by National Research Foundation of Korea grants funded by the Korean government (MSIP) (2008-0062618, 2012M3A9C4048818).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. E.-Y.K. and M.-S.C. collected data, contributed to the discussion, and wrote, reviewed, and edited the manuscript. U.J.J. collected data and reviewed and edited the manuscript. T.P. and J.W.Y. contributed to the discussion and reviewed and edited the manuscript. M.-S.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.