Concurrent Activation of Liver X Receptor and Peroxisome Proliferator-Activated Receptor Alpha Exacerbates Hepatic Steatosis in High Fat Diet-Induced Obese Mice

Liver X receptor (LXR) activation improves glucose homeostasis in obesity. This improvement, however, is associated with several side effects including hyperlipidemia and hepatic steatosis. Activation of peroxisome proliferator-activated receptor alpha (PPARα), on the other hand, increases fatty acid oxidation, leading to a reduction of hyperlipidemia. The objective of this study was to investigate whether concurrent activation of LXR/PPARα can produce synergistic benefits in treating obesity-associated metabolic disorders. Treatment of high fat diet-induced obese mice with T0901317, an LXR activator, or fenofibrate, the PPARα agonist, or in combination alleviated insulin resistance and improved glucose tolerance. The combined treatment dramatically exacerbated hepatic steatosis. Gene expression analysis in the liver showed that combined treatment increased the expression of genes involved in lipogenesis and fatty acid transport, including srebp-1c, chrebp, acc1, fas, scd1 and cd36. Histochemistry and ex vivo glycerol releasing assay showed that combined treatment accelerated lipid mobilization in adipose tissue. Combined treatment also increased the transcription of glut4, hsl, atgl and adiponectin, and decreased that of plin1, cd11c, ifnγ and leptin. Combined treatment markedly elevated the transcription of fgf21 in liver but not in adipose tissue. These results suggest that concurrent activation of LXR and PPARα as a strategy to control glucose and lipid metabolism in obesity is beneficial but could lead to elevation of lipid accumulation in the liver.


Introduction
Liver X receptors (LXR) are transcription factors belonging to the nuclear receptor superfamily. Since their initial discovery in 1995, LXRs have emerged as powerful metabolic regulators in different tissues and cell types. LXRs have been shown to regulate cholesterol, bile acid, triglyceride and glucose homeostasis as well as inflammation and intestinal lipid absorption [1]. Similar to LXR, peroxisome proliferator-activated receptor alpha (PPARa) is a ligand-activated transcription factor that belongs to the steroid hormone receptor superfamily. PPARa is expressed predominantly in tissues that have a high level of fatty acid catabolism, such as the liver, heart, and muscle [2]. PPARa regulates the expression of a number of genes critical for lipid and lipoprotein metabolism, and PPARa ligand fibrates are used for the treatment of dyslipidemia due to their ability to lower plasma triglyceride levels and elevate HDL cholesterol levels. Physiologically, both LXR and PPARa need to form heterodimers with retinoid X receptor (RXR) to initiate the expression of their target genes [3]. Therefore, a tight cross-talk exists between LXR and PPARa [4,5].
LXR activation produces a variety of beneficial effects in managing metabolic disorders. For example, previous studies by Cao et al. and Laffitte et al. show that LXR activation improves glucose tolerance in diabetic animal models [6,7]. Consistent with these studies, our recent work demonstrates that activation of LXR protects mice from high fat diet-induced obesity and insulin resistance [8]. In addition, murine studies have shown positive effects of LXR agonists on insulin resistance and atherosclerosis [9,10]. Due to these beneficial effects, LXR has been identified as attractive pharmacological target for management of metabolic disorders. Unfortunately, these beneficial effects are associated with several severe side effects including hyperlipidemia and hepatic steatosis [8,11].
On the other hand, activation of PPARa accelerates lipid absorption and increases fatty acid oxidation, leading to an improvement in lipid metabolism and a reduction of hyperlipidemia [12,13,14]. Moreover, PPARa activators have been shown to regulate obesity in rodents by both increasing hepatic fatty acid oxidation and decreasing levels of circulating triglycerides responsible for adipose cell hypertrophy and hyperplasia [15,16].
The focus of the current study is to assess the effects of concurrent activation of LXR and PPARa on systemic metabolism and hepatic fat accumulation under the status of obesity in which the metabolism of glucose and lipids are dysregulated. We demonstrate that combined treatment by T0901317, a potent activator of LXR, and fenofibrate, an agonist of PPARa, alleviated insulin resistance and improved glucose tolerance. Surprisingly, this combined treatment dramatically exacerbated hepatic steatosis in obese mice. Mechanistic studies suggest the exacerbation effect is caused by increased lipogenesis in the liver and accelerated lipid mobilization in the adipose tissue.

Ethics Statement
The use of animals in this study was in compliance with relevant federal guidelines and institutional policies and the animal protocol was approved by the IACUC of the University of Georgia.

Animals and Animal Treatments
Male C57BL/6 mice (23-25 g, Charles River, Wilmington, MA, USA) were fed a high fat diet (Bio-serv, F3282) for 12 weeks and became obese. These mice were then divided into four groups (5 each), including the control group, a T0901317-treated group, a fenofibrate-treated group and a group with a combined treatment of T0901317 and fenofibrate (Cayman Chemical, Ann Arbor, MI, USA). Mice in the control group were injected with carrier solution (dimethyl sulfoxide), and mice in treated groups were injected with T0901317 (2.5 mg/kg/day, i.p.), fenofibrate (25 mg/ kg/day, i.p.) or a combination of T0901317 and fenofibrate (T0901317, 2.5 mg/kg/day; fenofibrate, 25 mg/kg/day, i.p.), respectively. The injections were performed daily for 5 days while keeping animals on high fat diet. After the last injection, mice were fasted overnight and sacrificed using carbon dioxide.

Intraperitoneal Glucose Tolerance Test (IPGTT) and Insulin Tolerance Test (ITT)
IPGTT and ITT were performed using 4 groups of obese mice (5 each) treated with the same regimen. Mice employed in IPGTT were fasted for 6 h before the test. Glucose solubilized in phosphate buffered saline was injected (i.p.) at 1.5 g/kg, and the time-point was set as 0 min. Blood glucose levels at 0, 30, 60 and 120 min were measured using glucose meters and glucose test strips. Mice utilized in ITT were fasted for 4 h before the injection of insulin (Humulin, 0.75 U/kg) from Eli Lilly (Indianapolis, IN), and blood glucose levels were measured at 0, 30, 60 and 120 min after insulin injection.

Histochemical Study
For haematoxylin and eosin (H&E) staining, tissues of interest were dissected and fixed overnight in 10% neutral buffered formalin. The samples were dehydrated using gradient ethanol and embedded in paraffin. Tissue sections were cut at 6 mm in thickness and dried at 37uC for 2 h. H&E staining were performed using a commercial kit (#3500, BBC Biochemical). For the frozen section and Oil-red O staining, liver samples were frozen in liquid nitrogen and sectioned at 8 mm in thickness using a Cryostat. These sections were placed on slides and washed with 60% isopropanol before being stained with Oil-red O (Electron Microscopy Sciences) for 30 min and counterstained using haematoxylin. The tissue slides were examined using Nikon ECLIPSE-Ti optical microscope and pictures were taken and analyzed using Nikon NIS-Elements AR software.

Liver Triglyceride Assay
Liver triglyceride levels were determined following a previously reported method [17,18]. Briefly, liver samples (200-400 mg) were homogenized in a mixture of chloroform and methanol (2:1, volume ratio). The homogenates were kept at 4uC overnight before being centrifuged at 12,000 rpm for 20 min. Supernatants were transferred into new tubes, dried and then dissolved in 5% Triton-X100. The triglyceride concentration was determined using a commercial kit (#TR22203, Thermo-Scientific).

Biochemical Analysis of Blood Samples
Blood samples were collected from animals which were fasted overnight and centrifuged at 3,000 rpm for 10 min to isolate plasma. Plasma concentrations of glucose (Thermo-Scientific), insulin (Mercodia), triglyceride (Thermo-Scientific), cholesterol (Genzyme Diagnostics) and free fatty acids (BioAssay Systems) were determined using commercial kits. Based on fasting glucose and fasting insulin levels, we also calculated the value of Homeostasis Model of Assessment -Insulin Resistance (HOMA-IR) using the following formula: HOMA-IR = [fasting insulin (ng/ mL) 6 fasting plasma glucose (mg/dL)/405] [19].

Gene Expression Analysis
Liver and white adipose tissue samples were freshly collected and immediately frozen at 280uC until use. Total RNA was purified using TRIZOL reagent from Invitrogen (Grand Island, NY). The purified RNA samples were dissolved in RNase-free water and kept at 280uC and quantitative real-time PCR (qPCR) was performed using SYBR Green as detection reagent. The primers for qPCR were synthesized at Sigma (St. Louis, MO) and their sequences are listed in the Table S1. For all primers, melting curve analysis was conducted and confirmed as a single DNA duplex. The DDCt method was employed for quantitative examination and the GAPDH mRNA was used as an internal control [20].

Ex vivo Glycerol Releasing Assay
The ex vivo glycerol releasing assay was conducted following a previously reported method [21]. Briefly, epididymal white adipose tissue was dissected and cut into small pieces (,20 mg per piece). The cut tissues were put into Krebs-Ringer buffer (12 mM HEPES, 121 mM NaCl, 4.9 mM KCl, 1.2 mM MgSO 4 and 0.33 mM CaCl 2 ) with 3.5% fatty acid-free BSA and 0.1% glucose. Supernatants were collected at pre-determined timepoints and their glycerol levels were measured using a commercial kit purchased from Cayman Chemical (Ann Arbor, MI).

Statistical Analysis
Statistical analysis was performed using one-way analysis of variance. The results were expressed as the mean 6 SD. A p value below 0.05 (p,0.05) was considered significantly different.

Combined Treatment by T0901317 and Fenofibrate Exacerbated Hepatic Steatosis in High Fat Diet-induced Obese Mice
Compared to control and animals treated with either T0901317 or fenofibrate, combined treatment greatly enlarged liver size, which could be recognized even with the naked eye ( Figure 1A). The order of the number of vacuoles seen by H&E staining is: combination group. T0901317-treated group < fenofibratetreated group.. control group ( Figure 1A), suggesting that combined treatment significantly aggravated liver fat accumulation. A similar trend is evident by Oil-red O staining ( Figure 1A). To define the degree of hepatic steatosis among these animals, we extracted lipids from the livers and measured their triglyceride content. Activation of LXR, PPARa or both increased the liver triglyceride level from 40.066.2 (mg/g tissue) in control animals to 63.3611.9, 61.1610.5, and 85.3610.5 (mg/g tissue), respectively ( Figure 1B). The liver triglyceride level with the combined treatment was significantly higher compared to that of control group (p,0.01) and T0901317-treated group (p,0.05), suggesting that combined treatment had a synergistic effect on hepatic steatosis in high fat diet-induced obese mice. We also measured liver weight ( Figure 1C) and observed a similar trend: the mouse livers received combined treatment weighs ,6.6% of body weight, greater than ,5.5% of body weight for T0901317-treated animals < that of fenofibrate-treated animals (,5.3% of body weight). The average liver weight for control animals was approximately 4.2% of body weight of the obese mice. Additionally, we repeated this experiment using additional 4 groups of obese mice (5 each) without performing overnight fasting, and obtained similar results showing that combined treatment markedly exacerbated hepatic steatosis (data not shown).

Impact of LXR or/and PPARa Activation on Glucose and Lipid Levels in the Blood
To assess the impact of different treatments on glucose homeostasis, we measured fasting glucose and insulin levels. Figure 2A shows that T0901317 treatment reduced blood glucose concentration from 8.960.7 to 6.260.8 mM (p,0.01). Less impact was seen in animals treated with fenofibrate (,7.5 mM). The most significant impact was seen in animals who received the combined treatment reaching blood concentration at 4.960.7 mM (p,0.05 vs T0901317-treated group). Similarly, the combined treatment has the highest impact on insulin level with fasting insulin concentration of 1.960.5 ng/ml, comparing to 7.062.0 ng/ml in control animals, and 3.260.9 ng/ml and 4.761.0 ng/ml in animals treated with T0901317 or fenofibrate, respectively (p,0.01, Figure 2B), suggesting that the insulin resistance was greatly improved by the combined treatment. The HOMA-IR value, which reflects the degree of insulin resistance, showed ,85% reduction in animals with combined treatment ( Figure 2C). Regarding blood levels of triglyceride, cholesterol and free fatty acids, T0901317 treatment increased blood triglyceride by ,1.7-fold and combined treatment completely blocked the elevation ( Figure 2D). There was no difference in cholesterol level ( Figure 2E) among animals either treated or untreated. All treated animals showed a higher level of free fatty acids in the blood ( Figure 2F).

Combined Treatment by T0901317 and Fenofibrate Improved Glucose Tolerance and Alleviated Insulin Resistance
To further investigate the impact of combined treatment on glucose homeostasis, we performed IPGTT. Combined treatment markedly improved the glucose tolerance of obese mice in IPGTT, reducing the glucose peak level by ,27.6% compared to control ( Figure 3A). AUC calculations make this point clearer ( Figure 3B). Next we conducted ITT to assess the impact of combined treatment on insulin resistance. As expected, the insulin resistance of obese mice was greatly alleviated by combined treatment ( Figure 3C).

Impact on Gene Expression in the Liver
To confirm the activation of these nuclear receptors, we selected a set of target genes and measured their mRNA levels. As expected, T0901317 treatment greatly increased mRNA levels of cyp7a1, abcg5 and abcg8, 3 target genes of LXR, by ,4.3-fold, ,3.5-fold and ,4.1-fold respectively ( Figures S1A-C). Similarly, fenofibrate treatment markedly elevated mRNA levels of cyp4a10 and cyp4a14, 2 target genes of PPARa, by ,29.8-fold and ,32.1fold, respectively (Figures S1D-E). To explore the possible mechanism, we measured the expression of genes involved in lipid and glucose metabolism in the liver. The srebp-1c and chrebp are two master genes for lipogenesis in the liver. Combined treatment significantly increased the expression of srebp-1c and chrebp by 2.260.2-fold and 1.660.4-fold, respectively ( Figures 4A  and 4B), indicating that this treatment increased lipogenesis. We also measured the expression of acc1, fas and scd1, 3 vital genes involving lipogenesis, and found that the combined treatment increased their expression by 2.960.5, 5.760.8 and 34.866.3 folds, respectively ( Figures 4C, 4D and 4E). Collectively, these data suggest that combined treatment enhanced lipogenesis in the liver. Combined treatment also increased transcription of the cd36 gene encoding fatty acid transporter by 8.261.5 folds ( Figure 4F). The expression of cpt1a coding for CPT1a for fatty acid b-oxidation was significantly reduced by the combined treatment ( Figure 4G). In addition, combined treatment decreased the expression of pepck but not g6p ( Figures 4H and 4I), two crucial genes for gluconeogenesis in the liver. To further explore the regulation network, we measured mRNA levels of several nuclear receptors involved in lipid metabolism including lxra, ppara, pparc and ppard. Combined treatment did not significantly change mRNA levels of ppara, ppara and ppard while greatly reducing that of pparc ( Figures  S1F-I). Additionally, we determined the transcription of fgf21 whose expression is regulated by both PPARa and LXR. As expected, activation of either PPARa or LXR markedly increased the transcription of fgf21 ( Figure 5A) in the liver but the increase was much higher with combined treatment. However, none of the treatments affects fgf21 gene expression in the adipose tissue ( Figure 5B).

Combined Treatment Reduced Adipocyte Size
Combined treatment reduced the size of epididymal white adipose tissue (WAT) (Figure 6A), indicating this treatment accelerated lipid mobilization in WAT. Interestingly, combined treatment also reduced the size of white adipocytes surrounding intrascapular brown adipose tissue (BAT) ( Figure 6A). Similarly, combined treatment decreased the number and size of vacuolar structures in BAT as indicated by the white areas of the tissue section with H&E staining ( Figure 6A), suggesting that this treatment also reduced fat deposited in BAT. We measured the diameter of white adipocytes and found combined treatment significantly reduced their diameter by ,20 mm ( Figure 6B) and size ( Figure 6C), further indicating that the treatment stimulated lipid mobilization in WAT (p,0.01 vs control; p,0.05 vs T0901317-treated group).

Combined Treatment Accelerated Lipid Mobilization in WAT
A glycerol releasing assay was employed to confirm the effect of combined treatment on lipid mobilization in WAT. Results in Figure 7A show that the glycerol release rate in ex-vivo was ,15.8, ,33.8, ,26.6, and ,31.8 (mg/100 mg tissue/hour) for control, T0901317-treated, fenofibrate-treated animals, and animals with combined treatment, respectively. Similarly, the value of area under curve (AUC) also shows combined treatment increased glycerol releasing by ,2 folds ( Figure 7B).

Gene Expression in WAT
To investigate the impact of combined treatment on the transcription of genes involved in lipid and glucose transport and storage, including abca1, abcg1, glut4 and plin1, we measured their mRNA levels in WAT. Abca1 and abcg1 are vital for cholesterol reverse transport, and both are target genes of LXR. As expected, combined treatment significantly increased the expression of these two genes (Figures 8A and 8B). Glut4 is pivotal for intracellular transport of glucose and its expression level is tightly correlated with insulin sensitivity. Combined treatment increased the transcription of the glut4 gene ( Figure 8C). Expression of plin1 gene encoding a protein that coats lipid droplets and protects lipids against lipolysis was greatly decreased in animals treated with T0901317, fenofibrate or in combination ( Figure 8D). Meanwhile, mRNA levels of hsl and atgl were slightly but significantly increased by combined treatment (Figures 8E and 8F). We also determined mRNA levels of several adipokines including fgf21, leptin and adiponectin. Surprisingly, the transcription of fgf21 in adipose tissue was not significantly changed ( Figure 5). Combined treatment slightly down-regulated the transcription of leptin and up-regulated that of adiponectin ( Figure 9A). Interestingly, T0901317 increased mRNA level of lxra without significant change on that of ppara, pparc and ppard ( Figure 9B). In addition, combined treatment suppressed gene expression of cd11c and ifnc that are tightly correlated with the severity of chronic inflammation (Figures 9C-D).

Discussion
In this study, we demonstrate that activation of LXR by T0901317, PPARa by fenofibrate or in combination, has significant impact on glucose and lipid homeostasis and lipid accumulation in the liver and adipose tissues of high fat dietinduced obese mice. Concurrent activation of both nuclear receptors dramatically exacerbated hepatic steatosis in high fat diet-induced obese mice (Figure 1) but alleviated insulin resistance, improved glucose tolerance and blocked T0901317-induced hyperlipidemia (Figures 2 and 3). In the liver, combined treatment decreased the transcription of genes for gluconeogenesis and increased mRNA level of genes for lipogenesis and fatty acid transport ( Figure 4). In WAT the treatment increased the expression of genes for cholesterol and glucose transport and decreased the transcription of plin1 which protects lipid droplets from lipolysis ( Figure 8). Combined treatment accelerated lipid mobilization in WAT, as evidenced by reduced size of white adipocytes and enhanced glycerol release (Figures 6 and 7).
The exacerbation effect of combined treatment could be attributed to increased lipogenesis in the liver and accelerated lipid mobilization in WAT, as depicted in Figure 10. Physiologically, activated LXR or PPARa forms heterodimer with retinoid X receptor (RXR) and initiates the transcription of different sets of genes for cholesterol reverse transport and fatty acid metabolism, respectively [3]. Both nuclear receptors are involved in lipogenesis in the liver [22]. Activation of both LXR and PPARa significantly increased lipogenesis via elevating the transcription of srebp-1c, chrebp, acc1, fas and scd1 (Figure 4). Meanwhile, combined treatment accelerated lipid mobilization in WAT and released more free fatty acids into the circulation (Figures 2F, 6 and 7). The fatty acids entered hepatocytes primarily via CD36 whose transcription was up-regulated by the combined treatment ( Figure 3F). Collectively, the endogenously generated and exogenously acquired lipids aggregate in the liver and exacerbate hepatic steatosis in these obese mice.
Combined treatment by activators of LXR and PPARa exerted a synergistic effect in increasing lipogenesis in the liver via elevating the transcription of acc1, fas and scd1. Consistent with previous studies by Grefhorst et al. and Chisholm et al. showing that LXR activation leads to hepatic steatosis [11,23], our data also show LXR activation significantly increased the expression of genes involved in lipogenesis, including srebp-1c, chrebp, acc1, fas and scd1 (Figure 3), leading to elevated blood triglyceride and aggravated hepatic steatosis. Oosterveer et al. reported PPARa activation using fenofibrate simultaneously induces fatty acid oxidation and synthesis in the liver, resulting in increased hepatic triglycerides [24]. On the other hand, Larter et al. showed activation of PPARa by Wy14643, another agonist of PPARa, improved hepatic steatosis in female diabetic foz/foz mice [25]. Coinciding with the results of Oosterveer et al., our data also show fenofibrate increased the transcription of scd1 and aggravated liver fat accumulation ( Figure 3E). The discrepancy between our results and that of Larter et al. is likely resulted from the two different animal models employed, one employing wild type HFD-induced obese mice and the other using Alms1 mutant (foz/foz) mice [25,26]. Interestingly, simultaneous activation of LXR and PPARa by T0901317 and fenofibrate exerted a synergistic effect in elevating the transcription of acc1, fas and scd1 (Figures 3C to  3E), three key genes in the pathway of lipogenesis. This synergistic effect may play an important role in causing the aggravated hepatic steatosis in obese mice.
In agreement with the study by Stenson et al. showing LXR activation enhanced lipolysis in vitro [27], our data demonstrate that LXR activation accelerated lipid mobilization and glycerol release in WAT in vivo and ex vivo (Figures 6 and 7). The impact of LXR activation on lipid turnover in WAT is complicated. Based on the results of cell culture in vitro, some studies proposed LXR activation enhances lipid accumulation in adipocytes [28,29,30], but others did not [27,31]. In line with the study by Stenson et al. and Ross et al. [27,31], our data clearly show LXR activation by T0901317 accelerated lipid mobilization and glycerol release in vivo and ex vivo, probably via coordinately modulating the expression of plin1, hsl and atgl ( Figures 8D-F). In addition, since insulin is the primary hormone controlling fat storage in adipocytes, the reduced level of blood insulin ( Figure 2B) achieved by combined treatment may also contribute to lipolysis in adipose tissue. Besides WAT, BAT plays critical roles in lipid metabolism as well [32]. Interestingly, our histochemical examinations show LXR activation decreased fat accumulation in BAT ( Figure 6A), which is different from previous study by Korach-André et al.
showing LXR deficiency decreased BAT lipid accumulation in LXR gene knockout mice [33]. The discrepancy between these two studies is likely resulted from the two different animal models employed, one employing wild type obese mice and the other using LXR gene knockout mice. Alternatively, this discrepancy could be that the reduced BAT fat observed in this study is a consequence of systemic activation of LXR rather than a direct local effect. Collectively, this accelerated lipid mobilization in adipose tissue may contribute in exacerbating hepatic steatosis in the high fat diet-induced obese mice.
Accumulating evidence suggests the existence of a tight crosstalk between LXR and PPARa in multiple tissues [4,5,34,35,36]. In response to endogenous and exogenous signals, both LXR and PPARa heterodimerizes with RXR to initiate the expression of the target genes. Therefore, RXR is also involved in this network. For example, a previous study by Lenhard et al. showed that activation of RXR by LG100268, a potent agonist of this nuclear receptor, decreased blood glucose but markedly increased hepatic fat accumulation in db/db mice [37]. Using luciferase reporter gene assay and gel shift assay, Yamada and colleagues elegantly demonstrated that LXR-RXR-PPARa forms a network that tightly regulates lipid degradation and lipogenesis [4,5]. Subsequent investigations by Colin et al. and Ducheix et al. provide valuable information for this cross-talk in vivo, showing LXR-PPARa drives lipid metabolism in response to oxysterol and fatty acids [34,35]. Using chromatin immunoprecipitation-sequencing, Boergesen et al. further elucidated the extensive cross-talk between LXR and PPARa at the level of binding to shared genomic sites [36]. Consistent with these studies, our data show that simultaneous activation of LXR/PPARa synergistically modulated the expression of a set of genes involved in lipid and glucose metabolism in liver, including acc1, fas, scd1, cd36 and pepck ( Figure 4). Similar to these genes, fgf21 is also regulated by LXR and PPARa [38,39,40]. Recently, Dutchak et al. demonstrated that FGF21 is also an inducible autocrine factor in adipose tissue that functions to regulate PPARc [41]. Interestingly, our data show that simultaneous activation of LXR/PPARa greatly increased the transcription of fgf21 in liver but not in adipose tissue ( Figure 5), indicating that different regulatory mechanism may exist in these two tissues. In addition, because T0901317 may activate other nuclear receptors besides LXR [42], further studies are required to better elucidate this cross-talk at molecular level.
Macrophage infiltration and chronic inflammation are features of diet-induced obesity and contribute to development of obesity-associated metabolic disorders including glucose intolerance and fatty liver [43,44]. LXR activation has a welldefined role in suppressing inflammation, and activators of LXR are negative regulators of macrophage inflammatory gene expression [45,46]. Similar to that of LXR, PPARa activation  produces an anti-inflammatory effect as well [47]. Emerging evidence suggests that LXR and PPARa signaling influences multiple facets of inflammation and immunity, thereby providing important cross-talk between metabolism and immune system [48]. In this context, we determined the transcription of a set of genes involved in chronic inflammation, and found that simultaneous activation of LXR/PPARa slightly but significantly reduced mRNA levels of cd11c and ifnc ( Figure 9). This reduction may also contribute to the improvement in insulin resistance and glucose tolerance.
In summary, in this study we demonstrate that combined treatment by activators of LXR and PPARa alleviated insulin resistance and improved glucose homeostasis but dramatically exacerbated hepatic steatosis in high fat diet-induced obesity. This exacerbation effect is caused by increased lipogenesis in the liver and accelerated lipid mobilization in WAT. Collectively, these data suggest that cautions should be taken in considering activation of both LXR and PPARa as a strategy for treatment of obesity or obesity related diseases.     Author Contributions Figure 10. Depicted pathway for the exacerbated hepatic steatosis caused by combined treatment. Combined treatment activates LXR and PPARa, thereby increasing lipogenesis in the liver via elevating the transcription of srebp-1c, chrebp, acc1, fas and scd1. Meanwhile combined treatment accelerates lipid mobilization in the adipose tissue, thereby releasing more free fatty acids that are transported into hepatocytes by CD36 and further aggravate hepatic steatosis. doi:10.1371/journal.pone.0065641.g010