The marine‐derived furanone reduces intracellular lipid accumulation in vitro by targeting LXRα and PPARα

Abstract Recent studies have demonstrated that commercially available lipid‐lowering drugs cause various side effects; therefore, searching for anti‐hyperlipidaemic compounds with lower toxicity is a research hotspot. This study was designed to investigate whether the marine‐derived compound, 5‐hydroxy‐3‐methoxy‐5‐methyl‐4‐butylfuran‐2(5H)‐one, has an anti‐hyperlipidaemic activity, and the potential underlying mechanism in vitro. Results showed that the furanone had weaker cytotoxicity compared to positive control drugs. In RAW 264.7 cells, the furanone significantly lowered ox‐LDL‐induced lipid accumulation (~50%), and its triglyceride (TG)‐lowering effect was greater than that of liver X receptor (LXR) agonist T0901317. In addition, it significantly elevated the protein levels of peroxisome proliferator‐activated receptors (PPARα) and ATP‐binding cassette (ABC) transporters, which could be partially inhibited by LXR antagonists, GSK2033 and SR9243. In HepG2 cells, it significantly decreased oleic acid‐induced lipid accumulation, enhanced the protein levels of low‐density lipoprotein receptor (LDLR), ABCG5, ABCG8 and PPARα, and reduced the expression of sterol regulatory element‐binding protein 2 (~32%). PPARα antagonists, GW6471 and MK886, could significantly inhibit the furanone‐induced lipid‐lowering effect. Furthermore, the furanone showed a significantly lower activity on the activation of the expression of lipogenic genes compared to T0901317. Taken together, the furanone exhibited a weak cytotoxicity but had powerful TC‐ and TG‐lowering effects most likely through targeting LXRα and PPARα, respectively. These findings indicate that the furanone has a potential application for the treatment of dyslipidaemia.

receptors that regulate lipid metabolism via their transcriptional activity. They activate the expression of target genes by binding to specific repeat DNA response elements with their obligate heterodimeric partner retinoid X receptor (RXR). 8 PPARα modulates the transcription of genes encoding key enzymes in fatty acid catabolism pathway such as carnitine palmitoyltransferase 1A and 2, acyl-CoA dehydrogenase and 3-hydroxy-3-methylglutaryl-CoA synthase 2. 9 PPARs are activated by fatty acids and eicosanoids, and inactivated by synthetic compounds such as GW6471 and MK886.
PPARα agonists such as fibrates can lower TG and increase the levels of high-density lipoprotein cholesterol (HDL-C). 10,11 Thus, fibrates play key roles in the treatment of lipid disorders. 12 However, they also exhibit various side effects, for example, rhabdomyolysis, liver toxicity and nephrotoxicity. [13][14][15] Liver X receptors (LXRs) are another kind of nuclear receptor that plays key roles in the regulation of lipid metabolism. LXRs are composed of two isotypes, LXRα and LXRβ, all of them can form obligate heterodimers with RXRα and then bind to a specific DNA recognition sequence known as an LXR response element. 16,17 The LXRs are activated by endogenous ligands such as oxysterols, intermediate precursors in the cholesterol biosynthetic pathway such as desmosterol, and synthetic agonists such as T0901317. They are inactivated by inhibitors such as GSK2033 and SR9243. 16 Upon activation, LXRs improve the initial step of reverse cholesterol transport (RCT) by up-regulation of its target genes ATP-binding cassette (ABC) G1 and ABCA1 in peripheral cells such as macrophages. In hepatocytes, LXRα stimulates the expression of cholesterol 7 alpha-hydroxylase A1 (CYP7A1), a rate-limiting enzyme in the bile acid synthesis pathway, and up-regulates the biliary cholesterol excretion via directly activating the expression of ABCG5 and G8 in the liver. 16 Furthermore, LXRs accelerate cholesterol excretion in the small intestine via activation of the expression of ABCG5 and G8. However, a big problem for LXRs agonists is that they can accelerate lipogenesis in the liver mainly due to the transcriptional induction of lipogenic genes such as sterol regulatory element-binding protein (SREBP)-1c, fatty acid synthase (FAS), acetyl-CoA carboxylase 1 (ACC1), stearoyl-CoA desaturase 1 (SCD1) and diacylglycerol O-acyltransferase (DGAT), increasing hepatic TG levels. [16][17][18] Furanone, a five-membered heteroaromatic ring containing an oxygen atom, is considered as one of the pharmacophores of biologically active substances. Furanones are classified into three main types: 2(3H)-furanones, 2(5H)-furanones and 3(2H)-furanones.
Furanone is present in lots of natural products including food and has been reported to have various biological functions such as anticancer, antiviral, antifungal, antibacterial, anti-inflammatory, antioxidant, antiarthritic and anti-hyperlipidaemic. 19 The best known and most studied furanone is ascorbic acid (Vitamin C). During the past decades, researchers paid a great deal of attention to the synthesis of furanone derivatives and investigation of their pharmacological activities. 19, 20 Husain et al reviewed therapeutic potential of furanone derivatives in the literature from 1987 to 2018, focusing on their anti-inflammatory, anticancer and antimicrobial activities. 19 We have reported 14 new and 17 known metabolites from the fungus Setosphaeria sp SCSIO41009. 21 Here, we reported for the first time that the furanone named as 5-hydroxy-3-methoxy-5-methyl-4-butylfuran-2(5H)-one had an effective lipid-lowering activity via influencing multiple processes of lipid metabolism.

| Preparation of lipoproteins
Plasma was obtained from healthy volunteers at the Affiliated Hospital of Weifang Medical University. To obtain LDL fraction, plasma was subjected to sequential ultracentrifugation as previously described. 22,23 In brief, the plasma density was adjusted to 1.006 g/mL for ultracentrifugation at 10°C (400 000 × g for 24 hours). The upper layer containing very low-density lipoproteins was removed, and the density was re-adjusted to 1.063 g/mL for ultracentrifugation at 400,000 × g for an additional 24 hours to obtain the upper layer containing low-density lipoproteins (LDL). EDTA-2Na (0.1%, w/v) was added to chelate the metal ions, thereby reducing oxidation during ultracentrifugation. The protein content of the fractions was determined by the Bradford method. The LDL fraction was stored at 4°C until use.
Oxidized LDL (Ox-LDL) was prepared by the method described previously. 23 In brief, LDL (~10 mg/mL) was incubated with CuSO 4 (10 μmol/L) at 37°C for 24 hours; then, the reaction was stopped by addition of 500 μmol/L EDTA-2Na. The resulting ox-LDL was dialysed against 0.01 M phosphate-buffered saline (PBS, PH = 7.4) at 4°C for 24 hours and then filtered through a 0.22-μm filter and stored at 4°C until use. for 24 hours. Cell viability was evaluated by the MTT method. 24,25 In brief, after the treatment period, 20µL of 5.0 mg/mL MTT was added to each well. Cells were then incubated for an additional 2 hours for the formation of formazan crystals. After washing cells in PBS for 3 times, formazan was solubilized in 150 µL of DMSO, and the optical density at 570 nm was recorded using a SpectraMax i3x

| Cell culture
Multi-Mode Microplate Platform (Molecular Devices, San Jose, CA, USA). Controls were defined as groups with vehicle only treatment.

| Oil Red O staining
For Oil Red O staining, cells grown on glass coverslips in six-well plates were fixed with 4% (w/v) paraformaldehyde at room temperature for 30 minutes and then stained with filtered Oil Red O solution (5 mg/mL in 60% isopropanol) at room temperature for 1 hour.
Afterwards, the coverslips were washed with H 2 O for several times, dried and mounted on slides. 26,27 Lipid-stained area was captured and quantified using Axio Vert.A1 inverted microscope (Zeiss, Jena, Germany), and images were recorded with an Axiocam 506 colour camera (Zeiss). For the LXR inhibition experiments, 10 µmol/L of LXR antagonist GSK2033 or SR9243 was added 2 hours before the addition of the furanone. 27,28 The cells were washed in PBS for 3 times and treated with 0.2 mL of RIPA lysis buffer at 4°C for 30 minutes. 29,30 The obtained mixture was heated at 70°C for 10 minutes in a water bath and then centrifugated at 1500 × g for 5 minutes. The obtained supernatant was used to detect the lipid levels by assay kits according to the manufacturer's instructions. The absorbance at 505 nm was recorded by a SpectraMax i3x Multi-Mode Microplate Platform (Molecular Devices, San Jose, CA, USA).

| Measurement of intracellular lipid levels
HepG2 cells were incubated with 0.5 mmol/L oleic acid for 24 hours. Then, cells were washed with PBS for 3 times and treated with 5 µmol/L of the furanone, 1 µmol/L of T0901317 or 5 µmol/L of fenofibrate for an additional 24 hours. For PPARα inhibition experiments, 10 µmol/L of PPARα antagonist MK886 or GW6471 was added 2 hours before the addition of the furanone. 27,28 Afterwards, lipids in cells were extracted and measured using the same methods as described above. The same treatment was also used to investigate the changes in the levels of mRNA and protein.

| Fluorescent immunocytochemistry assay
The treated RAW 264.7 macrophages grown on glass coverslips were washed with PBS 3 times, fixed with 4% paraformaldehyde,

| Real-time quantitative PCR
Total RNAs were isolated from cells using TRIzol reagent (SparkJade, Qingdao, China) according to the manufacturer's instructions. The concentration and purity of total RNAs were determined spectrophotometrically by measuring the absorbance at 260 nm and 280 nm using a UV spectrophotometer, and cDNA was produced using an ABI Veriti 96-Well Thermal Cycler (Waltham, MA, USA) and FastQuant RT Kit with gDNase (Tiangen). Real-time PCR was performed in an ABI QuantStudio3 PCR System (Waltham, MA, USA) using SYBR Green qPCR Master Mix and gene-specific primers with an initial denaturation step at 95°C for 10 minutes followed by 40 cycles of 95°C for 15 seconds, 58°C for 30 seconds and 68°C for 60 seconds. The primers for qRT-PCR are listed in Table 1. The fold change in the expression of targets relative to the housekeeping gene GAPDH was calculated based on the 2 -ΔΔCt relative expression formula.

| Protein isolation, electrophoresis and Western blotting
Total proteins from the cells were extracted using RIPA lysis buffer with complete protease inhibitor according to the manufacturer's instructions. Equal amounts of protein were subjected to 6% or 10% SDS-PAGE and transferred onto polyvinylidene fluoride membranes by electroblotting. After blocking in Tris-buffered saline containing 0.1% Tween-20 and 5% non-fat dry milk for 2 hours at room temperature, the membranes were incubated with primary antibodies overnight at 4°C. After washing 3 times, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 2 hours at room temperature. Immunoblots were revealed by enhanced chemiluminescence reaction and visualized using a high-performance chemiluminescence film. Images were captured by Clinx ChemiScope 6000 Pro (Shanghai, China), and densitometry analysis was conducted using Clinx Image Analysis Software (Shanghai, China). The expression of the proteins was normalized by housekeeping protein β-actin. 24,31

| Data analysis
All the bioassay results were expressed as the mean ± standard deviation (SD) for at least three independent experiments. Statistical analysis was performed with one-way analysis of variance (ANOVA) followed by Tukey's test. Differences were considered to be significant at a P < .05.

| The purity of the furanone is suitable for pharmacological study
The furanone, 5-hydroxy-3-methoxy-5-methyl-4-butylfuran-2(5H)one, isolated from the fermentation broth of the fungus Setosphaeria sp SCSIO41009 exhibited a single peak on an Acquity UPLC BEH C18 column, and its purity was 98.1% as evaluated by the peak area ( Figure 1). This furanone showed a sodium adduct ion at m/z 223.0946 [M + Na] + , and its molecular formula was established as C 10 H 16 O 4 by the NMR data. 21 In general, the purity of a compound above 95% is suitable for pharmacological study. 32,33 Taken together, these data demonstrated that the furanone is pure enough for further pharmacological study. The detailed structure of the furanone as shown in Figure 1 was described in our previous publication. 21

| The furanone lowered lipid levels in ox-LDLladen RAW264.7 cells
The cytotoxicity of the furanone, fenofibrate and LXR agonist T0901317 was determined by the MTT method. As shown in Figure 2A, the furanone and fenofibrate did not show significant cytotoxicity within the concentration up to 10 μmol/L. However, T0901317 reduced cell viability by ~25% and 74% at the concentration of 5 μmol/L and 10 μmol/L, respectively. To avoid cytotoxicity, the concentration of the furanone and fenofibrate was set to 5 μmol/L, and the concentration of LXR agonist T0901317 was set to 1 μmol/L as described previously. 34 Oil Red O staining results indicated that ox-LDL at a concentration of 50 μg/mL can significantly elevate the lipid levels in   Figure 2B) and ~48% (P < .01, Figure 2B), respectively. Further, TC and TG analysis by assay kits showed that T0901317 markedly reduced the cellular levels of TC (~43%, P < .01, Figure 2D) and TG (~33%, P < .05, Figure 2E) compared to the model group. The furanone treatment reduced cellular TC and TG levels by ~49% (P < .01) and ~58% (P < .01), respectively. More importantly, the furanone's TG reduction effect was better than T0901317 ( Figure 2E, P < .05).

| The furanone improved the expression of transporters in RAW264.7 cells
It is well documented that ABCA1 mediates cholesterol efflux from peripheral cells to apolipoprotein A1 or pre-β HDL, while ABCG1 and SR-B1 mediate cholesterol efflux from peripheral cells to mature HDL. 22 In this study, LXR agonist T0901317 significantly enhanced the protein levels of ABCA1, ABCG1 and SR-B1 ( Figure 3A-D). It was of note that the furanone increased the expression of ABCA1 about threefold and 60% as compared to the vehicle and model, respectively, as determined by Western blot ( Figure 3A) and immunocytochemistry ( Figure 3B). Furthermore, the furanone increased the protein levels of ABCG1 and SR-B1 by ~1.8-fold and 1.5-fold, respectively, compared with vehicle group (Figure 3C and D, P < .01). However, the effect of T0901317 and furanone on the expression of ABCG1 did not show any significant difference compared to the model group ( Figure 3C).
Taken together, the effects of this furanone at the concentration of 5 μmol/L on enhancing cholesterol transporters were close to that of LXR agonist T0901317 at the concentration of 1 μmol/L.

| Addition of LXR antagonists partially inhibited the mRNA levels of LXRs and ABC transporters compared to those treated with furanone alone
Further investigation using LXRs antagonists demonstrated that GSK2033 and SR9243 inhibited the mRNA levels of LXRα and LXRβ by ~52% and ~40%, respectively, ( Figure 4A  for GSK2033 and ~30% for SR9243) and LXRβ (~28% for GSK2033 and SR9243) compared to the furanone alone treatment. That is to say, the addition of the furanone significantly increased the mRNA levels of LXRs compared with the treatment of LXR antagonists ( Figure 4A and B). More importantly, LXR antagonist addition reduced the mRNA levels of ABCA1 and ABCG1 by ~43% and 36%, respectively, compared with the group treated with the furanone alone. On the other hand, compared to the antagonist alone treatment, the addition of this furanone significantly enhanced the mRNA levels of ABCA1 and ABCG1 by ~38% and 50%, respectively ( Figure 4C and D).

| The furanone lowered lipid levels in macrophages partially via enhancing the LXRα/ ABC pathways
As shown in Figure 5A and B, LXRα agonist T0901317 significantly improved the protein levels of LXRα, but not LXRβ compared with the model group. It was worth noting that the furanone, like T0901317, significantly enhanced the expression of LXRα but not LXRβ ( Figure 5A and B). We also found that LXR antagonists GSK2033 and SR9243 inhibited the protein levels of ABCA1

| The TG-lowering effect of the furanone was mainly attributed to its effect on PPARα in RAW264.7 cells
Compared to the model group, PPARα agonist fenofibrate significantly improved the mRNA and protein levels of PPARα by ~2.5-fold and ~1.9-fold, respectively (P < .01, Figure 6A and B).
Furthermore, the furanone increased the mRNA and protein levels of PPARα by ~48% and 130%, respectively, compared to the model group (P < .01). It was of note that PPARα antagonists MK886 and GW6471 reduced the TC-and TG-lowering effect of the furanone from 55% to 34% and from ~40% to ~12%, respectively ( Figure 6C and D). These data suggest that PPAR antagonists reduce TC-and TG-lowering effect of the furanone by ~38% and 70%, respectively, and indicate that the PPARα pathway plays an important role in the lipid-lowering effect of the furanone in RAW264.7 cells.
To further understand why the furanone had a better TG reduction effect than T0901317, we measured the mRNA levels of genes involved in fatty acid biosynthesis, elongation and desaturation via RT-PCR. As shown in Figure 6E-H, LXRα agonist T0901317 significantly enhanced the mRNA levels of SREBP-1c, FAS, ACC1 and DGAT1 compared to the vehicle (P < .01 or P < .05). Although the furanone significantly increased the mRNA levels of ACC1 and DGAT1 by ~40% (P < .05), but not as much as T0901317, especially for that of ACC1 ( Figure 6G, P < .05 compared to the T0901317treated group). Additionally, the mRNA levels of SCD1 and DGAT2 were undetectable; Ct numbers were ~30.

| The furanone reduced lipid levels in oleic acidloaded HepG2 cells
The cytotoxicity effect of the furanone on HepG2 cells was similar to that on the RAW264.7 cells ( Figure 7A). Although palmitic acid alone or in combination with oleic acid could induce lipid accumulation, palmitic acid is not as effective as oleic acid and palmitic acid can induce the expression of PPARα. 34 Oleic acid alone is generally used to induce lipid accumulation in HepG2 cells. 35 Therefore, oleic acid was used in the present study. Oil Red O staining results demonstrated that 0.5 mmol/L oleic acid treatment significantly increased lipid accumulation in HepG2 cells (~3.3fold, P < .01, Figure 7B and C), and fenofibrate at 5 μmol/L can markedly decrease the cellular lipid levels compared to the model group ( Figure 7B and C). Further assays indicated that fenofibrate significantly decreased TC (~31.5%, Figure 7D, P < .01) and TG (~47.4%, Figure 7E, P < .01) levels in HepG2 cells when compared to the model group. Interestingly, the furanone also significantly reduced the levels of TC (~78.2%, P < .01, Figure 7D) and TG (~48.7%, P < .01, Figure 7E) in HepG2 cells when compared with the model group.

| The furanone enhanced LDLR and decreased SREBP-2 expression in HepG2 cells
SR-B1 and LDLR mediate the transfer of HDL cholesterol and non-HDL lipids from plasma to the liver for metabolism, respectively. In this study, we found that the furanone showed no significant effect on the mRNA and protein levels of SR-B1 ( Figure 8A and B). PCSK9 ( Figure 8E). Additionally, fenofibrate significantly increased SR-B1 mRNA but not protein levels.
SREBPs are important transcription factors involved in the regulation of lipid metabolism and homeostasis in the liver. As shown in Figure 8F-I, fenofibrate treatment showed no significant effect on the mRNA and protein levels of SREBP-1c and SREBP-2. In this study, the furanone significantly enhanced the mRNA levels of SREBP-1c (~27%, P < .05, Figure 8F) without any effect on the protein levels ( Figure 8G). Furthermore, the effect of the furanone on the mRNA levels of SREBP-1c was similar to that of fenofibrate ( Figure 8F), but to a much lesser extent than T0901317 (P < .05). It was worth noting that this furanone significantly reduced the mRNA (~35%, P < .01, Figure 8H) and protein levels of SREBP-2 (~29%, P < .05, Figure 8I).
Furthermore, the furanone exhibited a significant difference in the inhibitory effect of SREBP-2 compared to fenofibrate (~50%, P < .01, Figure 8H and I).

| The furanone increased the expression of ABCG5 and ABCG8 in HepG2 cells
CYP7A1 is an important rate-limiting enzyme in the biosynthesis process of bile acids. In this study, we found that fenofibrate did not affect the mRNA and protein levels of CYP7A1 ( Figure 9A and B). On the other hand, the furanone significantly elevated the mRNA levels of CYP7A1 compared to the model or fenofibrate group (~3-fold, Figure 9A, P < .01), but did not affect the protein levels of CYP7A1 ( Figure 9B). It was of note that the furanone significantly elevated the mRNA and protein levels of ABCG5 ( Figure 9C and D, ~2-fold, P < .01) and ABCG8 ( Figure 9E and F, ~1.8-fold) compared to the model group. The effects of the furanone on ABCG5 and ABCG8 were stronger than that of fenofibrate that showed no significant effects on the expression of either ABCG5 or ABCG8 ( Figure 9C-F, P < .01 or P < .05).
Although the effect of the furanone on the mRNA levels of PPARα was weaker than that of fenofibrate ( Figure 10C, P < .01), there was no significant difference in the protein levels between the two groups ( Figure 10B). To define the role of PPARα in the anti-hyperlipidaemic effect of the furanone, two antagonists were used in the following study. As shown in Figure 10C and D, the TC-lowering effect of the furanone was changed from 63% (without antagonist) to 35% (with antagonist, Figure 10C, P < .01), and the TG-lowering effect was changed from 45% (without MK886) to 16.0% (with MK886, Figure 9D, P < .05). These results indicated that PPARα antagonists inhibited the TC-and TG-lowering effect of the furanone To further investigate the TG-lowering effect of the furanone, the mRNA levels of genes involved in fatty acid biosynthesis, elongation and desaturation were investigated using RT-PCR. As shown in Figure 10E-H, LXRα agonist T0901317 significantly enhanced the mRNA levels of FAS, ACC1, SCD1 and DGAT1 compared to that of the vehicle (P < .01 or P < .05). The furanone promoted the mRNA levels of ACC1 (~47% increase, P < .05); however, the effect was lower than that of T0901317 ( Figure 10F). It was of note that the effects of the furanone on the mRNA levels of FAS and SCD1 were significantly lower compared to that of T0901317 ( Figure 10E and G, P < .01). Additionally, the mRNA levels of DGAT2 were undetectable.

| D ISCUSS I ON
RCT is a physiological process in which excess peripheral cholesterol is transported to the liver for excretion into the bile and then faeces. 22 It is believed that RCT participates in the reduction of hyperlipidaemia and cardiovascular disease. 36,37 To imitate the RCT Furthermore, SR9243 showed no effect on the TG accumulation in zebrafish hepatocytes. 40 These promiscuous activities of the LXR antagonists may be due to their unexpected effects on lipogenic genes. For instance, GSK2033 could significantly improve rather than suppress the expression of the lipogenic genes such as FAS and SREBP-1c and therefore had no effect on hepatic steatosis in a mouse model of non-alcoholic fatty liver disease. 41 Furthermore, these antagonists may target a number of other nuclear receptors, such as the glucocorticoid receptor, pregnane X receptor and farnesoid X receptor, all of which can definitely alter hepatic gene expression. 41 However, the underlying mechanisms of these antagonists on lipid metabolism need to be further investigated. The inconsistence of the effects on the mRNA and protein levels of SR-B1, SREBP-1c ( Figure 8) and CYP7A1 (Figure 9) may be attributed to the unknown post-transcriptional modifications.
SREBPs are important transcription factors involved in the regulation of lipid metabolism and homeostasis in the liver. 48  SREBP-2 specifically activates the transcription of genes involved in cholesterol metabolism, such as HMG-CoA reductase and LDLR, thereby regulating cholesterol biosynthesis. 48 The results of fenofibrate on SREBPs were consistent with previous reports. 51,52 Our data demonstrated that the lipid-lowering effect of this furanone may be partially attributed to the down-regulation of SREBP-2 in HepG2 cells. Fenofibrate did not significantly modulate the expression of SREBP-2 ( Figure 8I) in the present study, which was consistent with previous reports. 51,52 Furthermore, PPARα agonists, fibrates, have been successfully explored as anti-hyperlipidaemic drugs, especially for TG lowering. 53,54 We demonstrated that the furanone, like fenofibrate, increased the protein and mRNA levels of PPARα in both RAW264.7 and HepG2 cells. Interestingly, PPARα antagonist MK886 and GW6471 could only H, mRNA expression of DGAT1. Data are expressed as mean ± SD (n = 3). # means P < .05 vs vehicle; ## means P < .01 vs vehicle; && means P < .01 vs fenofibrate. In Fig. C and D, *means P < .05 vs OA alone; **means P < .01 vs OA alone; ^m eans P < .05 vs furanone (without antagonist) group; ^^m eans P < .01 vs furanone (without antagonist) group. In Fig. E-H, *means P < .05 vs T0901317; **means P < .01 vs T0901317 partially inhibit the lipid-lowering effect of the furanone. However, its anti-hyperlipidaemic effects could not be fully abolished. Therefore, the furanone may exert its lipid-lowering effect, especially on TG, via up-regulating PPARα in RAW 264.7 and HepG2 cells as that of fenofibrate, 47,54 but other unknown mechanisms also contribute to this effect. In addition, the PPARα inhibitors, MK886 and GW6471, had no effects on cellular lipid levels in HepG2 cells. Consistently, it has been reported that MK886 had no significant effect on oleic acid-induced lipid accumulation in cells. 30 The underlying mechanisms may be similar to those of LXR antagonists 41 as aforementioned.
Hundreds of furanone derivatives have been synthesized during the past decades, and these compounds exhibited various activities. 19 Based on our findings, the furanone may not induce severe lipid accumulation in the liver as those of LXR agonists due to its weak effect on the expression of TG biogenesis genes. However, the lipid-lowering mechanisms of the furanone need to be further investigated in other models and especially in vivo models.
We will synthesize a series of the furanone-based compounds and then investigate the association between their structure and activity.

ACK N OWLED G EM ENT
This work was supported by Natural Science Foundation of China (81770463, 31300639) and Shandong Provincial Natural Science Fund (ZR2013HQ014).

CO N FLI C T O F I NTE R E S T
The authors declare that there are no conflicts of interest in this manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on requirement.