Synthesis and anti-obesity effects in vivo of Crotadihydrofuran C as a novel PPARγ antagonist from Crotalaria albida

Crotadihydrofuran C (CC) from the herbs of Crotalaria albida is able to inhibit adipocyte differentiation and lipid accumulation. However, the effects of CC on obesity and metabolic disorders have not yet been elucidated. In our study, the first enantioselective synthesis of the 2-isopropenyl dihydrofuran isoflavone skeleton (CC) is described. The convenient and efficient synthetic protocols developed skilfully solve the problems of the ortho-para directing group and Suzuki coupling reaction using a boronic acid pinacol ester that was more stable and easy to obtain. Furthermore, CC treatment of high-fat diet (HFD)-fed obese mice remarkably reduced their body weight, fat mass, and lipid level as well as improved insulin resistance and non-alcoholic fatty liver disease (NAFLD). A TR-FRET assay showed that CC was specifically bound to PPARγ LBD, which was further confirmed by the molecular docking study. These results suggest that CC could be a useful and potential natural product for treating metabolic diseases, including obesity, hyperlipidemia insulin resistance and NAFLD, without toxic side-effects.


Crotadihydrofuran C counters obesity and improves HFD-induced obesity in mice.
Based on the results of our previous studies that CC inhibited PPARγ transactivity and suppressed adipocyte differentiation via reducing mRNA expression of PPARγ target genes in vitro 25 , we postulated that CC might have a treatment effect on obesity and metabolic disorders in obese mice. HFD feeding for three months led to a significant increase in the body weight and created an obesity model in C57/BL mice (Fig. 4a). In contrast, supplementation with CC decreased the weight gain and fat accumulation in a dose-dependent manner in HFD-fed mice (Fig. 4a,b). The reduction in weight gain of CC-treated mice was largely attributed to decreased overall fat mass, without any change in the lean mass (Fig. 4b). Histological analysis showed CC treatment caused a remarkable decrease in the size of the adipocytes in the visceral adipose tissue versus mice fed only a high-fat diet (Fig. 4c). In addition, we did not observe obvious clinical signs of toxicity or mortality, such as changes in skin, fur, eyes, gait, posture, response to handling and the bizarre behavior during the entire period of the study, as well as no toxicity to cell proliferation in our previous study. These results revealed that effect of CC on body weight and metabolism could not be caused by CC toxicity. Furthermore, the mean body temperature and food consumption did not vary significantly between the HF and CC groups (Fig. 4d,e), suggesting that the effects of CC on obesity parameters were not due to decreased food intake and energy expenditure. Hence, these data indicated that CC has a beneficial effect in reducing body weight gain and fat accumulation in obese mice. CC treatment ameliorates diabetes and hyperlipidemia in DIO mice. Consistent with reduced adiposity, CC-treated mice showed significant improvement in their glucose haemostasis (Fig. 5a,b). After 20 days of treatment with CC, mice exhibited remarkable reductions in blood glucose levels following intraperitoneal glucose injection compared with obese control mice, suggesting that CC enhanced glucose tolerance. Similar results were detected in an intraperitoneal insulin tolerance test. Hyperlipidemia is one hallmark of insulin resistance and type 2 diabetes. As shown in Fig. 5c, the serum insulin level of obese mice was higher than that of the low fat diet group. Compared with the model group, an obvious reduction of the insulin level was observed in the CC-treated group. Furthermore, CC remarkably down-regulated the level of leptin, which is the key regulator of body weight secreted by adipocytes, but it did not lead to any change in the adiponectin level (Fig. 5d,e). The above results demonstrated that CC attenuates serum insulin and leptin levels in HFD-induced obese mice, which may bring benefits for the improvement of obesity and insulin resistance.
To investigate the effect of CC on the improvement of lipid metabolism, the levels of total triglyceride (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-c) and low-density lipoprotein cholesterol (LDL-c) were determined in the serum (Fig. 5f). The TG levels of HFD mice increased significantly. After 20 days of treatment, the TG levels of the CC-treated group were effectively reduced and almost reached those of the chow group. However, the TC, HDL-c and LDL-c levels remained unchanged. These results indicated that CC treatment could reduce the serum lipid level and improve hyperlipidemia in DIO mice.
CC reverses high-fat diet-induced steatosis, steatohepatitis and liver fibrosis in mice. NAFLD is frequently associated with insulin resistance and is defined as a liver component of metabolic syndrome, which is related to excessive accumulation of hepatic fat and encompasses a spectrum of conditions ranging from steatosis alone to steatohepatitis with inflammation and fibrosis 26 . To investigate whether CC improves hepaticsteatosis, we compared the profiles of the livers of mice fed a chow diet, HFD alone or HFD containing CC. As shown in Fig. 6a, although HFD induced massive hepatic steatosis, CC decreased lipid accumulation in the liver, which was verified by quantification of the TG content of liver tissues (Fig. 6b). Compared with the HF group, CC lowered the liver TG content by approximately 30% and did not cause any difference in the TC content in obese mice. Next, we determined the contents of alanine aminotransferase (ALT) and aspartate aminotransferase (AST), which are characterized as liver diagnostic markers. The results showed that CC reduced the increase in the plasma concentration of AST induced by HFD. In contrast, CC did not lead to any significant change in the ALT level, but it showed a decreasing tendency (Fig. 6c). Furthermore, histological analyses of the mouse liver demonstrated that CC further ameliorated tissue structure and hepatic steatosis (Fig. 6d). These above results indicated that oral CC is effective in reducing HFD-induced hepatic steatosis in mice.  Heptaic steatosis is a precursor of more advanced liver disease and can progress to non-alcoholic steatohepatitis(NASH), which is characterized by inflammation and fibrosis [27][28][29] . In order to further confirm that CC adimistration could bring about sustained improvement of liver fibrosis, we next analysed the serum level of pro-inflammatory cytokines such as tumor necrosis factor α (TNF-α ), interleukin (IL)-1β and IL-6. As shown in Fig. 6e, the serum level of TNF-α , IL-1β and IL-6 was significantly increased in HF group compared with the chow group. CC treatment remarkably decreased the levels of TNF-α , IL-1β and IL-6. This anti-inflammatory effect was also reflected in reduced the mRNA expression of inflammation mediators including IL-1β , IL-6, TNFα , and cluster of differentiation (CD) 68 in adipose tissue, whereas the levels of IL-4 and IL-10 were not changed remarkably (Fig. 6f). These results suggest that CC may improve inflammatory state in DIO mice. Activation of hepatic stellate cells and extracellular matrix synthesis are responses to hepatic injury 30 . To further determine whether CC protects against HF diet-induced liver fibrosis, we then assessed the hepatic mRNA levels of several marker genes through quantitative PCR (qPCR). As shown in Fig. 6g, expression of transforming growth factor (TGF)-β 1, α -smooth muscle actin (SMA), type I collagen (Col1) a and type III collagen (Col3) a significantly increased in mice fed with HF diet. Treatment with CC reversed the increases of α -SMA, Col1a and Col3a expression, which indicate that anti-fibrogenic properties are displayed by CC. In addition to alteration of the expression of pro-fibrogenic genes, we also investigated expression of genes involved in mediating extracellular matrix (ECM) remodelling. Similar to the results of pro-fibrogenic genes, CC administration markedly suppressed unregulated levels of tissue inhibitors of metalloproteinase (TIMP)1, matrix metalloproteinase (MMP) 9 caused by hepatic fat accumulation. Taken together, these results demonstrate that CC effectively reversed HFD-induced hepatic steatosis, inflammation and even fibrosis.

CC functions as a PPARγ antagonist with a potent binding affinity and regulates PPARγ gene expression in vivo.
PPARγ is a master regulator of adipocyte differentiation, glucose and lipid metabolism, and inflammation. PPARγ agonists induce lipogenesis and adipocyte differentiation. In contrast, PPARγ antagonists have opposite effects. In our previous studies, several results revealed that CC may be a PPARγ antagonist 25 . First, a reporter assay showed that CC selectively reduced the transactivity of PPARγ induced by rosiglitazone. Second, CC inhibited adipocyte differentiation of 3T3-L1 cells, accompanying with the reduced expression of PPARγ and its downstream genes in vitro 25 . To confirm this hypothesis and further confirm the binding of CC to PPARγ , we next performed competitive binding assays using time-resolved fluorescence resonance energy transfer (TR-FRET). As shown in Fig. 7a, CC is capable of dose-dependently displacing the rosiglitazone from binding to PPARγ and showed a strong binding to human PPARγ (K i = 1.57 μ M, IC 50 = 3.65 μ M). Next, docking studies were performed to investigate the exact binding sites of CC to PPARγ . Since the CC is in the S configuration, the hydrogen atom of isopropenyl at the 2′ position forms a weak H-bond with His266. The same hydrogen atom could form a H-benzene interaction with Phe264, which enhances the binding stability. In addition, the carbonyl group of chromone forms a strong H-bond with Cys285. Reasonably, we speculate that CC with a S configuration has stronger activity than that of the R configuration. The predicted binding model showed that interaction between CC and the PPARγ ligand binding site is similar to the observation between PPARγ and its known agonist, TZDs. In our present studies, we have confirmed that CC competed with rosiglitazone to bind to the human PPARγ receptor with a Ki of 1.57 μ M. Furthermore, molecular docking studies suggested CC may interact with the ligand binding domain of PPARγ , where hydrogen bonds at His266, Phe264 and Cys285 are predicted to be formed. These presented evidences support a notion that CC is a novel PPARγ antagonist.
To reveal the mechanism underlying CC-regulated improvement in metabolic disorders, the gene expression profiles of liver and white adipose tissue were analysed by qPCR. Consistent with the results in 3T3-L1 adipocytes, CC markedly suppressed the expression of the classic PPARγ target gene involved in the insulin sensitivity and synthesis and transport of fatty acids in liver tissue, such as adipose fatty acid-binding protein 2 (aP2), fatty acid synthase (FAS), lipoprotein lipase (LPL) and acetyl-CoA carboxylase (ACC) (Fig. 7c). Similarly, the expressions of PPARγ , aP2, the cluster of differentiation 36 (CD36) and stearoyl-CoA desaturase (SCD) 1 in white adipose tissue were remarkably downregulated in DIO mice (Fig. 7d). These data indicate that CC has been shown to have PPARγ -independent action on fat accumulation and lipogenesis in vivo, supporting the idea that CC could be a potential PPARγ antagonist.
In summary, this paper outlines the first high-yielding enantioselective synthesis of the 2-isopropenyl dihydrofuran isoflavone skeleton using a palladium catalyzed Suzuki coupling reaction in the presence of aboronic acid pinacol ester. This is a first chiral synthesis to obtain a large enough enough amount to carry out animal experiments. Furthermore, we identified that CC reduces weight gain and fat accumulation as well as improves glucose homeostasis, hepatic lipid, inflammation and fibrosis as a novel natural antagonist of PPARγ . The potent effect of CC on obesity and metabolic diseases without apparent toxic side-effects makes CC a promising candidate in the development of anti-obesity pharmacotherapy.

2-Hydroxy-6-methoxybenzaldehyde (5).
To a mixture of 4 (50 g, 300 mmol), NaI (110 g, 734 mmol) in CH 3 CN (700 ml) and CH 2 Cl 2 (350 ml) at 0 °C AlCl 3 (100 g, 734 mmol) was added portion-wise over a period of 30 min. The resulting mixture was further stirred at 0 °C for 10 min. TLC showed that the reaction was complete. The mixture was washed with saturated aqueous Na 2 S 2 O 3 solution, the organic layers were separated and dried over saturated aqueous NaCl, and then, they were concentrated under reduced pressure to give desired products

Animals and diets.
Six-week-old female C57/BL6 mice were purchased from the SLAC laboratory and housed in a temperature-controlled and pathogen-free room under a 12:12-h light-dark cycle with free access to food and water until the start of the experiments. We fed thirty seven-week-old C57/BL6 mice with HFD to induce obesity for three months and ten mice with LFD as a chow group. The obese mice were randomly fed with Scientific RepoRts | 7:46735 | DOI: 10.1038/srep46735 CC-1 (50 mg/kg) or CC-2 (100 mg/kg) milled in HFD or regular HFD for 20 days (n = 10). At the end of the study period, all mice were fasted for 12 h and then scarified. Blood and tissue samples were collected and either fixed in 4% paraformaldehyde or snap frozen and stored at − 80 °C.
All animal experiments and protocols met the requirements of the Animal Ethics Committee and approved by Shanghai University of Traditional Chinese Medicine in accordance with the Guide for the Care and Use of Laboratory Animals (Approved Number: SZY201604003).
Metabolic Analyses. Body weight and food consumption were recorded every two days, and the body composition was measured by nuclear magnetic resonance (NMR) spectroscopy. For the intraperitoneal glucose tolerance test (ipGTT) and intraperitoneal insulin tolerance test (ipITT), all mice were received with 1 g/kg glucose and 0.75 U/kg insulin (Sigma, St. Louis, MO). The total cholesterol, TG, insulin, leptin, adiponectin, ALT and AST of serum and hepatic lipids were detected as described previously 20 . Inflammatory markers assay in serum. Serum TNF-α , IL-1β and IL-6 were measured using enzyme-linked immunosorbent assay according to the manufacturer's instructions (ebioscience, USA). All determinations were performed in duplicates.
Histochemistry. The fresh liver samples and white adipose tissue were fixed in 4% paraformaldehyde, embedded in paraffin, sectioned into 5 μ m and stained with haematoxylin and eosin (H&E). For studying lipid accumulation in the liver, liver samples were frozen, embedded in O.C.T. compound, sliced into 8 μ m, and then stained with Oil Red O. All protocols are in accordance with a standard procedure.

TR-FRET assay.
A LanthaScreen time-resolved fluorescence resonance energy transfer (TR-FRET) competitive binding assay was performed according to the manufacturer's (Invitrogen, Germany) protocol. In brief, the assay was conducted in black noncoated low-volume round-bottomed 384-well Coring plates. The interaction between the ligand-binding domain (LBD) of human nuclear receptors tagged with glutathione S-transferase (GST), the terbium labelled anti-GST antibody and the fluorescent small molecule was measured by a PerkinElmer Envision plate reader (PerkinElmer, Waltham, MA). For the TR-FRET ratios, the emission signal at 520 nm was divided by the emission signal at 495 nm.
Computational molecular docking. The crystal structure of PPARγ (PDB code 2I4J) was retrieved from the Research Collaborator for Structural Bioinformatics (RCSB) Protein Bank. Docking was conducted using MOE2012.10.
RNA isolation and quantitative real-time PCR. The total RNA from liver and adipose tissue was isolated using the Trizol reagent (TaKaRa, Japan). Oligo-dT-primed cDNA synthesis was carried out with 3 μ g of total RNA using a cDNA synthesis kit (Thermo, America). Quantitative real-time assays for mRNA measurements were performed using SYBR green in an ABI StepOnePlus Real-Time PCR system (Applied Biosystems, Foster City, CA) with primer sequences as listed in Table 1. All final expressions of target genes were calculated relative to the housekeeping gene β -actin.
Statistical Analysis. All data are presented as a mean ± SEM. Individual pairwise comparisons were analysed by paired or unpaired two tailed t-tests. Two-way analysis of variance (ANOVA) was used for multiple comparisons. Differences with P values < 0.05 were considered statistically significant.