Bauhiniastatin-1 alleviates diet induced obesity and lipid accumulation through modulating PPAR-γ/AMPK expressions: In-vitro, in-vivo and in-silico studies.

Background: Consumption of energy dense foods and sedentary lifestyles have led to high prevalence of obesity and associated disorders. Intensive research efforts have focussed to develop effective alternative therapeutics from plant sources. Bauhiniastatins have been reported to possess antineoplastic activity. In the present study, Bauhiniastatin-1 (BSTN1) was isolated and puried from Bauhinia purpurea and evaluated for its therapeutic ecacy against adipogenesis and obesity using high fat diet (HFD)-induced obese rodent model and 3T3-L1 cells. Methods: We performed in-vitro experiments like MTT assay, Oil Red O (ORO) stain, cellular lipid content, glycerol release and RT-PCR analysis in 3T3-L1 cells. In-vivo parameters like body weight gain, body composition, plasma adipokines, serum & liver lipid proles, liver marker enzymes, western blot analysis and histopathological examination were conducted in rat model. In addition, molecular docking studies were also performed to understand interaction of BSTN1 with peroxisome proliferator-activated gamma receptor (PPAR-γ) and AMP-activated protein kinase (AMPK) which supported our experimental results. Results: BSTN1 at 20 μM signicantly (p<0.001) inhibited cell differentiation and lipid accumulation of 3T3-L1 adipocytes. Mechanistic studies showed that mRNA expression of key adipogenic markers, PPARγ, fatty acid synthase (FAS) and sterol-regulatory element-binding protein-1 (SREBP1) were down-regulated while AMPK was up-regulated by BSTN1. Oral administration of BSTN1 (5 mg/kg. b.wt.) to HFD-induced obese rats substantially decreased body weight gain, fat mass, serum and liver lipid levels and promoted integrity of hepatic and adipose tissue architecture compared to HFD-control rats. In BSTN1 administered groups, decreased serum aspartate transaminase (AST) and alanine aminotransferase (ALT) levels, decreased plasma leptin but increased adiponectin levels were noted. Western blot analysis of adipose and hepatic tissues collected from BSTN1 treated rats showed decreased expression level of PPAR-γ but increase in AMPK expression relative to the untreated group. In-silico studies showed strong binding interactions of BSTN1 against PPAR-γ and AMPK, the key molecules of adipogenesis and obesity. Conclusions: Taken the

antimicrobial, analgesic and anti-cancerous activities and hence used in traditional and folk medicine to treat various diseases [3].
Obesity, once considered as a cosmetic issue, has drawn tremendous attention in recent times because of its critical role in metabolic syndrome disorders. The excess calories consumed than expended are stored in adipose tissue (AT), liver and muscle, and if it continues for longer periods leads to obesity, NAFLD and associated disorders [4]. Non-alcoholic fatty liver disease (NAFLD) impacts almost 25% of adults in the general population and includes a spectrum of diseases ranging from steatosis to non-alcoholic steatohepatitis. This condition is an independent risk factor for cardiovascular diseases, diabetes mellitus and all-cause mortality [5,6,7]. Moreover, recent reports have linked obesity to high severity in COVID-19 infected patients [8]. As there are very limited non-surgical therapeutic options and in view of the immense demand for drugs to treat obesity and NAFLD and considering the adverse effects associated with existing drugs, research efforts to develop natural product-based therapeutics have drawn much attention [9,10,11]. In our previous study, we reported the anti-hyperlipidemic activity of bark extract of Bauhinia purpurea Linn. [12]. In the present study, we have made an attempt to isolate and purify the active principle, Bauhiniastatin-1, evaluated its anti-NAFLD and anti-obesity activity and underlying molecular mechanisms using diet induced obese rat model. Adipogenic, lipogenic and lipolytic studies are complemented using the most widely used white adipogenesis cell culture model of 3T3-L1 preadipocytes. In addition, in silico molecular docking studies were carried out to understand the mechanisms of action of BSTN1.
Isolation and puri cation of Bauhiniastatin-1 (BSTN1) The bark of B. purpurea was collected from the Seshachalam forests, Tirupati, Andhra Pradesh, India. It is authenticated by taxonomist in the department of Botany, Sri Venkateswara University, Tirupati, voucher number is 136 and specimen was preserved in departmental herbarium. The bark of B. purpurea was powdered, extracted with ethanol and further fractionated by column chromatography using different solvents [12]. The collected fractions were subjected to LC-MS analysis on 6520 Accurate Q-TOF (Agilent Santa Clara, CA) mass spectrometer to identify the major compounds. Bauhiniastatin-1 was puri ed using an HPLC system equipped with a binary gradient system, a variable UV-VIS-detector and a Rheodyne Model 7725 injector with a loop size of 20 µL, and an integrator. Reverse phase chromatographic analysis was carried out in isocratic conditions using a C-18 reverse phase column (250 x 4.6 mm id., 5 µL C-18) at 40°C. Mobile phase consisted of methanol:water (20:80 v/v) with a ow rate of 1mL/min. The detection of compounds was performed at 220 nm. A single sharp peak at 5.942 min of retention time was identi ed as BSTN1 [13].

Cell culture and differentiation of adipocytes
The 3T3 L1 pre-adipocytes of American Type Culture Collection (ATCC) cells were maintained and cultured in DMEM supplemented with 10% FBS at 37°C in a humidi ed atmosphere with 5% CO 2 . For adipogenesis studies, 3T3-L1 were grown to con uence, cells were stimulated with adipogenesis differentiation medium of induction (DMI) consisting of DMEM, 5% FBS, 0.5 mM IBMX, 1 µM DEXA and 10 µg/mL insulin for 2 days followed by treating cells with differentiation medium (DMEM with 5% FBS and 10 µg/mL insulin) for additional 8-10 days [14]. All the media that we used contained 100 IU/mL penicillin and 100 mg/mL streptomycin. A volume of 0.01 % DMSO was used as vehicle control for invitro experiments. For evaluating anti-adipogenic effects of BSTN1, 3T3-L1 cells cultured and adipogenesis was induced in 12 well plates with different concentrations of BSTN1 (5, 10 and 20 µM) in differentiation medium. The adipocytes were stained for neutral lipids (lipid droplets) and observed under a bright eld microscope or used for other studies. Cytotoxicity studies / Cell viability assay (MTT assay) The 3T3 L1 preadipocytes were cultured in DMEM and cell viability assays were conducted as previously described [15,16]. Determination of adipocyte lipid content using quantitative Oil Red O staining Lipid contents in adipocytes were visualized as well as measured using Oil Red O (ORO) staining [17,18].

Lipolysis studies
The lipolysis studies were conducted by measuring glycerol levels released into the cell culture medium, using commercial kit (Lipolysis assay kit, ab185433, Abcam, Shangai) following manufacturer's instructions [19]. The glycerol content was expressed as nmol/well. RT-PCR studies Total RNA was isolated from 3T3-L1 cells by using tri-reagent (Sigma Aldrich, USA) according to manufacturer's protocol and reverse transcribed to obtain cDNA using cDNA synthesis kit (Applied Bio Systems, Foster City, USA) [20]. Two nano grams of cDNA was used for RT-PCR. The PCR ampli cation was performed with transcript speci c primers (Additional le S1). Animal studies Male WNIN rats (aged 5-6 weeks), normal pellet diet and high fat diets were obtained from National Institute of Nutrition (NIN), Hyderabad, India. After one-week acclimatization period, rats were fed with either normal diet or HFD, water ad-libitum and maintained at standard laboratory conditions (temperature: 23°C ± 2°C; humidity: 40-60%) for 18 weeks as described in experimental design. Among different types of HFDs used to induce obesity, a HFD with 60% calories from fat induces obesity most effectively comparable to western diet [21,22]. Animals were fed on 20% protein diet (normal pellet diet) (low fat control diet, percent of energy from carbohydrate-64%, protein-20% and fat-16%) (Additional le S2) or high fat diet (percent of energy from carbohydrate-20%, protein-20% and fat-60%) (Additional le S3) contained all the recommended macro and micronutrients. To test the therapeutic activity of BSTN1, 1.25, 2.5 and 5 mg/kg b.wt. of BSTN1 was suspended in 0.5% carboxymethylcellulose (CMC) and orally administered for 6 weeks from 13th week onwards using an intra-gastric tube. We selected these concentrations based on our initial pilot studies using solvent extracts [12]. The body composition, body weight, fat percent of each rat was measured by Total Body Electrical Conductivity (TOBEC) using a small animal body composition analysis system (EM-SCAN, Model SA-3000 Multi detector, Spring eld, USA). At the end of the experiment, animals were anesthetized using iso urane, blood was collected by heart puncture method. Plasma and/or serum were separated by centrifugation at 2500 rpm for 15 min. Various organs and tissues including abdominal adipose tissue and liver were dissected, and stored appropriately. For histology studies tissues were xed and processed as described in later sections. Plasma leptin and adiponectin levels Plasma leptin and adiponectin are key adipokines secreted by adipocytes. Adipokine levels were measured in experimental rats using enzyme-linked immunosorbent assay kits (Crystal Chem, Downers Grove, IL, USA). These assays were performed in duplicates (n = 6), as per the manufacturer's guidelines and adipokine levels were expressed in ng/mL [23]. Estimation of serum lipid pro le Serum total cholesterol (TC) was estimated by CHOD-PAP method, triglycerides (TGs) was estimated by GPO-TOPS method, HDL-cholesterol, VLDL-cholesterol, LDL-cholesterol were estimated by selective inhibition method (Agappe Diagnostics Ltd., Kerala, India), phospholipids (PLs) and free fatty acids (FFAs) were assessed as previously described [24].

Estimation of hepatic lipid levels
Lipids were extracted from the livers of experimental animals as described [25]. In brief, the tissues were rinsed with ice-cold physiological saline, homogenized in cold chloroform-methanol (2:1, v/v) and the contents were extracted for 24 h. The extraction was repeated four times. The combined ltrate was washed with 0.7% potassium chloride and the aqueous layer was discarded. The organic layer was made up to a known volume with chloroform and used for hepatic lipid analysis.

Measurement of AST and ALT activities
Hepatic marker enzymes, aspartate transaminase (AST) and alanine transaminase (ALT) activities were estimated at the end of the experiment by using commercially available kit (Agappe Diagnostics Ltd, Kerala, India) following the manufacturer's protocol.

Western blot analysis
Adipose and hepatic tissue proteins were extracted with lysis buffer (Sigma Aldrich, USA) and quanti ed using Bradford method [26]. Equal amount of protein (40 µg) was resolved on 10% SDS-PAGE gel and transferred onto a nitrocellulose membrane. To block non-speci c binding sites, blots were incubated at room temperature with 5% skimmed milk (v/v) for 1 h followed by overnight incubation with primary antibodies of anti-PPAR-γ, anti-AMPK and mouse anti-β actin (1:1,000 dilution, ABclonal Technology, USA) at 4°C. The immuno-reactive antigen was then recognized by incubation with HRP-conjugated secondary antibody (1:1,000 dilution, Abclonal Technology, USA). Immuno-reactive bands were visualized using chemiluminescence detection system.

Histopathological examination
Adipose and liver tissues were collected from both control and experimental rats and xed in formalin solution. A small piece of tissue was cut, trimmed, processed and prepared para n blocks. Then the para n blocks were sectioned (5-8 µm) using microtome and stained using haematoxylin and eosin (H&E) following standard histology protocol [27]. Molecular docking studies (Accession of target protein) The three-dimensional structure of FAS (6NNA), AMPK (6C9F), PPAR-γ (3WMH) and BSTN1 were downloaded from the RCSB protein Data Bank and Pub chem. The atomic coordinates of the ligand were geometrically optimized using Argus Lab 4.0.1. [28]. In-silico studies were carried against FAS (6NNA), AMPK (6C9F) and PPAR-γ (3WMH), with ligand (BSTN1) using the docking program Patchdock [29]. After the docking, protein-ligand complexes were studied using PyMol viewer tool (www.pymol.org)1. Protein and ligand interactions were analysed and visualized through PyMol viewer tool (www.pymol.org)1.

Statistical analysis
The results are expressed as the mean ± standard deviation (SD), and comparison was made by using one-way ANOVA programme followed by Tukey's post hoc tests to study the signi cance level (SPSS, version 17.0; SPSS Inc., Chicago, IL, USA).

LC-MS & HPLC analysis
The chloroform fraction of ethanolic extract of B. purpurea was subjected to LC-MS analysis. Seventeen compounds were identi ed from which BSTN1 (m.wt.: 283.0496) was isolated, puri ed and con rmed by HPLC (Figs. 1A-1C).
Cytotoxicity studies by MTT assay The effect of BSTN1 on cell viability of 3T3-L1 cells and cytotoxicity was analysed at 48 h using MTT assay. BSTN1 showed IC50 value of 118 µM in the dose range of 10-160 µM, dose was optimized to less than 10% for further in-vitro analyses (Fig. 2D).
Effect of BSTN1 on adipogenesis and lipid content BSTN1 at 20 µM markedly reduced adipogenesis and intracellular lipid levels of 3T3-L1 cells, as observed by Oil-Red-O-stained images ( Fig. 2A). We further quanti ed the lipid content by extracting the (using isopropanol) Oil-Red-O stain from 3T3-L1 adipocytes (Fig. 2B). To complement these studies, we veri ed the lipolytic capacity of BSTN1 by quantifying the levels of glycerol release from treated versus control and vehicle control 3T3-L1 adipocytes. Signi cantly high level of glycerol release was observed in adipocytes treated with BSTN1 than untreated cells and the maximum lipolytic activity was noticed at a concentration of 20 µM (Fig. 2C).

mRNA expression (RT-PCR) studies
The mRNA expression levels of key adipogenic transcriptional and lipogenic factors, FAS, SREBP1, AMPK and PPAR-γ in differentiated 3T3-L1 adipocytes were estimated in the absence and presence of BSTN1 (5, 10 and 20 µM) (Fig. 3). The expression of FAS, SREBP1 and PPAR-γ were down-regulated, while AMPK was up-regulated with increasing concentration of BSTN1 as represented in Figs. 3A to 3D.
Adipogenic and lipogenic protein markers expression in adipose and hepatic tissues Figure 4 shows the protein expression levels of AMPK and PPAR-γ in hepatic and adipose tissues of placebo and obese-treated rats. BSTN1 treated obese rats showed signi cant alteration in the key adipogenic and energy homeostasis proteins, PPAR-γ and AMPK respectively. The expression of these proteins was normalized to house-keeping gene β-actin. BSTN1 mediated normalized levels of AMPK and PPAR-γ proteins could reduce the fat accumulation in liver and adipose tissues, reduce non-alcoholic fatty liver disease and lead obesity ailments.
Body composition and body weight of obese rats Table 1 depicts the changes in body weight and body composition of experimental rats. Consumption of HFD for 18 weeks resulted in signi cant changes in body weights (486 ± 9.0 g) and total body fat levels (79 ± 7.8 g) in HFD control group, compared to normal control group of rats whose body weight and total fat were 248.1 ± 6.3 g and 10.2 ± 1.4 g, respectively. Oral administration of BSTN1 (1.25, 2.5 and 5 mg/kg. b.wt.) for 6 weeks (from 13 to 18 weeks) considerably reduced body weight gain and body composition parameters in a dose dependent manner with maximum reduction (in body weight 320 ± 13.9 g and total fat levels 24.9 ± 1.9 g) being noted in rats treated with 5 mg/kg. b.wt. of BSTN1.  Effect of BSTN1 on leptin and adiponectin levels and adipose tissue architecture Figure 5 shows the systemic levels of leptin and adiponectin in control and experimental obese rats. We observed markedly elevated levels of plasma leptin but decrease in adiponectin levels in HFD fed rats, when compared to the normal rats. Interestingly, treatment with BSTN1 has signi cantly (p < 0.01) decreased leptin levels, while the levels of adiponectin were increased. The H&E-stained adipose tissue sections of control rats showed the normal adipocyte structure (Fig. 5B). In contrast, adipocytes of HFDinduced obese rats are signi cantly larger (Fig. 5C) (hypertrophy) with indistinct cell walls compared to adipocytes of control rats. Relatively smaller sized adipocytes were visualized in BSTN1 treated rats indicating reduced lipid droplets, and this effect was more signi cant than even orlistat treated group (Figs. 5D and 5E).
Effect of BSTN1 on hepatic enzymes (AST and ALT) and hepatic tissue architecture AST and ALT levels (p < 0.01) in dose dependent manner when compared to HFD control rats. Hepatic tissues of untreated and treated rats were sectioned, stained with H&E staining that represents normal hepatic structure made up of healthy hepatic lobules and a central vein possessing radiating strands in untreated rats (Fig. 6B). But an accumulated lipid content around the central vein and in between the hepatocytes lead to in ated and disruption of lobules were seen in obese rats (Fig. 6C). Interestingly, the liver sections from obese rats treated with BSTN1 (Fig. 6D) were seem to be better/equally recovered than that of Orlistat treated group (Fig. 6E). However, central veins were found to be a little dilated in BSTN1 groups than normal but cells as such were found to be healthy.

Effect of BSTN1 on serum and liver lipid pro les
The HFD-induced obese rats exhibited signi cant alterations in serum and liver lipid levels (over their respective control rats). Treatment with BSTN1 (1.25, 2.5 and 5 mg/kg. b.wt.) had signi cantly normalized the altered levels of serum total cholesterol (TC), triglycerides (TGL), high-density lipoprotein (HDL), low-density lipoprotein (LDL), very low-density lipoproteins (VLDL), phospholipids (PLs) and free fatty acids (FFA) ( Table-2). Similarly, the liver tissue lipids including total cholesterol (TC), triglycerides (TGL) and free fatty acids (FFA) were also markedly normalized in the presence of BSTN1 (Table 3).

Discussion
In the present study, Bauhiniastatin-1, an oxepin derivative was isolated and puri ed from Bauhinia purpurea and evaluated for its anti-obesity and anti-NAFLD e cacy. Accumulation of lipid droplets in adipocytes plays a central role in the growth and hypertrophy of adipose tissue. For the prevention and management of obesity, inhibition of adipocyte differentiation and promotion of glycerol release by breakdown of triglycerides present in adipocytes is therefore essential [30]. Activated AMPK attenuates lipid accumulation during adipogenesis by inhibiting the expression of PPARγ, C/EBPα, SREBP1 and FAS [31]. AMPK also plays a critical role in regulating hepatic lipid metabolism, glucose transport and gluconeogenesis, hence, AMPK may also be considered as a target for the treatment of obesity and NAFLD [32]. In this study, as evident from the Oil Red O stain and glycerol estimation assay, the puri ed BSTN1 showed substantial inhibition of adipogenesis and lipid accumulation in 3T3-L1 cells compared to untreated cells ( Figs. 2A-2C). With BSTN1 treated adipocytes, down-regulation of gene expression of PPARγ, SREBP1 and FAS and up-regulation of AMPK was observed, indicating their role in energy homeostasis, attenuating adipogenesis and lipogenesis (Figs. 3A-3D). PPARγ consists of DNA binding domain, agonist-dependent-and agonist-independent-activation domain expressed ubiquitously but abundantly in adipocytes [33]. PPARγ activates fatty acid uptake leading to neutral lipid accumulation in adipocytes [34] and su cient for adipogenic induction [35,36] making it a master regulator of adipogenesis. PPARγ sensitizers, thiozolidinediones, induce adipocyte differentiation [37,38] but improve glycemic control, insulin sensitivity and β-cell function implicating PPARγ's role in metabolic regulation [39,40]. Overall, PPARγs are important transcriptional factors that play a central role in glucose homeostasis, fatty acid synthesis and adipogenesis [41]. Obese condition leads to insulin resistance which causes activation of PPAR-γ and triggers glucose transporters (Glut-2) in liver that favour glucose uptake by hepatocytes and synthesis of triacylglycerols (TAGs) in liver ultimately leading to the pathogenesis of NAFLD [42,43]. However, in our studies, BSTN1 down regulated PPARγ thus, might reduce ectopic fat accumulation in hepatic tissue and prevents non-alcoholic fatty liver condition.
In animal experiments we did analysis of body composition parameters. A substantially reduced body weight gain and total fat mass was noticed with BSTN1 treated rats over their HFD controls. Biochemical assays of serum AST, ALT, serum and liver lipid and plasma adipokine pro les indicated the therapeutic role of BSTN1 in attenuating HFD induced alterations in rats. Adipokines are adipose tissue-derived bioactive molecules that play important role in human health or disease. Leptin and adiponectin might promote lipolysis, fatty acid oxidation and inhibit lipogenesis in liver and adipose tissue through AMPK activation [44,45,46]. AMPK gets activated through phosphorylation of a conserved threonine in the activation loop of kinase domain in the α subunit [47,48] or binding of AMP that can be mimicked by ADP and can be antagonized by ATP [49]. Many of the AMPK activators including natural plant derived therapeutics like resveratrol and berberine activate AMPK by indirectly inhibiting ATP levels through increasing AMP and/or ADP [50,51,52].
Histological examination of liver and adipose tissue micrographs clearly demonstrated the antiadipogenic and anti-lipogenic or pro-lipolytic effect of BSTN1 (Figs. 5 and 6). Western blot and RT-PCR analysis of liver and adipose tissue showed down regulated expression of PPAR-γ but up-regulation of AMPK in BSTN1 treated groups con rming anti-adipogenic and anti-lipogenic roles of BSTN1 (Figs. 3 and  4). This is in agreement with the fact that, when AMPK is activated, it diminishes the expression of C/EBPα, PPARγ, SREBP1 and acyl-CoA carboxylase (ACC) in adipose and hepatic tissues and thus inhibits synthesis of fatty acids and triglycerides [53]. Green tea extract, capsaicin, guggulsterone, genistein and piperine were shown to regulate glucose homeostasis and fatty acid synthesis in liver as well as adipogenesis in adipose tissue through PPARγ and AMPK mediated signaling [54].
Binding of BSTN1 against PPAR-γ and AMPK showed favourable binding interaction compared to FAS and other adipogenic pathway transcriptional factors that we tested, demonstrating them as possible targets for BSTN1 to attenuate NAFLD, adipogenesis and obesity. Our docking studies also suggest that BSTN1 binds AMPK at the AMP binding site through hydrogen bonding with Thr86 and two other residues, Val127 and Pro129. BSTN1 binding with Thr86 implicates that BSTN1 potentially activates AMPK function as well as expression through the interactions or vice versa. MEKT75 show similar scaffold like MEKT-21 and 9i (antagonist), implicating direct inhibitory activity of BSTN1 on PPAR-γ function [55]. The observed reduction of PPAR-γ expression could be in part due to these interactions initiating negative feedback loop. In addition, potential binding prediction of BSTN1 with FAS indicates the inhibitory function of BSTN1 on FAS as well. BSNT1 binding at fatty acid binding site of FAS seem to be very interesting, but the lower negative energy of these interaction warrants to further investigate these enzyme substrate/inhibitor interactions.
Our studies further demonstrate that BSTN1 binds to AMPK near the adenosine monophosphate (AMP) binding site there by potentially mimicking AMP to activate AMPK. However, the binding of BSTN1 to the AMPK and PPAR-γ proteins seems to have positive and negative regulatory action on the transcription of AMPK and/or PPAR-γ respectively. Additionally, 5-aminoimidazole-4-carboxamide riboside (AMPK activator) mediated activation of AMPK was shown to inhibit PPAR-γ and PPAR-α which may not be mediated through RXR or PPAR/RXR binding to DNA implicating potential multiple level regulation of BSTN1 on PPAR-γ [56]. This suggests complex agonist/antagonist binding dynamics with AMPK and PPAR-γ explaining BSTN1 binding to these two key proteins.
MTT assay showed non-signi cant reduction in 3T3-L1 cell viability up to 48 hours, however the observed anti-adipogenic functions in our studies might also be partially due to additional BSTN1 mediated adipocyte speci c apoptotic pathway regulation as observed in Oil-Red-O stain photomicrographs. Cell viability (in vitro assays), histology studies and hepatic health markers status indicate the safety of BSTN1.

Conclusions
In this study, both in vivo and in vitro studies have clearly demonstrated the therapeutic e cacy of BSTN1 against adipogenesis, obesity and NAFLD via PPAR-γ and AMPK modulation. These observations are further complimented and con rmed by molecular docking studies. Therefore, BSTN1 can be considered as a potential therapeutic molecule for managing obesity ailments.

Declarations
Ethics approval and consent to participate This study was approved by the Institutional Animal Ethical Committee of Sri Venkateswara University. The institutional review board approved this procedure.

Consent for publication
All authors revised the data and approved the manuscript being submitted. And we have not used any third parties' data.
Availability of data and materials The analyzed and/or used datasets presented herein are available from the corresponding author upon reasonable request.

Competing interests
No competing interests among all the authors.     treated obese rat's show smaller adipocytes than the adipocytes of obese untreated group. # p<0.001 indicate signi cant difference between HFD control and normal control groups. * p<0.05 and ** p<0.01 indicates signi cant difference between HFD control (placebo) and treated groups.