Bidens pilosa and its active compound inhibit adipogenesis and lipid accumulation via down-modulation of the C/EBP and PPARγ pathways

Obesity and its complications are a major global health problem. In this study, we investigated the anti-obesity effect and mechanism of an edible plant, Bidens pilosa, and its active constituent. We first assessed the long-term effect of B. pilosa on body composition, body weight, blood parameters in ICR mice. We observed that it significantly decreased fat content and increased protein content in ICR mice. Next, we verified the anti-obesity effect of B. pilosa in ob/ob mice. It effectively and dose-dependently reduced fat content, adipocyte size and/or body weight in mice. Moreover, mechanistic studies showed that B. pilosa inhibited the expression of peroxisome proliferator activated receptor γ (PPARγ), CCAAT/enhancer binding proteins (C/EBPs) and Egr2 in adipose tissue. Finally, we examined the effect of 2-β-D-glucopyranosyloxy-1-hydroxytrideca-5,7,9,11-tetrayne (GHT) on adipogenesis in adipocytes. We found that B. pilosa significantly decreased the adipogenesis and lipid accumulation. This decrease was associated with the down-regulation of expression of Egr2, C/EBPs, PPARγ, adipocyte Protein 2 (aP2) and adiponectin. In summary, this work demonstrated that B. pilosa and GHT suppressed adipogenesis and lipid content in adipocytes and/or animals via the down-regulation of the Egr2, C/EBPs and PPARγ pathways, suggesting a novel application of B. pilosa and GHT against obesity.

of fullness 10 . Despite their efficacy, weight-loss drugs are often accompanied by undesirable side effects as well as cost-effectiveness concerns 11 . A wealth of information indicates that plants and their compounds can decrease food intake and fat absorption, increase lipid metabolism and stimulate energy expenditure 12 . Therefore, plants and their compounds are considered to be a natural, alternative way to control obesity.
B. pilosa is an Asteraceae plant. The Food and Agriculture Organization of the United Nations promoted its cultivation in the 1990 s because it is easy to grow, palatable and edible 13 . This plant is commonly used as a potherb or herbal medicine globally 14 . Currently, B. pilosa is reported to treat over 41 categories of diseases 15 . Over 200 phytochemicals have been identified from this plant, which may explain some of its pharmaceutical actions 15 . In addition to hypertension 16,17 , this plant 18,19 and/or its polyynes 20,21 has been demonstrated to treat diabetes. More recently, the use of B. pilosa as a nutraceutical was reported to be clinically effective against diabetes 22 . However, the anti-obesity mechanism of B. pilosa is not clear. In this study, we first investigated the effect of B. pilosa on food intake, fat content, body weight and/or adipocyte size in ICR and ob/ob mice. Further, we tested the effect of this plant on the expressional regulation of Egr2, C/EBPs and PPARγ in adipose tissues. Additionally, we examined the effect of B. pilosa on blood biochemistry. Finally, we studied the action of GHT, one active polyyne of B. pilosa, on adipogenesis and regulation of the gene expression of Egr2, C/EBPs and PPARγ in adipocytes.

Results
Long-term effect of B. pliosa on body weight, biochemical and hematological parameters and body composition in ICR mice. To explore the anti-obesity effects of B. pilosa, we first assessed its longterm effect on body weight, body composition, serum biochemistry and hematology in ICR mice. ICR male (Sup. Fig. S1a) and female (Sup. Fig. S1b) mice were randomly assigned into 4 groups with 5 mice per group. Four groups were fed with a standard diet (0% BP), and standard diet with 0.5% B. pilosa extract (BP), 1.5% BP and 2.5% BP for 24 weeks, respectively. No significant difference (P ≥ 0.05) in the body weight of ICR mice in either gender was observed before and after 24-wk treatment (Sup. Fig. S1a, S1b and Table 1). The food and water intake in control and B. pilosa-fed groups of male and female ICR mice were not statistically different (P ≥ 0.05, data not shown). Of note, body composition data showed that B. pilosa dependently reduced crude fat content in ICR females (Table 1). This reduction was more noticeable in males (Table 1). In addition, we also observed that B. pilosa dependently increased crude protein content in males to a greater extent than females (Table 1).
We also examined the effect of B. pilosa on biochemical and hematological parameters in the mouse blood of each group. The variation in biochemical and hematological parameters between the control and B. pilosa-fed groups was observed (Sup. Tables S1 and S2). However, these parameters were indeed within normal range. The results suggest the safety of B. pliosa.
Collectively, these data showed that B. pliosa had an impact on body composition in mice.
B. pilosa decreases body weight gain and fat content but increases lean tissue content in ob/ob mice. Next, the ob/ob mice, a mouse model of obesity, were used to further investigate the effect of B. pilosa on body weight and body composition. The 5-week-old males were randomly divided into 3 groups, 5 mice a group, fed standard diet (0% BP) and standard diet containing low (0.5% BP) or high (2.5% BP) dose of B. pilosa extract for 5 weeks. No significant difference (P ≥ 0.05) was observed in food and water consumption in the control and treatment groups of ob/ob mice (Fig. 1a). In contrast, B. pilosa dose-dependently decreased body weight (Fig. 1a), body fat (Fig. 1a) and serum lipids ( Table 2). For adipose tissues, B. pilosa significantly diminished the weight of visceral and subcutaneous fat but not brown fat (Fig. 1c). Akin to the data in ICR mice, NMR data showed that B. pilosa decreased fat content in ob/ob mice in a dose-dependent manner (Fig. 2a). B. pilosa also increased the Four groups of 5-week-old ICR males and females were fed standard diet and standard diet containing 0.5% B. pilosa extract (BP), 1.5% BP, and 2.5% BP for 24 wk. Body weight and composition were measured. The body composition including crude protein and fat were calculated as percentages in the dried carcass mass. The data from 5 mice per group are expressed as mean ± SEM. ANOVA was used to analyze the statistical significance. *P < 0.05 and **P < 0.01 are considered to be statistically significant when compared with control group (0% BP).   Table 2. Effect of B. pilosa on serum chemistry in ob/ob mice. Three groups of 5-week-old ob/ob mice were fed standard diet and standard diet containing 0.5% B. pilosa extract (BP) and 2.5% BP for 5 wk. Serum samples from 5 mice per group were collected for biochemical analysis and the data are expressed as mean ± SEM. *P < 0.05 is considered to be statistically significant when compared with control group (0% BP).
content of lean tissue in the mice (Fig. 2b). However, no difference (P ≥ 0.05) in the body fluid between control and B. pilosa-fed mice was observed (Fig. 2c).

B. pilosa reduces cell size and expression level of Egr2, C/EBPs and PPARγ in adipose tissues of obese mice.
To dissect the mechanism by which B. pilosa increased fat in mice, we first examined the effect of B. pilosa on adipocytes in brown, subcutaneous and visceral adipose tissue. The histochemical data revealed that B. pilosa dose-dependently reduced cell size of adipocytes in all three types of adipose tissue (Fig. 3a). The cell size distribution of these adipose tissues was quantified. A shift from large cell size to small cell size in adipose tissues was noted in B. pilosa-fed mice (Fig. 3b). Moreover, this shift was dependent on the dose of B. pilosa and more obvious in brown fat (Fig. 3b). Accordingly, B. pilosa dose-dependently diminished average adipocyte area in adipose tissues (Fig. 3c). Since B. pilosa effectively reduced fat accumulation and fat cell size, we then sought to find out which of the master gene(s) involved in adipogenesis and lipid metabolism regulated by B. pilosa were investigated. We first explored the impact of B. pilosa on the transcriptional and translational levels of Egr2, C/EBPγ , C/EBPβ , C/EBPα and PPARγ in the fat tissue of control and B. pilosa-fed mice. The data demonstrated that B. pilosa suppressed the expression of Egr2, C/EBPs and PPARγ in a dose-dependent fashion (Fig. 3d,e).

GHT inhibits adipogenesis but does not affect cell viability in (pre)adipocytes. Our previous
phytochemical studies and bioassays indicated that GHT is a major compound for glycemic control in B. pilosa extract 20,21 . To further explore the molecular mechanism of B. pilosa in adipogenesis in this study, we, on one hand, tried to identify the likely active compound(s) in B. pilosa by combining phytochemisty with adipogenesis  Fig. 1 were subjected to MRI analysis. Fat tissue, (a) lean tissue (b) and body fluid (c) in relation to body mass were measured and the data are presented as mean ± SEM. ANOVA was used to compare the difference between control and treatment groups and P ≥ 0.05 (NS) and P < 0.05 (*) are shown.  (Fig. 1) were collected. The adipose tissues were stained with hematoxylin and eosin (a). The size distribution (b) and average area (c) of adipocytes in every 100-mm 2 area range of adipose tissues were quantified using ImageJ software. The data, expressed as mean ± SEM, was analyzed using Student's t-test. P < 0.05 (*) are considered statistically significant. (d,e) Both mRNA (d) and protein (e) level of Egr2, C/EBPγ , C/EBPβ , C/ EBPα and PPARγ in visceral adipose tissue (VAT) of the ob/ob mice (Fig. 1) fed with different doses of B. pilosa was analyzed with RT-PCR and Western blot. The ratio of each gene product to that of internal control was calculated. assays in mouse 3T3-L1 pre-adipocytes. Based on this bioactivity-directed fractionation and isolation strategy (Sup. Fig. S2a), we found that B. pilosa extract and its butanol fraction, but not water fraction, inhibited adipogenesis in mouse 3T3-L1 pre-adipocytes (Sup. Fig. S2b). We also confirmed that GHT was one active compound for suppressing adipogenesis (Sup. Fig. S2b). Besides, GHT in B. pilosa extract was used for quality control among the batches of B. pilosa (Sup. Fig. S3). On the other hand, we investigated the molecular mechanism by which B. pilosa inhibited adipogenesis in mouse 3T3-L1 pre-adipocytes. Consistent with the data on the adiposity in ob/ob mice (Fig. 2), B. pilosa decreased the transcriptional and translational levels of Egr2, C/EBPγ , C/EBPβ , C/EBPα and PPARγ (Sup. Fig. S4) in differentiating adipocytes.
Furthermore, the effect of GHT on adipogenesis in mouse 3T3-L1 pre-adipocytes was investigated. GHT at 25 μM or less did not show cytotoxicity and a slight cytotoxicity was observed in GHT at 50 μM (Fig. 4a). Furthermore, the effect of GHT on the differentiation of 3T3-L1 pre-adipocytes into adipocytes was investigated. As expected, the inducer composed of Dex, IBMX and insulin initiated the production of lipid droplets in adipocytes compared to control cells (Vehicle vs Inducer, upper panel, Fig. 4b). Moreover, rosiglitazone enhanced this production (Inducer + RSG, upper panel, Fig. 4b). In sharp contrast, GHT dose-dependently inhibited this production (Inducer + GHT, upper panel, Fig. 4b). The inhibitory effect of GHT on the production of lipid droplets was also confirmed in human SGBS cells (lower panel, Fig. 4b).
In parallel, the effect of GHT on the transcriptional and translational levels of Egr2, C/EBPγ , C/EBPβ , C/ EBPα and PPARγ in the same 3T3-L1 cells as in Fig. 4b was investigated. The 3T3-L1 pre-adipocytes had a basal expression of Egr2, C/EBPs and PPARγ at the transcriptional (Fig. 4c) and translational (Fig. 4d) levels. In contrast, the inducer significantly up-regulated the expression of these genes at the transcriptional and translational levels. Further, rosiglitazone slightly increased this up-regulation (Fig. 4c,d). However, GHT decreased the up-regulation of the expression of these genes by the inducer (Fig. 4c,d). Subsequently, the expression level of aP2 and adiponectin, two downstream genes of C/EBPs and PPARγ were examined. As expected, the inducer up-regulated the expression level of aP2 and adiponectin. In contrast, GHT reduced this up-regulation (Fig. 4e).
Overall, the data showed that B. pilosa and its active constituent, GHT, reduced adipogenesis in adipocytes via down-regulation of Egr2, C/EBPs and PPARγ as well as their downstream genes, aP2 and adiponectin (Fig. 4f).

Discussion
Plants and compounds can exert anti-obesity action via reduction of appetite and fat digestion/absorption and/ or increase of lipid breakdown and energy expenditure 12 . Particularly, edible plants represent an extraordinary source of foods, nutraceuticals and pharmaceuticals against obesity. In this study, we demonstrated that B. pilosa reduces fat content in mouse models (Sup. Fig. S1 and Fig. 2). Its anti-obesity action involves the inhibition of Egr2, C/EBPs, and PPARγ pathways in adipose tissues (Fig. 3d,e). We also studied the likely molecular basis of B. pilosa and GHT in differentiating adipocytes (Fig. 4, Sup. Fig. S2, and Sup. Fig. S4). Consistent with the in vivo data (Fig. 3), we confirmed that B. pilosa and GHT inhibited adipogenesis via the reduced expression of Egr2, C/ EBPs, and PPARγ (Fig. 4). Although the exact molecular target of B. pilosa and GHT remains elusive, B. pilosa and GHT exerted their anti-adipogenic action via the down-regulation of the adipogenic transcriptional factors such as Egr2, C/EBPs and PPARγ as described in Fig. 4f. We also performed a bioavailability test for GHT. Following an oral administration, GHT was easily taken into the blood circulation, peaked by 30 minutes and declined within 6 hours (Sup. Fig. S5). The data suggest that GHT can be easily ingested and reach the circulation system.
B. pilosa is generally recognized as safe for ethnomedicinal or culinary use worldwide 13,15 . A dose of 400 mg B. pilosa extract per kilogram of body weight of three times per day, is considered safe in men 22 . Here, mice did not show any toxicity to a daily dose of B. pilosa extract at 27 g per kg body weight (data not shown). Consistently, this did not affect biochemical parameters and number of blood cells except for blood glucose in ICR mice (Sup. Tables S1 and S2) and blood glucose, insulin, lipids and lipoproteins in ob/ob mice (Table 2). Our data and other published literature suggest low safety risk of taking this plant orally. Aside from efficacy and palatability, B. pilosa is rather more cost-effective than ginseng and others due to its fast growth rate and low nutrient requirement.
High calorie diets and an inactive lifestyle drive adipocyte hyperplasia and hypertrophy, resulting in obesity. Thus, targeting adipogenesis and lipid metabolism is thought to be an anti-obesity strategy 12,23 . In this study, we showed that B. pliosa and its active compound, GHT, inhibited adipogenesis in adipocytes (Figs 3 and 4 and Sup. Fig. S4). The mechanistic studies showed that B. pliosa and GHT suppressed the up-regulation of gene expression of Egr2, C/EBPs, and PPARγ and their downstream genes, aP2 and adiponectin, during adipogenesis (Figs 3 and 4 and Sup. Fig. S4). Interestingly, we also observed an increase in protein content in ICR mice (Table 1) and ob/ ob mice (Fig. 2b). This observation might be due to the effect of B. pilosa on the energy shift from fat deposition to protein deposition since food intake did not show a significant difference in control and B. pilosa-fed animals (Fig. 1a). It has been shown that the energy required to deposit 1 kJ of protein and 1 kJ of fat was 2.25 and 1.36 kJ in Zucker rats, respectively 24 . Therefore, unequal conversion between fat and protein in ob/ob mice by B. pilosa may explain the reduction in their body weight.
Overall, our data taken alone with published literature confirm the beneficial function of the edible plant, B. pilosa, for metabolic syndromes such as diabetes 18,19,22 , hypertension 17   This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/