Flavonoid derivative exerts an antidiabetic effect via AMPK activation in diet-induced obesity mice

Abstract In our previous study, a derivative of tiliroside, 3-O-[(E)-4-(4-ethoxyphenyl)-2-oxobut-3-en-1-yl]kaempferol (Fla-OEt) significantly enhanced glucose consumption in insulin resistant HepG2 cells. This article deals with the antihyperglycemic and antihyperlipidemic effects of Fla-OEt in diet-induced obesity (DIO) mice. Daily administration of Fla-OEt significantly decreased oral glucose tolerance test, intraperitoneal insulin tolerance test and serum lipids. Hyperinsulinemic–euglycemic clamp and the ratio of high-density-lipoprotein/low-density-lipoprotein with Fla-OEt treatment were increased comparing with high-fat diet (HFD) group, so lipid metabolism was improved. Histopathology examination showed that the Fla-OEt restored the damage of adipose tissues and liver in DIO mice. Moreover, compared with HFD group, Fla-OEt treatment significantly increased the phosphorylation of AMPK and ACC in adiposity tissues, liver, and muscles. The mechanism of its action might be the activation of AMPK pathway. It appears that Fla-OEt is worth further study for development as a lead compound for a potential antidiabetic agent. Graphical abstract


Introduction
Type 2 diabetes (T2D) is a metabolic disease characterised by insulin resistance (IR) and elevated hepatic glucose output. The most risk factor for T2D is obesity, in which the chronic overconsumption of food leads to metabolic dysregulation such as hyperglycemia, hyperlipidemia and IR. These conditions are frequently associated with reduced insulin sensitivity and impaired glucose and lipid metabolism. Due to the increasing number of T2D and obesity cases in the world, more effective therapies for maintaining glucose homeostasis and improving insulin sensitivity is greatly needed.
AMP-activated protein kinase (AMPK) is well known as a major energy sensor and a master regulator for maintaining whole body metabolic homeostasis (Hardie 2008). In model systems, sustained decreases in AMPK activity accompany IR, whereas AMPK activation increases insulin sensitivity (Steinberg & Kemp 2009). Nearly, all of the physiological effects of peripheral AMPK activation would be beneficial for a patient with T2D. For this reason, the pharmacological activation of AMPK has been a seemingly promising target for drug discovery and development (Coughlan et al. 2014).
Our previous results suggest that tiliroside isolated from Potentilla chinensis Ser. (Rosaceae) exerted antihyperglycemic, antihyperlipidemic and antioxidant effects in vivo (Qiao et al. 2011). It is also reported that tiliroside treatment activated AMP-activated protein kinase in both the liver and skeletal muscles (Goto et al. 2012). Furthermore, a series of tiliroside derivatives were designed and synthesised by our group. Several flavonoid derivatives, including Fla-OEt ( Figure S1) reveal significant enhancing glucose consumption effects in IR HepG2 cells in comparison with metformin. In addition, an analogue of Fla-OEt, 3-O-[(E)-4-(4-cyanophenyl)-2-oxobut-3-en-1-yl]kaempferol, significantly activates AMPK activity and reduces ACC activity in HepG2 cells . This article deals with the antidiabetic effect of Fla-OEt in high-fat diet (HFD) induced obesity mice (DIO mice). In this study, the AMPK activation of Fla-OEt in vivo was also investigated.

Effects of Fla-OEt on physical parameters
Feeding with HFD, the mice in the HFD group developed obesity and had increased body and visceral fat pad (epididymis, mesenteric and perirenal adipose) weights compared to those in the LFD group ( Figures S2A and S2C). After Fla-OEt and metformin treatment for four weeks, body weights and visceral adipose tiusse weights decreased compared to those in the subjects of the HFD group. Furthermore, no obvious differences were observed between groups treated with Fla-OEt and metformin. At the same time, the amount of food intake was almost the same for all groups fed with HFD ( Figure S2B). Recent investigations have revealed several natural products which possessed weight-reducing effect, such as Plantago lanceolata leaves (Yoshida et al. 2012) and Moro orange extract, which contains flavone glycosides (Cardile et al. 2015). In this article, Fla-OEt, as a pure compound, also had a significant anti-obesity action.

Effect of Fla-OEt on glucose tolerance, insulin tolerance and hyperinsulinemic-euglycemia clamp test
Glucose tolerance and clearance were further carried out by oral glucose tolerance test (OGTT) and intraperitoneal insulin tolerance test (IPITT), respectively. In the OGTT, Fla-OEt and metformin administrations effectively improved glucose tolerances by decreasing the peak of the plasma glucose and reduced AUCs after oral glucose loading (Figures S3A and  S3B). After injection of insulin, blood glucose levels decreased more quickly in Fla-OEt-and metformin-treated mice compared to DIO mice that were not treated ( Figures S3C and S3D).
The glucose infusion rate (GIR), which was required to keep a steady state of euglycemia, was lower in mice on the HFD group than in the LFD group. The GIRs of subjects in the Fla-OEt and metformin groups were significantly higher than those of the subjects in the untreated HFD group ( Figure S3E). In addition, Fla-OEt treatment resulted in obvious reductions in the level of serum insulin in DIO mice (Table S1). These results indicated that Fla-OEt treatment successfully ameliorated the impaired insulin sensitivity induced by HFD.

Effect of Fla-OEt on serum general characteristics
As shown in Table S1, the serum levels of TC, TG and NEFA in mice fed with HFD were increased, while the level of high-density-lipoprotein (HDL)/low-density-lipoprotein (LDL) decreased compared to those in the normal control group. Supplementation with Fla-OEt significantly reduced the levels of serum TC, TG and NEFA. Although Fla-OEt and metformin shared a number of features in metabolic lipid disorder, the HFD + Fla-OEt group showed a significant enhancing effect on the ratio of HDL/LDL (p < 0.05), while HFD + Met group had no significant difference compared to the results for the HFD group (Table S1).

Effect of Fla-OEt on histopathological change in epididymal adipose tissues and the liver
After four weeks of supplementation, compared to the HFD group, the adipocyte sizes of subjects in the Fla-OEt group were significantly decreased. There were no obvious accumulations of lipid droplets apparent in the livers of the Fla-OEt-treated group compared to the HFD group ( Figure S4). Histological analysis of both the adipose tissue and the liver confirmed the development of Fla-OEt in DIO mice models.

Effect of Fla-OEt on the activation of AMPK pathway in adiposity tissue, muscle and liver
The activation of AMPK is beneficial for energy balance via the phosphorylation of various downstream substrates on a multitude of tissues. Once activated, AMPK stimulates hepatic fatty acid (FA) oxidation and reduces FA synthesis and glycolysis (Steinberg & Kemp 2009), inhibits gluconeogenesis, as well as cholesterol and triglyceride synthesis. It inhibits adipocyte lipolysis, stimulates skeletal muscle glucose uptake (Ruderman et al. 2013) and mitochondrial biogenesis, and inhibits protein and glycogen synthesis (Steinberg & Kemp 2009).
As shown in Figures S5A-S5C, compared to the HFD group, Fla-OEt treatments markedly enhanced the phosphorylation of AMPK and its downstream target ACC in the adiposity tissue, liver and muscle. Meanwhile, Fla-OEt treatment significantly decreased the TG, TC and NEFA accumulation and increased the HDL/LDL ratio in DIO mice. These findings may be explained by AMPK activation and upregulation of the expression of p-ACC caused by Fla-OEt treatment. Moreover, it was found that there were antidiabetic effects and molecular mechanisms involved in the beneficial effects of dietary flavonoids (Babu et al. 2013).
However, there have been a few studies about AMPK activation of flavonoids in vivo and piperlongumine in hepG2 cells (Ryu et al. 2014), such as dietary bilberry extract containing large amounts of anthocyanins ameliorates hyperglycemia and insulin sensitivity via the activation of AMP-activated protein kinase in KK-Ay mice (Takikawa et al. 2010). In the present study, Fla-OEt treatment with 5 mg/kg/day in vivo exhibited antidiabetic effect via AMPK activation compared with the large intake of diet flavonoids (Takikawa et al. 2010;Babu et al. 2013).

Conclusion
In conclusion, Fla-OEt, a derivative of tiliroside, reduces whole-body adiposity, ameliorates metabolic lipid disorder, improves insulin sensitivity and exhibited benefits for other disorders characterised by IR in DIO mice. The mechanism of its action, at least in part, might be the activation of AMPK in the target tissues of insulin resistance. It appears that Fla-OEt is worth further study for development as a lead compound for a potential antidiabetic agent.

Supplementary material
Experimental details relating to this article are available online alongside Figures S1-S5, and Table S1.

Disclosure statement
No potential conflict of interest was reported by the authors.