Bioorganic & Medicinal Chemistry Letters 25 (2015) 5237–5242

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry Letters

journal homepage: www.elsevier .com/ locate/bmcl

Saponarin activates AMPK in a calcium-dependent mannerand suppresses gluconeogenesis and increases glucoseuptake via phosphorylation of CRTC2 and HDAC5

http://dx.doi.org/10.1016/j.bmcl.2015.09.0570960-894X/� 2015 Elsevier Ltd. All rights reserved.

Abbreviations: 2-NBDG, 2-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-2-deoxyglucose; AMPK, AMP-activated protein kinase; CPT-cAMP, 8-(4-chlorophe-nylthio)-cAMP; CREB, cyclic AMP-responsive element-binding protein; CRTC2,CREB regulated transcription coactivator-2; ChIP, chromatin immunoprecipitation;FBS, fetal bovine serum; FoxO1, forkhead box O1; G6Pase, glucose-6-phosphatase;GLUT-4, glucose transporter-4; HDAC5, histone deacetylase-5; MEF2, myocyteenhancer factor-2; PEPCK, phosphoenolpyruvate carboxykinase; T2DM, type IIdiabetes mellitus.⇑ Corresponding author. Tel.: +82 2 3290 3029; fax: +82 2 3290 3653.

E-mail address: junelee@korea.ac.kr (S.-J. Lee).y Equal contribution.

Woo-Duck Seo a,y, Ji Hae Lee b,c,y, Yaoyao Jia b,c,y, Chunyan Wub,c, Sung-Joon Lee b,c,⇑aCrop Foundation Division, National Institute of Crop Science, Rural Development Administration, Wanju-Gun, Jeollabuk-do 565-851, Republic of KoreabDepartment of Biotechnology, College of Life Sciences & Biotechnology, Korea University, Seoul 136-713, Republic of KoreacDepartment of Food Bioscience and Technology, College of Life Sciences & Biotechnology, Korea University, Seoul 136-713, Republic of Korea

a r t i c l e i n f o

Article history:Received 7 May 2015Revised 19 September 2015Accepted 24 September 2015Available online 26 September 2015

Keywords:AMPKSaponarinGluconeogenesisGlucose uptake

a b s t r a c t

This study investigated the molecular mechanism of saponarin, a flavone glucoside, in the regulation ofinsulin sensitivity. Saponarin suppressed the rate of gluconeogenesis and increased cellular glucoseuptake in HepG2 and TE671 cells by regulating AMPK. Using an in vitro kinase assay, we showed thatsaponarin did not directly interact with the AMPK protein. Instead, saponarin increased intracellular cal-cium levels and induced AMPK phosphorylation, which was diminished by co-stimulation with STO-609,an inhibitor of CAMKKb. Transcription of hepatic gluconeogenesis genes was upregulated by nucleartranslocation of CRTC2 and HDAC5, coactivators of CREB and FoxO1 transcription factors, respectively.This nuclear translocation was inhibited by increased phosphorylation of CRTC2 and HDAC5 by sapo-narin-induced AMPK in HepG2 cells and suppression of CREB and FoxO1 transactivation activities in cellsstimulated by saponarin. The results from a chromatin immunoprecipitation assay confirmed thereduced binding of CRTC2 on the PEPCK and G6Pase promoters. In TE671 cells, AMPK phosphorylatedHDAC5, which suppressed nuclear penetration and upregulated GLUT4 transcription, leading to enhancedglucose uptake. Collectively, these results suggest that saponarin activates AMPK in a calcium-dependentmanner, thus regulating gluconeogenesis and glucose uptake.

� 2015 Elsevier Ltd. All rights reserved.

Type II diabetes mellitus (T2DM) is a chronic metabolic diseasecaused by insulin resistance, and its worldwide prevalence has dra-matically increased over the past few decades.1 Hyperglycemia andhyperinsulinemia in T2DM is primarily caused by uncontrolledhepatic gluconeogenesis and reduced glucose uptake in skeletalmuscles.2 Thus, appropriate control of hepatic gluconeogenesisand cellular glucose uptake are critical in the treatment andprevention of T2DM.

Biguanides, thiazolidinediones, and DPP-4 inhibitors have beenwidely prescribed for the treatment of T2DM.3 Biguanides areknown to activate AMP-activated protein kinase (AMPK) and showimprovement in the clinical symptoms of T2DM. As previouslyshown, AMPK is activated by a high cellular AMP-to-ATP ratio andfunctions as a master regulator of cellular energy homeostasis byregulating various target proteins via phosphorylation.4 In glucosemetabolism, AMPK suppresses hepatic gluconeogenesis andincreases glucose uptake in both skeletalmuscle and adipose tissue,thus contributing to improved blood glucose homeostasis.5 AMPKsuppresses hepatic glucose production through the phosphoryla-tion of two major proteins,6,7 cAMP responsive element-bindingprotein (CREB) regulated transcription coactivator-2 (CRTC2) andhistone deacetylase-5 (HDAC5). AMPK suppresses hepatic gluco-neogenesis through inhibitory phosphorylation of CRCT2 andHDAC5. These phosphorylation events induce the sequestrationof CRTC2 and HDAC5 in the cytoplasm. CRTC2 and HDAC5 arecoactivators of CREB and forkhead box O1 (FoxO1), two majortranscription factors that induce the transcription of phospho-enolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase

5238 W.-D. Seo et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5237–5242

(G6Pase), major genes in hepatic gluconeogenesis. Thus, AMPK-dependent phosphorylation of CRTC2 and HDAC5 leads to the sup-pression of gluconeogenesis.8 Alternatively, the uptake of cellularglucose is facilitated by glucose transporter-4 (GLUT4), the exocy-tosis to the plasma membrane and gene expression of which areregulated by the action of insulin. GLUT4 gene expression isregulated by the phosphorylation of HDAC5, which acts as itstranscriptional repressor. Transcriptional regulation of the GLUT4promoter includes the myocyte enhancer factor-2 (MEF2)-bindingdomain, which is inhibited by HDAC5. AMPK-dependent phospho-rylation of HDAC5 prevents its nuclear translocation, which in turnleads to the upregulation of GLUT4 gene expression.9,10

Synthetic prescribed drugs are efficient therapeutics for T2DM;however, their side effects can often be serious (e.g., biguanidescause lactic acidosis).11 Thus, using natural compounds with mod-erate activity could be a helpful alternative approach for the pre-vention of T2DM. The volume of publications on this subject hassuggested that flavonoids ameliorate the symptoms of T2DM. Forexample, the intake of flavan-3-ols rich cocoa for 18 weeksimproves the homeostatic model assessment index of insulin resis-tance, which can ameliorate diabetes in human trials.12

In this study, we investigate the effect of saponarin (apigenin-6-C-glucosyl-7-O-glucoside) (Fig. 1A), a flavone glucoside. Saponarinis found in diverse plants, including Tinospora cordifolia,13 which isused in an anti-diabetic drug,14 Aloe barbadensis,15 and barleyleaves.16 Recently, we found that the administration of barleysprout extract for 12 weeks lowered fasting glucose levels in the

A B

D HepG2p-AMPK

AMPK

p-A

A

Figure 1. Saponarin suppresses gluconeogenesis and increases glucose uptake by AMgluconeogenesis in HepG2 cells. Cells were stimulated with saponarin for 24 h, and glucEffects of saponarin on glucose uptake in TE671 cells. Intracellular concentration ofPhosphorylation activation of AMPK in cells stimulated with saponarin. ImmunoblottTreatments were performed for 4 h. The different letters denote significant differences aData are represented as the mean ± SEM. Con, vehicle control; CPT-cAMP at 100 lM; Minsulin at 100 nM; A769, A769662 at 100 lM; STO, STO-609 (CAMKK-b inhibitor) at 1 m

plasma and improved insulin sensitivity in mice fed a high-fatdiet.17 Barley sprout contained saponarin as a major flavonoid(1.1%, w/w),18 which could improve glucose metabolism andinduce insulin sensitivity. Thus, we examined the mechanism ofsaponarin on gluconeogenesis and glucose uptake using culturedhepatocytes and myocytes.

Saponarin improves gluconeogenesis and glucose uptake: Plasmaglucose homeostasis is regulated by two major mechanisms: glu-cose production in hepatocytes and glucose uptake in myocytes.The effects of saponarin on gluconeogenesis19,20 and glucoseuptake21,22 were investigated in HepG2 and TE671 cells, culturedhuman hepatocytes and myocytes,23 respectively (Fig. 1).

Glucose production in HepG2 cells stimulated with saponarin inglucose-free media was significantly decreased compared to that inthe controls. Glucose production was reduced by 56% in the pres-ence of 100 lM of saponarin (Fig. 1B), and this reduction wasgreater than the effect of metformin (1 mM). In myocytes, sapo-narin treatment for 24 h induced glucose uptake regardless of thepresence of insulin. For example, glucose uptake in TE671 cellsincreased by +72% at 100 lM of saponarin and glucose uptakewas further increased to +96% in the presence of insulin (Fig. 1C).These results suggest that saponarin has a synergistic effect withinsulin on glucose uptake.

Saponarin regulates AMPK through induction of intracellular cal-cium levels: Hepatic gluconeogenesis and muscular glucose uptakeare both regulated by AMPK activity. Thus, we investigated theeffects of saponarin on AMPK activation. Through immunoblotting

C

ConA76

9S50

S100

Con+STO

A769+

STO

S50+S

TO

S100+

STO0

200

400

600

800

a

b

ac c

a

cd

a aRel

ativ

e pr

otei

n (%

)

TE671MPK

MPK

cd c

PK activation. (A) Chemical structure of saponarin. (B) Effects of saponarin onose production in the cells was quantified by the enzymatic analysis of glucose. (C)2-NBDG was measured fluorimetrically after 24 h of saponarin stimulation. (D)ing analysis was carried out as described in the materials and methods section.mong the groups assessed by one-way ANOVA followed by the Tukey test (P <0.05).et, metformin at 1 mM; S50, saponarin at 50 lM; S100, saponarin at 100 lM, Ins,M.

W.-D. Seo et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5237–5242 5239

analysis, we observed that saponarin induced the phosphorylationof AMPK at Thr 172, which is known to lead to AMPK activation(Fig. 1D). Next, the mechanism of AMPK activation by saponarinwas investigated. First, using a cell-free kinase assay, we found thatsaponarin did not bind directly to AMPK to activate it, whereas A-769662, a synthetic agonist of AMPK, directly activated AMPKallosterically (EC50 = 576 nM, Fig. S1). Thus, saponarin is not anallosteric regulator of AMPK. Second, AMPK activity can be acti-vated by intracellular calcium levels through the activation of anupstream kinase, CAMKK-b.24 Our results assessed by a cellularcalcium assay25,26 showed that saponarin increased the calciumconcentration in both HepG2 and TE671 cells (Fig. 2A) and thatthe induction of AMPK phosphorylation by saponarin disappearedin the presence of STO-609, an inhibitor of CAMMK-b (Fig. 1D).These results suggest that saponarin activates AMPK phosphoryla-tion in a calcium-dependent manner.

Saponarin suppresses CREB and FoxO1 activity in cultured hepato-cytes:AMPK suppresses hepatic gluconeogenesis through inhibitoryphosphorylation of CRCT2 and HDAC5. These phosphorylationevents induce the sequestration of CRTC2 and HDAC5 in the cyto-plasm. CRTC2 and HDAC5 are coactivators of CREB and FoxO1, twomajor transcription factors that induce the transcription of PEPCKand G6Pase. Thus, AMPK-dependent phosphorylation of CRTC2and HDAC5 leads to the suppression of gluconeogenesis.8 In HepG2

Figure 2. Saponarin induces intracellular calcium concentration and phosphorylation of Clevels. (B) Phosphorylation and nuclear levels of CRTC2 and HDAC5 quantified by immugroups assessed by one-way ANOVA followed by the Tukey test (P <0.05). Data are repsaponarin at 50 lM; S100, saponarin at 100 lM.

cells stimulated with saponarin (100 lM) for 4 h, we observed sig-nificantly increased phosphorylation of both CRTC2 and HDAC5(229% and 187%, respectively) compared to the controls and thusreduced nuclear CRTC2 and HDAC5 (�51% and �50%, respectively)via immunoblotting analysis (Fig. 2B).27 The transactivation of CREBand FoxO1 assessed by luciferase asssay28 were suppressed in cellstreated with saponarin (�62% and�14% reduction, respectively, vs.controls, P <0.05, Fig. 3A). The effect of saponarin was greater onCREB activity. Finally, via a ChIP assay,29,30 we observed that sapo-narin stimulation reduced the binding of CRTC2 to both PEPCKand G6Pase promoters (-28% and -34% reduction, respectively, vs.controls, Fig. 3B); thus, mRNA expression31,32 of PEPCK and G6Pasewas downregulated by saponarin (Fig. 3C). These results suggestthat saponarin suppresses hepatic gluconeogenesis primarilythrough the activation of the AMPK–CREB signaling axis, whichdownregulates PEPCK and G6Pase gene expression.

Saponarin phosphorylates HDAC5 and induces GLUT4 expression incultured myocytes: GLUT4 expression is controlled by two tran-scription factors, GEF and MEF2A, both of which are stimulatedby AMPK. Specifically, hyperacetylation of the MEF2A promoteris facilitated by the AMPK-dependent phosphorylation of HDAC5.Saponarin-stimulated HDAC5 phosphorylation (+63% in the pres-ence of 100 lM of saponarin, Fig. 4A), thus reducing nuclearHDAC5 in TE671 cells. Accordingly, the mRNA expression of

RTC2 and HDAC5. (A) Saponarin stimulation immediately increases cellular calciumnoblotting analysis. The different letters denote significant differences among theresented as the mean ± SEM. Con, vehicle control; A769, A769662 at 100 lM; S50,

5240 W.-D. Seo et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5237–5242

GLUT431,32 was upregulated in cells stimulated with saponarin(Fig. 4B). These results suggest that saponarin induces glucoseuptake by activation of GLUT4 gene expression.

Saponarin is a flavone glucoside that is found in some plantleaves, including barley sprout, Aloe vera, Cucumis sativus, andmosses. Saponarin has been shown to be involved in several bio-logical activities, including hypoglycemic effects and the improve-ment of insulin sensitivity.14 However, its molecular mechanismsof action has been elusive. In this study, we suggest that saponarinstrongly activated AMPK phosphorylation via CAMKKb, which wasconfirmed by STO-609, a CAMKKb inhibitor. Saponarin did notinteract directly with the AMPK protein, thus it is not an allostericregulator. Activation of AMPK led to the phosphorylation oftranscription modulators CRTC2 and HDAC5, which preventedtheir transport into the nucleus. This prevented CREB- andFoxO1-dependent transcription of PEPCK and G6Pase in hepato-cytes and suppression of GLUT4 by HDAC5 in myocytes. The resultsof this study revealed that saponarin induced phosphorylation ofCRTC2 and HDAC5, which decreased the expression of gluconeoge-nesis genes and increased the expression of GLUT4 in myocytes.

Figure 3. Effect of saponarin on the transactivation of CREB and FoxO1 in HepG2 cells. (ACRTC2 to the promoter of PEPCK and G6Pase genes assessed by the ChIP assay. (C) mRNexpression levels were compared with those in the control group as a reference. The diffeANOVA followed by the Tukey test (P <0.05). Data are represented as the mean ± SEMsaponarin at 100 lM.

Activation of AMPK enhances insulin sensitivity, stimulatingglucose uptake in muscle and adipose tissues and inhibitingglucose production in the liver; thus, this is an important strategyin the control of type II diabetes.33 The biguanide metformin is awidely prescribed insulin sensitizer that activates AMPK in multi-ple tissues. Metformin lowers plasma glucose and lipid levelsthrough the reduction of hepatic gluconeogenesis and inductionof glucose uptake in muscle tissue.34 Our results strongly indicatedthat saponarin activates AMPK via the calcium–CAMKKb pathwayand phosphorylates AMPK to stimulate its downstream targets.

PEPCK and G6Pase are the key genes in hepatic gluconeogenesis,and their expression is positively correlated with the developmentof type II diabetes.35 CREB and FoxO1 regulate these two genes.AMPK phosphorylates CRTC2 and HDAC5, transcriptional coactiva-tors of CREB and FoxO1, respectively, in gluconeogenesis. Phospho-rylated CRCT2 and HDAC5 are blocked from nuclear translocationand thus cannot participate in CREB- and FoxO1-dependent tran-scription of PEPCK and G6Pase. Thus, activation of AMPK reduceshepatic gluconeogenesis and glucose output to circulation. Weconfirmed that saponarin suppressed gluconeogenesis in cultured

) Transactivation of CREB and FoxO1 measured by the luciferase assay. (B) Binding ofA expression of the PEPCK and G6Pase genes assessed by qPCR analysis. The mRNArent letters denote significant differences among the groups performed by one-way. Con, vehicle control; A769, A769662 at 100 lM; S50, saponarin at 50 lM; S100,

Figure 4. Effect of saponarin on the phosphorylation of HDAC5 and GLUT4 gene expression in TE671 cells. (A) Phosphorylation and nuclear levels of HDAC5 in TE671 cellsstimulated with saponarin. (B) mRNA expression of GLUT4 assessed by qPCR. The different letters denote significant differences among the groups performed by one-wayANOVA followed by the Tukey test (P <0.05). Data are represented as the mean ± SEM. Con, vehicle control; A769, A769662 at 100 lM; S50, saponarin at 50 lM; S100,saponarin at 100 lM.

W.-D. Seo et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5237–5242 5241

hepatocytes following AMPK phosphorylation of CRTC2 andHDAC5, which decreased its nuclear levels and thus the transactiva-tion of CREB and FoxO1. Saponarin showed greater suppression onCREB activity compared with that of FoxO1.

GLUT4 is a key gene in glucose uptake in skeletal muscle cells.Its gene expression is regulated by two transcription factors, GEFand MEF2A, the activity of which is primarily controlled by AMPK.AMPK phosphorylates GEF directly to activate GLUT4 transcriptionand phosphorylates HDAC5, which hyperacetylates the MEF2Apromoter and induces GLUT4 transcription. The results of thisstudy confirmed that saponarin induces HDAC5 phosphorylationand decreases nuclear HDAC5, thus upregulating GLUT4 mRNAexpression. In addition to AMPK-dependent upregulation ofGLUT4, CAMKKb also induces GLUT4 via activating phosphoryla-tion of HDAC5. Because studies have suggested that the activationof both AMPK and CAMKKb reinforces GLUT4 transcription, theinduction of intracellular calcium and activation of CAMKKb bysaponarin may additionally contribute to the induction of GLUT4gene expression. We found that saponarin significantly increasedGLUT4 gene expression levels within 24 h.

Although this study suggests a novel mechanism of saponarin inthe regulation of insulin sensitivity in vitro, there are somelimitations. Saponarin induced AMPK phosphorylation in a cal-cium-dependent manner, and not by direct interaction with AMPKproteins. Thus, the molecular target of saponarin in the regulationof intracellular calcium concentrations has not been elucidated inthis study. There are many different ways to regulate intracellularcalcium concentration, including activation of Gaq in the G-proteincoupled receptor pathways and activation of ligand-gated calciumchannels. The mechanism of calcium regulation by saponarinshould be investigated further. In addition, the effects of saponarinneed to be confirmed in animal and human studies. These remain-ing questions and experiments should be pursued in future studies.

In conclusion, saponarin has a hypoglycemic effect throughthe activation of the CAMMKb–AMPK–CREB signaling axis inhepatocytes. Activation of AMPK and the induction of GLUT4 gene

expression contribute to the increase in glucose uptake. Theseevents could collectively improve glucose homeostasis and insulinsensitivity. Future studies in animals and humans will aim toconfirm these observations.

Acknowledgements

This work was carried out with the support of the ‘CooperativeResearch Program for Agriculture Science & TechnologyDevelopment (Project No. PJ01052801 and PJ010090032015)’,Rural Development Administration (RDA), Republic of Korea.

Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.bmcl.2015.09.057.

References and notes

1. Schwarz, P. E. H.; Gallein, G.; Ebermann, D.; Muller, A.; Lindner, A.; Rothe, U.;Nebel, I. T.; Muller, G. Diabetes Res. Clin. Pract. 2013, 100, 11.

2. Kahn, B. B. Cell 1998, 92, 593.3. Inzucchi, S. E. JAMA 2002, 287, 360.4. Hardie, D. G. Int. J. Obes. 2008, 32, S7.5. Viollet, B.; Lantier, L.; Devin-Leclerc, J.; Hebrard, S.; Amouyal, C.; Mounier, R.;

Foretz, M.; Andreelli, F. Front. Biosci. 2009, 14, 3380.6. Koo, S. H.; Flechner, L.; Qi, L.; Zhang, X. M.; Screaton, R. A.; Jeffries, S.; Hedrick,

S.; Xu, W.; Boussouar, F.; Brindle, P.; Takemori, H.; Montminy, M. Nature 2005,437, 1109.

7. Jitrapakdee, S. Int. J. Biochem. Cell Biol. 2012, 44, 33.8. Mihaylova, M. M.; Shaw, R. J. Nat. Cell Biol. 2011, 13, 1016.9. Ojuka, E. O.; Jones, T. E.; Nolte, L. A.; Chen, M.; Wamhoff, B. R.; Sturek, M.;

Holloszy, J. O. Am. J. Physiol. Endocrinol. Metab. 2002, 282, E1008.10. Mcgee, S. L.; van Denderen, B. J. W.; Howlett, K. F.; Mollica, J.; Schertzer, J. D.;

Kemp, B. E.; Hargreaves, M. Diabetes 2008, 57, 860.11. Stang, M.; Wysowski, D. K.; Butler-Jones, D. Diabetes Care 1999, 22, 925.12. van Dam, R. M.; Naidoo, N.; Landberg, R. Curr. Opin. Lipidol. 2013, 24, 25.13. Grover, J. K.; Vats, V.; Rathi, S. S. J. Ethnopharmacol. 2000, 73, 461.14. Sengupta, S.; Mukherjee, A.; Goswami, R.; Basu, S. J. Enzyme Inhib. Med. Chem.

2009, 24, 684.

5242 W.-D. Seo et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5237–5242

15. Keyhanian, S.; Stahl-Biskup, E. Planta Med. 2007, 73, 599.16. Benedet, J. A.; Umeda, H.; Shibamoto, T. J. Agric. Food Chem. 2007, 55, 5499.17. Lee, J. H.; Lee, S. Y.; Kim, B.; Seo, W. D.; Jia, Y.; Wu, C.; Jun, H. J.; Lee, S. J. Food

Res. Int. 2015, 72, 174.18. Seo, K. H.; Park, M. J.; Ra, J. E.; Han, S. I.; Nam, M. H.; Kim, J. H.; Lee, J. H.; Seo, W.

D. Food Funct. 2014, 5, 3005.19. HepG2 cells were seeded in 6-well plates (5 � 105 cells/well) in DMEM with

10% FBS. The following day, the media was replaced with FBS-free media andthe cells were incubated with 8-(4-chlorophenylthio)-cAMP (CPT-cAMP,100 lM), metformin (1 mM), or saponarin (50 and 100 lM) for another 24 h.The media was then replaced with 1 mL of glucose production bufferconsisting of glucose-free DMEM (Gibco, BRL, Eggenstein, Germany) withoutphenol red supplemented with 20 mM sodium lactate and 2 mM sodiumpyruvate. After a 6-h incubation, the media was collected and the glucoseconcentration was measured using an enzymatic colorimetric assay (Roche,Basel, Switzerland). The glucose concentrations were normalized to the totalprotein content determined from the whole cell lysates.

20. Yoon, J. C.; Puigserver, P.; Chen, G. X.; Donovan, J.; Wu, Z. D.; Rhee, J.; Adelmant,G.; Stafford, J.; Kahn, C. R.; Granner, D. K.; Newgard, C. B.; Spiegelman, B. M.Nature 2001, 413, 131.

21. Zou, C. H.; Wang, Y. J.; Shen, Z. F. J. Biochem. Biophys. Methods 2005, 64,207.

22. A cellular glucose uptake assay was performed according to the manufacturer’sinstructions (Cayman, Ann Arbor, MI, USA). Briefly, TE671 cells were seeded in96-well black plates (5 � 104 cells/well) in DMEM with 10% FBS and were co-treated with a test substance—saponarin, 1 mM metformin, or 100 nMinsulin—and 2-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-2-deoxyglucose(2-NBDG) for 24 h. At the end of the treatment, the media was removed andassay buffer was added. The amount of 2-NBDG in the cells wasfluorimetrically measured at 485/535 nm (Victor 2, Perkin Elmer, Norwalk,CT, USA).

23. HepG2 and TE671 cells were obtained from the Korean Cell Line Bank (Seoul, K.a. g. i. D. H., Logan, UT, USA) supplemented with 10% heat-inactivated fetalbovine serum (FBS) and 1% penicillin/streptomycin at 37 �C in 5% CO2. The cellswere seeded into 6- or 24-well culture plates and subsequently treated withmetformin (1 mM) or A-769662 (100 lM) and saponarin (50 and 100 lM).

24. Racioppi, L.; Means, A. R. J. Biol. Chem. 2012, 287, 31658.25. Gee, K. R.; Brown, K. A.; Chen, W. N. U.; Bishop-Stewart, J.; Gray, D.; Johnson, I.

Cell Calcium 2000, 27, 97.26. Cells were seeded at a density of 30,000 cells/well in black 96-well plates. The

Fluo-4 NW dye (Invitrogen, Carlsbad, CA, USA) was dissolved in assay buffercontaining HBSS, 20 mM HEPES, and probenecid. Dye-loading buffer wasaliquoted onto the cultured cells, which were incubated at 37 �C for 30 min andthen at room temperature for 30 min. Saponarin was dissolved in assay bufferand added into the 96-well plate. Fluorescence (494/516 nm) was directlymeasured in 10-s intervals using a Victor 2 fluorometer (Perkin Elmer).A-23187 (10 lM) was used as a positive control.

27. Cells were lysed with a 1% protease inhibitor cocktail at 4 �C. The proteinconcentration was determined using the Bradford reagent using bovine serumalbumin as the standard. Proteins (40 lg) in Laemmli sample buffer wereheated for 5 min prior to SDS–PAGE (8–10%). The separated proteins weretransferred onto a nitrocellulose membrane (Schleicher & Schuell Bioscience,Dassel, Germany) at 100 V for 1 h. Nonspecific binding was blocked with 5%non-fat dry milk in TBS-T buffer for 1 h at room temperature. The membraneswere incubated with the primary antibody overnight at 4 �C. The followingday, the membranes were washed with TBS-T for 40 min and incubated withthe secondary antibody for 1 h at room temperature. The immunoreactiveprotein bands were visualized with an enhanced chemiluminescence system(Animal Genetics, Seoul, Korea), and the image was obtained using a ChemiDocXRS+ System (Bio-Rad, USA) and quantified using Gel-Pro Analyzer software(Media Cybernetics Inc., Rockville, MD, USA).

28. HEK293T cells were seeded in 24-well plates at a density of 1 � 105 cells/well.The following day, the cells were co-transfected with a reporter vector (FoxO1and CREB) that contained the firefly luciferase gene under the control of atranscriptional binding site and a renilla luciferase vector for normalizationusing Lipofectamine 2000 (Invitrogen, USA). After 24 h of transfection, the cellswere stimulated with saponarin for 24 h and lysed using the Luciferase LysisBuffer (Promega, Madison, WI, USA) and then the luciferase assay was carriedout. The luminescence from the cell lysate was measured using a dual-luciferase system kit (Promega, USA).

29. Nelson, J. D.; Denisenko, O.; Bomsztyk, K. Nat. Protoc. 2006, 1, 179.30. HepG2 cells were fixed with 4% formaldehyde after saponarin stimulation. The

cells were collected by centrifugation (2000�g for 5 min at 4 �C) and washedtwice with cold PBS. An IP buffer containing protease inhibitors was added andthe nuclear pellet was collected by centrifugation (12,000�g for 1 min at 4 �C).To shear the chromatin, the nuclear pellet was sonicated (EpiShear probesonicator, Active Motif, Carlsbad, CA USA; 5 rounds of 15-s pulses at 50% poweroutput) and immunoprecipitated using CRTC2 antibodies. DNA was isolatedusing the Chelex 100 slurry (Bio-Rad) and PCR was performed. FLAG antibodywas used as a control.

31. Jia, Y.; Kim, S.; Kim, J.; Kim, B.; Wu, C.; Lee, J. H.; Jun, H. J.; Kim, N.; Lee, D.; Lee,S. J. Mol. Nutr. Food Res. 2015, 59, 344.

32. Total RNA was extracted from livers and HepG2 cells using RNAiso Plus(Takara, Shiga, Japan). cDNA was synthesized from 2 lg of total RNA using M-MLV Reverse Transcriptase (Mbiotech, Seoul, Korea) and oligo(dT) primers.Gene expression levels were measured using the iQ5 Real-Time PCR DetectionSystem (Bio-Rad, Hercules, CA) and RealMasterMix SYBR ROX reagent (5 Prime,Hamburg, Germany). The relative levels of gene expression were calculatedusing iQ5 Optical System Software version 2 (Bio-Rad), with the expression ofeach target gene being normalized to that of cyclophilin. The primer sequencesused to amplify the genes of interest are shown in Table S1.

33. Hardie, D. G.; Ross, F. A.; Hawley, S. A. Nat. Rev. Mol. Cell Biol. 2012, 13, 251.34. Kirpichnikov, D.; McFarlane, S. I.; Sowers, J. R. Ann. Intern. Med. 2002, 137, 25.35. Barthel, A.; Schmoll, D. Am. J. Physiol. Endocrinol. Metab. 2003, 285, E685.