ORIGINAL ARTICLE

Total polyphenol of Anemarrhena asphodeloidesameliorates advanced glycation end products-inducedendothelial dysfunction by regulation of AMP-KinaseQianwen ZHAO,1 Yan SUN,1 Yu JI,1 Lihua XU,2 Kang LIU,1 Baolin LIU,1 and Fang HUANG1

1Department of Pharmacology of Chinese Materia Medic, China Pharmaceutical University, Nanjing, and 2Department of Pharmacology ofChinese Materia Medic, Suzhou University, Suzhou, China

Correspondence

Fang Huang, Department ofPharmacology of Chinese Materia Medica,China Pharmaceutical University,Longmian Road, Nanjing 211198, China.Tel: +86 139 5184 7264Fax: 00 8625 8332 8071Email: huangfang.cpu@gmail.com

Received 14 July 2013; revised 6November 2013; accepted 12 November2013.

doi: 10.1111/1753-0407.12111

Abstract

Objective: Anemarrhena asphodeloides Bunge is widely used in China for thetreatment of diabetes and the polyphenol components are responsible for itsanti-diabetic action. This study aimed to investigate the effect of total poly-phenol of Anemarrhena asphodeloides (TPAA) on endothelial dysfunction andto elucidate underlying mechanisms.Methods: We stimulated endothelial cells with advanced glycation end prod-ucts (AGEs) to establish the model of endothelial dysfunction in vitro andobserved the effect of TPAA (10, 30, or 100 μg/mL) on AMP-Kinase (AMPK)activation implicated in regulation of nitric oxide (NO) and endothelin-1(ET-1) production. Meanwhile, nuclear factor-κB (NF-κB) activation, intrac-ellular reactive oxygen species (ROS) production, mitochondrial membranepotential (ΔΨm) and eNOS expression were investigated by western blot,fluorescence microscopy and real time-quantitative PCR analysis, respectively.Results: Total polyphenol of Anemarrhena asphodeloides enhanced AMPKphosphorylation and promoted the basal NO production along with theinhibition of ET-1 secretion in endothelial cells. TPAA inhibited NF-κBactivation by attenuating p65 phosphorylation and suppressed ROS produc-tion, well demonstrating its action in inhibition of ROS-associated inflamma-tion in the endothelium. Meanwhile, TPAA protected mitochondrial functionand endothelial homeostasis against AGEs insult by restoring ΔΨm andmRNA expression of eNOS. AGEs stimulation inhibited AMPK activationand induced the loss of NO production together with increased secretion ofET-1, but these changes were reversed by TPAA in a concentration-dependentmanner. Compound C, an AMPK inhibitor, attenuated the effects of TPAAmentioned above, indicating the involvement of AMPK.Conclusions: Total polyphenol of Anemarrhena asphodeloides inhibitedAGEs-induced ROS-associated inflammation and ameliorated endothelialdysfunction through beneficial regulation of AMPK activation.

Keywords: advanced glycation end products, AMP-Kinase, endothelialdysfunction, total polyphenol of Anemarrhena asphodeloides.

Significant findings of the study: TPAA inhibited ROS-associated inflammation and protected mitochondrialfunction against AGEs insult in endothelial cells by beneficial regulation of AMPK activity.What this study adds: Providing insight into the molecular action of TPAA in amelioration of endothelialdysfunction associated with diabetes and cardiovascular risks.

bs_bs_banner

Journal of Diabetes •• (2014) ••–••

1© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

Introduction

Endothelial dysfunction is characterized by impairedvasodilation with sustained low-grade inflammatorystate, which has been demonstrated in type 2 diabetesand is responsible for the increased cardiovascular risksin diabetes.1,2 While high blood glucose level alone mayhave detrimental effects on endothelial function,3,4

enhanced formation of glycated species includingadvanced glycation end products (AGEs), which is afurther consequence of hyperglycemia, exacerbatesendothelial dysfunction. AGEs are heterogeneous andcomplex group modifications, which are formed as aresult of a non-enzymatic biochemical reaction initiatedas the Maillard reaction.5 It has been well demonstratedthat AGEs are implicated in diabetes-associated neph-ropathy, retinopathy, neuropathy and pancreatic betacell dysfunction.6–8 Emerging evidence also demonstratesthat AGEs formation contributes to many of the delete-rious vascular effects of diabetes.9–11 Binding to AGEreceptors, AGEs evoke inflammatory responses by acti-vating nuclear factor-κB (NF-κB) pathway and increas-ing production of reactive oxygen species (ROS), leadingto endothelial dysfunction.12

As an energy sensor, AMP-activated protein kinase(AMPK) regulates glucose and lipid homeostasis inmuscle and adipose tissue and it also regulates endothe-lial homeostasis in vascular disease. Metformin, an anti-diabetes agent, promotes nitric oxide (NO) productionby activating AMPK in the endothelium.13 Epigallocat-echin gallate, a natural AMPK activator from tea, is alsoshown to stimulate NO production from endothelial cellsthrough PI3K/Akt/eNOS pathways.14 AMPK activatorinhibits inflammation in macrophages15,16 and adipo-cytes,17 and recent studies also demonstrated that AMPKactivator reduced endothelin-1 (ET-1) expression andsuppressed inflammation in the endothelium.18,19 Thisevidence indicates the involvement of AMPK in theregulation of endothelial homeostasis.

Anemarrhena asphodeloides Bunge is widely used inChina for the treatment of diabetes. People have con-firmed its beneficial effects on diabetes, inflammationand platelet aggregation,20–22 but studies are limited intimosaponin and other constituents. Mangiferin, a poly-phenol substrate in Anemarrhena asphodeloides Bunge, isreported to show anti-diabetic and cardioprotectiveeffects.23,24 This evidence indicates that polyphenol sub-strates, which are abundant in Anemarrhena asphodelo-ides Bunge, may be the major components that areresponsible for the beneficial effects of Anemarrhenaasphodeloides Bunge on diabetes. In the present study,we extracted total polyphenol from Anemarrhenaeasphodeloides Bunge and observed its chemoprotection

against AGEs-induced damage in endothelial cells. Itsuggested that total polyphenol of Anemarrhena aspho-deloides (TPAA) ameliorated endothelial dysfunction byinhibiting ROS-associated inflammation in an AMPK-dependent manner.

Methods

Drugs and reagents

AGE-BSA was provided by Biovision (Milpitas, CA,USA). The stock solution of AGE-BSA was diluted to aconcentration of 100 μg/mL with using Dulbecco’s modi-fied Eagle medium (DMEM; Gibco, USA) and 10% fetalbovine serum (FBS; PAA, Cölbe, Germany). Other drugsand reagents used in the present study were obtained asfollows: sodium salicylate (Tianjin Kemiou ChemicalAgent Center, Tianjin, China); 5-Aminoimidazole-4-carboxamide1-β-D-ribofuranoside (AICAR; BeyotimeInstitute of Biotechnology, Shanghai, China); compoundC and EGTA (Sigma, St. Louis, MO, USA); ET-1 enzymelinked immunosorbent assay (ELISA) kits (R&DSystems, Minneapolis, MN, USA). Anti-p-NF-κB p65(Ser536), anti-NF- κB p65, anti-AMPKα, anti-phospho-AMPKα (T172) were obtained from Cell Signaling Tech-nology (Beverly, MA, USA). Glutathione (GSH),mitoquinone mesylate (Mito Q) and the ECL PlusWestern blotting detection reagent were purchased fromBeyotime Institute of Bio-technology (Shanghai, China).The following kits were obtained from Beyotime Instituteof Biotechnology (Haimeng, China): DAF-FM DA,JC-1, Reactive Oxygen Species Assay Kit (DCFH-DA).

Preparation of total polyphenol fromAnemarrhenae asphodeloides

The crude aqueous extract was obtained from 1.0 kg ofthe Anemarrhena asphodeloides Bunge (AA) rhizomewhich were collected from He Nan province of China.Then the crude aqueous extract was extracted and puri-fied with a LSA-10 macroporouse absorbent resin (Xi’anLan Xiao Technology, Xi’an, China), eluting with 0.5%NaOH (2 L). Subsequently, the elution was concen-trated, dried and dissolved in 80% alcohol (1 L). Finally,the alcohol extraction was spray-dried and the percent ofTPAA in the dried preparation was about 70% (deter-mined by UV, 540 nm; Ferrous sulfate reagent). Themajor component of TPAA was analyzed by an Agilent1200 series liquid chromatograph (Agilent Technologies,Palo Alto, CA, USA). Chromatographic separation wasachieved on YMC ODS-A column (4.6 mm × 250 mm,5 μm) maintained at 30°C. The mobile phase wasmethanol-water (40:60,v/v) containing 0.2% triethylam-

TPAA ameliorates endothelial dysfunction Q. ZHAO et al.

2 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

ine and 0.4% acetic acid at a flow rate 1.0 mL/min. Thedetection wavelength was 258 nm. Mangiferin was themain component of TPAA, compared with mangiferinstandard (Fig. 1).

Cell culture

EA.hy926 cells, a cell line of human umbilical veinendothelial cell, were purchased from the Type CultureCollection of the Chinese Academy of Sciences (Shang-hai, China). Cells were grown in Dulbecco’s modifiedeagle medium (DMEM) supplemented with 10% (v/v)FBS, streptomycin (100 U/mL), and penicillin (100 U/mL) under an atmosphere of 5% CO2 and 95% humidi-fied air at 37°C. The medium was renewed every 2 daysuntil cells reached confluence. Cells were starved inserum-free medium for 4 h before experiments.

Western blot analysis

The phosphorylation of AMPK and p65 in endothelialcells were detected by Western blot. Cells were incubatedfor 30 min with TPAA (10, 30, or 100 μg/mL), AICAR(500 μmol/L) or Salicylate (Sal; 5 mmol/L) respectively,and then stimulated with or without 100 μg/mL AGEsfor a further 30 min. Cells were washed with ice-coldphosphate-buffered saline (PBS) and then proteins wereextracted with lysis buffer for 30 min on ice. The super-natants were collected and centrifuged for 10 min at aspeed of 12 000 rpm (4°C). Proteins were quantified withthe Bicinchoninic Acid (BCA) Protein Assay kit (BioskyBiotechnology, Nanjing, China) according to the manu-facturer’s instructions. Protein samples were separatedon a 12% sodium dodecyl sulfate-polyacrylamide gelelectrophoresis (SDS-PAGE) gel and transferred to

(a)

(b)

Figure 1 HPLC spectrum of total polyphe-nol of Anemarrhena asphodeloides (TPAA)and mangiferin standard. (a) HPLC spectrumof TPAA; (b) HPLC spectrum of mangiferinstandard.

Q. ZHAO et al. TPAA ameliorates endothelial dysfunction

3© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

polyvinylidene fluoride (PVDF) membranes (0.45 μm;Millipore, Billerica, MA, USA). Membranes wereblocked for 2 h at room temperature with 10% nonfatmilk in TBST (5 mmol/L Tris-HCl, pH 7.6, 136 mmol/LNaCl, 0.05% Tween-20) and incubated with the corre-sponding antibodies (1:800 dilution in TBST) overnightat 4°C. The membranes were incubated with secondaryantibody (HRP-labeled anti-rabbit or anti-mouse IgG;1:2000 dilution in TBST) at 37°C for 2 h. Signals weredetected using ECL Western Detection Reagents andquantified densitometrically using Image-Pro Plus 6.0(Media Cybernetics, Rockville, MD, USA).

Detection of intracellular NO withfluorescence microscopy

Nitric oxide production in endothelial cells was detectedby using the NO-specific fluorescent dye 3-amino,4-aminomethyl-2′,7′-difluorescein diacetate (DAF-FMDA; Beyotime, Haimeng, China). Cells were incubatedwith TPAA (10, 30, or 100 μg/mL), TPAA plus com-pound C (25 μmol/L) or AICAR (500 μmol/L) for30 min. After washing with PBS, cells were loaded with5 mmol/L DAF-FM DA and then they were kept for30 min at 37°C in the dark. Cells were rinsed three timesbefore being fixed in 2% paraformaldehyde (v/v) at 4°Cfor 5 min and observed with an Olympus IX81 invertedmicroscope accompanying with attached charge-coupleddevice camera (Retiga Exi, Burnaby, British Columbia,Canada) using appropriate filters with a peak excitationwavelength of 495 nm and a peak emission wavelengthof 515 nm respectively. The fluorescence density wasquantized with Image-Pro Plus 6.0 (IPP 6.0) software.

ET-1 analysis

Cells were planted in 6-well plates when they reached theproper density, and treated with TPAA (10, 30, or100 μg/mL), TPAA plus compound C (25 μmol/L) orAICAR (500 μmol/L). The basal secretion of ET-1 wasdetermined after 8 h incubation. In the case of AGEschallenge, ET-1 production was determined 12 h afterAGEs stimulation. The supernatants were obtained andstored in a freezer (–70°C), and the content of ET-1 wasdetected by ELISA according to the manufacturers’instructions.

Detection of the intracellular ROS production

Reactive oxygen species in the cells were measuredby using the ROS-specific fluorescent dye 2′,7′-dichlorofluorescein diacetate (DCFH-DA; Beyotime,China). Cells were pretreated with TPAA (10, 30, or100 μg/mL), TPAA plus compound C (25 μmol/L) GSH

(1 mmol/L) or Mito Q (0.1 μmol/L) for 30 min and thenchallenged with AGEs (100 μg/mL) for another 30 min.Cells were loaded with 10 μmol/L DCFH-DA at 37°C for30 min, washed three times with PBS, and fixed in 2%paraformaldehyde (v/v) at 4°C for 5 min. The fixed cellswere examined under an inverted microscope (IX81;Olympus, Tokyo, Japan) with an attached charge-coupled device (CCD) camera (Retiga Exi, Burnaby,Canada), with a peak excitation wavelength of 488 nmand a peak emission wavelength of 525 nm. Images werequantified densitometrically using Image-Pro Plus 6.0(Media Cybernetics, Silver Spring, MD, USA).

Mitochondrial membrane potential (ΔΨm) measurement

The fluorescent dye JC-1 has the ability to detect mito-chondrial membrane potential (ΔΨm). Cells were starvedin serum-free medium for 4 h and then incubated withTPAA (10, 30, or 100 μg/mL), TPAA plus compound C(25 μmol/L), GSH (1 mmol/L) or Mito Q (0.1 μmol/L)for 30 min and then challenged with AGEs (100 μg/mL)for another 30 min. Cells washed with PBS and loadedwith 25 μg/mL JC-1 at 37°C and kept in the dark. 30 minlater, cells were washed three times with PBS before fixedin 2% paraformaldehyde (v/v) at 4°C for 5 min. Fixedcells can be detected under an inverted microscope(IX81) with an attached CCD camera using appropriatefilters and peak excitation of 585 and 514 nm and emis-sion wavelengths of 590 and 529 nm, respectively.Images were quantified densitometrically using Image-Pro Plus 6.0, with the ratio of red to green fluorescenceused to assess ΔΨm.

Real time-quantitative PCR analysis foreNOS expression

Cells were incubated for 30 min with TPAA (10, 30, or100 μg/mL), AICAR (500 μmol/L) or Salicylate (Sal;5 mmol/L) respectively, and then stimulated with orwithout AGEs (100 μg/mL) for a further 24 h. Aftertreatment as mentioned above, cells were washed threetimes with PBS. Total RNA was extracted by usingTRIzol (Roche, USA) according to the manufacturer’sinstructions. The extracted RNA was dissolved in diet-hypyrocarbonate (DEPC)-treated water, and the concen-trations were determined by optical density measurementat 260 nm on a spectrophotometer. Two micrograms ofRNA was used for reverse transcriptase, cDNA was syn-thesized using a PrimeScript RT reagent (Bio-Rad, Her-cules, CA, USA) according to the manufacturer’sinstructions. One microliter of cDNA mixed with DEPC-treated water, Ssofast EvaGreen Supermix (Bio-Rad),Forward Primer and Reverse Primer (Sangon Biotech,Shanghai, China) in 20 μL reaction mixtures was trans-

TPAA ameliorates endothelial dysfunction Q. ZHAO et al.

4 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

ferred to different PCRs. The primers used in RT-qPCRwere as follows: eNOS: (Forward Primer: 5′- CTCCAGCCCCGGTACTACTC-3′; Reverse Primer: 5′-TTAGCCACGTGGAGCAGACT-3′); GAPDH: (Forward Pri-mer: 5′-GGTGAAGGTCGGTGTGAACG-3′; ReversePrimer: 5′- CTCGCTCCTGGAAGATGGTG-3′).

Detection of cytotoxicity

When cells had reached confluence in 96-well plates, theywere incubated with TPAA (10, 30, or 100 μg/mL) for24 h in serum-free medium. Cells were subsequentlyincubated with 5 mg/mL 3-(4,5-dimethyl-2 thiazoyl)-2,5-diphenyl- 2H-tetrazolium bromide (MTT) 20 μL perwell for 4 h, then culture medium was got rid of andMTT formazan crystals were dissolved by dimethyl-sulfoxide (DMSO) (150 μL per well) for 10 min. Theabsorbance of dissolved formazan at an optical density(OD) of 490 nm was detected by using a MK3 MultiskanMirco-plate reader (Thermo, Boston, MA, USA).

Statistical analysis

All values are expressed as the mean ± standard devia-tion (SD). The significance between the mean values wasevaluated by two-tailed unpaired Student’s t-test oranalysis of variance (anova) followed by Student’s t-test.P < 0.05 was considered significant.

Results

TPAA enhanced AMPK phosphorylation inendothelial cells

We first investigated AMPK activation in response toTPAA. As shown in Fig. 2a, TPAA, at concentrationsranging from 10 to 100 μg/mL, effectively phosphory-lated AMPK. TPAA (100 μg/mL) enhanced AMPKphosphorylation and this action was not affected in thepresence of EGTA, a calcium chelator (Fig. 2b). Whencells were exposed to AGEs, the phosphorylation ofAMPK was reduced, but this alteration was effectivelyreversed by TPAA in a concentration-dependentmanner (Fig. 2c). AMPK activator AICAR also signifi-cantly increased AMPK phosphorylation under bothconditions.

TPAA promoted NO production

As a vasodilator substance, NO is produced fromendothelial cells. As shown in Fig. 3, TPAA induced NOproduction in endothelial cells, whereas this action wasblocked by co-treatment with AMPK inhibitor com-pound C. Similar to TPAA, AMPK activator AICAR

also promoted NO production in endothelial cells, indi-cating the involvement of AMPK.

TPAA inhibited basal ET-1 production

Furthermore, we investigated the effect of TPAA on thebasal secretion of ET-1 in endothelial cells and observedthat TPAA significantly inhibited ET-1 secretion at con-centrations of 30 and 100 μg/mL. Whereas, this effectwas abolished by co-culturing with AMPK inhibitorcompound C. The result was shown in Fig. 4.

TPAA inhibited AGEs-induced NF-κB activation

We then investigated the effect of TPAA on NF-κBphosphorylation. As shown in Fig. 5a, AGEs challengeinduced NF-κB activation evidenced by increasing p65phosphorylation, while this action was blocked by pre-treatment of TPAA and salicylate, respectively. Interest-ingly, we also observed that the inhibitory effect ofTPAA was reversed when co-treated with compound C(Fig. 5b).

TPAA suppressed ROS production and restoredmitochondrial membrane potential (ΔΨm) inendothelial cells

Advanced glycation end products might induce ROSproduction and the confusion of mitochondrial mem-brane potential (ΔΨm) in endothelial cells. As shown inFig. 6a, pretreatment of TPAA at concentrations of 30and 100 μg/mL significantly suppressed ROS produc-tion while the effect was reversed by compound C treat-ment. Fluorescent dye JC-1 accumulates in coupledmitochondria as red aggregates; thus, cells exhibit highred fluorescence of the J-aggregates. Following collapseof the ΔΨm, JC-1 remains in a monomeric form andfluoresces green. Thus, the ratio of red to green fluo-rescence is used to assess ΔΨm. AGEs insult resulted inΔΨm collapse evidenced by reduced red fluorescenceand increased green fluorescence. Treatment of TPAAreversed the alteration, but the effect was abolished bycompound C (Fig. 6b). GSH and Mito Q have similareffects as TPAA on ROS production and ΔΨmcollapse.

TPAA upregulated eNOS expression and promoted NOproduction against AGEs insult

eNOS mRNA expressions were determined byRT-qPCR. As shown in Fig. 7a, AGEs stimulationresulted in the downregulation of mRNA for eNOS,whereas this alternation was reversed by treatment withTPAA in a concentration-dependent manner. Salicylatealso effectively upregulated eNOS expressions. Mean-

Q. ZHAO et al. TPAA ameliorates endothelial dysfunction

5© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

while, we also observed that TPAA restored NO produc-tion when endothelial cells were exposed to AGEs, butthis effect was abolished by co-incubation with com-pound C. AICAR as well as salicylate also effectivelyincreased NO productions. Results are shown in Fig. 7b.

TPAA inhibited AGEs induced ET-1 selection inendothelial cells

We investigated the effect of TPAA on the secretion ofET-1 in the presence of AGEs. AGEs stimulationinduced ET-1 secretion from the basal level (45.9 ng/mL)to 103.0 ng/mL. TPAA significantly reduced the elevated

level of ET-1 to 79.4 ng/mL at the concentration of100 μg/mL, and the inhibition rate reached to 41.33%. Inaddition, salicylate (5 mmol/L) also inhibited ET-1 secre-tion by 62.52% (Fig. 8).

Effects of TPAA on cell viability

To rule out the potential cytotoxic effect of TPAA on itsinhibitory action in the present study, we determined cellviability in the presence of TPAA ranging from 10 to300 μg/mL. Although 300 μg/mL TPAA reduced cellviability, cell viability was not affected by TPAA at con-centrations ranging from 10 to 100 μg/mL, which were

(a) (b)

(c)

Figure 2 Total polyphenol of Anemarrhenae asphodeloides (TPAA) increased AMPK phosphorylation in endothelial cells.(a) Cells were treated with TPAA (10, 30, or 100 μg/mL) or AICAR (500 μmol/L) for 30 min. (b) Cells were treated with TPAA (100 μg/mL) for30 min after pretreated with EGTA (500 μmol/L) for 30 min. (c) Cells were pretreated with TPAA (10, 30, or 100 μg/mL) or AICAR (500 μmol/L)for 30 min and then stimulated with AGEs (100 μg/mL) for another 30 min. Results were shown a representative of experiments that wererepeated independently three times. *P < 0.05 compared with control; §P < 0.05 compared with control.

TPAA ameliorates endothelial dysfunction Q. ZHAO et al.

6 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

the working concentrations used in the present study.The result was shown in Fig. 9.

Discussion

It is well established that AGEs accumulation in theendothelium is tightly associated with endothelial dys-function, which is responsible for the development ofcardiovascular complications in diabetes. In the presentstudy, we demonstrated that TPAA upregulated AMPKactivity and further showed that this action contributedto the amelioration of endothelial dysfunction inducedby AGEs stimulation.

AMP-kinase regulates systemic energy balance andmetabolism, and accumulating evidence demonstrates itsprotective effects on the endothelium.25,26 Therefore, wefirst investigated the effect of TPAA on AMPK activity.TPAA enhanced basal AMPK phosphorylation andrestored AMPK phosphorylation when endothelial cellswere exposed to AGEs insult, well demonstrating itspositive regulation of AMPK activity. As mangiferin isthe major component in the polyphenol fraction, thisresult was consistent with the recently published studywhich showed that mangiferin promoted AMPK phos-phorylation in muscle cells.27 Two upstream kinases,LKB1 and CaMKKβ, activate AMPK by phosphorylat-ing a threonine residue (Thr172) on its catalyticα-subunit.28 Because CaMKKβ activates AMPK in aCa2+-dependent manner, TPAA remained positive regu-lation of AMPK phosphorylation in the presence ofcalcium chelator EGTA, indicating that its action wasCaMKKβ-independent. Whether TPAA, especially itsmajor component mangiferin, regulates AMPK activitythrough LKB1 or other ways remains to be clarified.Differently from NO, ET-1 is a potent vasoconstrictor inthe endothelium. The balance between endothelium-derived NO and ET-1 plays an important role in themaintenance of endothelial homeostasis. TPAA pro-moted NO production and inhibited ET-1 secretion inendothelial cells under basal conditions, suggesting itspotential regulation of endothelial function. As aninhibitor of AMPK, compound C abolished both actionsof TPAA, indicating the involvement of AMPK in theregulation of NO and ET-1 by TPAA. This action wasalso consistent with published reports which showed thatAMPK promoted NO production and inhibited ET-1secretion in the endothelium.13,18

Total polyphenol of Anemarrhena asphodeloides regu-lated AMPK activity and it was tempting to knowwhether the action contributed to ameliorating endothe-lial dysfunction. Endothelial dysfunction is characterizedby inflammatory responses evidenced by increasedexpression of pro-inflammatory cytokines.29 NF-κB is atranscription factor and the p65, a subunit of NF-κB, is

Figure 3 Total polyphenol of Anemarrhenae asphodeloides (TPAA)promoted NO production in endothelial cells. Cells were treated withTPAA (10, 30, or 100 μg/mL), TPAA plus compound C (25 μmol/L) orAICAR (500 μmol/L) for 30 min. Intracellular NO production wasviewed with fluorescence microscopy. The result was expressed asthe mean ± SD of three independent experiments. *P < 0.05 com-pared with control; †P < 0.05 compared with TPAA (100 μg/mL) inthe absence of compound C.

Figure 4 Total polyphenol of Anemarrhenae asphodeloides (TPAA)inhibited endothelin-1 (ET-1) secretion in endothelial cells. Cells wereincubated with TPAA (10, 30, or 100 μg/mL), or TPAA (100 μg/mL)plus compound C (25 μmol/L) for 8 h. The content of ET-1 wasdetected by ELISA. Data were expressed as means ± SD. n = 4;*P < 0.05 compared with control; †P < 0.05 compared with TPAA(100 μg/mL) in the absence of compound C.

Q. ZHAO et al. TPAA ameliorates endothelial dysfunction

7© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

a central control for the activation of numerous inflam-matory genes. By binding to AGE receptor, AGEsinduced NF-κB-dependent inflammation in the endothe-lium.30 The inhibitory effect of TPAA on AGEs-inducedp65 phosphorylation was attenuated by AMPK inhibitorcompound C, indicating that TPAA inhibited endothe-lial inflammation in an AMPK-dependent manner.AMPK inhibits inflammation in macrophages andadipocytes15–17 and our work further confirmed thatTPAA inhibited inflammation via regulation of AMPKin the endothelium. Although some studies demonstratethe anti-inflammatory activity of mangiferin in the brainand airway.31,32 As a main component in the polyphenolextracts, mangiferin might be responsible for the poten-tial effects of TPAA on endothelial inflammation.

Inflammation is tightly associated with oxidativestress. Emerging evidence demonstrates the impact ofAGEs in the endothelial cells.33,34 The induction of intra-cellular reactive oxygen species (ROS) is linked to theactivation of the NADPH oxidase system and the mito-chondrial electron-transport system.35 The action ofMito Q is to promote electron transport through themitochondrial respiratory chain. In the present research,Mito Q effectively reduced AGEs-induced ROS produc-tion, indicating that AGEs-induced ROS production was

mainly derived from mitochondrial dysfunction. As asecondary result, we observed that ΔΨm collapseoccurred when cells were exposed to AGEs challenge.Mitochondrial function is essential for the maintenanceof endothelial homeostasis. eNOS is a biomarker forendothelial integrity and its action is to regulate NOproduction. AGEs insult reduced gene expression foreNOS, leading to endothelial dysfunction, as evidencedby the loss of NO production and increased ET-1 secre-tion. In the present study, TPAA restored NO produc-tion and inhibited ET-1 secretion, but these beneficialeffects were abolished by AMPK inhibitor compound C.In this regard, it was reasonable to believe that TPAAameliorated endothelial dysfunction through an AMPK-dependent pathway, at least in part.

In the present experiment, we demonstrated thatTPAA inhibited ROS-associated inflammation and pro-tected mitochondrial function against AGEs insult inendothelial cells. TPAA beneficially regulated AMPKactivity and this action contributed to amelioratingendothelial dysfunction. Better understanding of TPAAaction in the regulation of AMPK activity in endothe-lium could be beneficial for its possible applications incontrolling the increased cardiovascular risks in diabeticpatients.

(a) (b)

Figure 5 Total polyphenol of Anemarrhenae asphodeloides (TPAA) inhibited NF-κB activation in endothelial cells. Cells were starved inserum-free medium for 4 h, and then incubated with TPAA (10, 30, or 100 μg/mL), Salicylate (Sal; 5 mmol/L), or TPAA (100 μg/mL) pluscompound C (25 μmol/L) for 30 min, following by stimulation with AGEs (100 μg/mL) for another 30 min. Phosphorylation of p65 wasdetermined by Western blot. Data were expressed as means ± SD of three independent experiments. *P < 0.05 compared with control;†P < 0.05 compared with TPAA (100 μg/mL) in the absence of compound C; §P < 0.05 compared with control.

TPAA ameliorates endothelial dysfunction Q. ZHAO et al.

8 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

(a)

(b)

Fig

ure

6To

talp

olyp

heno

lofA

nem

arrh

enae

asph

odel

oide

s(T

PAA

)inh

ibite

dre

activ

eox

ygen

spec

ies

(RO

S)p

rodu

ctio

nan

dre

vers

esth

eco

llaps

eof

mito

chon

dria

lmem

bran

epo

tent

ial(

ΔΨm

)in

AG

Es-

stim

ulat

eden

doth

elia

lcel

ls.

Cel

lsw

ere

incu

bate

dw

ithTP

AA

(10,

30,

or10

0μg

/mL)

,re

duce

dgl

utat

hion

e(G

SH

;1

mm

ol/L

),m

itoqu

inon

em

esyl

ate

(Mito

Q;

0.1

μmol

/L),

orTP

AA

(100

μg/m

L)pl

usco

mpo

und

C(2

5μm

ol/L

).A

GE

s(1

00μg

/mL)

was

adde

dfo

rano

ther

30m

in.R

esul

tsw

ere

expr

esse

das

mea

ns±

SD

ofth

ree

inde

pend

ente

xper

imen

ts.*

P<

0.05

com

pare

dw

ithco

ntro

l(C

on);

†P<

0.05

com

pare

dw

ithTP

AA

(100

μg/m

L)in

the

abse

nce

ofco

mpo

und

C;

§P<

0.05

com

pare

dw

ithco

ntro

l.

Q. ZHAO et al. TPAA ameliorates endothelial dysfunction

9© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

(a)

(b)

Fig

ure

7To

talp

olyp

heno

lof

Ane

mar

rhen

aeas

phod

eloi

des

(TPA

A)u

preg

ualte

deN

OS

expr

essi

onan

dre

stor

edN

Opr

oduc

tion

agai

nst

AG

Es

insu

lt.C

ells

wer

epr

e-tr

eate

dw

ithTP

AA

(10,

30,

or10

0μg

/mL)

,AIC

AR

(AIC

;500

μmol

/L),

Sal

icyl

ate

(Sal

;5m

mol

/L),

orTP

AA

(100

μg/m

L)pl

usco

mpo

und

C(2

5μm

ol/L

)for

30m

in.(

a)C

ells

wer

est

imul

ated

with

AG

Es

(100

μg/m

L)fo

r24

h,an

dth

eex

pres

sion

ofeN

OS

mR

NA

was

dete

cted

byR

ealT

ime-

Qua

ntita

tive

PC

R(R

T-qP

CR

).(b

)Cel

lsw

ere

stim

ualte

dw

ithA

GE

s(1

00μg

/mL)

for

30m

in,a

ndin

trac

ellu

lar

NO

prod

uctio

nw

asvi

ewed

with

fluor

esce

nce

mic

rosc

opy.

Dat

aw

ere

expr

esse

das

mea

ns±

SD

ofth

ree

inde

pend

ent

expe

rimen

ts.

*P<

0.05

com

pare

dw

ithco

ntro

l(C

on);

†P<

0.05

com

pare

dw

ithTP

AA

(100

μg/m

L)in

the

abse

nce

ofco

mpo

und

C;

§P<

0.05

com

pare

dw

ithco

ntro

l.

TPAA ameliorates endothelial dysfunction Q. ZHAO et al.

10 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

Acknowledgments

This work is funded by the National Natural ScienceFoundation of China (Grant No. 81072976), the

National Natural Science Foundation of China (GrantNo. 81173623) and the Fundamental Research Funds forthe Central Universities (Program No. JKZ2011018).The work in our department was supported by the 2011Program for Excellent Scientific and TechnologicalInnovation Team of Jiangsu Higher Education.

Disclosure

The authors declare no conflict of interests.

References

1. McVeigh GE, Brennan GM, Johnston GD et al.Impaired endothelium dependent and independent vaso-dilation in patients with type 2 (non-insulin-dependent)diabetes mellitus. Diabetologia. 1992; 35: 771–6.

2. Calles-Escandon J, Cipolla M. Diabetes and endothelialdysfunction: A clinical perspective. Endocr Rev. 2001; 22:36–52.

3. Paneni F, Beckman JA, Creager MA, Cosentino F. Dia-betes and vascular disease: Pathophysiology, clinical con-sequences, and medical therapy: Part I. Eur Heart J.2013; 34: 2436–43.

4. Hartge MM, Kintscher U, Unger T. Endothelial dysfunc-tion and its role in diabetic vascular disease. EndocrinolMetab Clin North Am. 2006; 35: 551–60.

5. Brownlee M, Cerami A, Vlassara H. Advanced glycosy-lation end products in tissue and the biochemical basis ofdiabetic complications. N Engl J Med. 1988; 318: 1315–21.

6. Win MT, Yamamoto Y, Munesue S et al. Regulation ofRAGE for attenuating progression of diabetic vascularcomplications. Exp Diabetes Res. 2012; 2012: 894605.

7. Yamaqishi S. Role of advanced glycation end products(AGEs) and receptor for AGEs (RAGE) in vasculardamage in diabetes. Exp Gerontol. 2011; 46: 217–24.

8. Lin N, Zhang H, Su Q. Advanced glycation end-productsinduce injury to pancreatic beta cells through oxidativestress. Diabetes Metab. 2012; 38: 250–7.

9. Negre-Salvayre A, Salvayre R, Augé N et al. Hypergly-cemia and glycation in diabetic complications. AntioxidRedox Signal. 2009; 11: 3071–109.

10. Guo ZJ, Niu HX, Hou FF et al. Advanced oxidationprotein products activate vascular endothelial cells via aRAGE-mediated signaling pathway. Antioxid RedoxSignal. 2008; 10: 1699–712.

11. Sena CM, Matafome P, Crisostomo J et al. Methylgly-oxal promotes oxidative stress and endothelial dysfunc-tion. Pharmacol Res. 2012; 65: 497–506.

12. Daroux M, Prévost G, Maillard-Lefebvre H et al.Advanced glycation end-products: Implications for dia-betic and non-diabetic nephropathies. Diabetes Metab.2010; 36: 1–10.

13. Davis BJ, Xie Z, Viollet B, Zou MH. Activation of theAMP-activated kinase by antidiabetes drug metforminstimulates nitric oxide synthesis in vivo by promoting theassociation of heat shock protein 90 and endothelial nitricoxide synthase. Diabetes. 2006; 55: 496–505.

Figure 8 Total polyphenol of Anemarrhenae asphodeloides (TPAA)inhibite AGEs-induced endothelin-1 (ET-1) secretion in endothelialcells. Cells were incubated with TPAA (10, 30, or 100 μg/mL), Salicy-late (Sal; 5 mmol/L) or TPAA (100 μg/mL) plus compound C(25 μmol/L) for 8 h in the presence of AGEs (100 μg/mL). The con-tents of ET-1 was detected by ELISA. Data were expressed asmeans ± SD. n = 4; *P < 0.05 compared with control (Con);†P < 0.05 compared with TPAA (100 μg/mL) in the absence of com-pound C; §P < 0.05 compared with control.

Figure 9 The effect of Total polyphenol of Anemarrhenae asphode-loides (TPAA) on cell viability in endothelial cells. Cells were withTPAA (10, 30, or 100 μg/mL) for 24 h and cell viability wasdetermined using the 3-(4,5-dimethyl-2 thiazoyl)-2,5-diphenyl-2H-tetrazolium bromide. Results were expressed as means ± SD of fourindependent experiments. *P < 0.05 compared with control.

Q. ZHAO et al. TPAA ameliorates endothelial dysfunction

11© 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd

14. Kim JA, Formoso G, Li Y et al. Epigallocatechin gallate,a green tea polyphenol, mediates NO-dependent vasodi-lation using signaling pathways in vascular endotheliumrequiring reactive oxygen species and Fyn. J Biol Chem.2007; 282: 13–736.

15. Yang Z, Kahn BB, Shi H, Xue BZ. Macrophage alpha1AMP-activated protein kinase (alpha1AMPK) antago-nizes fatty acid-induced inflammation through SIRT1.J Biol Chem. 2010; 285: 19–051.

16. Galic S, Fullerton MD, Schertzer JD et al. Hematopoi-etic AMPK β1 reduces mouse adipose tissue macrophageinflammation and insulin resistance in obesity. J ClinInvest. 2011; 121: 4903–15.

17. Lihn AS, Pedersen SB, Lund S, Richelsen B. The anti-diabetic AMPK activator AICAR reduces IL-6 and IL-8in human adipose tissue and skeletal muscle cells. MolCell Endocrinol. 2008; 292: 36–41.

18. Reiter CE, Kim JA, Quon MJ. Green tea polyphenolepigallocatechin gallate reduces endothelin-1 expressionand secretion in vascular endothelial cells: Roles forAMP-activated protein kinase, Akt, and FOXO1. Endo-crinology. 2010; 151: 103–14.

19. Schuhmacher S, Foretz M, Knorr M et al. α1AMP-activated protein kinase preserves endothelial functionduring chronic angiotensin II treatment by limiting Nox2upregulation. Arterioscler Thromb Vasc Biol. 2011; 31:560–6.

20. Xie W, Zhao Y, Zhang Y. Traditional Chinese medicinesin treatment of patients with type 2 diabetes mellitus. EvidBased Complement Alternat Med. 2011; 2011: 726723.

21. Xu LL, Wang L, Wang YQ. Effects of radix salviae mil-tiorrhiae, radix aconiti lateralis preparata and rhizomaanemarrhena on nitric oxide systems of endotoxemiamice. Zhongguo Zhong Xi Yi Jie He Za Zhi. 2006; 26:102–7.

22. Lu WQ, Qiu Y, Li TJ, Tao X, Sun LN, Chen WS. Anti-platelet and antithrombotic activities of timosaponinB-II, an extract of Anemarrhena asphodeloides. Clin ExpPharmacol Physiol. 2011; 38: 430–4.

23. Miura T, Ichiki H, Hashimoto I et al. Antidiabetic activ-ity of a xanthone compound, mangiferin. Phytomedicine.2001; 8: 85–7.

24. Prabhu S, Jainu M, Sabitha KE, Shyamala Devi CS.Effect of mangiferin on mitochondrial energy productionin experimentally induced myocardial infarcted rats.Vascul Pharmacol. 2006; 44: 519–25.

25. Towler MC, Hardie DG. AMP-activated protein kinasein metabolic control and insulin signaling. Circ Res. 2007;100: 328–41.

26. Zou MH, Wu Y. AMP-activated protein kinase activa-tion as a strategy for protecting vascular endothelial func-tion. Clin Exp Pharmacol Physiol. 2008; 35: 535–45.

27. Wang F, Yan J, Niu Y et al. Mangiferin and its aglycone,norathyriol, improve glucose metabolism by activation ofAMP-activated protein kinase. Pharm Biol. 2014; 52:68–73.

28. Carling D, Sanders MJ, Woods A. The regulation ofAMP-activated protein kinase by upstream kinases. Int JObes. 2008; 32: S55–9.

29. Libby P. Inflammation in atherosclerosis. Nature. 2002;420: 868–74.

30. Yan SD, Schmidt AM, Anderson GM et al. Enhancedcellular oxidant stress by the interaction of advanced gly-cation end products with their receptors/binding proteins.J Biol Chem. 1994; 269: 9889–97.

31. Márquez L, García-Bueno B, Madrigal JL, Leza JC. Man-giferin decreases inflammation and oxidative damage inrat brain after stress. Eur J Nutr. 2012; 51: 729–39.

32. Rivera DG, Hernández I, Merino N et al. Mangiferaindica L. extract (Vimang) and mangiferin reduce theairway inflammation and Th2 cytokines in murine modelof allergic asthma. J Pharm Pharmacol. 2011; 63: 1336–45.

33. Ahmed N. Advanced glycation endproducts-role inpathology of diabetic complications. Diabetes Res ClinPract. 2005; 67: 3–21.

34. Zhang FL, Gao HQ, Shen L. Inhibitory effect of GSPEon RAGE expression induced by advanced glycation endproducts in endothelial cells. J Cardiovasc Pharmacol.2007; 50: 434–40.

35. Basta G, Lazzerini G, Del Turco S et al. At least 2 distinctpathways generating reactive oxygen species mediate vas-cular cell adhesion molecule-1 induction by advanced gly-cation end products. Arterioscler Thromb Vasc Biol. 2005;25: 1401–7.

TPAA ameliorates endothelial dysfunction Q. ZHAO et al.

12 © 2013 Ruijin Hospital, Shanghai Jiaotong University School of Medicine and Wiley Publishing Asia Pty Ltd