Food and Chemical Toxicology 57 (2013) 99–105

Contents lists available at SciVerse ScienceDirect

Food and Chemical Toxicology

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

Effect of quercetin-rich onion peel extracts on arterial thrombosis in rats

0278-6915/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fct.2013.03.008

Abbreviations: OPE, onion peel extract; SD rat, Sprague-Dawley rat; PT,prothrombin time ; aPPT, activated partial thromboplastin time; PRP, platelet richplasma; PPP, platelet poor plasma; HUVECs, human umbilical vein endothelial cells;ERK, extracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; MAPK,mitogen-activated protein kinase.⇑ Corresponding author. Tel.: +82 2 940 2857; fax: +82 2 940 2850.

E-mail address: mjshin@korea.ac.kr (M.-J. Shin).

Seung-Min Lee a, Jiyoung Moon b, Ji Hyung Chung c, Yong-Jun Cha d, Min-Jeong Shin b,⇑a Department of Food and Nutrition,College of Human Ecology, Yonsei University, Seoul 120-749, Republic of Koreab Department of Food and Nutrition, Korea University, Seoul 136-703, Republic of Koreac Cardiology Division, Yonsei Cardiovascular Hospital and Cardiovascular Research Institute, Yonsei University College of Medicine, Seoul 120-752, Republic of Koread Department of Food and Nutrition, Changwon National University, Changwon 641-773, Republic of Korea

a r t i c l e i n f o

Article history:Received 19 October 2012Accepted 10 March 2013Available online 21 March 2013

Keywords:Onion peel extractQuercetinArterial thrombosisBlood coagulationaPTT

a b s t r a c t

The aim of this study was to examine whether oral supplementation of quercetin-rich onion peel extract(OPE) influences blood coagulation and arterial thrombosis in Sprague-Dawley (SD) rats. 24 male rats,5 weeks old, were divided into three groups with different diets (C: control, 2 mg OPE: chow diet with2 mg OPE supplementation, 10 mg OPE: chow diet with 10 mg OPE supplementation) for 6 weeks. Bloodcoagulation parameters including prothrombin time (PT), activated partial thromboplastin time (aPTT)and platelet aggregation were examined. The OPE did not affect blood cholesterol levels but significantlydecreased blood triglyceride and glucose levels. PT, aPTT and platelet aggregation were not significantlydifferent among all tested groups. However, in vivo arterial thrombosis was significantly delayed ingroups that were fed 2 mg and 10 mg OPE diets compared to the control group. In addition, the OPEgreatly diminished thrombin-induced expression of tissue factor in human umbilical vein endothelialcells (HUVECs), a coagulation initiator. In addition, extracellular signal-regulated kinase (ERK) and c-Jun N-terminal kinase (JNK) signaling pathways activated by thrombin treatment were prevented bythe OPE pre-treatment. These results indicate that OPE may have anti-thrombotic effects throughrestricting the induced expression of tissue factor via down-regulating mitogen-activated protein kinase(MAPK) activation upon coagulation stimulus, leading to the prolongation of time for arterial thrombosis.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Regulation of homeostasis is important to maintain the integ-rity of blood vessel system (Furie and Furie, 2008). To achieve this,dynamic balancing between blood coagulation and anticoagulationoccurs to ensure the blood flows in a well controlled manner (Furieand Furie, 2008). Thrombosis results from extravasation of bloodsignals for coagulation and recruitment of circulating platelets tothe injured sites as part of normal process during recovery from in-jury in vessel walls. When procoagulants come into contact withthe circulating leukocytes, the aggregation of platelets is initiated.Factor VIIa binds to tissue factor (TF) that is present on fibroblastsand leukocytes and activates Factors IX and X. Prothrombins arecleaved by these factors to form thrombin. Then thrombin cleavesfibrinogen to fibrin to form a fibrin clot along with the stabilizationof the platelet plug. Thrombin activates platelets at the site of

thrombus formation and drive the growth of thrombus (Chesebroet al., 1995). Activation of platelet activation plays an importantrole in the formation of arterial thrombosis. Fibrin and aggregatedplatelets are main constituents of intra-arterial thrombi (Heems-kerk et al., 2002). Therefore inhibition of thrombin-induced plate-let activation is considered to be important for the treatment ofmyocardial infarction (Braunwald, 2003). A pathological status ofocclusive thrombosis such as myocardial infarct can also occurwhen an excess of blood clot forms due to hyperactivation of bloodcoagulating factors or platelet hyperactivity (Falk and Fernandez-Ortiz, 1995; Thompson et al., 1995). An excessively large thrombuscan obstruct blood flow by adhering to the blood vessel andincreasing the viscosity of blood, which ultimately will hinderthe transfer of oxygen or nutrients to the tissues (Ruggeri, 2002;Wu and Thiagarajan, 1996). If an atherosclerotic plaque ruptures,platelets are activated by the exposed collagen in the extracellularmatrix to the blood, which can lead to the blockage of blood flow inan artery. In this regard, a proper regulation of thrombus formationis considered to be essential to maintain normal physiological pro-cesses (Briggs et al., 2001).

Onion (Allium cepa L.) is a biennial plant botanically belongingto the Liliaceae (Griffiths et al., 2002). It contains a rich source ofphytochemicals such as dietary flavonoids including quercetin

100 S.-M. Lee et al. / Food and Chemical Toxicology 57 (2013) 99–105

and kaempferol (Miean and Mohamed, 2001) and organosulfurcompounds like allyl propyl disulfide and diallyl disulfide (Block,1985). According to earlier in vitro and ex vivo studies, these indi-vidual constituents along with onion have been implicated inanti-thrombotic effects by inhibiting platelet aggregation (Aliet al., 1999; Briggs et al., 2001; Hubbard et al., 2006; Phillips andPoyser, 1978; Srivastava, 1984; Yamada et al., 2004). While thefavorable effects of onion on the thrombotic state in vitro havebeen accumulated, it is still unclear whether the short term effectcan last longer in animals and the long term ingestion of onion canimprove thrombotic activities in these settings. To our knowledge,there have been two in vivo studies to test the effects of onion onin vivo thrombus formation, performed using rats and dogs (Briggset al., 2001; Yamada et al., 2004). However, the authors only testedshort term effects of onion within a range of hours and adaptedmethods of irradiation or mechanical damage to induce thrombusin arteries (Briggs et al., 2001; Yamada et al., 2004).

In the present study, we hypothesized that long term intake ofquercetin-rich OPE could protect against hyperactivation of throm-bosis by modulating blood coagulation. To prove this, we tested theeffects of the OPE on ex vivo blood coagulation and in vivo throm-bosis after 6 weeks of supplementation.

2. Materials and methods

2.1. Preparation of onion peel extracts

Onion peel extract (OPE) was prepared with onion peels purchased from Non-ghyup (Changnyeong, Korea). After washing onion peels three times in tap water,OPE was extracted with 60% aqueous ethanol solution for 3 h at 50 �C in an extrac-tor (1 kL, Hansung F&C Co., Ltd., Korea) and then filtered by filter press (HankookIndustry Co. Ltd., Korea). The filtrates were concentrated to 2.4� Brix as percent sol-uble solid which was measured using a refractometer (Atago Co. Ltd., Tokyo, Japan)in a vacuum concentrator (1 kL, Hansung F&C Co., Ltd., Korea). The concentrateswere processed to a powder with a freeze dryer (SFDTS-200 kg Samwon IndustryCo., Ltd., Korea), then passed through a #40 mesh to be a fine powder, which finallycontains 250 mg quercetin/g of OPEs.

2.2. Animals and diets

Five weeks old male Sprague-Dawley (SD) rats were purchased from Koatech(Pyungtek, Korea). The experiments consisted of two separate sets. One (n = 24)was used for analysis of prothrombin time per activated partial thromboplastintime (PT/aPTT), platelet aggregation and the other (n = 24) was used to analyze arte-rial thrombus formation model. After 1 week adaptation period, the animals wererandomly assigned to one of the experimental groups which received regular chowdiet (Harlan, USA) and grown for 6 weeks. Each experimental group consisted of acontrol group and two OPE groups. Two OPE groups were orally galvaged with OPEsolution containing 2 mg (equivalent to 0.5 mg quercetin) and 10 mg OPE (equiva-lent to 2.5 mg quercetin), respectively. The rats were maintained in a pathogen-freeenvironment where temperature (18–24 �C) and humidity (50–60%) are controlled.All the experimental procedures were approved by the Committee on AnimalExperimentation and Ethics of Yonsei University College of Medicine.

2.3. Animal blood collection

After 7 weeks including the adaption period, the rats were sacrificed after star-vation for 12 h. After they were anesthetized with zoletile (Virbac, France) 30 mg/kgmixed with rompun (Bayerkorea, Korea) 10 mg/kg, blood was collected from theabdominal inferior vena cava and transferred in an EDTA containing vacutainer tubefor biochemical measurements. Also some aliquot was transferred into Na-citrate(anti-coagulating agent) containing polystyrene tubes for PT, aPTT and plateletaggregation assay. Whole blood samples in Na-citrate containing tube were centri-fuged at 220g for 15 min at 4 �C to obtain platelet rich plasma (PRP) and then at2010g for 5 min at 4 �C to obtain platelet poor plasma (PPP).

2.4. Measurements of PT, aPTT and platelet aggregation analysis

To assess blood coagulation, prothrombin time (PT) and activated partialthromboplastin time (aPTT) levels were obtained with using ACL-100 (Instrumen-tation Laboratory, Spain). For platelet aggregation analysis, the whole blood sam-ples in Na-citrate containing tube were centrifuged and separated into a PRP.Excess of PRP was also further centrifuged at 13,572g for 3 min to separate a PPP.In aggregation experiments adenosine diphosphate (10 ll) (ADP, Chrono-log,

USA) was added to the PRP (490 ll) in platelet aggregometer cuvette and changedin light transmission were recorded in a platelet aggregometer (Chrono-log, USA)set with PPP as 100% and PRP as the baseline.

2.5. In vivo measurement of arterial thrombus formation

At the end of the experiments, the anesthetized rats were placed on a servo-controlled, heated operating table, which temperature was maintained at 37 �C. Apart of the right carotid artery was exposed and isolated from the vagus nerveand neighboring tissues. Blood flow rate in the artery was monitored using an aorticblood flow rate probe (Transonic Flowprobe, Transonic systems Inc., USA). 50% fer-ric chloride solution in distilled water was used as arterial thrombus inducer. A2 mm2 piece of Whatman filter paper was saturated with 70% of ferric chloridesolution and was put around the carotid artery for 10 min. The flowmeter probeplaced near the Whatman filter paper measured the flow for up to 60 min anddetermined the time required for thrombotic arterial occlusion to occur.

2.6. Cell viability assay

Human umbilical vein endothelial cells (HUVEC) (1 � 106 cells) were seeded oneach well of 6-well culture plate. After 24 h, the cells were serum-starved overnightprior to the addition of OPE. Next day, they were treated with varying amounts ofOPE (0, 1, 5, 10, 50, and 100 lM) for 6 h. Cell viability was measured by adding1 mg/mL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)to each well and incubating at 37 �C for 1 h. After incubation, absorbance was mea-sured at a wavelength of 570 nm with Infinite� M200 NanoQuant microplate reader(Tecan Trading AG, Switzerland). This assay was repeated three times.

2.7. RNA extraction and semi-quantitative RT-PCR

To analyze gene expression, the HUVEC cells (1 � 106 cells) were seeded oneach well of 6 well culture plates. After 24 h, the cells were serum-starved over-night prior to the addition of the OPE. The starved HUVEC cells were cultured in ser-um-free medium with various concentrations (0, 50, and 100 lM) of OPE for 1 h andthen stimulated with 3 U/ml thrombin (Sigma Aldrich, St. Louis, MO) for 5 h at37 �C. At the end of the treatment time, medium in all wells was suctioned carefullyand each well was added with lysis buffer 400 lL. Total RNA was extracted fromHUVEC cells using RibospinTM Kit (GeneAll, Korea) according to the manufacturer’sprotocol. cDNA was synthesized from 1 lg of RNA using oligo-dT and Superscrip-tTM II reverse transcriptase (Invitrogen, USA). 1 lg of cDNA was used for PCR.PCR primers for Tissue Factor: (forward primer 50-CGCCAACTGGTAGACATGG-30

and reverse primer 50-GACTTGATTGACGGGTTTGG-30) and GAPDH (forward primer50-TCCACCACCCTGTTGCTGTA-30 and reverse primer 50-ACCACAGTCCATGCCATCAC-30). The PCR performed by Step One Plus (Applied Biosystems, USA) and conditionswere as followed: 15 min at 95 �C, followed by 40 cycles of 94 �C for 30 s, 58 �C for20 s and 72 �C for 30 s. GAPDH was used as the control in the comparative (CT)method.

2.8. Western blot analysis

The HUVEC cells (1 � 106 cells) were seeded on each well of 6 well cultureplates. After 24 h, the cells were serum-starved overnight prior to the addition ofOPE. The starved HUVEC cells were cultured in serum-free medium with variousconcentrations (0, 50, and 100 lM) of OPE for 1 h and then stimulated with 3 U/ml thrombin (Sigma–Aldrich, St. Louis, MO) for 5 h at 37 �C. At the end of the treat-ment time, medium in all wells was suctioned carefully and each well was addedwith radioimmunoprecipitation assay (RIPA) buffer (40 mM HEPES pH 7.5,120 mM NaCl, 1 mM EDTA, 1% Triton X-100) containing a protease inhibitor cocktail(Roche Diagnostics, Mannheim, Germany). Protein concentrations were determinedby BCA method (Sigma–Aldrich, St. Louis, MO). Equal amounts of protein lysateswere mixed with 5X loading buffer (1 M Tris-HCl, pH 6.8, a trace amount of bromo-phenol blue, 50% glycerol, 10% SDS, and DW) and lysis buffer and denatured at 95 �Cfor 5 min. Samples were loaded onto 10% sodium dodecyl sulfate (SDS)–polyacryl-amide gel (PAGE). After electrophoresis, proteins were electrophoretically trans-ferred from the gel onto PVDF membrane in buffer (2.5 mM Tris, 19.2 mM glycinepH 8.3) at 0.3 mA/cm2 for 1 h 30 min at RT. Residual binding sites on the membranewas blocked by incubating the membrane in TBS (pH7.6) containing 0.1% Tween 20and 5% nonfat dry milk for 1 h at RT. The blots were washed in TBS containing 0.1%Tween 20 and then incubated with appropriate antibody overnight at 4 �C. Afterwashing, the membrane was incubated with anti-rabbit or mouse IgGAb conjugatedwith HRP, and bands were visualized using enhanced chemiluminescence (ECL,Young In Frontier, Seoul, Korea) and quantified by densitometery using an Alpha-view� software (Alpha Innotech, USA). Western blot analysis was performed usingspecific antibodies for Tissue factor (American Diagnostica, Stamford, CT), JNK,phospho-JNK, ERK1/2, phospho-ERK1/2, and b-actin (Santa Cruz Biotechnology,USA).

S.-M. Lee et al. / Food and Chemical Toxicology 57 (2013) 99–105 101

2.9. Statistical analysis

Statistical analysis used SPSS-PC + (Statistical Package for the Social Science,SPSS Inc., Chicago, IL, USA). The values were presented as means ± S.E. Statisticalanalysis between the experimental groups was assessed by Student’s t-test orone-way analysis of variance (ANOVA). Null hypotheses of no difference were re-jected if p-values were less than 0.05.

Fig. 1. Effects of OPE supplementation on serum total cholesterol (A), LDL-cholesterol (B), HDL-cholesterol (C), triglycerides (D), and glucose (E) levels in ratsfed with a normal diet for 6 weeks. Tested by analysis of variance (ANOVA) withDuncan’s multiple range test. Data were expressed as the mean ± S.E. p < 0.05. CTL(n = 8), 2 mg OPE (n = 7), 10 mg OPE (n = 7); failure to obtain one blood sample in2 mg OPE group; one of 10 mg OPE group has died. Values with the samesuperscript letter within the column are not significantly different.

3. Results

Table 1 showed body weight, food intakes and baseline bio-chemical measurements of the animals. No differences in theseparameters were observed among the groups.

3.1. The effects of the OPE on blood lipid parameters and glucose levels

The animals fed on a regular diet with oral supplementation ofthe OPE for 6 weeks were examined for the blood lipid profilesand glucose levels. The blood levels of total cholesterol, HDL-choles-terol, and LDL-cholesterol in the OPE-fed animals were not differentfrom those of the control rats without the OPE (Fig. 1A–C). However,rats which had 10 mg OPE supplementation daily showed signifi-cantly reduced blood triglyceride levels when compared to thecontrol (Fig. 1D). In addition, blood glucose levels decreased signif-icantly in the group fed with 10 mg OPE supplementation daily,which was not shown in control rats (Fig. 1E). These results indicatethat a daily dose of 10 mg of the OPE could lower blood triglycerideand glucose levels even in a normal diet whereas blood cholesterollevels were not likely affected by the OPE.

3.2. The effects of the OPE on PT, aPTT and platelet functions

To identify the effects of the OPE on platelet aggregation, we as-sessed prothrombin time (PT) and activated partial thromboplastintime (aPTT) and performed platelet aggregation assay in rats afterproviding a diet supplemented either with 2 mg or 10 mg of theOPEs for 6 weeks. PT and aPTT values and platelet aggregation fromthe OPE-fed groups were similar with those of control rats (Fig. 2Aand B). The percentage of aggregated platelet obtained in a plateletaggregation assay in OPE-provided rats was similar to those of con-trol rats (Fig. 2C). Therefore, OPE does not appear to affect PT andaPTT nor platelet aggregation.

3.3. The effects of the OPE on in vivo arterial thrombosis

We further investigated the effects of the OPE on thromboticstatus in an oxidation-induced arterial thrombosis in vivo. In thistest in vivo thrombosis was induced by ferric chloride in artery.The time required to produce arterial thrombus was measured in

Table 1Effect of OPE supplementation on final body weight, FER and biochemical p

Groups CTL (n = 8) 2 mg O

Final body weight (g) 354.5 ± 9.7 340.4 ±Food intake (g/day) 24.9 ± 1.8 24.7 ±FERa 0.18 ± 0.01 0.16 ±BUN (mg/dl)b 26.8 ± 0.5 27.7 ±Cr (mg/dl)c 0.51 ± 0.02 0.48 ±GOT (U/l)d 81.0 ± 4.1 96.4 ±GPT (U/l)e 43.4 ± 3.4 50.0 ±

Values are represented as Mean ± S.D.a FER: Food efficiency ratio.b BUN: Blood urea nitrogen.c Cr: Creatinine.d GOT: Glutamic oxaloacetic transferase.e GPT: Glutamic pyruvic transferase.f Tested by analysis of variance (ANOVA) with Duncan’s multiple range t

significantly different (p > 0.05).

animals. A significant increase in time for arterial thrombosiswas shown in both OPE-fed animal groups with either 2 mg or10 mg daily as compared to the control (Fig. 3). These findings sug-gest that the OPE intake prolonged the time to initiate arterialthrombosis and may delay thrombus formation in the artery.

3.4. The down-regulation of tissue factor expression by the OPE

Based on the above observations of in vivo thrombotic effects ofthe OPE and no ex vivo coagulation effect of the OPE, we hypothe-sized that the OPE could influence in vivo thrombosis by

arameters.

PE (n = 8) 10 mg OPE (n = 8) p-Valuef

16.7 350.7 ± 14.6 NS1.7 25.1 ± 2.2 NS0.02 0.18 ± 0.02 NS0.6 28.3 ± 1.2 NS0.02 0.49 ± 0.02 NS5.3 91.7 ± 6.3 NS3.5 49.7 ± 6.2 NS

est. Values with the same superscript letter within the column are not

Fig. 2. Effect of OPE supplementation on PT, aPTT, and platelet aggregation. Testedby analysis of variance (ANOVA) with Duncan’s multiple range test. Mean ± S.E. CTL(n = 8), 2 mg OPE (n = 7), 10 mg OPE (n = 7); failure to obtain one blood sample in2 mg OPE group; one of 10 mg OPE group has died. Prothrombin time (PT), activatedpartial thromboplastin time (aPTT).

Fig. 3. Effect of OPE supplementation on arterial thrombus formation in vivo. Ratcarotid arteries were subjected to chemical injury by placing a 1 mm2 piece ofWhatman filter paper saturated with 70% of ferric chloride around the carotidartery for 5 min. Blood flow was then measured with a blood flowmeter. Tested byanalysis of variance (ANOVA) with Duncan’s multiple range test. Mean ± S.E. CTL(n = 8), 2 mg OPE (n = 8), 10 mg OPE (n = 8).

102 S.-M. Lee et al. / Food and Chemical Toxicology 57 (2013) 99–105

modulating the levels of a coagulation initiator. To test this idea,we utilized human umbilical cord vein endothelial cells (HUVECs)and measured the expression levels of tissue factor, an initiator ofblood coagulation. MTT assay was first performed to examine if the

OPE treatment influence the cell viability. The concentrations of upto 100 lg/ml of OPE did not significantly affect the viability of HU-VEC cells (data not shown). Incubation with thrombin greatly in-creased transcript and protein levels of tissue factors in HUVECcells. Pretreatment with 50 lg/ml or 100 lg/ml of the OPE signifi-cantly lowered thrombin-induced increase in mRNA and proteinlevels of tissue factor in a dose-dependent manner (Fig. 4A andB). The phosphorylation of JNK and ERK were detected upon thethrombin treatment (Fig. 4C and D). The pretreatment of the OPEat 100 lg/ml of concentration greatly reduced phosphorylated lev-els of JNK and ERK. These findings indicate that the OPE preventedthe over-expression of tissue factor that is induced by thrombin,which could be in part mediated by decreasing the activation ofMAPK signaling pathway.

4. Discussion

In the current study, we examined the effects of OPE as a foodsource containing high amount of quercetin on blood coagulationand thrombosis. Since excessive blood coagulation can be harmfulin the regulation of homeostasis, we first examined any significanteffect of the OPE on soluble clotting factors as measured by PT andaPTT and the tendency of platelet aggregation. PT or aPTT resultsindicate the efficiency of clotting pathway initiated in extrinsicor intrinsic manner, respectively. A deficiency in coagulation fac-tors prolongs PT or aPTT depending on their types of factors. Inour study, PT and aPTT were not affected by the 6 weeks of oralsupplementation of OPE, indicating that the OPE did not affectthe coagulation factors to alter blood coagulation. Platelet aggrega-tion was also examined as the percentage of aggregated platelets,which were compared between the control group and the OPEgroups. Adenosine diphosphate (ADP) was used to activate plateletaggregation in our study. G-protein-coupled purinergic receptorsmediate signaling for ADP-mediated platelet aggregation. 6 weeksof OPE ingestion did not affect ADP-induced platelet aggregation inresponse to stimulation.

Our results do not appear to be in accord with the previous re-ports (Hubbard et al., 2003, 2006) demonstrating that onion orquercetins a main bioactive component in it effectively reducedplatelet aggregation. The discrepancy could result from severalcauses. First, the concentrations of the OPE used in our studymay not be high enough to reach blood concentrations of quercetinto exert an effect in ex vivo platelet aggregation tests, resulting inlack of efficacy with doses of OPE used for the study. In anotherstudy showing quercetin effects on platelet aggregation, quercetinwas administered via the tail vein of mice (Mosawy et al., 2012). Insuch a way any loss of quercetin that may occur through intestinalabsorption could be avoided. Because we provided OPE orally, theuptake of quercetin into the system could be less than what wasprovided. Second, the OPE used for our study may have less effec-tive forms of quercetin possibly due to differences in modificationsof quercetin or the presence of unidentified inhibitors that couldlimit its effectiveness. In comparison, the quercetin effect shownby Hubarrd et al was derived from the use of purified quercetin(Hubbard et al., 2003). Thus, the degree of the effectiveness ofOPE may not be the same as those of a purified quercetin. Third,there is a chance to miss proper time points to detect the maxi-mum effects of OPE in platelet aggregation. Onion soup demon-strated its inhibitory effects on platelet aggregation greatly at 3 hafter ingestion (Hubbard et al., 2006). These effects seemed to betime-dependent, as judged by plasma concentration of quercetinthat peaked at 2 h and declined to the baseline levels within 24 hafter ingestion of onion soup (Hubbard et al., 2006). Likewise what-ever short term effects of the OPE might have lasted for only sev-eral hours, which were not timely detected in our long term

Fig. 4. OPE effects on expression of tissue factor and activation of JNK and ERK signaling in HUVEC cells. (A) mRNA expression of tissue factor in HUVEC cells was examinedupon 1 h pretreatment of various amounts (0, 50, and 100 lM) of OPE followed by thrombin-induced stimulation for 5 h, using real-time PCR. (B) Tissue factor protein levelsin HUVEC cells were examined upon 1 h pretreatment of various amounts (0, 50, and 100 lM) of OPE followed by thrombin-induced stimulation for 5 h, usingimmunoblotting. (C and D) Phosphorylated and unphosphorylated forms of JNK and ERK protein levels in HUVEC cells were examined upon 1 h pretreatment of variousamounts (0, 50, and 100 lM) of OPE followed by thrombin-induced stimulation for 5 h, using immunoblotting. The values from the independent experiments were quantified,normalized to GAPDH expression level and expressed as fold changes (A). b-actin was used as loading control. The representative image was shown (B, C, and D).

S.-M. Lee et al. / Food and Chemical Toxicology 57 (2013) 99–105 103

study probably due to rapid decomposition of biologically activecomponents during the metabolism. Lastly, long term intake ofthe OPE might have adjusted the system to make it more sensitiveto initiate platelet aggregation upon signal as a way to feedbackmechanism. On the other hand, it is also possible that the animalsfed on the regular diets maintained their normal blood coagulationsystem and the long term intake of OPE did not disturb this well-maintained blood coagulation system. Any significant effect ofthe OPE might have been detected better in the case of hypercoag-ulation state.

Although the OPE could not be effective in ex vivo platelet test, itis possible that the OPE affects the earlier step to trigger the produc-tion of coagulation initiator in blood coagulation process, which wasmissing in the ex vivo platelet aggregation test. In addition, in vivoplatelet aggregation can occur by various agonist besides ADP,which include thrombin, collagens, epinephrine, arachidonic acid(Zhou and Schmaier, 2005). Thus, it is not clear whether the OPEmight inhibit effectively platelet aggregation initiated by otherstimuli such as thrombin. In order to clarify the anti-thromboticfunction of the OPE in vivo, we next performed the in vivo arterialthrombosis test. A physical contact of ferric chloride solution withthe outside surface of artery is required to trigger coagulation (Kurz

et al., 1990). Surprisingly, the intake of the OPE with both concentra-tions caused a significant delay in in vivo arterial thrombosis. It isconsistent with the previous studies showing that quercetinpredominant in onion inhibited ferric chloride-induced arterialthrombosis (Mosawy et al., 2012). In addition, structurally querce-tin-related flavonol was also shown to inhibit in vivo arterial throm-bosis induced by laser (Jasuja et al., 2012). Our results are distinctfrom the earlier findings in that the use of the long-term querce-tin-rich OPE supplementation was shown to be effective to lowerthe risk of excess thrombus formation in the artery in vivo, furtherraising the questions about the anti-thrombotic mechanism.

Tissue factor is a crucial initiator of blood coagulation. The sub-stantial production of tissue factor is observed in the early stageof coagulation (Furie and Furie, 2008). Any delay in the productionof tissue factor during coagulation process may adversely affect aproper response to blood injury. On the other hand, hyper-activityof blood coagulation due to over-expression of tissue factor can beharmful, causing an increase in thrombotic risks (Korte et al.,2000; Tripodi et al., 2004). In addition, tissue factors in macrophagefoam cells are known to trigger the production of thrombin and fi-brin (Furie and Furie, 2008); thereafter thrombin also initiatesplatelet activation. Activated platelets could adhere to the vessel

104 S.-M. Lee et al. / Food and Chemical Toxicology 57 (2013) 99–105

walls causing certain disease conditions such as artery disease orstroke (Becker, 1999). Therefore, the expression of tissue factorneeds to be well-balanced in order to avoid hyper-activated bloodcoagulation and related thrombotic conditions while maintaininga blood coagulation role when needed. For instance, aberrantup-regulation of tissue factor expression is often related to thedevelopment of cardiovascular disorders (Chu, 2006). Based onthese considerations, the modulation of tissue factor expressionby the OPE could be one of the mechanisms of anti-thrombotic ac-tions of the OPE. In the present study, we exploited the cell culturesystem using human vascular endothelial cells (HUVECs) to test theeffect of the OPE on expression levels of tissue factor, because of thelimited amount of arterial tissues to be examined for the proteinlevels of tissue factor. Our results demonstrated that the pre-treat-ment of the OPE on HUVECs lowered thrombin-induced expressionof tissue factor. While the exact mechanism by which OPE exerts ananti-thrombotic effect in vivo is not known, we propose that throm-bin-enhanced tissue factor expression could be prevented by theOPE through inactivating MAPK signaling pathways. In the presentstudy, OPE pretreatment lowered phosphorylated levels of ERK andJNKs, suggesting that these MAPK pathways were down-regulated.Previously, chemical inhibitors specific to EKR, p38, and JNK signal-ing pathways were shown to block thrombin-induced tissue factorexpression (Liu et al., 2004; Stahli et al., 2006). Thus OPE treatmentwhich was effective to inhibit JNK and ERK signaling appears to in-hibit blood coagulation cascade partly through inactivating MAPKsignaling pathway and ultimately reduce thrombosis. Further stud-ies to identify the molecular pathway involved in this process areneeded.

Although the potential mechanism of an antithrombotic effect ofOPE is proposed in this study, the identification of the bioactivecomponents in OPE responsible for the observed effects still re-mains to be determined. Among the possible components containedin OPE, one could guess that the effects of the OPE might be mainlyderived from the quercetin in it. Earlier studies demonstrated thatquercetin, a predominant polyphenol contained in the OPE de-creased tissue factor expression in endothelial and mononuclearcells (Di Santo et al., 2003; Kaur et al., 2007). Overall, it appears thatthe quercetin-rich OPE when taken orally for at least 6 weeks mayreduce the risk of arterial thrombus formation by restricting hy-per-expression of tissue factor even without an effect on plateletfunction. It is also possible that OPE may modulate glycoproteinsignaling and expression of adhesion molecules and/or influenceother platelet activation signals other than ADP to diminish throm-bus formation in in vivo arterial thrombosis assay. Quercetin, whichis rich in the OPE, was proven to affect various steps in signaling re-lated to changes in intracellular calcium concentration and path-ways triggered by the collagen receptor, glycoprotein VI, whichresulted in the restriction of collagen-mediated activation of plate-let (Hubbard et al., 2003). Inhibition of glycoprotein VI by quercetinwas also proven in human studies, showing the healthy subjectssupplemented with quercetin or quercetin rich onion loweredphosphorylation of glycoprotein VI (Hubbard et al., 2006, 2004).Quercetin also inhibited inducible expression of intracellular adhe-sion molecule-1 (ICAM-1) that aid the recruitment of coagulationmolecules in human endothelial cells (Kobuchi et al., 1999). Onthe other hand, flavonoids such as genistein were proven to com-pete for binding to thromboxane A(2) (TxA2) receptor as a way toinhibit platelet function (Guerrero et al., 2005). Thus quercetin, akind of flavonoid, could also inhibit TxA2-mediated platelet aggre-gation. In addition, quercetin-impaired platelet aggregation wasshown to be related to its inhibition of kinase activity (Navarro-Nu-nez et al., 2009). In this respect, the OPE enriched with quercetinmight prevent arterial thrombosis not only by affecting tissue factorexpression but also by modulating glycoprotein VI signaling andexpression of adhesion molecules, as well as other agonist-activated

platelet function and kinase activity, which needs to be experimen-tally elucidated. However, one cannot exclude the possibility thatthe antithrombotic effects were derived from other compoundsthan quercetin. Indeed, it was demonstrated that sulfur-containingcompounds in onion inhibited platelet aggregation (Briggs et al.,2000; Morimitsu and Kawakishi, 1991; Morimitsu et al., 1992).Therefore, the identification of bioactive components in OPE dis-playing antithrombotic effects needs further study.

Finally, homeostatic abnormalities are associated with insulinresistance which are characterized by glucose intolerance (Wan-namethee et al., 2005) and dyslipidemia including hypercholester-olemia and hypertriglyceridemia (Lacoste et al., 1995). It can bealso speculated that the favorable effects of OPE on blood lipid pro-files shown in the present study together with the previous result(Lee et al., 2011) could favorably contribute to the anti-thromboticeffects of the OPE as measured by in vivo assay.

To conclude, the present data suggest that long-term OPE sup-plementation would be active in the suppression of thrombosis.The mechanism of action of the OPE could be partly through dimin-ishing MAPK signaling activation by thrombin and thereby block-ing the hyper-expression of tissue factor. Taken together, ourdata may provide an information to support the idea that long-term oral supplementations of the OPE could ameliorate arterialthrombosis and abnormal homeostasis.

5. Conflict of Interest

The authors state no conflict of interest.

Acknowledgements

This research was supported by Basic Science Research Programthrough the National research Foundation of Korea (NRF) fundedby the Ministry of Education, Science and Technology (2012-0002119) and High Value-added Food Technology DevelopmentProgram, Ministry for Food, Agriculture, Forestry and Fisheries,Republic of Korea.

References

Ali, M., Bordia, T., Mustafa, T., 1999. Effect of raw versus boiled aqueous extract ofgarlic and onion on platelet aggregation. Prostaglandins Leukot. Essent FattyAcids 60, 43–47.

Becker, R.C., 1999. Thrombosis and the role of the platelet. Am. J. Cardiol. 83, 3E–6E.Block, E., 1985. The chemistry of garlic and onions. Sci. Am. 252, 114–119.Braunwald, E., 2003. Application of current guidelines to the management of

unstable angina and non-ST-elevation myocardial infarction. Circulation 108,III28–III37.

Briggs, W.H., Xiao, H., Parkin, K.L., Shen, C., Goldman, I.L., 2000. Differentialinhibition of human platelet aggregation by selected allium thiosulfinates. J.Agric. Food Chem. 48, 5731–5735.

Briggs, W.H., Folts, J.D., Osman, H.E., Goldman, I.L., 2001. Administration of rawonion inhibits platelet-mediated thrombosis in dogs. J. Nutr. 131, 2619–2622.

Chesebro, J.H., Toschi, V., Lettino, M., Gallo, R., Badimon, J.J., Fallon, J.T., Fuster, V.,1995. Evolving concepts in the pathogenesis and treatment of arterialthrombosis. Mt. Sinai J. Med. 62, 275–286.

Chu, A.J., 2006. Tissue factor upregulation drives a thrombosis-inflammation circuitin relation to cardiovascular complications. Cell Biochem. Funct. 24, 173–192.

Di Santo, A., Mezzetti, A., Napoleone, E., Di Tommaso, R., Donati, M.B., De Gaetano,G., Lorenzet, R., 2003. Resveratrol and quercetin down-regulate tissue factorexpression by human stimulated vascular cells. J. Thromb. Haemostasis: JTH 1,1089–1095.

Falk, E., Fernandez-Ortiz, A., 1995. Role of thrombosis in atherosclerosis and itscomplications. Am. J. Cardiol. 75, 3B–11B.

Furie, B., Furie, B.C., 2008. Mechanisms of thrombus formation. New Engl. J. Med.359, 938–949.

Griffiths, G., Trueman, L., Crowther, T., Thomas, B., Smith, B., 2002. Onions–a globalbenefit to health. Phytother. Res. 16, 603–615.

Guerrero, J.A., Lozano, M.L., Castillo, J., Benavente-Garcia, O., Vicente, V., Rivera, J.,2005. Flavonoids inhibit platelet function through binding to the thromboxaneA2 receptor. J. Thromb. Haemostasis 3, 369–376.

Heemskerk, J.W., Bevers, E.M., Lindhout, T., 2002. Platelet activation and bloodcoagulation. Thromb. Haemostasis 88, 186–193.

S.-M. Lee et al. / Food and Chemical Toxicology 57 (2013) 99–105 105

Hubbard, G.P., Stevens, J.M., Cicmil, M., Sage, T., Jordan, P.A., Williams, C.M.,Lovegrove, J.A., Gibbins, J.M., 2003. Quercetin inhibits collagen-stimulatedplatelet activation through inhibition of multiple components of theglycoprotein VI signaling pathway. J. Thromb. Haemostasis 1, 1079–1088.

Hubbard, G.P., Wolffram, S., Lovegrove, J.A., Gibbins, J.M., 2004. Ingestion ofquercetin inhibits platelet aggregation and essential components of thecollagen-stimulated platelet activation pathway in humans. J. Thromb.Haemostasis 2, 2138–2145.

Hubbard, G.P., Wolffram, S., de Vos, R., Bovy, A., Gibbins, J.M., Lovegrove, J.A., 2006.Ingestion of onion soup high in quercetin inhibits platelet aggregation andessential components of the collagen-stimulated platelet activation pathway inman: a pilot study. Br. J. Nutr. 96, 482–488.

Jasuja, R., Passam, F.H., Kennedy, D.R., Kim, S.H., van Hessem, L., Lin, L., Bowley, S.R.,Joshi, S.S., Dilks, J.R., Furie, B., Furie, B.C., Flaumenhaft, R., 2012. Protein disulfideisomerase inhibitors constitute a new class of antithrombotic agents. J. Clin.Invest. 122, 2104–2113.

Kaur, G., Roberti, M., Raul, F., Pendurthi, U.R., 2007. Suppression of human monocytetissue factor induction by red wine phenolics and synthetic derivatives ofresveratrol. Thromb. Res. 119, 247–256.

Kobuchi, H., Roy, S., Sen, C.K., Nguyen, H.G., Packer, L., 1999. Quercetin inhibitsinducible ICAM-1 expression in human endothelial cells through the JNKpathway. Am. J. Physiol. 277, C403–C411.

Korte, W., Clarke, S., Lefkowitz, J.B., 2000. Short activated partial thromboplastintimes are related to increased thrombin generation and an increased risk forthromboembolism. Am. J. Clin. Pathol. 113, 123–127.

Kurz, K.D., Main, B.W., Sandusky, G.E., 1990. Rat model of arterial thrombosisinduced by ferric chloride. Thromb. Res. 60, 269–280.

Lacoste, L., Lam, J.Y., Hung, J., Letchacovski, G., Solymoss, C.B., Waters, D., 1995.Hyperlipidemia and coronary disease. Correction of the increased thrombogenicpotential with cholesterol reduction. Circulation 92, 3172–3177.

Lee, K.H., Park, E., Lee, H.J., Kim, M.O., Cha, Y.J., Kim, J.M., Lee, H., Shin, M.J., 2011.Effects of daily quercetin-rich supplementation on cardiometabolic risks inmale smokers. Nutr. Res. Pract. 5, 28–33.

Liu, Y., Pelekanakis, K., Woolkalis, M.J., 2004. Thrombin and tumor necrosis factoralpha synergistically stimulate tissue factor expression in human endothelialcells: regulation through c-Fos and c-Jun. J. Biol. Chem. 279, 36142–36147.

Miean, K.H., Mohamed, S., 2001. Flavonoid (myricetin, quercetin, kaempferol,luteolin, and apigenin) content of edible tropical plants. J. Agric. Food Chem. 49,3106–3112.

Morimitsu, Y., Kawakishi, S., 1991. Optical resolution of 1-(methylsulfinyl)propylalk(en)yl disulfides, inhibitors of platelet aggregation isolated from onion. Agric.Biol. Chem. 55, 889–890.

Morimitsu, Y., Morioka, Y., Kawakishi, S., 1992. Inhibitors of platelet aggregationgenerated from mixtures of allium species and/or S-alk(en)yl-l-cysteinesulfoxides. J. Agric. Food Chem. 40, 368–372.

Mosawy, S., Jackson, D.E., Woodman, O.L., Linden, M.D., 2012. Inhibition of platelet-mediated arterial thrombosis and platelet granule exocytosis by 30 ,40-dihydroxyflavonol and quercetin. Platelets.

Navarro-Nunez, L., Rivera, J., Guerrero, J.A., Martinez, C., Vicente, V., Lozano, M.L.,2009. Differential effects of quercetin, apigenin and genistein on signallingpathways of protease-activated receptors PAR(1) and PAR(4) in platelets. Br. J.Pharmacol. 158, 1548–1556.

Phillips, C., Poyser, N.L., 1978. Inhibition of platelet aggregation by onion extracts.Lancet 1, 1051–1052.

Ruggeri, Z.M., 2002. Platelets in atherothrombosis. Nat. Med. 8, 1227–1234.Srivastava, K.C., 1984. Aqueous extracts of onion, garlic and ginger inhibit platelet

aggregation and alter arachidonic acid metabolism. Biomed. Biochim. Acta 43,S335–S346.

Stahli, B.E., Camici, G.G., Steffel, J., Akhmedov, A., Shojaati, K., Graber, M., Luscher,T.F., Tanner, F.C., 2006. Paclitaxel enhances thrombin-induced endothelialtissue factor expression via c-Jun terminal NH2 kinase activation. Circ. Res. 99,149–155.

Thompson, S.G., Kienast, J., Pyke, S.D., Haverkate, F., van de Loo, J.C., 1995.Hemostatic factors and the risk of myocardial infarction or sudden deathin patients with angina pectoris. European concerted action on thrombosisand disabilities angina pectoris study group. New Engl. J. Med. 332, 635–641.

Tripodi, A., Chantarangkul, V., Martinelli, I., Bucciarelli, P., Mannucci, P.M., 2004. Ashortened activated partial thromboplastin time is associated with the risk ofvenous thromboembolism. Blood 104, 3631–3634.

Wannamethee, S.G., Lowe, G.D., Shaper, A.G., Rumley, A., Lennon, L., Whincup, P.H.,2005. The metabolic syndrome and insulin resistance: relationship tohaemostatic and inflammatory markers in older non-diabetic men.Atherosclerosis 181, 101–108.

Wu, K.K., Thiagarajan, P., 1996. Role of endothelium in thrombosis and hemostasis.Annu. Rev. Med. 47, 315–331.

Yamada, K., Naemura, A., Sawashita, N., Noguchi, Y., Yamamoto, J., 2004. An onionvariety has natural antithrombotic effect as assessed by thrombosis/thrombolysis models in rodents. Thromb. Res. 114, 213–220.

Zhou, L., Schmaier, A.H., 2005. Platelet aggregation testing in platelet-rich plasma:description of procedures with the aim to develop standards in the field. Am. J.Clin. Pathol. 123, 172–183.